4. Kay is studying the variables Daily Total Sunshine (\(x\)) and Daily Total Rainfall (\(y\)) from the large data set for Leeming in 2015
Kay starts with 5th May and then selects every 10th day thereafter.
(a) State the name of the sampling technique Kay uses. (1)
Kay wants to find the regression line of \(y\) on \(x\) for these data.
(b) Using your knowledge of the large data set, explain how Kay might need to clean these data before finding the equation of the regression line. (1)
The equation of the regression line Kay finds is \(y = 0.741 + 0.199x\)
(c) Using your knowledge of the large data set,
(i) state the units of the gradient of the regression line,
(ii) give an interpretation of the \(y\)-intercept of the regression line. (2)
Kay’s teacher claimed that the greater the amount of sunshine in a day the lower the amount of rain there should be.
(d) State, giving a reason, whether the teacher’s claim is true for Kay’s data. (1)
The teacher used all the data for these variables from the large data set for Leeming in 2015, as a sample. The teacher calculated the product moment correlation coefficient for \(x\) and \(y\) to be \(-0.160\)
In a suitable test to determine whether there is evidence to support the teacher’s claim, the \(p\)-value was 0.015
(e) Using a 5% level of significance, state the hypotheses and conclusion for this test. (2)
(f) For the test in part (e) describe
(i) the sample,
(ii) a possible population. (2)
Mark scheme (a)
Scheme
Marks
AO
Systematic (sampling)
B1
1.2
(1)
Notes
B1: for systematic (ignore other non-contradictory descriptions). Condone misspelling. If a clear choice is given e.g. stratified or systematic we take the final answer.
Mark scheme (b)
Scheme
Marks
AO
The (Daily Total) Rainfall data may contain “tr” entries (these will need a suitable value substituted before calculations can take place.)
B1
2.4
(1)
Notes
B1: for mention of “tr” or “trace” entries in rain(fall) [ or \(y\)] data. Ignore “n/a”. Ignore any comment about what to do with trace.
Mark scheme (c)
Scheme
Marks
AO
(i) mm/h (o.e.) e.g. \(\mathrm{mm\,h^{-1}}\) or \(\mathrm{mm\,hrs^{-1}}\) or \(\dfrac{\text{mm}}{\text{hours}}\)
B1
1.1b
(ii) When there is no sun(shine) there is (on average) 0.7(41) (mm) of rain or the amount of rain when there is no sun(shine) (o.e.)
B1
2.4
(2)
Notes
(i) B1: for mm/h or equivalent in words.
(ii) B1: for the idea of amount of rain when no sun. “Minimum rain” is B0 Don’t need value or units but if given must be correct or consistent with (i)
Mark scheme (d)
Scheme
Marks
AO
e.g. Not consistent (since); Kay’s line says positive correlation or gradient or regression coefficient or regression line or 0.199 is positive
B1
2.4
(1)
Notes
B1: for a suitable reason and saying not consistent (o.e.) e.g. “false” or “untrue” or “no” Reason only needs to be about the line and may be a description e.g. as \(x\) increases \(y\) increases or at least 2 values of \(x\) substituted and \(y\) values evaluated
[\(p\)-value < 5% so significant result] there is evidence to support the teacher’s claim
B1
2.2b
(2)
Notes
B1: for both hypotheses correct in terms of \(\rho\) (condone attempt at \(\rho\) that looks like \(p\))
B1: for a correct conclusion in context with no contradiction. Using words “support” and “teacher’s claim” or “negative correlation” and “sun or \(x\)” and “rain or \(y\)” [If a comparison is given it must be \(0.015 < 0.05\)]
Mark scheme (f)
Scheme
Marks
AO
(i) The sample is rainfall (\(y\)) and sunshine (\(x\)) for May~Oct in (Leeming) in 2015
B1
1.1b
(ii) e.g. (rainfall and sunshine) for: (Leeming) for all of 2015, or Leeming anytime (not just May~Oct) or Leeming for May~October for other years too or May~Oct in UK etc
B1
3.3
(2)
(9 marks)
Notes
(i) B1: for recognising that all the data/days for these variables in LDS for 2015 is the sample. Condone omitting “Leeming” B0 for “sunshine and rainfall for Leeming in 2015” It could be the population.
(ii) B1: for a suitable attempt to describe a population. The teacher’s sample must be a subset.
2. Amar is studying the flight of a bird from its nest.
He measures the bird’s height above the ground, \(h\) metres, at time \(t\) seconds for 10 values of \(t\) Amar finds the equation of the regression line for the data to be \(h = 38.6 - 1.28t\)
(a) Interpret the gradient of this line. (1)
The product moment correlation coefficient between \(h\) and \(t\) is \(-0.510\)
(b) Test whether or not there is evidence of a negative correlation between the height above the ground and the time during the flight. You should
state your hypotheses clearly
use a 5% level of significance
state the critical value used
(3)
Jane draws the following scatter diagram for Amar’s data.
(c) With reference to the scatter diagram, state, giving a reason, whether or not the regression line \(h = 38.6 - 1.28t\) is an appropriate model for these data. (1)
Jane suggests an improved model using the variable \(u = (t - k)^2\) where \(k\) is a constant.
She obtains the equation \(h = 38.1 - 0.78u\)
(d) Choose a suitable value for \(k\) to write Jane’s improved model for \(h\) in terms of \(t\) only. (1)
Mark scheme (a)
Scheme
Marks
AO
e.g. The height (\(h\))decreases by about 1.28 m for each second of the flight
B1
3.4
(1)
Notes
B1 for a suitable interpretation in context [value can be 1.3 or 1.28 or “just over 1”] per sec Must have underlined words (o.e.) and units “m” or metres and “s” or seconds NB “descends” implies “height decreases” Condone e.g. “decreases by \(-\,1.28\) m”
[\(r = -0.510\) not sig] there is insufficient (o.e.) evidence of a negative correlation between height (or \(\underline{h}\)) and time (or \(\underline{t}\))
A1
2.2b
(3)
Notes
B1 for both hypotheses correct in terms of \(\rho\) [accept a \(p\) or p but not \(r\) or r] Must be attached to \(\mathrm{H_0}\) and \(\mathrm{H_1}\)
M1 for a critical value corresponding to their \(\mathrm{H_1}\): 1-tail: awrt \(\pm\,0.549\) or 2-tail (B0 scored for \(\mathrm{H_1}\)): awrt \(\pm\,0.632\) (tables 0.6319) If hypotheses are in words and can deduce whether one or two-tail then use their words. If no hypotheses or their \(\mathrm{H_1}\) is not clearly one or two-tail assume one-tail
A1 a correct conclusion in context mentioning correlation and height and time A comparison or statement such as “not sig” is not needed but if seen must be correct. Do NOT award this A mark if contradictory comments or working seen e.g. “reject \(\mathrm{H_0}\)” or comparison of 0.510 with significance level of 0.05 or e.g. \(-0.549 \gt -0.510\)
NB Can award B0M1A1
SCB0(for 2-tail) M0(for cv = \(\pm\,0.549\)) scored: Allow 1 mark (score as B0M0A1) for conclusion such as: “insufficient evidence of (negative) correlation between height and time of flight”
Mark scheme (c)
Scheme
Marks
AO
No – since points seem to follow a curve/quadratic (rather than a line) or since points are “non-linear” but regression line/ model is linear or e.g. between (\(t = 5\) and 7) height drops by much more than 2.56 m or e.g. gradient is positive up to \(t = 3.5\) (line gradient \(\lt 0\)) or e.g. gradient is positive initially (line gradient \(\lt 0\)) or e.g. gradient is positive and then negative
B1
2.4
(1)
Notes
B1 for saying no and giving a suitable supporting reason Don’t allow “correlation” on its own instead of “gradient” B0 for simply saying “points don’t lie close to a straight line” Need mention of curve or some other feature of scatter plot that differs from regression line. B0 for just “non-linear” without mention of the model being linear B0 for simply comparing 1 or 2 points – need a comment about general pattern
Mark scheme (d)
Scheme
Marks
AO
[\(h = 38.1 - 0.78(t - k)^2\) with] a suitable \(k\) i.e. in the range 3~4.5
B1
3.3
(1)
(6 marks)
Notes
B1 for a value of \(k\) in the range [3, 4.5] Do not need \(k = \ldots\) Accept a value embedded in Jane’s model. ISW any errors in multiplying out bracket.
6. Anna is investigating the relationship between exercise and resting heart rate. She takes a random sample of 19 people in her year at school and records for each person
their resting heart rate, \(h\) beats per minute
the number of minutes, \(m\), spent exercising each week
Her results are shown on the scatter diagram.
(a) Interpret the nature of the relationship between \(h\) and \(m\) (1)
Anna codes the data using the formulae
\[\begin{gathered}x = \log_{10} m\\ y = \log_{10} h\end{gathered}\]
The product moment correlation coefficient between \(x\) and \(y\) is \(-0.897\)
(b) Test whether or not there is significant evidence of a negative correlation between \(x\) and \(y\) You should
state your hypotheses clearly
use a 5% level of significance
state the critical value used
(3)
The equation of the line of best fit of \(y\) on \(x\) is
\[y = -0.05x + 1.92\]
(c) Use the equation of the line of best fit of \(y\) on \(x\) to find a model for \(h\) on \(m\) in the form\[h = am^k\]where \(a\) and \(k\) are constants to be found. (5)
Mark scheme (a)
Scheme
Marks
AO
eg As the number of minutes exercise (\(m\)) increases the resting heart rate (\(h\)) decreases or the gradient of the curve is becoming flatter with increasing \(m\): diminishing effect of each additional minute of exercise
B1
2.4
(1)
Notes
B1: eg Idea as one increases the other decreases (in context). Allow use of \(m\) and \(h\) eg As \(m\) increases \(h\) decreases. Do not allow negative correlation with no context or \(\rho \lt 0\)
Allow there is a negative correlation/association/relationship/exponential between minutes exercise(\(m\)) and resting heart rate (\(h\)) oe
There is evidence that the product moment correlation is less than 0/ there is a negative correlation
A1
2.2b
(3)
Notes
B1: Both hypotheses correct in terms of \(\rho\) (allow p)
M1: For the cv of \(-0.3887\) or any cv such that \(0.3 \lt \lvert\text{cv}\rvert \lt 0.5\)
A1: Independent of hypotheses. Correct conclusion that implies reject \(\mathrm{H}_0\) on basis of seeing \(-0.3887\) or if they give 0.3887 we must see the comparison \(0.3887 \lt 0.897\) and which mentions “pmcc/correlation/relationship” and less than 0/ negative or \(\boldsymbol{\rho \lt 0}\)
\(\log_{10} h = -0.05\,\log_{10} m + 1.92\) or \(h = am^k \to \log_{10} h = \log_{10} am^k\)
M1
1.1b
\(\log_{10} h = -\log_{10} m^{0.05} + 1.92\) or \(\log_{10} h = \log_{10} m^{-0.05} + 1.92\) or \(h = 10^{1.92 - 0.05\log_{10} m}\) oe or \(\log_{10} h = \log_{10} a + \log_{10} m^k\) or \(\log_{10} a = 1.92\)
M1
2.1
\(\log_{10} hm^{0.05} = 1.92\) or \(\log_{10}\left(\dfrac{h}{m^{-0.05}}\right) = 1.92\) or \(h = 10^{1.92} \times 10^{-0.05\log_{10} m}\) oe or \(\log_{10} h = \log_{10} a + k\,\log_{10} m\)
M1
1.1b
\(hm^{0.05} = 10^{1.92}\) or \(\dfrac{h}{m^{-0.05}} = 10^{1.92}\) or \(h = 10^{1.92} \times 10^{\log_{10} m^{-0.05}}\) or \(\log_{10} a = 1.92\) and \(k = -0.05\)
M1
1.1b
\(h = 10^{1.92}m^{-0.05}\) or \(h = 83.17\ldots m^{-0.05}\) or \(a =\) awrt 83.17 and \(k = -0.05\)
A1
1.1b
(5)
(9 marks)
Notes
In this part once M0 is scored no more marks can be scored. Condone no base
M1:May be implied by 2nd M1 mark
Method 1: Correct substitution for both \(x\) and \(y\) Method 2: Taking the log of both sides
M1:May be implied by 3rd M1 mark
Method 1: Correct use of the power log rule or making \(h\) the subject
Method 2: Correct use of the addition/subtraction log rule
M1:This line implies M1M1M1
Method 1: Correct use of the addition/subtraction log rule or eqn in the form \(h = 10^{1.92} \times 10^{-0.05\log m}\)
Method 2: A second correct step for correct use of the power log rule
M1:This line implies M1M1M1M1
Method 1: Correct removal of logs or \(h = 10^{1.92} \times 10^{\log m^{-0.05}}\) Method 2: Log \(a\) (or \(a\)) and \(k\) correct
A1: Allow \(h =\) awrt \(83.2m^{-0.05}\)
NB award 5/5 for \(a =\) awrt 83.2 and \(k = -0.05\) or \(h =\) awrt \(83.2\ldots m^{-0.05}\) or \(h = 10^{1.92}m^{-0.05}\)
2. Marc took a random sample of 16 students from a school and for each student recorded
the number of letters, \(x\), in their last name
the number of letters, \(y\), in their first name
His results are shown in the scatter diagram on the next page.
(a) Describe the correlation between \(x\) and \(y\). (1)
Marc suggests that parents with long last names tend to give their children shorter first names.
(b) Using the scatter diagram comment on Marc’s suggestion, giving a reason for your answer. (1)
The results from Marc’s random sample of 16 observations are given in the table below.
\(x\)
3
6
8
7
5
3
11
3
4
5
4
9
7
10
6
6
\(y\)
7
7
4
4
6
8
5
5
8
4
7
4
5
5
6
3
(c) Use your calculator to find the product moment correlation coefficient between \(x\) and \(y\) for these data. (1)
(d) Test whether or not there is evidence of a negative correlation between the number of letters in the last name and the number of letters in the first name. You should
state your hypotheses clearly
use a 5% level of significance
(3)
Mark scheme (a)
Scheme
Marks
AO
Negative
B1
1.2
(1)
Notes
B1 for “negative” Allow “slight” or “weak” etc Allow a description e.g. “as \(x\) increases \(y\) decreases” or in context e.g. “people with longer last names tend to have shorter first names” A comment of “negative skew” is B0
Need to see distinct or separate responses for (a) and (b)
Mark scheme (b)
Scheme
Marks
AO
Marc’s suggestion is compatible because it’s negative correlation
B1
2.4
(1)
Notes
B1 for a comment that suggests data is compatible with the suggestion and a suitable reason such as “there is negative correlation” or a description in \(x\) and \(y\) or in context or the points lie close to a line with negative gradient or draw line \(y = x\) and state that more points below the line so supports (or is compatible with) his suggestion A reason based on just a single point is B0 e.g. “11 letters in last name has only 5 in first name”
(significant result / reject \(\mathrm{H}_0\)) There is evidence of negative correlation between the number of letters in (or length of) a student’s last name and their first name
A1
2.2b
(3)
(6 marks)
Notes
B1 for both hypotheses correct in terms of \(\rho\)
M1 for a critical value compatible with their \(\mathrm{H}_1\): 1-tail: awrt \(\pm\) 0.426 (condone \(\pm\) 0.425) or 2-tail (B0 scored for \(\mathrm{H}_1\)) : awrt \(\pm\) 0.497 If hypotheses are in words and can deduce whether one or two-tail then use their words. If no hypotheses or their \(\mathrm{H}_1\) is not clearly one or two tail assume one-tail
A1 for compatible signs between cv and \(r\) and a correct conclusion in context mentioning correlation and number of letters or length and name (ft their value from (c)) Do NOT award this A mark if contradictory comments or working seen e.g. “accept \(\mathrm{H}_0\)” or comparison of 0.426 with significance level of 0.05 etc
NB The M1A1 can be scored independently of the hypotheses
2. A random sample of 15 days is taken from the large data set for Perth in June and July 1987.
The scatter diagram in Figure 1 displays the values of two of the variables for these 15 days.
Figure 1
(a) Describe the correlation. (1)
The variable on the \(x\)-axis is Daily Mean Temperature measured in °C.
(b) Using your knowledge of the large data set,
(i) suggest which variable is on the \(y\)-axis,
(ii) state the units that are used in the large data set for this variable. (2)
Stav believes that there is a correlation between Daily Total Sunshine and Daily Maximum Relative Humidity at Heathrow.
He calculates the product moment correlation coefficient between these two variables for a random sample of 30 days and obtains \(r = -0.377\)
(c) Carry out a suitable test to investigate Stav’s belief at a 5% level of significance. State clearly
your hypotheses
your critical value
(3)
On a random day at Heathrow the Daily Maximum Relative Humidity was 97%
(d) Comment on the number of hours of sunshine you would expect on that day, giving a reason for your answer. (1)
Mark scheme (a)
Scheme
Marks
AO
Negative
B1
1.2
(1)
Notes
B1: for stating negative. “Negative skew” is B0 though
Mark scheme (b)
Scheme
Marks
AO
(i) Rainfall or Pressure
B1
2.2b
(ii) mm or hPa or Pascals or hectopascals or mb or millibars
B1ft
1.1b
(2)
Notes
(b)(i) B1: for mentioning “rainfall” (allow “rain” or “precipitation”) or “pressure” (if more than 1 answer both must be correct) NB the other quantitative variable for Perth is: Daily Mean Wind Speed and scores B0 [Not allowed “wind speed” since \(r = +0.15\) and in winter might expect wind to raise temp]
(ii) B1ft: for giving the correct units. If Daily Mean Wind Speed (kn) or knots “Wind speed” and “knots” would score B0B1 but any other variable scores B0B0
\(r \lt -0.3610\) so significant result and there is evidence of a correlation between Daily Total Sunshine and Daily Maximum Relative Humidity
A1
2.2b
(3)
Notes
B1: for both hypotheses correct in terms of \(\rho\)
M1: for the correct critical value compatible with their H1: allow \(\pm 0.361(0)\) If the hypotheses are 1-tail then allow cv of \(\pm 0.3061\) e.g. Alternative hypothesis with \(r \lt \pm 0.377\) implies a one-tail test or H0 and H1 in words saying “H0: there is no correlation, H1: there is correlation” is two-tail If there are no hypotheses (or they are nonsensical) assume 2-tail so M1 for \(\pm 0.361(0)\)
A1: for a correct conclusion in context based on comparing \(-0.377\) with their cv. Condone incorrect inequality e.g. \(-0.3610 \lt -0.377\) as long as they reject H0 Do not accept contradictory statements such as “accept H0 so there is evidence of …” Can say “support for Stav’s belief”(o.e.e.g. “claim”) or “evidence of a correlation between sunshine and humidity” condone “negative correlation” or comments such as “if humidity is high amount of sunshine will be low”
Mark scheme (d)
Scheme
Marks
AO
Humidity is high and there is evidence of correlation and \(r \lt 0\) So expect amount of sunshine to be lower than the average for Heathrow(oe)
B1
2.2b
(1)
(7 marks)
Notes
B1: for stating low amount of sunshine (o. e.) and some reference to \(r \lt 0\) or fog Check for the following 2 features: (i) low sunshine: allow \(\leqslant 5\) hrs (LDS mean for 2015 is 5.3, humidity 97% is 4.1, \(\geqslant 97\)% is 3.1) (ii) negative correlation may be described in words e.g. “high humidity gives low sunshine” orfog (LDS says >95% humidity is foggy) so less sunshine
3. Barbara is investigating the relationship between average income (GDP per capita), \(x\) US dollars, and average annual carbon dioxide (CO2) emissions, \(y\) tonnes, for different countries.
She takes a random sample of 24 countries and finds the product moment correlation coefficient between average annual CO2 emissions and average income to be 0.446
(a) Stating your hypotheses clearly, test, at the 5% level of significance, whether or not the product moment correlation coefficient for all countries is greater than zero. (3)
Barbara believes that a non-linear model would be a better fit to the data. She codes the data using the coding \(m = \log_{10} x\) and \(c = \log_{10} y\) and obtains the model \(c = -1.82 + 0.89m\)
The product moment correlation coefficient between \(c\) and \(m\) is found to be 0.882
(b) Explain how this value supports Barbara’s belief. (1)
(c) Show that the relationship between \(y\) and \(x\) can be written in the form \(y = ax^n\) where \(a\) and \(n\) are constants to be found. (5)
\((0.446 \gt 0.3438)\) so there is evidence that the product moment correlation coefficient (pmcc) is greater than 0/there is positive correlation
A1
2.2b
(3)
Notes
B1: for both hypotheses correct in terms of \(\rho\)
M1: for the critical value: sight of 0.3438 or any cv such that \(0.25 \lt |\text{cv}| \lt 0.45\)
A1: a comment suggesting a significant result/\(\mathrm{H}_0\) is rejected on the basis of seeing +0.3438 and which mentions “pmcc/correlation/relationship” and “greater than 0/positive” (not just \(\rho \gt 0\)) or an answer in context e.g. ‘as “income”(o.e.) increases, “CO2/emissions”(o.e.) increases’ A contradictory statement scores A0 e.g. ‘Accept H0, therefore positive correlation’
Mark scheme (b)
Scheme
Marks
AO
The value is close(r) to 1 or there is strong(er) (positive) correlation
B1
2.4
(1)
Notes
B1: for suitable reason e.g. \(r\) is close(r) to 1 or “strong(er)”/“near perfect” “correlation” Do not allow ‘association’
Mark scheme (c)
Scheme
Marks
AO
Method 1 \(\log_{10} y = -1.82 + 0.89(\log_{10} x)\)
\(y = ax^n \to\) \(\log_{10} y = \log_{10}(ax^n)\)
M1
1.1b
\(\log_{10} y = \log_{10} a + \log_{10} x^n\)
M1
2.1
\(\log_{10} y = \log_{10} a + n\log_{10} x\) \([\log_{10} a = -1.82,\ n = 0.89]\)
M1
1.1b
\(y = 0.015x^{0.89}\)
A1A1
1.1b 1.1b
(5)
For both methods, once an M0 is scored, no further marks can be awarded and condone missing base 10 throughout
Method 1: (working to the model)
M1: Correct substitution for both \(c\) and \(m\) (may be implied by 2nd M1 mark)
M1: Making \(y\) the subject to give an equation in the form \(y = 10^{a + b(\log_{10} x)}\) (may be implied by 3rd M1 mark)
M1: Correct multiplication to give an equation in the form \(y = 10^a \times 10^{b(\log_{10} x)}\) (this line implies M1M1M1 provided no previous incorrect working seen)
Method 2: (working from the model)
M1: Taking the log of both sides (may be implied by 2nd M1 mark)
M1: Correct use of addition rule (may be implied by 3rd M1 mark)
M1: Correct multiplication of power (this line implies M1M1M1 provided no previous incorrect working seen)
Both methods
A1: \(n = 0.89\) or \(a = \text{awrt } 0.015\) or \(y = ax^{0.89}\) or \(y = \text{awrt } 0.015x^n\) (dep on M3)
A1: \(n = 0.89\) and \(a = \text{awrt } 0.015\) / \(y = \text{awrt } 0.015x^{0.89}\) (dep on M3) do not award the final A1 if answer is given in an incorrect form e.g. \(y = 0.015 + x^{0.89}\)
2. Tessa owns a small clothes shop in a seaside town. She records the weekly sales figures, £\(w\), and the average weekly temperature, \(t\)°C, for 8 weeks during the summer.
The product moment correlation coefficient for these data is \(-0.915\)
(a) Stating your hypotheses clearly and using a 5% level of significance, test whether or not the correlation between sales figures and average weekly temperature is negative. (3)
(b) Suggest a possible reason for this correlation. (1)
Tessa suggests that a linear regression model could be used to model these data.
(c) State, giving a reason, whether or not the correlation coefficient is consistent with Tessa’s suggestion. (1)
(d) State, giving a reason, which variable would be the explanatory variable. (1)
Tessa calculated the linear regression equation as \(w = 10\,755 - 171t\)
(e) Give an interpretation of the gradient of this regression equation. (1)
Critical value: \(-0.6215\) (Allow any cv in range \(0.5 \lt |\text{cv}| \lt 0.75\))
M1
1.1a
\(r \lt -0.6215\) so significant result and there is evidence of a negative correlation between \(w\) and \(t\)
A1
2.2b
(3)
Notes
B1 for both hypotheses in terms of \(\rho\)
M1 for the critical value: sight of \(\pm 0.6215\) or any cv such that \(0.5 \lt |\text{cv}| \lt 0.75\)
A1 must reject \(\mathrm{H}_0\) on basis of comparing \(-0.915\) with \(-0.6215\) (if \(-0.915 \lt 0.6215\) is seen then A0 but may use \(|r|\) o.e. which is fine) and mention “negative”, “correlation/relationship” and at least “\(w\)” and “\(t\)”
Mark scheme (b)
Scheme
Marks
AO
e.g. As temperature increases people spend more time on the beach and less time shopping (o.e.)
B1
2.4
(1)
Notes
B1 for a suitable reason to explain negative correlation using the context given. e.g. “As temperature drops people are more likely to go shopping (than to the beach)” e.g. “As temperature increases people will be outside rather than in shops” A mere description in context of negative correlation is B0
SO e.g. “As temperature increases people don’t want to go shopping/buy clothes” is B0 e.g. “Less clothes needed as temp increases” is B0
Mark scheme (c)
Scheme
Marks
AO
Since \(r\) is close to \(-1\), it is consistent with the suggestion
B1
2.4
(1)
Notes
B1 for a suitable reason e.g. “strong”/“significant”/“near perfect” “correlation”, \(|r|\) close to 1 and saying it is consistent with the suggestion. Allow “yes” followed by the reason.
Mark scheme (d)
Scheme
Marks
AO
\(t\) will be the explanatory variable since sales are likely to depend on the temperature
B1
2.4
(1)
Notes
B1 For identifying \(t\) and giving a suitable reason. Need idea that “\(w\) depends on \(t\)” or “\(w\) responds to \(t\)” or “\(t\) affects \(w\)” (o.e.) Allow \(t\) (temperature) affects the other variable etc Just saying “\(t\) is the independent variable” or “\(t\) explains change in \(w\)” is B0 N. B. Suggesting causation is B0 e.g. “\(t\) causes \(w\) to decrease”
Mark scheme (e)
Scheme
Marks
AO
Every degree rise in temperature leads to a drop in weekly earnings of £171
B1
3.4
(1)
(7 marks)
Notes
B1 for a description that conveys the idea of rate per degree Celsius. Must have 171, condone missing “£” sign.
6. A group of climbers collected information about the height above sea level, \(h\) metres, and the air temperature, \(t\) °C, at the same time at 8 different points on the same mountain.
The data are summarised by
\[\sum h = 6370 \qquad \sum t = 61 \qquad \sum th = 31\,070 \qquad \sum t^2 = 693\]
(a) Show that \(\mathrm{S}_{th} = -17\,501.25\) and \(\mathrm{S}_{tt} = 227.875\) (3)
The product moment correlation coefficient for these data is \(-0.985\)
(b) State, giving a reason, whether or not this value supports the use of a regression equation to predict the air temperature at different heights on this mountain. (1)
(c) Find the equation of the regression line of \(t\) on \(h\), giving your answer in the form \(t = a + bh\). Give the value of your coefficients to 3 significant figures. (7)
(d) Give an interpretation of your value of \(a\). (1)
One of the climbers has just stopped for a short break before climbing the next 150 metres.
(e) Estimate the drop in temperature over this 150 metre climb. (2)
\((\mathrm{S}_{th}) = -17\,501.25\) and \((\mathrm{S}_{tt}) = 227.875\) (*)
A1cso
(3)
Notes
1st M1 for a correct expression for \(\mathrm{S}_{th}\) 2nd M1 for a correct expression for \(\mathrm{S}_{tt}\) Allow 1 slip e.g. 6730; \(61 \times 6370\) or 388570; \(61^2\) or 3721. Consistent use of \(n \neq 8\) M0M1
A1cso for both answers correct and both Ms scored.
Mark scheme (b)
Scheme
Marks
\(|r|\) close to 1 or \(r\) is close to \(-1\) therefore it does support the linear model
B1
(1)
Notes
B1 for correct and relevant comment about the value of \(r\) and saying it does support or “yes” Allow “ it is..”“strong” or “near perfect” correlation BUT B0 for “perfect”or “highly negative”
Mark scheme (c)
Scheme
Marks
\(\left[r = \dfrac{\mathrm{S}_{yx}}{\sqrt{\mathrm{S}_{yy} \times \mathrm{S}_{xx}}}\right]\) so \(r = \dfrac{\mathrm{S}_{th}}{\sqrt{\mathrm{S}_{tt} \times \mathrm{S}_{hh}}}\) or \(r^2 = \dfrac{(\mathrm{S}_{th})^2}{\mathrm{S}_{tt} \times \mathrm{S}_{hh}}\) or \(\mathrm{S}_{hh} = \dfrac{(\mathrm{S}_{th})^2}{r^2 \times \mathrm{S}_{tt}}\) or substitute 1 value
In (c) condone \(x\) for \(h\) and \(y\) for \(t\) except in 4th A1
1st M1 for the sight of the formula for \(r\) and an attempt to do something useful with it
1st A1 for a correct numerical expr’n in \(\mathrm{S}_{hh}\) or \(\sqrt{\mathrm{S}_{hh}}\) Accept with 3sf values (ignore – signs)
2nd A1 for awrt 1 390 000 (3sf gives 1 384 422.948 but scores 1st A1 and 2nd A0)
2nd M1 for a correct expression for \(b\) seen (ft their values to 3sf) Use of \(\mathrm{S}_{tt} \to -76.8\) is M0
3rd A1 for awrt – 0.013 (candidates using 3sf for \(\mathrm{S}_{hh}\) should therefore get this)
Beware \(\frac{\mathrm{S}_{tt}}{\mathrm{S}_{th}} = \frac{227.875}{-17501.25} = -0.0130\ldots\) but is 2nd M0 3rd A0 Ans only of – 0.0126..is M1A1A1M1A1
3rd M1 for a correct use of \(\bar{t}\) and \(\bar{h}\) to find \(a\) ft their \(b\) (allow letter \(b\) or even \(b = -0.985\) )
4th A1 for a correct equation with \(a\) = awrt 17.7 and \(b\) = awrt – 0.0126 [No \(y\) and \(x\)]
Mark scheme (d)
Scheme
Marks
\(a\) is an estimate of the temperature at sea level is (17.7 °C)
B1
(1)
Notes
B1 for stating or implying that it is the temperature (value not needed) at sea level
M1 for a correct expression equivalent to \((\mp)\) \(150b\). Can use letter \(b\) or ft their value(s).
A1 for awrt 2 (°C not required) Allow \(\pm\) can give if “\(a\)” incorrect or “\(b\)” from M0A0 in (c) Common wrong answer of 11520 can score M1A0 even if no working seen.
1. A clothes shop manager records the weekly sales figures, £\(s\), and the average weekly temperature, \(t\) °C, for 6 weeks during the summer. The sales figures were coded so that
(a) Find \(\mathrm{S}_{wt}\) and \(\mathrm{S}_{tt}\) (3)
(b) Write down the value of \(\mathrm{S}_{ss}\) and the value of \(\mathrm{S}_{st}\) (2)
(c) Find the product moment correlation coefficient between \(s\) and \(t\). (2)
The manager of the clothes shop believes that a linear regression model may be appropriate to describe these data.
(d) State, giving a reason, whether or not your value of the correlation coefficient supports the manager’s belief. (1)
(e) Find the equation of the regression line of \(w\) on \(t\), giving your answer in the form \(w = a + bt\) (3)
(f) Hence find the equation of the regression line of \(s\) on \(t\), giving your answer in the form \(s = c + dt\), where \(c\) and \(d\) are correct to 3 significant figures. (2)
(g) Using your equation in part (f), interpret the effect of a 1 °C increase in average weekly temperature on weekly sales during the summer. (1)
M1 for a correct expression using their values provided \(\mathrm{S}_{tt}\) and \(\mathrm{S}_{ss}\) both > 0
A1 for awrt – 0.801 (Correct ans. only M1A1, – 0.80 with no working M1A0)
Mark scheme (d)
Scheme
Marks
\(r\) is close to – 1 or \(|r|\) is close to 1 or “strong” (o.e.) [negative] correlation … so “yes” ordoes support the belief
B1ft
(1)
Notes
B1ft for a correct comment that uses their value of \(r\) as support, provided \(0.5 \leqslant |r| \leqslant 1\) For \(|r| \lt 0.5\) comment must be “does not support”, because “weak” (o.e.) correlation.
NB “points lie close to a straight line” is B0 unless supported by mention of their value of \(r\)
Mark scheme (e)
Scheme
Marks
\(b = \dfrac{\text{"}-49\text{"}}{\text{"}74.8\dot{3}\text{"}} = [-0.6547\ldots],\ \ a = \dfrac{42}{6} - b \times \dfrac{119}{6} = [19.9866\ldots]\) or \(a = 7 - b \times 19.8\dot{3}\)
M1, M1
So \(w = 20.0 - 0.655t\)
A1
(3)
Notes
1st M1 for a correct expression for \(b\) or awrt \(-0.66\) or \(-0.65\) Ft their answers from (a)
2nd M1 for a correct expression for \(a\) ft their value for \(b\)
A1 for a correct equation in \(w\) and \(t\) only with \(a = 20\) or awrt 20.0 and \(b\) = awrt – 0.655 (No fractions) If their \(a\) and \(b\) are given to more than 3 sf, accept answers in (f) to 3sf or better.
Mark scheme (f)
Scheme
Marks
\(s = 20\,000 - 655t\) or \(c = 20\,000\) and \(d = -655\)
B1ft B1ft
(2)
Notes
1st B1 ft for correct \(c\) or “their 20.0”\(\times 1000\) 2nd B1ft for correct \(d\) or their “– 0.655”\(\times 1000\) Values can be in an \(s\), \(t\) eq’n or \(c =\), \(d =\) (Their \(a\) and \(b\) needn’t be to 3 sf and ft their letter for \(t\))
Mark scheme (g)
Scheme
Marks
Decrease in sales of [£] 655 ( ignore any minus sign)
B1ft
(1)
(14 marks)
Notes
B1ft for stating clearly both decrease (o.e.) and [£] 655. Ft their \(d\) and allow “increase” if \(d \gt 0\)
3. Before going on holiday to Seapron, Tania records the weekly rainfall (\(x\) mm) at Seapron for 8 weeks during the summer. Her results are summarised as
\[\sum x = 86.8 \qquad \sum x^2 = 985.88\]
(a) Find the standard deviation, \(\sigma_x\), for these data. (3)
Tania also records the number of hours of sunshine (\(y\) hours) per week at Seapron for these 8 weeks and obtains the following
(b) Show that \(\mathrm{S}_{yy} = 716\) (correct to 3 significant figures) (1)
(c) Find \(\mathrm{S}_{xy}\) (2)
(d) Calculate the product moment correlation coefficient, \(r\), for these data. (2)
During Tania’s week-long holiday at Seapron there are 14 mm of rain and 70 hours of sunshine.
(e) State, giving a reason, what the effect of adding this information to the above data would be on the value of the product moment correlation coefficient. (2)
M1 for a correct expr’ for st. dev or variance (ignore label)[may be implied by 2.35 or 5.5125]
1st A1 for a correct expression for st. dev (must have square root) can ignore label
2nd A1 for awrt 2.35 (allow \(s = 2.5099\ldots\) or awrt 2.51). If they have \(\sigma^2 = 2.35\) score A0 but condone no label
Mark scheme (b)
Scheme
Marks
\(\mathrm{S}_{yy} = 8 \times \sigma_y^{\,2} = 716\) (3 sf) but may see \(1136.584 - \dfrac{58^2}{8}\) or \(27628(.084168) - \dfrac{464^2}{8}\) or 716.08... (= 716 to 3 sf) (*)
B1cso
(1)
Notes
B1cso for a correct expression or sight of at least 716.08… (NB limits: 716.00~716.16) Do not allow verification. Beware circular arguments: \(716 \to \Sigma y^2 \to\) expr’ \(\to 716\)
M1 for a correct expression for \(\mathrm{S}_{xy}\) (NB \(\Sigma y = 464\))
A1 for \(-133.9\) or awrt \(-134\) [No fractions] (Answer only 2/2)
Mark scheme (d)
Scheme
Marks
\(r = \dfrac{-133.9/8}{\sigma_x \times \sigma_y}\) or \(\dfrac{-133.9}{\sqrt{44.1 \times 716}}\)
M1
\(= \) awrt \(-\)0.753 or \(-\)0.754
A1
(2)
Notes
M1 for a correct expression for \(r\) (ft their values for \(\mathrm{S}_{xy}\) and \(\sigma_x\) or \(\mathrm{S}_{xx}\))[Allow ft of \(\mathrm{S}_{yy}\)]
A1 for awrt \(-0.753\) or \(-0.754\) (Answer only 2/2)
Mark scheme (e)
Scheme
Marks
\(r \lt 0\) means high sunshine and low rain; this is high sunshine high rain
B1
[this is not in keeping with the trend so] \(r\) is closer to 0 or \(|r|\) decreases
B1
(2)
(10 marks)
Notes
If they do not have an answer to (d) or their value of \(r\) is > 0 or \(|r| \gt 1\) score B0B0 here
1st B1 for a suitable reason contradicting \(r \lt 0\) e.g. new value is not in keeping with trendor both \(14 \gt \bar{x}\) and \(70 \gt \bar{y}\) or saying both above average. Allow for \(-0.48 \lt\) new \(r \lt -0.47\)
2nd B1 for a correct statement about \(r\) getting closer to zero e.g. \(|r|\) decreases A comment that \(r\) decreases or \(r\) is smaller or \(r\) is “less negative” is B0 “\(r\) increases” is B0 unless they also say that it gets closer to 0
1. A biologist is studying the behaviour of bees in a hive. Once a bee has located a source of food, it returns to the hive and performs a dance to indicate to the other bees how far away the source of the food is. The dance consists of a series of wiggles. The biologist records the distance, \(d\) metres, of the food source from the hive and the average number of wiggles, \(w\), in the dance.
Distance, \(d\) m
30
50
80
100
150
400
500
650
Average number of wiggles, \(w\)
0.725
1.210
1.775
2.250
3.518
6.382
8.185
9.555
[You may use \(\sum w = 33.6 \quad \sum dw = 13\,833 \quad \mathrm{S}_{dd} = 394\,600 \quad \mathrm{S}_{ww} = 80.481\) (to 3 decimal places)]
(a) Show that \(\mathrm{S}_{dw} = 5601\) (2)
(b) State, giving a reason, which is the response variable. (1)
(c) Calculate the product moment correlation coefficient for these data. (2)
(d) Calculate the equation of the regression line of \(w\) on \(d\), giving your answer in the form \(w = a + bd\) (4)
A new source of food is located 350 m from the hive.
(e)
(i) Use your regression equation to estimate the average number of wiggles in the corresponding dance.
(ii) Comment, giving a reason, on the reliability of your estimate. (2)
Mark scheme (a)
Scheme
Marks
\(\mathrm{S}_{dw} = 13833 - \dfrac{\text{"}1960\text{"} \times 33.6}{8}\) or \(13833 - \dfrac{65856}{8}\) (But 13833 – 8232 is M0)
M1
\(= \)5601 (*)
A1 cso
(2)
Notes
M1 for clear attempt to find \(\Sigma d\) and use in a correct formula. Accept \(1300 \lt \Sigma d \lt 2500\) For the M1 we can condone a single slip e.g. using 1383 instead of 13833 etc
A1cso for correct \(\Sigma d\) and 5601 only. Must see the formula and so have scored M1
Mark scheme (b)
Scheme
Marks
\(w\), since the number of wiggles depends on the distance or \(w\) depends on \(d\)
B1
(1)
Notes
B1 Must select \(w\) (or wiggles) and reason based on the idea that \(w\) is dependent on \(d\) Allow \(w\) “changes according to”/ “is determined/affected by” Must mention \(w\) and \(d\) B0 for “\(w\) is measured” or “\(d\) is explanatory/indep’t” or “\(w\) can’t be controlled” or “\(w\) responds to \(d\)”
1st M1 for a correct expression for \(b\). (Allow ft of their incorrect \(\mathrm{S}_{dw}\))
1st A1 for awrt 0.014 No fractions. [Answer only 2/2] Can be given at final equation. [Must come from correct formula not gradient of line from e.g. (650, 9.555) to (30, 0.725)]
2nd M1 for a correct method for \(a\). Follow through their value of \(b\) and their \(\Sigma d\)
2nd A1 for a correct equation for \(w\) and \(d\) with \(a\) = awrt 0.722 and \(b\) = awrt0.0142 No fractions Equation in \(x\) and \(y\) is A0 Answer only 4/4
(ii) Reliable since 350 m is in the range of the data
B1
(2)
(11 marks)
Notes
1st B1 for awrt 5.7 or awrt 5.6
2nd B1 for a reason citing 350 (m) or mentioning \(d\) is in the range of the data and stating reliable. Allow “Interpolation (or not extrapolation) therefore reliable”. Saying “5.7 (or \(w\) or just “it”) is in the range” is B0 “accurate” instead of “reliable” is B0 “strong correlation” (without mention of interpolation o.e.) is B0 Apply ISW if a correct comment is seen.
4. Statistical models can provide a cheap and quick way to describe a real world situation.
(a) Give two other reasons why statistical models are used. (2)
A scientist wants to develop a model to describe the relationship between the average daily temperature, \(x\) °C, and her household’s daily energy consumption, \(y\) kWh, in winter.
A random sample of the average daily temperature and her household’s daily energy consumption are taken from 10 winter days and shown in the table.
\(x\)
\(-0.4\)
\(-0.2\)
0.3
0.8
1.1
1.4
1.8
2.1
2.5
2.6
\(y\)
28
30
26
25
26
27
26
24
22
21
[You may use \(\sum x^2 = 24.76 \qquad \sum y = 255 \qquad \sum xy = 283.8 \qquad \mathrm{S}_{xx} = 10.36\)]
(b) Find \(\mathrm{S}_{xy}\) for these data. (3)
(c) Find the equation of the regression line of \(y\) on \(x\) in the form \(y = a + bx\) Give the value of \(a\) and the value of \(b\) to 3 significant figures. (4)
(d) Give an interpretation of the value of \(a\) (1)
(e) Estimate her household’s daily energy consumption when the average daily temperature is 2 °C (2)
The scientist wants to use the linear regression model to predict her household’s energy consumption in the summer.
(f) Discuss the reliability of using this model to predict her household’s energy consumption in the summer. (2)
Mark scheme (a)
Scheme
Marks
To simplify (or represent) a real world problem (o.e.) To improve understanding (o.e.) To analyse a real world problem or can change variables/replicate easily (oe) To make predictions or find estimates (o.e.)
B1g B1h
(2)
Notes
Make sure reasons refer to models and not tests
1st B1g (be fairly generous) for a sensible reason not using “quick”, “cheap” or “describe”
2nd B1h (be slightly harder) for two convincing reasons (both based on the list above) Use professional judgement and mark as B0B0 or B1B0 or B1B1 do not use B0B1
M1 for a correct expression for \(b\) (ft their \(\mathrm{S}_{xy} \neq 283.8\))
A1 for awrt \(-2.1\) (allow \(-15/7\))
M1 for a correct expression for \(a\) and ft their 12 (allow use of a letter \(b\))
A1 for \(y = 28.1 - 2.14x\) (awrt 28.1 and awrt \(-2.14\)) Must be \(y\) and \(x\) and no fractions
Mark scheme (d)
Scheme
Marks
(28.1 kWh) of energy are used when the temperature is 0[°C]
B1
(1)
Notes
B1 for a contextualised interpretation e.g. the amount of energy used when temperature is 0[°C] or [28.1] kWh used when temp. is 0[°C] [Can ft their 28.1]Need temp or °sign [B0 for “value of \(y\) when \(x = 0\)” since no context in words]
Mark scheme (e)
Scheme
Marks
\(y = 28.1 - 2.14(2) = \)
M1
awrt 23.8
A1
(2)
Notes
M1 for substituting \(x = 2\) into their equation
Mark scheme (f)
Scheme
Marks
The regression model is based on temperatures from the winter,
B1
so not reliable in the summer. Stating it is reliable (whatever the reason) is B0B0
dB1
(2)
(14 marks)
Notes
B1 for reasoning to suggest that temperatures are different in summer or the model was based only on data from the winter. Allow mention of extrapolation (o.e.)
M1 for a correct expression (condone missing \(-\))
A1 for awrt \(-0.749\) (\(-0.75\) or awrt 0.749 with no working scores M1 A0).
Mark scheme (c)
Scheme
Marks
\(-0.749\)
B1ft
(1)
Notes
B1ft for \(-0.749\) or ft their answer to (b) to at least 2sf. Must be in the range \(-1 \lt \text{'}(b)\text{'} \lt 1\)
Mark scheme (d)
Scheme
Marks
House H: 156 400/85 = [£1840/m\(^2\) or \(q = 4\)] House J: 172 900/95 = [£1820/m\(^2\) or \(q = 2\)]
M1
Since (\(r = -0.749\),) there is negative correlation. or The higher the price (per square metre), the lower the distance from the train station.
dM1
Therefore…..House H is likely to be closer.
A1
(3)
(8 marks)
Notes
M1 for calculating price/square metre for both \(H\) and \(J\). Can be implied by sight of 1840 and 1820 (so OK if not labelled or mis-labelled) These may be seen in the table in the question. Allow comment like “\(H\) is £20/square metre more than \(J\)”
dM1 dependent on 1st M1 for a statement that correlation is negative or a contextualised interpretation of the negative correlation.
\(r \gt 0\) If \(r \gt 0\) allow equivalent statements about positive correlation
A1 (dependent on both Ms) for House H is likely to be closer (No ft if \(r \gt 0\))
3. A large company is analysing how much money it spends on paper in its offices every year. The number of employees, \(x\), and the amount of money spent on paper, \(p\) (£ hundreds), in 8 randomly selected offices are given in the table below.
\(x\)
8
9
12
14
7
3
16
19
\(p\) (£ hundreds)
40.5
36.1
30.4
39.4
32.6
31.1
43.4
45.7
(You may use \(\sum x^2 = 1160 \qquad \sum p = 299.2 \qquad \sum p^2 = 11\,422 \qquad \sum xp = 3449.5\))
(a) Show that \(S_{pp} = 231.92\) and find the value of \(S_{xx}\) and the value of \(S_{xp}\) (5)
(b) Calculate the product moment correlation coefficient between \(x\) and \(p\). (2)
The equation of the regression line of \(p\) on \(x\) is given in the form \(p = a + bx\).
(c) Show that, to 3 significant figures, \(b = 0.824\) and find the value of \(a\). (4)
(d) Estimate the amount of money spent on paper in an office with 10 employees. (2)
(e) Explain the effect each additional employee has on the amount of money spent on paper. (1)
Later the company realised it had made a mistake in adding up its costs, \(p\). The true costs were actually half of the values recorded. The product moment correlation coefficient and the equation of the linear regression line are recalculated using this information.
1st B1 for \(\sum x = 88\) seen. May be in a correct formula or implied by 192 or 158.3
2nd B1cso for a correct expression for \(S_{pp}\)
M1 for a correct expression for \(S_{xx}\) or \(S_{xp}\) (ft their \(\Sigma x\)). If we don’t see an explicit \(\Sigma x = k\) but consistent use of \(k\) instead of 88 in \(S_{xp}\) and \(S_{xx}\) then award M1
1st A1 for \(S_{xx} = 192\) 2nd A1 for \(S_{xp}\) = awrt 158
3. The table shows data on the number of visitors to the UK in a month, \(v\) (1000s), and the amount of money they spent, \(m\) (£ millions), for each of 8 months.
M1 for a correct expression for \(r\). Ans only of 0.96 or awrt 0.96 is M1A0. Ans only of 0.962 or awrt 0.962 is M1A1. Do not allow fractions for A1
Mark scheme (b)
Scheme
Marks
\(r\) is close to 1 or a strong correlation. [“points are close to a straight line” isB0] [Just “positive” correlation is B0] [Use of “relationship” or “skew” not “correlation” is B0]
B1
(1)
Notes
B1 for comment implying strong correlation. (e.g. big/high/clear etc) B0 if \(|r| \gt 1\)
M1 for 1326.25 – (‘their \(b\)’ \(\times\) 2423.75) Condone fractions or awrt 1330 for \(\bar{m}\) and awrt 2420 for \(\bar{v}\)
A1 for a correct equation in \(m\) and \(v\) with \(a\) = awrt \(-467\) and \(b\) = awrt 0.74. Condone \(\frac{2521}{3407}\) for \(b\) and \(\frac{-1591740}{3407}\) for \(a\). [Equation in \(y\) and \(x\) is A0]
Mark scheme (e)
Scheme
Marks
\(b\) is the money(spent) per visitor. (i.e. definition of a rate in words.)[ignore values]
B1
So each 1000 visitors generates an extra £0.74 million or each visitor spends £740 oe
B1ft
(2)
Notes
1st B1 for a correct definition of the rate in words. Must state or imply “money per visitor”. Allow alternative words or symbols e.g. £ or “pounds” for money, “people” for visitors etc
2nd B1ft for a correct numerical rate (ft their value of \(b\))
e.g. “each visitorspends £740” is B1B1, “\(b\) is the extra money spent per visitor” is B1B0 [no values] “\(b\) is increase of £0.74 million in \(m\) as \(v\) increases by 1000” is B0B1[£ for money but no “visitors”] “increase in \(m\) as \(v\) increases” is B0B0 [Idea of rate but letters not words and no numerical value of rate]
Mark scheme (f)
Scheme
Marks
\(m = -467 + 0.74\times 2500\)
M1
\(m = 1383\) (£ million) awrt 1380
A1
(2)
Notes
M1 sub. \(v = 2500\) into their equation. Simply substituting 2 500 000 is M0 (unless adjusted eqn)
A1 awrt 1380 units (£ and million not required)
Mark scheme (g)
Scheme
Marks
As 2500 is within the range of the data set or it involves interpolation.
B1
The value of money spent is reliable
dB1
(2)
(13 marks)
Notes
1st B1 for 2500 or 2 500 000 or visitors or \(v\) is in range. “it” is B0 unless \(v\) clearly implied
2nd dB1 for stating it is reliable. Dependent on previous B mark being awarded
“both \(v\) and \(m\) in range” or “1380 in range” is B0 but use ISW so “interpolation since both in range” scores B1 for the “interpolation”. “Not extrapolation” counts as “interpolation”
5. A researcher believes that parents with a short family name tended to give their children a long first name. A random sample of 10 children was selected and the number of letters in their family name, \(x\), and the number of letters in their first name, \(y\), were recorded.
(a) Find \(\mathrm{S}_{yy}\) and \(\mathrm{S}_{xy}\) (3)
(b) Calculate the product moment correlation coefficient, \(r\), between \(x\) and \(y\). (2)
(c) State, giving a reason, whether or not these data support the researcher’s belief. (2)
The researcher decides to add a child with family name “Turner” to the sample.
(d) Using the definition \(\mathrm{S}_{xx} = \sum(x - \bar{x})^2\), state the new value of \(\mathrm{S}_{xx}\) giving a reason for your answer. (2)
Given that the addition of the child with family name “Turner” to the sample leads to an increase in \(\mathrm{S}_{yy}\)
(e) use the definition \(\mathrm{S}_{xy} = \sum(x - \bar{x})(y - \bar{y})\) to determine whether or not the value of \(r\) will increase, decrease or stay the same. Give a reason for your answer. (2)
M1 for a correct expression for \(r\) – ft their 20.9 (provided it is > 0) and their 16. Use of 382 for 16 or 393 for 20.9 is M0
A1 for awrt 0.661
Mark scheme (c)
Scheme
Marks
Researcher’s belief suggests negative correlation, data suggests positive correlation
B1
So data does not support researcher’s belief
dB1
(2)
Notes
1st B1 for a suitable reason contrasting belief with data. They must state the sign (positive or negative) of the correlation of data or the belief and imply the other is opposite
2nd dB1 Dependent on a correct reason for saying it does not support the claim e.g. State “does not support the belief because data has positive correlation” scores B1B1 BUT State “does support the belief because data has positive correlation” scores B0B0
Mark scheme (d)
Scheme
Marks
New \(x\) equals \(\bar{x} = 6\)
B1
Since \(\mathrm{S}_{xx} = \sum(x - \bar{x})^2\) the value of \(\mathrm{S}_{xx}\) is the same = 28
dB1
(2)
Notes
1st B1 for clearly stating that new value of \(x = (6 =)\) mean
2nd dB1 Dep. on 1st B1 for a reason that shows \(\mathrm{S}_{xx}\) is unchanged e.g. extra term is 0 so \(\mathrm{S}_{xx}\) is the same
ALT 1st B1 for seeing \(\sum x = 66\) and new \(\sum x^2 = 424\) (or \(388 + 6^2\)) and attempt at \(\mathrm{S}_{xx}\) 2nd B1 for showing \(\mathrm{S}_{xx} = 28\) with \(n = 11\) and no incorrect working seen and a final comment
Mark scheme (e)
Scheme
Marks
\(\mathrm{S}_{xy} = \sum(x - \bar{x})(y - \bar{y}) = \sum(x - \bar{x})y\) so the new term will be zero (since mean = \(x\)) and since \(\mathrm{S}_{yy}\) increases
B1
So \(r\) will decrease
dB1
(2)
(11 marks)
Notes
1st B1 for a clear reason that mentions \(\mathrm{S}_{xy}\) is the same and the increase in \(\mathrm{S}_{yy}\). Saying that \(r\) increases or stays the same is B0B0
2nd dB1 Dependent on 1st B1 for saying \(r\) will decrease.
1. Sammy is studying the number of units of gas, \(g\), and the number of units of electricity, \(e\), used in her house each week. A random sample of 10 weeks use was recorded and the data for each week were coded so that \(x = \dfrac{g - 60}{4}\) and \(y = \dfrac{e}{10}\). The results for the coded data are summarised below
(a) Find the equation of the regression line of \(y\) on \(x\) in the form \(y = a + bx\). Give the values of \(a\) and \(b\) correct to 3 significant figures. (4)
(b) Hence find the equation of the regression line of \(e\) on \(g\) in the form \(e = c + dg\). Give the values of \(c\) and \(d\) correct to 2 significant figures. (4)
(c) Use your regression equation to estimate the number of units of electricity used in a week when 100 units of gas were used. (2)
1st M1 for substitutions into their equation to get an equation in \(e\) and \(g\). Need \(y = \frac{e}{10}\) and \(x = \frac{g - 60}{4}\)
2nd dM1 Dep. on 1st M1 for an attempt to simplify (at least removing fractions). Allow one slip
1st A1 for an equation \(e\) = awrt 46 \(\pm\) …..
2nd A1 for an equation \(e = \)….+ awrt \(0.15g\)
ALT 1st M1 for use of \(d = \dfrac{10\times\text{"their } b\text{"}}{4}\) or sight of 0.15 used as gradient 2nd dM1 Dep. on 1st M1 for use of \(\bar{e} = 10\times\text{"their } \bar{y}\text{"}\) or sight of 58 and use of \(\bar{g} = 4\times\text{"their } \bar{x}\text{"} + 60\) or sight of 79.2 and use of these values to find \(c\) in \(c = \bar{e} - d\bar{g}\)
1. A meteorologist believes that there is a relationship between the height above sea level, \(h\) m, and the air temperature, \(t\) °C. Data is collected at the same time from 9 different places on the same mountain. The data is summarised in the table below.
\(h\)
1400
1100
260
840
900
550
1230
100
770
\(t\)
3
10
20
9
10
13
5
24
16
[You may assume that \(\sum h = 7150\), \(\sum t = 110\), \(\sum h^2 = 7\,171\,500\), \(\sum t^2 = 1716\), \(\sum th = 64\,980\) and \(\mathrm{S}_{tt} = 371.56\)]
(a) Calculate \(\mathrm{S}_{th}\) and \(\mathrm{S}_{hh}\). Give your answers to 3 significant figures. (3)
(b) Calculate the product moment correlation coefficient for this data. (2)
(c) State whether or not your value supports the use of a regression equation to predict the air temperature at different heights on this mountain. Give a reason for your answer. (1)
(d) Find the equation of the regression line of \(t\) on \(h\) giving your answer in the form \(t = a + bh\). (4)
(e) Interpret the value of \(b\). (1)
(f) Estimate the difference in air temperature between a height of 500 m and a height of 1000 m. (2)
M1 for at least one correct expression (condone transcription error)
1st A1 for \(\mathrm{S}_{hh} =\) awrt 1 490 000 or \(\mathrm{S}_{th} =\) awrt \(-22\,400\) (Condone \(S_{xx}\) or \(S_{xy} = \ldots\) or even \(S_{yy} = \ldots\))
2nd A1 for \(\mathrm{S}_{th} = -22\,400\) and \(\mathrm{S}_{hh} = 1\,490\,000\) only. [This mark is assessing correct rounding] (Allow no labels but mis-labelling \(\mathrm{S}_{th}\) as \(\mathrm{S}_{hh}\) etc loses the final A1)
M1 for attempt at correct formula. Allow minor transcription errors of 2 or 3 digits. Must have their \(\mathrm{S}_{hh}\), \(\mathrm{S}_{th}\) and given \(\mathrm{S}_{tt}\) (3sf or better) in the correct places. Condone missing “−” Award M1A0 for awrt \(-0.95\) with no expression seen. M0 for \(\dfrac{64980}{\sqrt{7171500\times 1716}}\) (corrected from the printed mark scheme, which has 7.864 in place of 1716)
Mark scheme (c)
Scheme
Marks
Yes as \(r\) is close to \(-1\) (if \(-1 \lt r \lt -0.5\)) or Yes as \(r\) is close to 1 (if \(1 \gt r \gt 0.5\)) [ If \(-0.5 \leqslant r \leqslant 0.5\) allow “no since \(r\) is close to 0”] [ If \(|r| \gt 1\) award B0]
B1ft
(1)
Notes
B1ft must comment on supporting and state: high/strong/clear (negative or positive) correlation “points lie close to a straight line” is B0 since there is no evidence of this.
\(= \pm 7.5\) (awrt \(\pm 7.5\)) (only ft a value < 100)
A1ft
(2)
(13 marks)
Notes
M1 for a correct expression seen based on their equation. Allow transcription error of 1 digit. If answer is \(500\times\) their \(b\) to 2sf and < 100 (M1A1), If answer is \(500\times\) their \(b\) to 2sf and \(\geqslant 100\) (M1A0)
3. A biologist is comparing the intervals (\(m\) seconds) between the mating calls of a certain species of tree frog and the surrounding temperature (\(t\) °C). The following results were obtained.
\(t\) °C
8
13
14
15
15
20
25
30
\(m\) secs
6.5
4.5
6
5
4
3
2
1
(You may use \(\sum tm = 469.5\), \(\mathrm{S}_{tt} = 354\), \(\mathrm{S}_{mm} = 25.5\))
(a) Show that \(\mathrm{S}_{tm} = -90.5\) (4)
(b) Find the equation of the regression line of \(m\) on \(t\) giving your answer in the form \(m = a + bt\). (4)
(c) Use your regression line to estimate the time interval between mating calls when the surrounding temperature is 10 °C. (1)
(d) Comment on the reliability of this estimate, giving a reason for your answer. (1)
Mark scheme (a)
Scheme
Marks
\(\sum t = 140\) (or \(\bar{t} = 17.5\)) and \(\sum m = 32\) (or \(\bar{m} = 4\))
1st B1 for 140 seen in correct context or correctly labelled
2nd B1 for 32 seen in correct context or correctly labelled. (allow a fully correct expression – not “…+…+…”) 4480 used correctly is B1B1
M1 for attempting a correct expression. Follow through their 140 and their 32 You may see attempt at \(\sum(t - \bar{t})(m - \bar{m})\). This must have all the products seen.
A1cso requires a correct expression seen and no incorrect working leading to \(-90.5\)
So equation of the line is \(\boldsymbol{m = 8.47 - 0.256t}\) (allow \(m = \frac{11999}{1416} - \frac{181}{708}t\))
A1
(4)
Notes
1st M1 for a correct expression for \(b\). Follow through their \(\mathrm{S}_{tm}\). Condone missing “−”
1st A1 for awrt \(-0.26\) or condone \(-0.25\)
2nd M1 for a correct method for \(a\). Follow through their sums from part (a) and their value of \(b\)
2nd A1 for a correct equation for \(m\) and \(t\) with \(a\) = awrt 8.47 and \(b\) = awrt \(-0.256\). Must be an equation in \(m\) and \(t\), use of \(x\) or \(y\) scores A0 here.
Mark scheme (c)
Scheme
Marks
\((8.47 - 0.256\times 10 =)\ 5.9\ldots\) awrt 5.9
B1
(1)
Notes
B1 for awrt 5.9 Accept 6 if the correct expression (awrt \(8.47 - 10\times\) awrt 0.256) is seen
Mark scheme (d)
Scheme
Marks
Should be reliable since 10 is in the range (of the data)
B1
(1)
(10 marks)
Notes
B1 for suggesting it is reliable and mentioning 10 within the range (of the data.) or suggesting it is reliable since interpolating or not extrapolating
NB “it is reliable since it is in the range” is B0 since “it” is not explicit enough Condone extra non-relevant comments but penalise contradictory comments. e.g. “near the extreme so not reliable but not extrapolated so reliable” is B0 since contradicts “reliable since 10 is within the range (of temps) and 5.9 within range of times” is B1 since irrelevant
1. A teacher asked a random sample of 10 students to record the number of hours of television, \(t\), they watched in the week before their mock exam. She then calculated their grade, \(g\), in their mock exam. The results are summarised as follows.
(a) Find \(\mathrm{S}_{tt}\) and \(\mathrm{S}_{gt}\) (3)
(b) Calculate, to 3 significant figures, the product moment correlation coefficient between \(t\) and \(g\). (2)
The teacher also recorded the number of hours of revision, \(v\), these 10 students completed during the week before their mock exam. The correlation coefficient between \(t\) and \(v\) was \(-0.753\)
(c) Describe, giving a reason, the nature of the correlation you would expect to find between \(v\) and \(g\). (2)
M1 for attempt at correct formula. Must have their \(\mathrm{S}_{tt}\), \(\mathrm{S}_{gt}\) and given \(\mathrm{S}_{gg}\) in the correct places. Condone missing “−” Award M1A0 for awrt \(-0.71\) with no expression seen M0 for \(\dfrac{1550.2}{\sqrt{8702\times 7.864}}\) Correct answer only is 2/2
Mark scheme (c)
Scheme
Marks
Positive
B1
e.g. high \(v\) corresponds to low \(t\) and low \(t\) corresponds to high \(g\) so expect high \(v\) to corresponds to high \(g\) or expect more revision to result in a better grade
B1
(2)
(7 marks)
Notes
1st B1 for saying "positive". Ignore mention of skew.
2nd B1 for suitable reason that mentions at least \(v\) and \(g\) and supports positive correlation. e.g. “the less revision done the lower the grade” is B1 “should do better with more revision” is B0 since does not mention grades “both coefficients are similar” or two sketches of negative correlation with labelled axes is B1 since \(v\), \(t\) and \(g\) are implied Allow use of letters \(v\) and \(g\) Allow equivalent terms e.g. “study” instead of “revision” or “score” instead of “grade”
3. A scientist is researching whether or not birds of prey exposed to pollutants lay eggs with thinner shells. He collects a random sample of egg shells from each of 6 different nests and tests for pollutant level, \(p\), and measures the thinning of the shell, \(t\). The results are shown in the table below.
\(p\)
3
8
30
25
15
12
\(t\)
1
3
9
10
5
6
[You may use \(\sum p^2 = 1967\) and \(\sum pt = 694\)]
(a) Draw a scatter diagram on the axes below to represent these data. (2)
(b) Explain why a linear regression model may be appropriate to describe the relationship between \(p\) and \(t\). (1)
(c) Calculate the value of \(S_{pt}\) and the value of \(S_{pp}\). (4)
(d) Find the equation of the regression line of \(t\) on \(p\), giving your answer in the form \(t = a + bp\). (4)
(e) Plot the point \((\bar{p}, \bar{t})\) and draw the regression line on your scatter diagram. (2)
The scientist reviews similar studies and finds that pollutant levels above 16 are likely to result in the death of a chick soon after hatching.
(f) Estimate the minimum thinning of the shell that is likely to result in the death of a chick. (2)
Mark scheme (a)
Scheme
Marks
Use overlay
B1 B1
(2)
Notes
B2 for all 6 data points plotted correctly. B1 for any 5 correct. Points not wholly outside the circles.
Mark scheme (b)
Scheme
Marks
Points (appear to) lie close to a (straight) line or “strong /high correlation”
B1
(1)
Mark scheme (c)
Scheme
Marks
\(\sum p = 93\) and \(\sum t = 34\) (may be seen in table)
1st M1 for attempting \(\sum p\) and \(\sum t\). Allow \(80 \lt \sum p \lt 100\) and \(30 \lt \sum t \lt 40\)
2nd M1 for one correct expression for \(S_{pt}\) or \(S_{pp}\), f.t. their \(\sum p\) and \(\sum t\).
1st A1 for \(S_{pt}\)
2nd A1 for \(S_{pp}\)
Mark scheme (d)
Scheme
Marks
\(b = \left[\dfrac{S_{pt}}{S_{pp}} =\right] \dfrac{\text{"}167\text{"}}{\text{"}525.5\text{"}} = [0.31779\ldots]\) (check their answer if expression not seen)
\(\boldsymbol{t = 0.741 + 0.318p}\) (Accept \(a = \frac{2336}{3153}\) and \(b = \frac{334}{1051}\) in their equation)
A1
(4)
Notes
B1ft for correct expression for the gradient, f.t. their 167 and 525.5 from (c)
M1 for correct use of \(a = \bar{t} - b\bar{p}\) f.t. their values. Condone 5.6 for \(\bar{t}\)
1st A1 for awrt 0.74 NB use of 526 gives 0.745566… and gets A0
2nd A1 for a correct equation for \(t\) in terms of \(p\) with \(a\) and \(b\) awrt 3sf. An equn in \(y\) or \(x\) is A0
Mark scheme (e)
Scheme
Marks
\((\bar{p}, \bar{t}) = (15.5, 5.7)\) plotted on the graph (not wholly outside the circle)
B1
Correct line plotted as per overlay. For \(p = 5\); \(2 \lt t \lt 3\) and for \(p = 30\); \(10 \lt t \lt 11\) Their line must stretch roughly as far as the points and go through the \((\bar{p}, \bar{t})\) circle
2. A bank reviews its customer records at the end of each month to find out how many customers have become unemployed, \(u\), and how many have had their house repossessed, \(h\), during that month. The bank codes the data using variables \(x = \dfrac{u - 100}{3}\) and \(y = \dfrac{h - 20}{7}\)
The results for the 12 months of 2009 are summarised below.
Marks for part (a) must be seen in (a), do not award if only seen in (b)
B1 for a correct expression for \(S_{xy}\)
M1 for correct attempt at \(r\) f.t. their 3990 but \(\dfrac{23070}{\sqrt{5606.25\times 4244}}\) is M0
A1 for awrt 0.818 If an answer of 0.82 only is seen then B1M1A0 can be given
Mark scheme (b)
Scheme
Marks
0.818
B1ft
(1)
Notes
B1ft for awrt 0.818 or f.t. their answer to part (a) for \(|r| \lt 1\). Allow 2sf or 1sf follow through. Answer in (b) must be correct or match one of their answers in (a). Must be a number.
Mark scheme (c)
Scheme
Marks
Positive correlation or value of \(r\) is close to 1 or value of \(r \gt 0\) (NOT “high/ strong correlation”)
B1
So there is support for the bank’s claim or “increase in unemployment is accompanied by increase in house repossessions”
B1
(2)
(6 marks)
Notes
1st B1 for a reason of positive correlation (allow even if \(r \gt 1\)). “positive skew” or “positive gradient” is B0 but 2nd B1 is still possible
2nd B1 for a comment that suggest this supports the claim. Marks in (c) are independent but first B1 requires some idea of positive correlation
(c) SC If \(|r| \lt 0.2\) allow this alternative to the mark scheme: 1st B1 for saying there is no or little correlation 2nd B1 for a comment that says this does not support the bank’s claim
(a) Find \(S_{tt}\) and \(S_{tw}\) for these data. (3)
(b) Calculate, to 3 significant figures, the product moment correlation coefficient between \(t\) and \(w\). (2)
(c) Find the equation of the regression line of \(w\) on \(t\) in the form \(w = a + bt\) (4)
(d) State, with a reason, which variable is the explanatory variable. (2)
(e) Using this model, estimate
(i) the weight of a coin which is 5 years old,
(ii) the effect of an increase of 4 years in age on the weight of a coin.
(2)
It was discovered that a coin in the original sample, which was 5 years old and weighed 20 grams, was a fake.
(f) State, without any further calculations, whether the exclusion of this coin would increase or decrease the value of the product moment correlation coefficient. Give a reason for your answer. (2)
M1 require ‘their -5.03’ as numerator and /their 191.6’ as denominator. A1 awrt -0.026 M1 for use of correct formula with \(b\) or ‘their \(b\)’; require \(- -\) or + and values in the correct place. A1 for equation as written with values awrt 3 sf. with \(w\) and \(t\). Accept fractional answers that are accurate to 3sf when evaluated as decimals
Mark scheme (d)
Scheme
Marks
The explanatory variable is the age of each coin. This is because the age is set and the weight varies.
B1 B1
(2)
Notes
B1 for ‘Age’ or \(t\) or ‘years’ B1 for ‘you use age / t to predict w’ or ‘you can control t/ age’ or ‘weight depends on age’ or similar
Mark scheme (e)
Scheme
Marks
(i) awrt 11.5
B1
(ii) Decrease(in weight of coin of 0.1052 g) = 0.1 or -0.1 or increase of -0.1 awrt(-0.1)
B1
(2)
Notes
B1 awrt 11.5 B1 awrt -0.1 but ‘decrease of -0.1’ is B0.
Mark scheme (f)
Scheme
Marks
Decrease; removing the fake will result in a better linear fit so \(r\) will be closer to -1
B1;B1
(2)
(15 marks)
Notes
B1 for Decrease only but ‘mod r increases’ explicitly stated in words or symbols award B1. B1 accept ‘stronger correlation’ or ‘increase in correlation’ or ‘better linear fit’ or ‘ \(r\) closer to -1’ or ‘points are closer to a straight line’ or ‘point is an outlier’ or equivalent
Special Case 1 Attempt to calculate \(S_{tw}\) \(\sum tw = 1660.62,\ \sum t = 153,\ \sum w = 91.75\) or \(S_{tw} = 1660.62 - \dfrac{153 \times 91.75}{9}\) or awrt 101 (corrected from the printed mark scheme: \(\sum tw = 1669.62\)) or \(S_{tw} \gt 0\) with some calculation B1 “Increase” B1 (2)
Special Case 2 Attempt to calculate \(S_{ww}\) \(\sum w^2 = 1248.96625 - 400 = 848.96625\) or awrt 849 or \(S_{ww} = 848.96625 - \dfrac{91.75^2}{9}\) or awrt -86.4 or \(S_{ww} \lt 0\) B2 (2)
Special Case 3 Argument based on standard deviation. e.g. \(\sigma_w \approx 0.126\) and \(\bar{w} = 11.175\) so fake coin is over 69 sds away from the mean B1 ‘(very) unlikely’ or ‘impossible’ B1 (2)
7. A teacher took a random sample of 8 children from a class. For each child the teacher recorded the length of their left foot, \(f\) cm, and their height, \(h\) cm. The results are given in the table below.
\(f\)
23
26
23
22
27
24
20
21
\(h\)
135
144
134
136
140
134
130
132
(You may use \(\sum f = 186 \quad \sum h = 1085 \quad \mathrm{S}_{ff} = 39.5 \quad \mathrm{S}_{hh} = 139.875 \quad \sum fh = 25291\))
(a) Calculate \(\mathrm{S}_{fh}\) (2)
(b) Find the equation of the regression line of \(h\) on \(f\) in the form \(h = a + bf\). Give the value of \(a\) and the value of \(b\) correct to 3 significant figures. (5)
(c) Use your equation to estimate the height of a child with a left foot length of 25 cm. (2)
(d) Comment on the reliability of your estimate in (c), giving a reason for your answer. (2)
The left foot length of the teacher is 25 cm.
(e) Give a reason why the equation in (b) should not be used to estimate the teacher’s height. (1)
1st M1 for a correct expression for \(b\), ft their part (a) but not \(\mathrm{S}_{fh}\) = 25291 1st A1 for awrt 1.6 2nd M1 for use of \(a = \bar{h} - b \times \bar{f}\), ft their value for \(b\). Must use \(\bar{h}\) and \(\bar{f}\) not values from table. 2nd A1 for awrt 97.5 [NB \(a\) = 135 – 1.63\(\times\)23 = 97.51 but M0A0 since not using \(\bar{h}\) and \(\bar{f}\) ] 3rd A1ft for an equation for \(h\) and \(f\) with their coefficients to 3sf. Dependent on both Ms Must be 3sf not awrt. Give this mark if seen in (c). Equation must be in \(h\) and \(f\) not \(y\) and \(x\).
M1 for using their equation and \(f\) = 25 to find \(h\) A1 for their final answer in [138, 139]. Can give if they have 137.7... but round to 138
Mark scheme (d)
Scheme
Marks
Should be reliable, since 25 cm(or \(f\) or footlength) is within the range of the data
B1, B1
(2)
Notes
1st B1 for suggesting it is reliable 2nd B1 for mentioning that 25 cm is within range of data. “interpolation”or“not extrapol’B1 Use of “it” or a comment that height is in range is B0 but apply ISW
Mark scheme (e)
Scheme
Marks
Line is for children – a different equation would apply to adults or Children are still growing, height will increase more than foot length
B1
(1)
(12 marks)
Notes
B1 for some comment that states a difference between children and teachers(adults) Must mention teacher/adults and children e.g. ".teacher is not in same age group as the children", "equation is for children not adults" “children and adults are different populations” "teacher will be taller" is B0 since no mention of children. “equation is only valid for children” is OK since “only” implies not suitable for adults Or Reference to different growth rates
1. On a particular day the height above sea level, \(x\) metres, and the mid-day temperature, \(y\)°C, were recorded in 8 north European towns. These data are summarised below
(b) Calculate, to 3 significant figures, the product moment correlation coefficient for these data. (2)
(c) Give an interpretation of your coefficient. (1)
A student thought that the calculations would be simpler if the height above sea level, \(h\), was measured in kilometres and used the variable \(h = \dfrac{x}{1000}\) instead of \(x\).
(d) Write down the value of \(\mathrm{S}_{hh}\) (1)
(e) Write down the value of the correlation coefficient between \(h\) and \(y\). (1)
Mark scheme (a)
Scheme
Marks
\(\mathrm{S}_{yy} = 4305 - \dfrac{181^2}{8}\)
M1
= 209.875 (awrt 210)
A1
(2)
Notes
M1 for a correct expression. Allow one slip e.g. 4350 for 4305
M1 for a correct expression for \(r\), follow through their answer to (a). Condone no “–” Allow M1 for \(\pm\) 0.87 with no working. (−0.871 is M1A1)
Mark scheme (c)
Scheme
Marks
Higher towns have lower temperature or temp. decreases as height increases
B1
(1)
Notes
B1 Must mention temperature (o.e.) and height (above sea level) and interpret the relationship between them. Must be a correct and sensible comment. e.g. "As temperature increases the height of the sea decreases" is B0. BUT simply stating "As temperature increases the height decreases" is B1 although "As height increases the temperature decreases" would be better. Treat mention of 0.87... as ISW "strong negative correlation between height and temp" is B0 (no interpretation) " as \(x\) increases \(y\) decreases" is B0 (no mention of height and temperature)
B1ft for awrt −0.871 or ft their final answer to part (b) to the same accuracy (or 3 sf) provided \(-1 \lt r \lt 1\) Answer to part (e) must be a number “it’s the same” is B0
(a) Find the equation of the regression line of \(p\) on \(v\) in the form \(p = a + bv\). (4)
(b) Using your regression line estimate the annual yield of peas per acre when the annual rainfall is 85 cm. (2)
Mark scheme (a)
Scheme
Marks
\(b = \dfrac{1.688}{5.753} = 0.293\)
M1A1
\(a = 3.22 - 4.42 \times 0.293 = 1.9231\ldots\)
M1
\(p = 1.92 + 0.293v\)
A1
(4)
Notes
Can ignore (a) and (b) labels here
1st M1 for a correct expression for \(b\). \(\dfrac{1.688}{1.168}\) is M0 1st A1 for awrt 0.29
2nd M1 for use of \(a = \bar{p} - b\bar{v}\) follow through their value of \(b\)(or even just the letter \(b\)) 2nd A1 for a complete equation with \(a\) = awrt 1.92 and \(b\) = awrt 0.293 \(y\) or \(p\) = 1.92 + 0.293\(x\) is A0 Correct answer with no working is 4/4
Mark scheme (b)
Scheme
Marks
\(v = \dfrac{85 - 5}{10} = 8\)
M1
\(p = 1.92 + 0.293 \times 8 = 4.3\) (awrt 4.3)
A1
(2)
(6 marks)
Notes
M1 for an attempt to find the value of \(v\) when \(x\) = 85 ( at least 2 correct terms in \(\pm\dfrac{85 - 5}{10}\) ) or for an attempt to find an equation for \(p\) in terms of \(x\) and using \(x\) = 85 Attempt at equation of \(p\) in \(x\) requires \(p = 1.92 + 0.293\dfrac{(x - 5)}{10}\)
A1 for awrt 4.3 (award when first seen and apply ISW) N.B. \(p = 1.92 + 0.293 \times 85\) (o.e.) is M0A0
M1 for at least one correct expression 1st A1 for \(S_{ll} =\) awrt 3420 (Condone \(S_{xx} = \ldots\) or even \(S_{yy} = \ldots\)) 2nd A1 for \(S_{lw} =\) awrt 570 (Condone \(S_{xy} = \ldots\) )
M1 for attempt at correct formula. Must have their \(S_{ll}\), \(S_{lw}\) and given \(S_{ww}\) in the correct places If \(S_{ll}\), \(S_{lw}\) are correct and an answer of awrt 0.57 is seen then award M1A0 M0 for \(\dfrac{29330.5}{\sqrt{327754.5 \times 289.6}}\)
Mark scheme (c)
Scheme
Marks
As the length of the salmon increases the weight increases
B1ft
(1)
(6 marks)
Notes
B1ft for a comment mentioning “length” and “weight”, not just \(l\) and \(w\), and the idea of longer salmon weighing more. e.g. “positive correlation between weight and length” is B0 since the idea of positive correlation is not explained. Allow “larger” instead of “heavier” or “longer” Ignore any spurious values mentioned such as 0.572 If their \(r\) is negative (but must be \(r \gt -1\)) ft an appropriate comment. Condone \(r \gt 1\) if comment is correct. If \(|r| \lt 0.4\) allow a comment of no or little relationship between weight and length but for \(0 \lt r \lt 0.4\) the printed answer is still acceptable too.
Treat mention of “skewness” as ISW if a correct interpretation is given
6. A travel agent sells flights to different destinations from Beerow airport. The distance \(d\), measured in 100 km, of the destination from the airport and the fare £\(f\) are recorded for a random sample of 6 destinations.
Destination
\(A\)
\(B\)
\(C\)
\(D\)
\(E\)
\(F\)
\(d\)
2.2
4.0
6.0
2.5
8.0
5.0
\(f\)
18
20
25
23
32
28
[You may use \(\sum d^2 = 152.09 \quad \sum f^2 = 3686 \quad \sum fd = 723.1\)]
(a) Using the axes below, complete a scatter diagram to illustrate this information. (2)
(b) Explain why a linear regression model may be appropriate to describe the relationship between \(f\) and \(d\). (1)
(c) Calculate \(S_{dd}\) and \(S_{fd}\) (4)
(d) Calculate the equation of the regression line of \(f\) on \(d\) giving your answer in the form \(f = a + bd\). (4)
(e) Give an interpretation of the value of \(b\). (1)
Jane is planning her holiday and wishes to fly from Beerow airport to a destination \(t\) km away. A rival travel agent charges 5p per km.
(f) Find the range of values of \(t\) for which the first travel agent is cheaper than the rival. (2)
Mark scheme (a)
Scheme
Marks
See overlay
B1 B1
(2)
Notes
1st B1 for at least 4 points correct (allow \(\pm\) one 2mm square) 2nd B1 for all points correct (allow \(\pm\) one 2 mm square
Mark scheme (b)
Scheme
Marks
The points lie reasonably close to a straight line (o.e.)
B1
(1)
Notes
Ignore extra points and lines Require reference to points and line for B1.
Mark scheme (c)
Scheme
Marks
\(\sum d = 27.7,\quad \sum f = 146\) (both, may be implied)
\(a = \dfrac{146}{6} - b \times \dfrac{27.7}{6} = 14.97\ldots\) so \(f = 15.0 + 2.03d\)
M1 A1
(4)
Notes
1st M1 for a correct expression for \(b\) - can follow through their answers from (c) 2nd M1 for a correct method to find \(a\) - follow through their \(b\) and their means 2nd A1 for \(f\) =.... in terms of \(d\) and all values awrt given expressions. Accept 15 as rounding from correct answer only.
Mark scheme (e)
Scheme
Marks
A flight costs £2.03 (or about £2) for every extra 100km or about 2p per km.
B1ft
(1)
Notes
Context of cost and distance required. Follow through their value of \(b\)
M1 for an attempt to find the intersection of the 2 lines. Value of \(t\) in range 500 to 505 seen award M1. Value of \(d\) in range 5 to 5.05 award M1. Accept \(t\) greater than 500 to 505 inclusive to include graphical solution for M 1A1
1. Gary compared the total attendance, \(x\), at home matches and the total number of goals, \(y\), scored at home during a season for each of 12 football teams playing in a league. He correctly calculated:
(a) Calculate the product moment correlation coefficient for these data. (2)
(b) Interpret the value of the correlation coefficient. (1)
Helen was given the same data to analyse. In view of the large numbers involved she decided to divide the attendance figures by 100. She then calculated the product moment correlation coefficient between \(\dfrac{x}{100}\) and \(y\).
(c) Write down the value Helen should have obtained. (1)
M1 for a correct expression, square root required Correct answer award 2/2
Mark scheme (b)
Scheme
Marks
Teams with high attendance scored more goals (oe, statement in context)
B1
(1)
Notes
Context required (attendance and goals). Condone causality. B0 for ‘strong positive correlation between attendance and goals’ on its own oe
Mark scheme (c)
Scheme
Marks
0.76(3)
B1ft
(1)
(4 marks)
Notes
Value required. Must be a correlation coefficient between -1 and +1 inclusive. B1ft for 0.76 or better or same answer as their value from part (a) to at least 2 d.p.
M1 for attempt at correct formula and at least one correct value (or correct ft) M0 for \(\dfrac{42948}{\sqrt{106397 \times 18181}}\)
A1ft All values correct or correct ft. Allow for an answer of 0.7 or 0.70 Answer only: awrt 0.701 is 3/3, answer of 0.7 or 0.70 is 2/3
Mark scheme (c)
Scheme
Marks
(Pmcc shows positive correlation.) Older patients have higher blood pressure
B1
(1)
Notes
B1 for comment in context that interprets the fact that correlation is positive, as in scheme. Must mention age and blood pressure in words, not just “\(t\)” and “\(p\)”.
Mark scheme (d)
Scheme
Marks
(d) Points plotted correctly on graph: -1 each error or omission (within one square of correct position) * see diagram below for correct points
B2
(2)
Notes
Record 1 point incorrect as B1B0 on epen. [NB overlay for (60, 135) is slightly wrong]
1. As part of a statistics project, Gill collected data relating to the length of time, to the nearest minute, spent by shoppers in a supermarket and the amount of money they spent. Her data for a random sample of 10 shoppers are summarised in the table below, where \(t\) represents time and £\(m\) the amount spent over £20.
\(t\) (minutes)
£\(m\)
15
−3
23
17
5
−19
16
4
30
12
6
−9
32
27
23
6
35
20
27
6
(a) Write down the actual amount spent by the shopper who was in the supermarket for 15 minutes. (1)
(b) Calculate \(S_{tt}\), \(S_{mm}\) and \(S_{tm}\).
(You may use \(\Sigma t^2 = 5478 \quad \Sigma m^2 = 2101 \quad \Sigma tm = 2485\))
(6)
(c) Calculate the value of the product moment correlation coefficient between \(t\) and \(m\). (3)
(d) Write down the value of the product moment correlation coefficient between \(t\) and the actual amount spent. Give a reason to justify your value. (2)
On another day Gill collected similar data. For these data the product moment correlation coefficient was 0.178
(e) Give an interpretation to both of these coefficients. (2)
(f) Suggest a practical reason why these two values are so different. (1)
M1 for attempt at correct formula, \(\dfrac{2485}{\sqrt{2101 \times 5478}}\) scores M1A0A0
A1ft f.t. their values for \(S_{tt}\) etc from (b) but don’t give for \(S_{tt} = 5478\) etc (see above)
Answer only (awrt 0.914) scores 3/3, 0.913 (i.e. truncation) can score M1A1ft by implication.
Mark scheme (d)
Scheme
Marks
0.914
B1f.t. (\(|r| \lt 1\))
e.g. linear transformation, coding does not affect coefficient (or recalculate)
dB1
(2)
Notes
B1f.t. (Must be the same as (c) or awrt 0.914)
2ndB1 dependent on 1st B1 Accept \(\sum m = 261, \sum m^2 = 8541, \sum tm = 6725 \to 0.914\)
Mark scheme (e)
Scheme
Marks
0.914 suggests longer spent shopping the more spent. (Idea more time, more spent)
B1
0.178 different amounts spent for same time.
B1
(2)
Notes
One mark for a sensible comment relating to each coefficient
For 0.178 allow “little or no link between time and amount spent”. Must be in context.
Just saying 0.914 is strong +ve correlation between amount spent and time shopping and 0.178 is weak correlation …scores B0B0.
Mark scheme (f)
Scheme
Marks
e.g. might spend short time buying 1 expensive item OR might spend a long time checking for bargains, talking, buying lots of cheap items.
B1g
(1)
(15 marks)
Notes
B1g for a sensible, practical suggestion showing that other factors might affect the amount spent. E.g. different day (weekend vs weekday) or time of day (time spent queuing if busy)
3. A metallurgist measured the length, \(l\) mm, of a copper rod at various temperatures, \(t\,{}^\circ\mathrm{C}\), and recorded the following results.
\(t\)
\(l\)
20.4
2461.12
27.3
2461.41
32.1
2461.73
39.0
2461.88
42.9
2462.03
49.7
2462.37
58.3
2462.69
67.4
2463.05
The results were then coded such that \(x = t\) and \(y = l - 2460.00\).
(a) Calculate \(S_{xy}\) and \(S_{xx}\). (You may use \(\Sigma x^2 = 15965.01\) and \(\Sigma xy = 757.467\)) (5)
(b) Find the equation of the regression line of \(y\) on \(x\) in the form \(y = a + bx\). (5)
(c) Estimate the length of the rod at \(40\,{}^\circ\mathrm{C}\). (3)
(d) Find the equation of the regression line of \(l\) on \(t\). (2)
(e) Estimate the length of the rod at \(90\,{}^\circ\mathrm{C}\). (1)
(f) Comment on the reliability of your estimate in part (e). (2)
3. A manufacturer stores drums of chemicals. During storage, evaporation takes place. A random sample of 10 drums was taken and the time in storage, \(x\) weeks, and the evaporation loss, \(y\) ml, are shown in the table below.
\(x\)
3
5
6
8
10
12
13
15
16
18
\(y\)
36
50
53
61
69
79
82
90
88
96
(a) On graph paper, draw a scatter diagram to represent these data. (3)
(b) Give a reason to support fitting a regression model of the form \(y = a + bx\) to these data. (1)
(c) Find, to 2 decimal places, the value of \(a\) and the value of \(b\).
(You may use \(\Sigma x^2 = 1352\), \(\Sigma y^2 = 53\,112\) and \(\Sigma xy = 8354\).)
(7)
(d) Give an interpretation of the value of \(b\). (1)
(e) Using your model, predict the amount of evaporation that would take place after
(i) 19 weeks,
(ii) 35 weeks. (2)
(f) Comment, with a reason, on the reliability of each of your predictions. (4)
3. A long distance lorry driver recorded the distance travelled, \(m\) miles, and the amount of fuel used, \(f\) litres, each day. Summarised below are data from the driver’s records for a random sample of 8 days.
The data are coded such that \(x = m - 250\) and \(y = f - 100\).
\[\Sigma x = 130 \qquad \Sigma y = 48 \qquad \Sigma xy = 8880 \qquad S_{xx} = 20\,487.5\]
(a) Find the equation of the regression line of \(y\) on \(x\) in the form \(y = a + bx\). (6)
(b) Hence find the equation of the regression line of \(f\) on \(m\). (3)
(c) Predict the amount of fuel used on a journey of 235 miles. (1)
1. The scatter diagrams below were drawn by a student.
The student calculated the value of the product moment correlation coefficient for each of the sets of data.
The values were
\[0.68 \qquad -0.79 \qquad 0.08\]
Write down, with a reason, which value corresponds to which scatter diagram. (6)
Mark scheme
Scheme
Marks
Diagram \(A\) : \(y\) & \(x\) : \(r = -0.79\); As \(x\) increases, \(y\) decreases or most points lie in the 2nd and 4th quadrant.
B1;B1dep
Diagram \(B\) : \(v\) & \(u\) : \(r = 0.08\); No real pattern. Several values of \(v\) for one value of \(u\) or points lie in all four quadrants, randomly scattered.
B1;B1dep
Diagram \(C\) : \(t\) & \(s\) : \(r = 0.68\); As \(s\) increases, \(t\) increases or most points lie in the 1st and 3rd quadrants
16 A medical student believes that, in adults, there is a negative correlation between the amount of nicotine in their blood stream and their energy level.
The student collected data from a random sample of 50 adults.
The correlation coefficient between the amount of nicotine in their blood stream and their energy level was −0.45
Carry out a hypothesis test at the 2.5% significance level to determine if this sample provides evidence to support the student’s belief.
For \(n = 50\), the critical value for a one-tailed test at the 2.5% level for the population correlation coefficient is 0.2787 [4 marks]
Compares ±0.45 or \(\lvert -0.45 \rvert\) and ±0.2787 May be seen on a diagram
M1
3.5a
States \(-0.45 \lt -0.2787\) or \(0.45 \gt 0.2787\) or \(\lvert -0.45 \rvert \gt 0.2787\) or \(\lvert -0.45 \rvert \gt \lvert 0.2787 \rvert\) and Infers \(\mathrm{H_0}\) rejected Condone accept \(\mathrm{H_1}\)
A1
2.2b
Concludes, from a fully correct comparison, in context by referring to negative correlation between the amount of nicotine in the blood stream and the energy level in adults Conclusion must not be definite, eg use of ‘suggest’, ‘support’ etc To be awarded R1, marks M1A1 must be scored as the minimum
There is sufficient evidence to suggest the student’s belief that, in adults, there is a negative correlation between the amount of nicotine in their blood stream and their energy level
16 An educational expert found that the correlation coefficient between the hours of revision and the scores achieved by 25 students in their A-level exams was 0.379
Her data came from a bivariate normal distribution.
Carry out a hypothesis test at the 1% significance level to determine if there is a positive correlation between the hours of revision and the scores achieved by students in their A-level exams.
The critical value of the correlation coefficient is 0.4622 [4 marks]
Mark scheme
Scheme
Marks
AO
States both hypotheses correctly for one-tailed test Do not accept \(r = 0\) and \(r \gt 0\) Accept \(\rho\) in words eg. \(\mathrm{H_0}\):No correlation/coefficient is 0 \(\mathrm{H_1}\):Positive correlation/coefficient > 0
B1
2.5
Compares given critical value 0.4622 with the given correlation 0.379 coefficient Condone 0.462 or 0.46 or 0.5 for 0.4622 Condone 0.38 or 0.4 for 0.379
M1
3.5a
Compares given critical value 0.4622 with the given correlation 0.379 coefficient correctly and infers \(\mathrm{H_0}\) is not rejected Allow reference to \(\mathrm{H_1}\)
A1
2.2b
Concludes correctly in context that there is insufficient evidence to suggest that there is positive correlation between the hours of revision and the scores achieved by them in their A level exams
There is insufficient evidence to suggest that there is a positive correlation between the hours of revision and the scores achieved by them in their A level exams
15 Jamal, a farmer, claims that the larger the rainfall, the greater the yield of wheat from his farm.
He decides to investigate his claim, at the 5% level of significance.
He measures the rainfall in centimetres and the yield in kilograms for a random sample of ten years.
He correctly calculates the product moment correlation coefficient between rainfall and yield for his sample to be 0.567
The table below shows the critical values for correlation coefficients for a sample size of 10 for different significance levels, for both 1- and 2-tailed tests.
1-tailed test significance level
5%
2.5%
1%
0.5%
2-tailed test significance level
10%
5%
2%
1%
Critical value
0.549
0.632
0.716
0.765
Determine what Jamal’s conclusion to his investigation should be, justifying your answer. [3 marks]
Mark scheme
Scheme
Marks
AO
Identifies critical value = 0.549
B1
1.1b
Compares 0.567 correctly to their critical value chosen from the table
M1
3.5a
Makes correct inference eg there is sufficient/significant evidence that the larger the rainfall, the greater the yield/positive correlation between the two FT their critical value chosen from the table
R1F
2.2b
(3 marks)
Typical solution
\[0.567 \gt 0.549\]
There is sufficient evidence that the larger the rainfall, the greater the yield.
10 Each month, the manager of a large store records the number, \(c\), of customers who visit the store, and the amount, £\(h\), spent on heating during that month. The manager wants to test whether there is linear correlation between \(c\) and \(h\).
For a randomly chosen year the value of Pearson’s product-moment correlation coefficient, \(r\), between \(c\) and \(h\) was \(-0.798\), correct to 3 significant figures.
(a) Using the table below, carry out the test at the 1% significance level. [5]
(b) Describe briefly two main features of a scatter diagram that could be drawn to illustrate the values of \(c\) and \(h\) for this year. There is no need to draw a diagram. [2]
(c) The manager makes the following statement. “The value of \(r\) shows that when we spend more on heating, fewer customers visit the store. So we should spend less on heating.” Comment briefly on this statement, making reference to the context. [2]
(d) Give a statement about a probability to explain the meaning of the value 0.7155 in the table below. [2]
Critical values of Pearson’s product-moment correlation coefficient
1-tail test
5%
2.5%
1%
0.5%
2-tail test
10%
5%
2%
1%
\(n\)
10
0.5494
0.6319
0.7155
0.7646
11
0.5214
0.6021
0.6851
0.7348
12
0.4973
0.5760
0.6581
0.7079
13
0.4762
0.5529
0.6339
0.6835
Mark scheme (a)
Scheme
Marks
AO
\(\mathrm{H}_0: \rho = 0\)
B1
1.1
\(\mathrm{H}_1: \rho \neq 0\) where \(\rho\) is the correlation coefficient for the population or where \(\rho\) is the correlation coefficient between amount spent (\(h\)) and no. of customers (\(c\))
B1
2.5
\(0.798 \gt 0.7079\) oe
B1FT
1.1
Reject \(\mathrm{H}_0\)
M1
1.1
Sufficient evidence for a (linear) correlation between amount spent (\(h\)) and no. of customers (\(c\)) oe
A1
2.2b
[5]
Notes
B1 B1: Subtract B1 for each error:
Undefined \(\rho\): B1B0
1-tail: B1B0
Allow other letters (but not \(r\), \(c\) or \(h\): B1B0)
Hypotheses in words (no parameter): B1B0
\(\mathrm{H}_0\): There is no correlation
\(\mathrm{H}_1\): There is correlation
Do not allow “negative” or “positive” correlation for \(\mathrm{H}_1\): B0B0
Accept “pmcc” for correlation coefficient
B1FT: FT their setup/hypotheses
e.g. for \(\mathrm{H}_1: \rho \lt 0\), compare 0.798 with 0.6581
Must use \(n = 12\) and specify a corresponding value from the table
0.6581 or 0.7079 only (NB not 0.6851 from \(n = 11\))
Must compare this with 0.798 with the same sign
Condone \(-0.798 \lt -0.7079\) or \(|-0.798|\)
Do not accept \(-0.798 \lt 0.6581\)
NB this is the only mark that can be scored with no hypotheses
M1: This step must be seen, consistent with their hypotheses and their comparison. Condone Accept \(\mathrm{H}_1\)
A1: Conclusion must be in context, not definite and consistent with their hypotheses and comparison.
Disregard any mention of “negative” or “positive”
“Relationship” A0
“Prove(d)” A0
Condone “there is evidence of a linear correlation between \(h\) and \(c\)”
Condone “significant” for “sufficient”
Must conclude that there is evidence for a correlation.
Mark scheme (b)
Scheme
Marks
AO
Points (fairly) close to a (straight) line
B1
1.2
with negative gradient oe
B1
1.2
[2]
Notes
B1: For a statement about the relative strength of the linear correlation:
Accept “points form a line” or
Accept “points lie (relatively) close to the line”
Not “points are close together” or “close to each other”
B1: For a statement about the appearance of the negative correlation:
Accept “line from 2nd to 4th quadrant”
Accept “two clusters in top left and bottom right”
Accept “line will be downwards sloping”
Not “negatively correlated” (must be a feature of the scatter diagram)
This mark only may be implied by a sketch (showing a scatter diagram with negative correlation, with or without a line of best fit).
Mark scheme (c)
Scheme
Marks
AO
Correlation does not imply causation
B1
2.3
A suggested 3rd factor affecting both \(c\) & \(h\) e.g. time of year, temperature, weather
B1
2.4
[2]
Notes
B1: oe, may be implied (but do not allow “independent”)
B1: Any sensible comment about the statement but must be in context:
Accept “Some people may not visit if the store is too cold”
Accept “the shop being too warm may mean customers don’t want to go inside”
Not “There may be a third factor affecting both \(c\) and \(h\)” (a possible factor must be specified)
Accept “more people in the store may mean there is less need for heating” (so the implication might be the other way around) or equivalent statements about causation
Mark scheme (d)
Scheme
Marks
AO
If no (linear) correlation in the population, then for (a sample of) 10 (pairs)
10 A researcher plans to carry out a statistical investigation to test whether there is linear correlation between the time (\(T\) weeks) from conception to birth, and the birth weight (\(W\) grams) of new-born babies.
(a) Explain why a 1-tail test is appropriate in this context. [1]
The researcher records the values of \(T\) and \(W\) for a random sample of 11 babies. They calculate Pearson’s product-moment correlation coefficient for the sample and find that the value is 0.722.
(b) Use the table below to carry out the test at the 1% significance level. [5]
Critical values of Pearson’s product-moment correlation coefficient.
1-tail test
5%
2.5%
1%
0.5%
2-tail test
10%
5%
2.5%
1%
\(n\)
10
0.5494
0.6319
0.7155
0.7646
11
0.5214
0.6021
0.6851
0.7348
12
0.4973
0.5760
0.6581
0.7079
13
0.4762
0.5529
0.6339
0.6835
Mark scheme (a)
Scheme
Marks
Very likely weight will increase with time oe or He is only looking for positive correlation
B1
[1]
Notes
Or eg "Expect weight to increase with time" oe "Foetuses grow" oe Ignore all else
Mark scheme (b)
Scheme
Marks
\(\mathrm{H}_0: \rho = 0\) Allow other letters
B1
\(\mathrm{H}_1: \rho \gt 0\) where \(\rho\) is the correlation coefficient for the population or where \(\rho\) is the correlation coefficient between time and weight
There is evidence of (positive linear) correlation between time from conception to birth and weight of new-born babies
Or eg It appears that birth weight increases with time (from conception to birth)
A1
[5]
Notes
B1B0 for 1 error, eg undefined \(\rho\) or 2-tail
For hypotheses in words, not using parameter: \(\mathrm{H}_0\): There is no correlation between time and weight \(\mathrm{H}_1\): There is positive correlation between time and weight B1B0 But omission of “positive”: B0B0
M1: May be implied by conclusion
A1: Allow without "positive” and without “linear" In context, not definite
11 A trainer was asked to give a lecture on population profiles in different Local Authorities (LAs) in the UK. Using data from the 2011 census, he created the following scatter diagram for 17 selected LAs.
He selected the 17 LAs using the following method. The proportions of people aged 18 to 24 and aged 65+ in any Local Authority are denoted by \(P_{\text{young}}\) and \(P_{\text{senior}}\) respectively. The trainer used a spreadsheet to calculate the value of \(k = \dfrac{P_{\text{young}}}{P_{\text{senior}}}\) for each of the 348 LAs in the UK. He then used specific ranges of values of \(k\) to select the 17 LAs.
(a) Estimate the ranges of values of \(k\) that he used to select these 17 LAs. [2]
(b) Using the 17 LAs the trainer carried out a hypothesis test with the following hypotheses. \(\mathrm{H}_0\): There is no linear correlation in the population between \(P_{\text{young}}\) and \(P_{\text{senior}}\). \(\mathrm{H}_1\): There is negative linear correlation in the population between \(P_{\text{young}}\) and \(P_{\text{senior}}\). He found that the value of Pearson’s product-moment correlation coefficient for the 17 LAs is \(-0.797\), correct to 3 significant figures.
(i) Use the table below to show that this value is significant at the 1% level. [2]
The trainer concluded that there is evidence of negative linear correlation between \(P_{\text{young}}\) and \(P_{\text{senior}}\) in the population.
(ii) Use the diagram to comment on the reliability of this conclusion. [2]
(c) Describe one outstanding feature of the population in the areas represented by the points in the bottom right hand corner of the diagram. [1]
(d) The trainer’s audience included representatives from several universities. Suggest a reason why the diagram might be of particular interest to these people. [1]
Critical values of Pearson’s product-moment correlation coefficient
1-tail test
5%
2.5%
1%
0.5%
2-tail test
10%
5%
2%
1%
\(n\)
1
-
-
-
-
2
-
-
-
-
3
0.9877
0.9969
0.9995
0.9999
4
0.9000
0.9500
0.9800
0.9900
5
0.8054
0.8783
0.9343
0.9587
6
0.7293
0.8114
0.8822
0.9172
7
0.6694
0.7545
0.8329
0.8745
8
0.6215
0.7067
0.7887
0.8343
9
0.5822
0.6664
0.7498
0.7977
10
0.5494
0.6319
0.7155
0.7646
11
0.5214
0.6021
0.6851
0.7348
12
0.4973
0.5760
0.6581
0.7079
13
0.4762
0.5529
0.6339
0.6835
14
0.4575
0.5324
0.6120
0.6614
15
0.4409
0.5140
0.5923
0.6411
16
0.4259
0.4973
0.5742
0.6226
17
0.4124
0.4821
0.5577
0.6055
18
0.4000
0.4683
0.5425
0.5897
19
0.3887
0.4555
0.5285
0.5751
20
0.3783
0.4438
0.5155
0.5614
21
0.3687
0.4329
0.5034
0.5487
22
0.3598
0.4227
0.4921
0.5368
23
0.3515
0.4132
0.4815
0.5256
24
0.3438
0.4044
0.4716
0.5151
25
0.3365
0.3961
0.4622
0.5052
26
0.3297
0.3882
0.4534
0.4958
27
0.3233
0.3809
0.4451
0.4869
28
0.3172
0.3739
0.4372
0.4785
29
0.3115
0.3673
0.4297
0.4705
30
0.3061
0.3610
0.4226
0.4629
Mark scheme (a)
Scheme
Marks
AO
\(k \gt 1.4\) (allow \(k \gt 1.1\) to 1.6) \(k \lt 0.25\) (allow \(k \lt 0.2\) to 0.3)
B1 B1
2.2b 2.2b
or \(1.4 \lt k \lt \ldots\ldots\) B1 \(\ldots\ldots \lt k \lt 0.25\) (ranges as above) B1
[2]
Notes
Allow \(\geqslant\) and \(\leqslant\) SC: \(0.25 \lt k \lt 1.4\): B1B0 (ranges as on left) Allow "\(x\)"
Mark scheme (b)
Scheme
Marks
AO
(i) \(0.797 \gt 0.5577\) or \(-0.797 \lt -0.5577\) or \(|-0.797| \gt 0.5577\)
B2
3.1b 3.2a
[2]
(ii) There are clusters (or groups etc.)
B1*
2.3
Apparent good correlation caused by clusters or Two clusters with no –ve corr'n within them or a comment similar to one of the above. AND Conclusion unreliable or Value of \(r\) is misleading oe
B1 dep B1*
3.5b
[2]
Notes
(b)(i) B2: \(0.797 \gt 0.6055\) or \(-0.797 \lt -0.6055\) B1 \(\pm 0.5577\) B1 Allow \(\geqslant\) or \(\leqslant\)
(b)(ii) or Not bivariate normal distribution B1 so use of tables for \(r\) not valid B1 NOT Too scattered Not represent whole pop Small sample Clusters not on reg line B1B0
Mark scheme (c)
Scheme
Marks
AO
High prop of 65+ or Low prop of 18-24 Prop of young very similar, or \(\approx 0.06\) Proportion of senior to young is high
B1
2.2b
[1]
Notes
B1: If consider only one age-group, must be proportion not number If consider both age-groups, allow eg Higher number of seniors than young or Many seniors, few young NOT: Similar proportions of 65+ Population is elderly
Mark scheme (d)
Scheme
Marks
AO
Top left points contain high prop of 18-24s. (So these LAs may be areas where there are universities or where they can recruit)
B1
2.2b
[1]
Notes
B1: Shows places where large nos of 18-24s Shows where to focus recruiting. So universities can recruit. 18-24s are their target group. No need to specify "Top left group" Allow "students" or "young" instead of "18-24s" Any implication that diagram enables you to see information about location of young people
11 Christa used Pearson’s product-moment correlation coefficient, \(r\), to compare the use of public transport with the use of private vehicles for travel to work in the UK.
(i) Using the pre-release data set for all 348 UK Local Authorities, she considered the following four variables.
Number of employees using public transport
\(x\)
Number of employees using private vehicles
\(y\)
Proportion of employees using public transport
\(a\)
Proportion of employees using private vehicles
\(b\)
(a) Explain, in context, why you would expect strong, positive correlation between \(x\) and \(y\). [1]
(b) Explain, in context, what kind of correlation you would expect between \(a\) and \(b\). [2]
(ii) Christa also considered the data for the 33 London boroughs alone and she generated the following scatter diagram.
One London Borough is represented by an outlier in the diagram.
(a) Suggest what effect this outlier is likely to have on the value of \(r\) for the 32 London Boroughs. [1]
(b) Suggest what effect this outlier is likely to have on the value of \(r\) for the whole country. [1]
(c) What can you deduce about the area of the London Borough represented by the outlier? Explain your answer. [1]
Mark scheme (i)(a)
Scheme
Marks
AO
Both the number of employees using public transport and the number of employees using private vehicles depend on the LA population.
E1
2.1
[1]
Notes
E1: or similar, but must be in context. Ignore all else NOT No. using pt is prop to no. using pv
Mark scheme (i)(b)
Scheme
Marks
AO
Negative
E1ind
2.2b
If a large prop use public transport then a smaller prop drive (and vice versa)
E1ind
2.4
[2]
Notes
E1ind: Ignore "strong" or "slight" etc
E1ind: or similar in context
NOT Inverse prop'n NOT "as \(a\) increases \(b\) decreases" unless in context
Mark scheme (ii)(a)
Scheme
Marks
AO
Decrease the size of \(r\) or Make \(r\) less negative
E1
2.2b
[1]
Notes
E1: Make (value of) \(r\) increase \(r\) closer to 0 Ignore eg "greatly" Ignore all else
NOT Make \(r\) decrease NOT Weaken the corr'n NOT Make corr'n less
Mark scheme (ii)(b)
Scheme
Marks
AO
Little effect (because the population of the LA is small compared with the whole population)
E1
2.2b
[1]
Notes
E1: or No effect or similar Ignore all else
Mark scheme (ii)(c)
Scheme
Marks
AO
Ignore all reference to public transport Type 1 answers People don't travel far to work Jobs are close High proportion walk (or cycle)
E1
2.4
[1]
Notes
Type 2 answers Any suggested reason why few drive eg Few garages; Parking expensive or similar in context
15 A personal trainer is investigating whether, in the general population, there is any association between resting pulse rate in beats per minute and mean hours per week spent running.
One week he collects a sample by asking 25 of his clients for the relevant data.
(a)
(i) State the name of the sampling technique used by the personal trainer. [1]
(ii) Explain why this sampling technique might introduce bias. [1]
A biologist researching the same topic collects a random sample of size 47. She represents the data using the scatter diagram shown below.
(b) Describe the association between resting pulse rate and mean hours per week spent running. [1]
(c) The biologist identifies two distinct regions on the scatter diagram. Identify these regions on the copy of the scatter diagram in the Printed Answer Booklet by drawing an appropriate vertical line between them. [1]
According to medical research, the normal resting pulse rate for an adult is between 60 and 110 beats per minute.
The biologist separates the data into a group to the left of the vertical line on the scatter diagram and a group to the right of the vertical line on the scatter diagram. She calculates Spearman’s rank correlation coefficient, \(r_s\), and the associated \(p\)-value for each group.
The results are shown in the table.
Group
\(r_s\)
\(p\)-value
To the left of the vertical line
0.209 36
0.419 98
To the right of the vertical line
–0.613 49
0.000 31
(d) With reference to the two groups identified by the biologist and to the values in the table, explain what may be inferred about the association between resting pulse rate and mean number of hours per week spent running. [6]
Mark scheme (a)
Scheme
Marks
AO
(i) opportunity (sampling)
B1
2.5
[1]
(ii) only people known to the personal trainer / his clients sampled oe so eg they are fitter than typical members of the population oe eg sample may not be representative oe
B1
2.3
[1]
Notes
(i) B1: allow convenience (sampling) do not allow eg convenient sampling; eg opportunistic sampling
(ii) B1: allow so any eg which alludes to sample possibly giving unrepresentative values
Mark scheme (b)
Scheme
Marks
AO
negative (association)
B1
2.4
[1]
Notes
B1: ignore adjectives such as weak, strong etc allow eg as mean hours spent running increases, resting pulse rate decreases; eg negative relationship; do not allow eg (negative) correlation; eg as pulse rate increases, mean hours per week spent running decreases
Mark scheme (c)
Scheme
Marks
AO
B1
2.2b
[1]
Notes
B1: approximately vertical line at about 2.5; B0 if more than one line
Mark scheme (d)
Scheme
Marks
AO
LH region 0.20936 is (fairly) small oe isw
B1
3.4
0.41998 is not close to 0 oe isw
M1
3.4
insufficient evidence to suggest association (between mean time spent running and resting pulse rate) oe
A1
2.2b
RH region –0.61349 is closer to –1 than 0 oe isw
B1
3.4
0.00031is close to 0 oe
M1
3.4
sufficient evidence to suggest (some) [negative] association (between mean time spent running and resting pulse rate) oe
A1
2.2b
[6]
Notes
B1: allow eg \(r_s\) is closer to 0 than to 1; eg low value of \(r_s\); eg allow \(r_s\) close to 0
M1: allow eg \(p\)-value \(\gt\) 0.05 (or 0.01 or 0.1) eg \(p\)-value not small do not allow eg \(p\)-value (relatively) large eg high \(p\)-value
A1: may infer no association
B1: allow eg \(r_s\) close to \(-1\) eg \(r_s\) is closer to –1 than 0; eg \(r_s\) is not small B0 for eg \(r_s\) is large oe
M1: allow eg \(p\)-value \(\lt\) 0.05 (or 0.01 or 0.10) allow eg \(p\)-value is (very) small/low
11 A householder is investigating whether there is any relationship between his monthly cost of gas and his monthly cost of electricity, both measured in pounds (£). The householder collects a random sample of monthly costs and presents them in the scatter diagram below.
One of the points on the diagram represents the energy costs in a month when the householder was away on holiday for three weeks. The other points represent the energy costs in months when the householder did not go away on holiday.
(a) On the copy of the diagram in the Printed Answer Booklet, circle the point which represents the month when the householder was most likely to have been away on holiday for three weeks. [1]
(b) With reference to the diagram, describe the relationship between the cost of gas and the cost of electricity. [1]
The householder decides to test whether there is evidence to suggest that there is any association between the monthly cost of gas and the monthly cost of electricity. The value of Spearman’s rank correlation coefficient for this sample is 0.4359 and the associated \(p\)-value is 0.091 95.
(c) Determine whether there is any evidence to suggest, at the 5% level, that there is any association between the monthly cost of gas and the monthly cost of electricity. [3]
Mark scheme (a)
Scheme
Marks
AO
B1
2.2b
[1]
Mark scheme (b)
Scheme
Marks
AO
(weak) positive association or (weak) positive correlation
B1
1.1
[1]
Notes
B1: allow eg as cost of electricity increases, cost of gas increases oe
Mark scheme (c)
Scheme
Marks
AO
0.09195 correctly compared with 0.05 or 0.025 only
M1
3.4
\(0.09195 \gt 0.025\)
A1
1.1
insufficient evidence [at the 5% level] to suggest any association between cost of gas and cost of electricity isw
A1
2.2b
[3]
Notes
M1: allow eg “\(p\)-value \(\gt 0.05\)”
A1: allow \(p\)-value \(\gt 0.025\)
A1:A0 if refers to correlation rather than association; dependent on award of previous A1
14 The pre-release material contains information concerning the median income of taxpayers in £ and the percentage of all pupils at the end of KS4 achieving 5 or more GCSEs at grade A*–C, including English and Maths, for different areas of London.
Some of the data for 2014/15 is shown in Fig. 14.1.
Fig. 14.1
Median Income of Taxpayers in £
Percentage of Pupils Achieving 5 or more A*–C, including English and Maths
City of London
61 100
#N/A
Barking and Dagenham
21 800
54.0
Barnet
27 100
70.1
Bexley
24 400
55.0
Brent
22 700
60.0
Bromley
28 100
68.0
A student investigated whether there is any relationship between median income of taxpayers and percentage of pupils achieving 5 or more GCSEs at grade A*–C, including English and Maths.
(a) With reference to Fig. 14.1, explain how the data should be cleaned before any analysis can take place. [1]
After the data was cleaned, the student used software to draw the scatter diagram shown in Fig. 14.2.
Fig. 14.2
The student calculated that the product moment correlation coefficient for these data is 0.3743.
(b) Give two reasons why it may not be appropriate to use a linear model for the relationship between median income of taxpayers in £ and the percentage of all pupils at the end of KS4 achieving 5 or more GCSEs at grade A*–C. [2]
The student carried out some further analysis. The results are shown in Fig. 14.3.
Fig. 14.3
median income of taxpayers in £
percentage of pupils achieving 5+ A*–C
mean
27 216
61.0
standard deviation
4177.5
5.32
The student identified three outliers in total.
(c)
Use the information in Fig. 14.3 to determine the range of values of the median income of taxpayers in £ which are outliers.
Use the information in Fig. 14.3 to determine the range of values of the percentage of all pupils at the end of KS4 achieving 5 or more GCSEs at grade A*–C which are outliers.
On the copy of Fig. 14.2 in the Printed Answer Booklet, circle the three outliers identified by the student.
[4]
The student decided to remove these outliers and recalculate the product moment correlation coefficient.
(d) Explain whether the new value of the product moment correlation coefficient would be between 0.3743 and 1 or between 0 and 0.3743. [1]
Mark scheme (a)
Scheme
Marks
AO
discard City of London (as part of the data not available) or discard any regions where one or more pieces of data are missing oe
B1
2.4
[1]
Notes
B1: LDS advantage do not allow if answer spoiled eg because it’s an anomaly, eg because it’s an outlier,
Mark scheme (b)
Scheme
Marks
AO
scatter does not look linear oe
B1
3.4
pmcc not close to 1 oe
B1
3.4
[2]
Notes
B1: ignore extra comments unless they contradict an otherwise correct answer
B1: ignore extra comments unless they contradict an otherwise correct answer
15 The pre-release material includes information on life expectancy at birth in countries of the world. Fig. 15.1 shows the data for Liberia, which is in Africa, together with a time series graph.
1960
1970
1980
1990
2000
2010
34.67
39.25
46.00
47.18
52.42
59.63
Fig. 15.1
Sundip uses the LINEST function on a spreadsheet to model life expectancy as a function of calendar year by a straight line.
The equation of this line is \(L = 0.473y - 892\), where \(L\) is life expectancy at birth and \(y\) is calendar year.
(a) Use this model to find an estimate of the life expectancy at birth in Liberia in 1995. [1]
According to the model, the life expectancy at birth in Liberia in 2025 is estimated to be 65.83 years.
(b) Explain whether each of these two estimates is likely to be reliable. [2]
(c) Use your knowledge of the pre-release material to explain whether this model could be used to obtain a reliable estimate of the life expectancy at birth in other countries in 1995. [1]
Fig. 15.2 shows the life expectancy at birth between 1960 and 2010 for Italy and South Africa.
Fig. 15.2
(d) Use your knowledge of the pre-release material to
Explain whether series 1 or series 2 represents the data for Italy.
Explain how the data for South Africa differs from the data for most developed countries.
[2]
Sundip is investigating whether there is an association between the wealth of a country and life expectancy at birth in that country. As part of her analysis she draws a scatter diagram of GDP per capita in US$ and life expectancy at birth in 2010 for all the countries in Europe for which data is available. She accidentally includes the data for the Central African Republic. The diagram is shown in Fig. 15.3.
Fig. 15.3
(e) On the copy of Fig. 15.3 in the Printed Answer Booklet, use your knowledge of the pre-release material to circle the point representing the data for the Central African Republic. [1]
Sundip states that as GDP per capita increases, life expectancy at birth increases.
(f) Explain to what extent the information in Fig. 15.3 supports Sundip’s statement. [2]
Mark scheme (a)
Scheme
Marks
AO
51.635 or 51.64 or 51.6
B1
3.4
[1]
Mark scheme (b)
Scheme
Marks
AO
1995 estimate (probably) reliable since it is interpolation
B1
2.2b
2025 estimate (probably) not reliable since it is extrapolation
B1
2.2b
[2]
Notes
B1: allow eg the first estimate..
B1: allow eg the second estimate…
Mark scheme (c)
Scheme
Marks
AO
No, because trends in life expectancy at birth may vary considerably between nations
B1
2.4
[1]
Notes
B1: LDS advantage
Mark scheme (d)
Scheme
Marks
AO
series 2 (the top one) is Italy – life expectancy (generally) higher in Europe (than Africa)
B1
2.4
the values are decreasing (from 1990) in South Africa (– unusual since most show an upward trend) or little (or no) overall increase in South Africa (since 1970) or South Africa has lower life expectancy (than most developed countries)
B1
2.4
[2]
Notes
B1: LDS advantage
B1: LDS advantage
Mark scheme (e)
Scheme
Marks
AO
B1
1.1
[1]
Notes
B1: Point at (700, 47.56) ringed LDS advantage
Mark scheme (f)
Scheme
Marks
AO
the diagram supports this statement for values of GDP per capita from \(k\) to \(n\) where \(0 \lt k \leqslant 20\,000\) and \(40\,000 \leqslant n \leqslant 60\,000\) since there appears to be positive correlation oe
B1
2.3
for values of GDP per capita \(\geqslant K\) where \(40\,000 \leqslant K \leqslant 60\,000\) there appears to be no association between GDP per capita and life expectancy at birth so the diagram does not support Sundip’s statement for these values
B1
2.2b
[2]
Notes
B1: must give specific range of values ; must say supports statement oe
B1: the range may be implied by reference to a specific range identified for the first mark; must say does not support statement oe
13 The pre-release material contains information concerning median house prices, recycling rates and employment rates. Fig. 13.1 shows a scatter diagram of recycling rate against employment rate for a random sample of 33 regions.
Fig. 13.1
The product moment correlation coefficient for this sample is 0.37154 and the associated \(p\)-value is 0.033.
Lee conducts a hypothesis test at the 5% level to test whether there is any evidence to suggest there is positive correlation between recycling rate and employment rate. He concludes that there is no evidence to suggest positive correlation because \(0.033 \approx 0\) and \(0.37154 > 0.05\).
(a) Explain whether Lee’s reasoning is correct. [2]
Fig. 13.2 shows a scatter diagram of recycling rate against median house price for a random sample of 33 regions.
Fig. 13.2
The product moment correlation coefficient for this sample is \(-0.33278\) and the associated \(p\)-value is 0.058.
Fig. 13.3 shows summary statistics for the median house prices for the data in this sample.
Statistics
n
33
Mean
465467.9697
σ
201236.1345
s
204356.2606
Σx
15360443
Σx2
8486161617387
Min
243500
Q1
342500
Median
410000
Q3
521000
Max
1200000
Fig. 13.3
(b) Use the information in Fig. 13.3 and Fig. 13.2 to show that there are at least two outliers. [2]
(c) Describe the effect of removing the outliers on
the product moment correlation coefficient between recycling rate and median house price,
the \(p\)-value associated with this correlation coefficient,
in each case explaining your answer. [2]
All 33 items in the sample are areas in London. A student suggests that it is very unlikely that only areas in London would be selected in a random sample.
(d) Use your knowledge of the pre-release material to explain whether you think the student’s suggestion is reasonable. [1]
Mark scheme (a)
Scheme
Marks
AO
Lee is wrong because he should make the comparison of 0.033 with 0.05
B1
2.2a
he should make the comparison of 0.37154 with 0
B1
2.2b
[2]
Notes
B1: allow he should have compared \(r\) with the critical value
if B0B0SC1 for Lee has confused \(r\) with \(p\) or for 0.37154 suggests positive correlation
Mark scheme (b)
Scheme
Marks
AO
\(465467 + 2 \times 204356\)
M1
2.1
awrt 874180 (or 867940 from use of 201236) from scatter diagram the outliers are approximately 920 000, 1 200 000
A1
2.2b
[2]
Notes
M1: condone use of 201236 instead of 204356; ignore work relating to lower tail or \(521000 + 1.5 \times (521000 - 342500)\)
A1: numerical values must be mentioned or 788750 in which case accept two or three outliers identified extra one is approximately 800 000 (corrected from the printed mark scheme: “867940 from use of 210236” is printed; 867940 comes from 201236)
Mark scheme (c)
Scheme
Marks
AO
the pmcc would (probably) be closer to 0 because the scatter is less well modelled by a straight line
B1
2.2b
the \(p\)-value would increase because a value which is closer to 0 is more likely assuming there is no correlation
B1
2.2b
[2]
Notes
if B0B0 allow SC1 for \(r\) closer to 0 and \(p\)-value larger
Mark scheme (d)
Scheme
Marks
AO
the student’s suggestion is reasonable, since there are other regions defined in the LDS
11 The pre-release material contains information concerning median house prices over the period 2004 – 2015. A spreadsheet has been used to generate a time series graph for two areas: the London borough of “Barking and Dagenham” and “North West”. This is shown together with the raw data in Fig. 11.1.
Year
Barking and Dagenham
North West
2004
160 000
107 000
2005
163 000
118 000
2006
168 000
127 000
2007
185 000
134 750
2008
190 000
129 950
2009
160 000
130 000
2010
171 000
130 000
2011
170 000
127 000
2012
174 995
130 000
2013
180 995
131 000
2014
215 000
138 500
2015
243 500
140 000
Fig. 11.1
Dr Procter suggests that it is unusual for median house prices in a London borough to be consistently higher than those in other parts of the country.
(a) Use your knowledge of the large data set to comment on Dr Procter’s suggestion. [1]
Dr Procter wishes to predict the median house price in Barking and Dagenham in 2016. She uses the spreadsheet function LINEST to find the equation of the line of best fit for the given data. She obtains the equation
\(P = 4897Y - 9\,657\,847\), where \(P\) is the median house price in pounds and \(Y\) is the calendar year, for example 2015.
(b) Use Dr Procter’s equation to predict the median house price in Barking and Dagenham in
2016
2017.
[2]
Professor Jackson uses a simpler model by using the data from 2014 and 2015 only to form a straight-line model.
(c) Find the equation Professor Jackson uses in her model. [2]
(d) Use Professor Jackson’s equation to predict the median house price in Barking and Dagenham in
2016
2017.
[2]
Professor Jackson carries out some research online. She finds some information about median house prices in Barking and Dagenham, which is shown in Fig. 11.2.
2016
2017
£290 000
£300 000
Fig. 11.2
(e) Comment on how well
Dr Procter’s model fits the data,
Professor Jackson’s model fits the data.
[2]
(f) Explain which, if any, of the models is likely to be more reliable for predicting median house prices in Barking and Dagenham in 2020. [1]
Mark scheme (a)
Scheme
Marks
AO
house prices are generally higher in London boroughs (than elsewhere in the country), so Dr Procter’s suggestion is probably wrong
B1
2.2a
[1]
Mark scheme (b)
Scheme
Marks
AO
214 505
B1
3.4
219 402
B1
1.1
[2]
Mark scheme (c)
Scheme
Marks
AO
\(P = 28\,500Y - 57\,184\,000\) (where \(Y\) is the calendar year)
B1
3.3
or \(P = 28\,500y + 215\,000\) (where \(y\) is the number of years after 2014)
B1
1.1
[2]
Notes
B1: gradient
B1: intercept
allow both marks for correct equation in any form isw allow eg \(y = 28\,500x - 57\,184\,000\)
Mark scheme (d)
Scheme
Marks
AO
2016 272 000
B1
3.4
2017 300 500
B1
1.1
[2]
Notes
FTtheir straight line model provided this gives values > 250 000
Mark scheme (e)
Scheme
Marks
AO
Dr Procter’s model is a (very) poor fit
B1
2.2a
Prof Jackson’s is a good fit, or works well for 2017, but not 2016
B1
2.2a
[2]
Notes
B1: dependent on correct values in (b)
B1:FT comment for their values > 250 000 this mark is dependent on having calculated values in part (d)
5. The World Health Organisation (WHO) compiles data on life expectancy and other factors.
A hobbyist statistician is exploring the data he collects from the WHO. The following screenshot shows part of his work.
Country
Life expectancy
Adult mortality
Hepatitis B
Polio
GDP
Population
Years of schooling
Life expectancy
Benin
60
249
82
78
784
13 958 086
10·7
60
Botswana
65·7
256
95
96
6533
2 705 113
12·6
65·7
Comoros
63·5
227
91
92
728
777 424
11·1
63·5
Costa Rica
79·6
AB
92
92
1146
5 235 470
14·2
79·6
Saudi Arabia
74·5
88
98
97
2733
16·1
74·5
Seychelles
73·2
168
98
97
1539
93 419
14·1
73·2
(a) Give two ways in which the data could be cleaned. [2]
The graph below shows the life expectancy and years of schooling for 16 randomly selected countries.
(b) Interpret the gradient of the regression line in context. [1]
(c) Comment on the causal nature or otherwise of the correlation. [1]
(d)
(i) Calculate an estimate of the life expectancy for a country that has 14·6 years of schooling. [1]
(ii) Comment on the reliability of your answer. [1]
The entire data set was used to produce the following scatter diagram.
(e) State the curious feature of the graph and what might be done about it. [1]
(f) The summary data for life expectancy, \(y\), in years, for the entire data set are given as follows.\[\sum y=13\,105{\cdot}9 \qquad \sum y^2=950\,615{\cdot}47 \qquad n=183\]
(i) Show that the mean is 71·6 years, correct to 3 significant figures, and calculate the standard deviation. [3]
(ii) The hobbyist statistician claims that 71·6 years represents average global life expectancy. Comment on this claim. [1]
Mark scheme (a)
Scheme
Marks
Comments
Valid way to clean data.
e.g. Population figure for Saudi Arabia could be found.
e.g. Add the population figure for Saudi Arabia.
e.g. AB in the Adult Mortality column for Costa Rica is clearly incorrect. This could be found out or the data could be ignored, depending on the context.
e.g. Remove duplicate columns. e.g. Check the accuracy of the 60 figure for Benin.
e.g. Check the Polio figure for Benin. e.g. Investigate to see if polio in Benin is an outlier.
E2
E1 each correct one.
Do not allow ‘Don’t use Saudi Arabia’ or ‘Don’t use Costa Rica’ unless justified e.g. ‘Finding the Adult mortality for Costa Rica might be difficult.’ E0 for ‘should use percentages’. E0 for ‘round to \(X\) significant figures’. E0 for ‘don’t use decimals.’ E0 for ‘group diseases together’. E0 for ‘polio in Benin is an outlier therefore ignore it’. E0 for ‘Use same number of significant figures for all data’. E0 for referencing lack of units.
[2]
Mark scheme (b)
Scheme
Marks
Comments
For each additional year of schooling, the life expectancy goes up 2.95 (or 3) years on average.
E1
[1]
Mark scheme (c)
Scheme
Marks
Comments
Valid comment. e.g. It could be causal. More education is likely to lead to better life choices and thus increases life expectancy.
e.g. There may be a third factor like wealthy countries are likely to have more education and also better access to health care.
E1
E0 for ‘life expectancy and schooling don’t correlate’.
B1 for 74.98. B1 for 75 with correct calculation. B0 for 75 with no calculation.
(ii) Valid comment E.g. Likely to be reasonably reliable because 14.6 is within the range of the data and the pattern of the data seems to be roughly linear. e.g. Reliable because we’re interpolating.
E1
Condone without ‘roughly linear’.
E0 for unreliable. E0 for ’75 is within the range of the data.’ E0 sample is too small.
[2]
Mark scheme (e)
Scheme
Marks
Comments
Valid comment. Eg. There seems to be lots of countries with 0 years of schooling. This is unlikely to be accurate. One could either ignore the data in any analysis or further research could be carried out to find the number of years of schooling for the countries with the missing data.
Accept answers rounding to 8.1, 8.2, 8.3, 8.25 Accept calculation of \(s=\sqrt{\dfrac{1}{182}\left(950615.47-\dfrac{13105.9^2}{183}\right)}=8.1237\ldots\) M1A1
[3]
(ii) Valid comment Eg. The mean should be weighted by country population size.
E1
E0 ‘May not include all countries of the world.’ E0 Only a small number/handful of countries. E0 I support this claim. E0 Shows life expectancy of people who didn’t go to school. E0 for implying we need to take into account the whole population. E0 for should use median or mode.
3. A city councillor is investigating the decline in the number of bus drivers in the city. During her research, she finds an article claiming that delivery companies are actively recruiting bus drivers to work for them.
She collects data on the number of bus drivers and the number of delivery drivers from 8 randomly selected cities of a similar size to her city. She calculates the correlation coefficient in order to carry out a hypothesis test.
(a) Explain why a one-tailed test could be appropriate. [1]
The product moment correlation coefficient for the data is \(-0{\cdot}8206\).
(b) Carry out a one-tailed significance test at the 5% significance level. [5]
(c)
(i) Suggest a reason why the city councillor might conclude that there is a causal link between the number of delivery drivers and the number of bus drivers.
(ii) What can the city councillor conclude about the recruitment of bus drivers by delivery companies over time? [2]
Mark scheme (a)
Scheme
Marks
Comments
Valid explanation e.g. The councillor is concerned about the decline of bus drivers being linked with an increase of delivery company drivers.
E1
For E1 must address decrease in bus drivers and increase in delivery drivers. Accept E1 Specifically looking for negative correlation. E1 Likely to be negative correlation. E0 Graph shows negative correlation. E0 Claim only talks about the decline, a negative statement. If it was research about positive or negative it would be 2-tailed. E0 Councillor is concerned about the decline of bus drivers.
[1]
Mark scheme (b)
Scheme
Marks
Comments
\(H_0:\rho=0\qquad H_1:\rho<0\)
B1
B0 for \(H_0\): correlation = 0. Population must be stated or implied. B0 for omission of \(\rho\) or use of \(r\). B0 for non-strict inequality.
\(\text{TS}=-0.8206\)
B1
B1 Labelled as test statistic or used in comparison with a critical value (not a significance level). If positive, only award once compared with \(0.6215\).
\(\text{CV}=(\pm)0.6215\)
B1
Since \(|\text{TS}|>0.6215\) (or \(-0.8206<-0.6215\)), there is sufficient evidence to reject \(H_0\).
B1
Must account for \(-\) at some point. e.g. \(-0.8206<0.6215\) is B0 FT their CV. Comparison may be implied by a diagram.
There is evidence to suggest that there is negative correlation between the number of delivery drivers and the number of bus drivers.
E1
cso Must have context.
[5]
Mark scheme (c)
Scheme
Marks
Comments
(i) Valid reason. e.g. There is only a limited number of drivers so a delivery company could offer more money to a driver and then that would lead to a decrease in bus drivers. e.g. Bus driver is becoming a less attractive job and delivery driver is becoming a more attractive job. e.g. Conditions of work may be better as a delivery driver compared to those of a bus driver.
E1
E1 Better job/pay/hours as a delivery driver.
E0 Number of bus drivers and delivery drivers are connected.
E0 for stating there is a causal link because bus drivers are becoming delivery drivers.
(ii) Valid reason. e.g. The city councillor cannot reach any conclusion because she does not have any data over time. She only has a snapshot at one point in time for 8 cities.
Does using cash for transactions affect people’s financial behaviour?
She collects data on ‘Cash transactions as a % of all transactions’ and ‘Household debt as a % of net disposable income’ from a random sample of 25 countries. The table below shows the data she collected. There are missing values, \(p\) and \(q\), for Malta and Denmark respectively.
Country
Cash transactions as a % of all transactions \(x\)
Household debt as a % of net disposable income \(y\)
Country
Cash transactions as a % of all transactions \(x\)
Household debt as a % of net disposable income \(y\)
Malta
92
\(p\)
France
68
120
Mexico
90
–14
Luxembourg
64
177
Greece
88
107
Belgium
63
113
Spain
87
110
Finland
54
137
Italy
86
87
Estonia
48
82
Austria
85
91
The Netherlands
45
247
Portugal
81
131
UK
42
147
Slovenia
80
56
Australia
37
214
Germany
80
95
USA
32
109
Ireland
79
154
Sweden
20
187
Slovakia
78
74
South Korea
14
182
Lithuania
75
46
Denmark
\(q\)
261
Latvia
71
43
The summary statistics and scatter diagram below are for the other 23 countries.
(a) Using the summary statistics for the 23 countries, calculate and interpret Pearson’s product moment correlation coefficient. [3]
(b) Calculate the equation of the least squares regression line of Household debt as a % of net disposable income (\(y\)) on Cash transactions as a % of all transactions (\(x\)). [5]
The regression line \(x\) on \(y\) is given below.
\[x=-0{\cdot}24y+91{\cdot}92\]
(c) By selecting the appropriate regression line in each case, estimate the values of \(p\) and \(q\) in the table. [2]
(d) Comment on the reliability of your answers in part (c). [1]
(e) Interpret the negative value of \(y\) for Mexico. [1]
Valid interpretation. e.g. The negative value of 0.628 implies that on average the more people rely on cash, the less debt they have as a proportion of their household income. e.g. The correlation is fairly strong between household debt and the use of cash for transactions. e.g. As the percentage of cash transactions increases, net household debt as a percentage of disposable income tends to decrease. e.g. Household debt and percentage of cash transactions are negatively correlated.
Valid explanation e.g. Not reliable because it’s extrapolation, rather than interpolation. e.g. May not be reliable because the values are very close to the extremes of the graph. e.g. May be reliable because there is fairly strong correlation and the values are only just outside the range of the graph.
E1
Mark scheme (e)
Scheme
Marks
Comments
Valid explanation. e.g. Net disposable income may be negative. e.g. Household incomes are less than outgoings leading to a negative value for net disposable income. e.g. Maybe an erroneous value.
3. Awena has a large data set of body measurements, and she wants to investigate relationships between body dimensions. In this particular investigation, she is testing for a correlation between forearm girth and bicep girth. The diagrams below show how to measure these.
(a) Why is it appropriate for Awena to use a one-tailed test? [1]
Awena takes a random sample of size 11 from her data set and plots the following scatter diagram.
(b) Using the computer output above, carry out a one-tailed significance test on the sample product moment correlation coefficient at the 0·5% level. [5]
(c) Blodwen also has access to the same large data set. She decides to do the same test using all of the 507 available data points. Her results are shown below.
(i) State the problem Blodwen will encounter when attempting to use statistical tables for her test.
(ii) How should Blodwen deal with this problem? [2]
Mark scheme (a)
Scheme
Marks
Comments
Valid reason, e.g.,
It is obvious that, in the general population, people with bigger biceps would have bigger forearms.
It is very unreasonable to think that people with bigger forearms have smaller biceps in general.
No need to check for negative correlation.
She knows there will be some kind of positive correlation.
E1
Condone “negative correlation is impossible.”
E0 for the diagram shows positive correlation. E0 for “it should be a two-tailed test.” E0 for “specifically looking for a positive or negative correlation” (both positive and negative stated). E0 for implying a causal relationship, e.g. “increase in forearm girth will cause an increase in bicep girth”.
[1]
Mark scheme (b)
Scheme
Marks
Comments
(Let \(\rho\) denote the population correlation coefficient between forearm girth and bicep girth.) \(H_0\colon\rho=0\) \(H_1\colon\rho\gt 0\)
B1
Allow other letters if defined. Allow worded hypotheses
B0 for \(H_0\): correlation = 0. Population must be stated or implied. B0 for omission of \(\rho\) or use of \(r\) B0 for a non-strict inequality in \(H_1\)
TS \(\left(=\sqrt{0.9412}\right)=0.97015\ldots\)
B1
Labelled as TS or used in correct comparison with a critical value (not with significance level). B0 for \(r=0.97015\) without label as TS or used in comparison. B0 for \(TS=\pm 0.97015\) unless the positive value correctly used later.
CV = 0.7348
B1
B0 for 0.7545 from Spearman’s rank. Condone 0.7348 stated as the CV even if using a two-tailed test.
Since 0.970 > 0.7348 (there is sufficient evidence to) Reject \(H_0\).
B1
FT for using 0.9412 and their CV. (Comparison may be implied by a diagram.)
Sufficient evidence to suggest there is a positive correlation between bicep girth and forearm girth.
E1
cso Do not allow categorical statements
E0 for omission of the word positive (unless positive implied by contextualised comment, e.g. E1 for sufficient evidence to suggest that the bigger the bicep girth the bigger the forearm girth.)
[5]
Mark scheme (c)
Scheme
Marks
Comments
(i) Her \(n\) is not in the table.
E1
oe Condone “likely not in the table”.
(ii) Valid comment, e.g.
Use a software package to carry out the test.
Use a software package to find the CV.
Use a software package to find the p-value.
Realise that the CVs are lower than those for \(n=100\) in the tables.
Realise that there is no need to carry out a test for correlation when the scatter diagram clearly shows very strong correlation (and is bivariate normal).
E1
Condone “Use an online table that goes to 507.” Allow “Use a calculator to find the CVs.” E0 for “use the calculator” without reference to a possible use of the calculator (e.g. find the CV). E0 for “use a smaller sample size.” E0 for “take a sample so that you can use tables.” Allow “work out critical values using a t-distribution.”
5. A medical researcher is investigating possible links between diet and a particular disease. She selects a random sample of 22 countries and records the average daily calorie intake per capita from sugar and the percentage of the population who suffer from this disease.
There are 22 data points and the product moment correlation coefficient is 0·893.
(a) Stating your hypotheses clearly, show that these data could be used to suggest that there is a link between the disease and sugar consumption. [5]
The medical researcher realises that her data is from the year 2000. She repeats her investigation with a random sample of 13 countries using new data from the year 2020. She produces the following graph.
(b) How should the researcher interpret the new data in the light of the data from 2000? [2]
Mark scheme (a)
Scheme
Marks
Comments
(Let \(\rho\) denote the population correlation coefficient between per capita daily calories from sugar consumption and percentage of the population with the disease.) \(H_0:\rho=0\qquad H_1:\rho\neq 0\)
B1
Allow other letters if defined. Allow worded hypotheses provided the population is stated or implied. B0 for \(H_0\): correlation = 0. Population correlation must be stated or implied. B0 for omission of \(\rho\), or \(r\) in place of \(\rho\). B0 for a one-tailed test.
\(\mathrm{TS}=0.893\)
B1
Labelled as TS or used in correct comparison
CV = (\(\pm\)) 0.5368 (at the 1% level) (CR is \(R\lt -0.5368\) or \(R\gt 0.5368\))
B1
Accept use of different significance levels. OR Award this mark for stating “the test statistic is greater than all critical values”.
Since \(\mathrm{TS}\gt\ 0.5368\), there is sufficient evidence to reject \(H_0\).
B1
FT their CV. Correct comparison required (may be implied through a diagram).
This suggests that there is a link between the disease and sugar consumption.
E1
CSO Do not allow categorical statements (condone categorical statements if “sufficient evidence” seen in m1 statement). Statement must match two tailed hypothesis. E0 for a one-tailed conclusion, e.g. there is evidence of a positive correlation between…
Mark scheme (b)
Scheme
Marks
Comments
Valid comment on correlation in 2000 and 2020. e.g. Although there is (evidence of) a (strong) positive correlation in 2000 there seems to be no correlation in 2020.
E1
Valid comment on the ambiguity of the studies. e.g. This would imply that sugar consumption is only coincidentally correlated with the disease in 2000. e.g. There must be a different cause for the disease other than sugar. e.g. \(n\) is too small to make meaningful comparisons. e.g. different samples lead to different conclusions, which may explain the difference. e.g. the link between the disease and sugar consumption has diminished over the 20 years.
E1
E1 for identifying a drawback of the conclusions or identifying a reason for the change. E0 for “this is because there is no link between the disease and sugar consumption” without substantiating with a valid reason.
OR Valid comment on third factor correlated with sugar consumption and the disease in 2000 no longer relevant in 2020. e.g. An additive may have been added to many products containing sugar in 2000 which was contributing to the prevalence of the disease but has been banned by 2020. e.g. Advancements in medical care for the disease from 2000 to 2020 are likely to explain the change in relationship.
(E2)
E1 for correctly identifying a possible confounding variable. E1 for explaining why this may lead to the change in relationship seen.
Total [7]
Corrected by Westie's Workshop
The board's mark scheme printed “would substantiating”. Changed to “without substantiating” because “would” does not make sense there; the sentence means a bare claim with no valid reason gets E0.
Original: see the board's mark scheme PDF (link above).
(a) Give two circumstances where it may be more appropriate to use Spearman’s rank correlation coefficient rather than Pearson’s product moment correlation coefficient. [2]
(b) A farmer needs a new tractor. The tractor salesman selects 6 tractors at random to show the farmer. The farmer ranks these tractors, in order of preference, according to their ability to meet his needs on the farm. The tractor salesman makes a note of the price and power take-off (PTO) of the tractors.
Tractor
Farmer’s rank
PTO (horsepower)
Price (£1000s)
A
1
77·5
80
B
6
87·9
45
C
5
53·0
47
D
4
41·0
53
E
2
112·0
60
F
3
90·0
61
Spearman’s rank correlation coefficient between the farmer’s ranks and the price is 0·9429.
(i) Test at the 5% significance level whether there is an association between the price of a tractor and the farmer’s judgement of the ability of the tractor to meet his needs on the farm. [4]
(ii) Calculate Spearman’s rank correlation coefficient between the farmer’s rank and PTO. [4]
(iii) How should the tractor salesman interpret the results in (i) and (ii)? [2]
Mark scheme (a)
Scheme
Marks
Comments
Two appropriate circumstances. e.g. When the data are ordinal. e.g. When the data are not bivariate normal.
B2
B1 for one circumstance.
Mark scheme (b)
Scheme
Marks
Comments
(b)(i)
\(H_0\): there is no association between the rank given by the farmer and the price. \(H_1\):there is an association between the rank given by the farmer and the price.
B1
Both. Do not allow correlation.
5% two tailed critical value = (\(\pm\))0.8286
B1
Since \(0.9429>0.8286\) there is sufficient evidence to reject \(H_0\).
B1
Comparing TS to CV FT their CV
There is sufficient evidence to suggest that there is an association between the rank given by the farmer and price.
E1
Comment in context Only award if previous 3 B1 have been awarded
(ii)
Tractor
Farmer’s rank
PTO rank (highest to lowest)
PTO rank (lowest to highest)
A
1
4
3
B
6
3
4
C
5
5
2
D
4
6
1
E
2
1
6
F
3
2
5
B1
Correct values for one set of PTO ranks
\(\sum d^2=24\) or \(\sum d^2=46\)
B1
si
\(r_s=1-\frac{6\times24}{6\times35}\) or \(r_s=1-\frac{6\times46}{6\times35}\)
M1
FT their ‘24’ or ‘46’
\(r_s=\pm0.3143\) or \(\pm\frac{11}{35}\)
A1
(iii)
Two valid comment. e.g. The PTO variable isn’t very strongly associated with the farmer’s ranks so may not be worth analysing.
e.g. The salesman doesn’t need to work that hard because the farmer already prefers the most expensive tractors.
e.g. because the number of tractors in the sample is small, the salesman should not place too much reliance on the results.
2. For a set of 30 pairs of observations of the variables \(x\) and \(y\), it is known that \(\sum x=420\) and \(\sum y=240\). The least squares regression line of \(y\) on \(x\) passes through the point with coordinates (19, 20).
(a) Show that the equation of the regression line of \(y\) on \(x\) is \(y=2{\cdot}4x-25{\cdot}6\) and use it to predict the value of \(y\) when \(x=26\). [6]
(b) State two reasons why your prediction in part (a) may not be reliable. [2]
Mark scheme (a)
Scheme
Marks
Comments
\(\bar{x}=14\) \(\bar{y}=8\)
B1
Both
\(b=\dfrac{20-8}{19-14}\)
M1
\(b=2.4\)
A1
convincing
Equation of regression line is
\(y-20=2.4(x-19)\)
M1
oe
\(y=2.4x-25.6\) *ag
A1
convincing
When \(x=26\) \(y=2.4\times26-25.6\) \(y=36.8\)
B1
1st ALTERNATIVE METHOD
Scheme
Marks
Comments
\(y=mx+c\)
\(20=2.4\times19+c\)
(M1)
oe
\(c=-25.6\)
\(y=2.4x-25.6\) *ag
(A1)
convincing
2nd ALTERNATIVE METHOD
Scheme
Marks
Comments
\(y=bx+a\) \(20=19b+a\)
(M1)
For 1st equation
\(\dfrac{240}{30}=\dfrac{420}{30}b+a\)
(B1)
For sight of \(\frac{240}{30}\) and \(\frac{420}{30}\)
(A1)
For correct equations
Solve simultaneously to get \(a=-25.6\) and \(b=2.4\)
(M1)
\(y=2.4x-25.6\) *ag
(A1)
Mark scheme (b)
Scheme
Marks
Comments
Comment on linearity e.g. Scatter diagram may not be linear. Correlation might be weak.
E1
Comment on range of \(x\) values. e.g. 26 may be beyond the range of observed \(x\) values.
7. Data from a large dataset shows the percentage of children enrolled in secondary education and the percentage of the adult population who are literate. The following graphs show data from 30 randomly selected regions from each of the Arab World, Africa and Asia. In each case, the least squares regression line of ‘% Literacy’ on ‘% Enrolled in Secondary Education’ is shown.
(a) Calculate the equation of the least squares regression line of ‘% Literacy’ (\(y\)) on ‘% Enrolled in Secondary Education’ (\(x\)) for Asia, given the following summary statistics. [5]\[\textstyle\sum x=2850{\cdot}836\qquad\sum y=2738{\cdot}656\qquad S_{xx}=88{\cdot}42142\]\[S_{yy}=204{\cdot}733\qquad S_{xy}=96{\cdot}60984\qquad n=30\]
(b) The Arab World, Africa and Asia each contain a region where 70% are enrolled in secondary education. The three regression lines are used to estimate the corresponding % Literacy. Which of these estimates is likely to be the most reliable? Clearly explain your reasoning. [2]
5. A government statistician wishes to investigate whether there is positive correlation between the variables Average score in the National Reading Test and Average House Price. She collects data on a randomly selected sample of 21 areas within Wales.
(a) Carry out a 5% significance test for correlation between the two variables, using the computer output below. [5]
After seeing this output, the editor of a newspaper leads with the front-page headline,
“Buy a more expensive house to boost your children’s reading ability!”
(b) Discuss whether the data support the headline. [2]
(c) Give a possible alternative explanation for this correlation. [1]
Mark scheme (a)
Scheme
Marks
Comments
(Let \(\rho\) denote the population correlation coefficient between average house price and average score in the national reading test). \(H_0:\rho=0\) \(H_1:\rho\gt 0\)
B1
Allow other letters if defined. Allow worded hypotheses. B0 for \(H_0\): correlation = 0. Population must be stated or implied. B0 for omission of \(\rho\) or use of \(r\) B0 for a non-strict inequality in \(H_1\)
TS = 0.86371…
B1
Labelled as TS or used in comparison B0 for \(TS=\pm 0.86371\) unless the positive value correctly used later.
CV = 0.3687
B1
FT their hypotheses (e.g., 0.4329 for two-tailed)
Since TS > 0.3687, there is sufficient evidence to reject \(H_0\).
B1
FT for using 0.746 FT their CV
Sufficient evidence to suggest there is positive correlation between the average house price and average national reading test score.
E1
CSO (correct solution only). E0 for categorical statements or omission of the word positive (unless positive implied by contextualised comment). E0 for conclusion not in context
Part (a) Total
[5]
Mark scheme (b)
Scheme
Marks
Comments
Valid comment saying the two variables are linked, i.e., giving a reason for the headline. e.g., The data support the idea that the more expensive houses are correlated with better reading scores.
E1
Valid comment saying why the headline is unreasonable. e.g., It’s unreasonable to suggest that a more expensive house will improve a child’s reading ability.
E1
E0 for correlation does not imply causation unless explained in context. Condone responses that give a valid alternative explanation for the correlation.
Part (b) Total
[2]
Mark scheme (c)
Scheme
Marks
Comments
Possible explanation. e.g., parents who can afford better houses may have a better education so are more likely to help their children to read.
E1
E0 for comments such as “those with higher reading scores can afford better houses”. Do not accept “small sample size” or “it is a coincidence”. Condone a repetition of a valid alternative explanation that was given in (b). Do not condone a valid alternative explanation given in (b) only.
4. Jessica is studying the relationship between hip girth, \(h\) cm, and thigh girth, \(t\) cm, for American adults who are physically active. She takes a random sample of 11 people from a very large dataset which she has downloaded into a spreadsheet software package. The results are shown below.
\(h\) (cm)
98·6
112·1
97·9
110·2
89·2
111·7
87·0
94·7
100·4
104·0
88·4
\(t\) (cm)
48·3
87·2
55·2
68·0
48·5
63·2
49·5
55·7
59·1
64·0
52·4
(a) Jessica notes that, for the thigh girth data, the lower quartile is 49·5 and the upper quartile is 64·0.
(i) Show that 87·2 should be classified as an outlier for \(t\).
(ii) Give a reason why Jessica might exclude the outlier.
(iii) Give a reason why Jessica might include the outlier. [4]
Jessica decides to exclude the outlier and produces the following scatter diagram.
(b) Interpret, in context, the correlation in the data shown in the diagram. [1]
The equation of the regression line of \(t\) on \(h\) for this sample is
\[t=0{\cdot}69h-11{\cdot}26.\]
(c) Interpret the gradient of the regression line in this context. [1]
(d) Use your knowledge of large data sets and spreadsheet software packages to suggest a way in which Jessica could improve her investigation. [1]
(ii) Valid reason. e.g. May distort calculations and therefore inferences about population.
E1
Condone “so that the mean will not be affected by outliers.” Allow “there may have been an error in measurement”. Allow “may skew the data”. E0 for “median does not change”. E0 for “make calculations more reliable”.
(iii) Valid reason. e.g., It’s only just an outlier. Still a valid measurement so should include it.
E1
Condone “to use all of the data”.
Mark scheme (b)
Scheme
Marks
Comments
The larger the hip girth, the larger the thigh girth, on average.
E1
oe Condone ‘tends’ or ‘in general’ in place of on average. Penalise first omission of on average only.
Mark scheme (c)
Scheme
Marks
Comments
Each increase of 1 cm of hip girth corresponds to a 0.69 cm increase in thigh girth on average.
E1
Penalise first omission of on average only. Watch out for hip girth and thigh girth in the wrong order.
Mark scheme (d)
Scheme
Marks
Comments
Using all of the data instead of a sample would lead to more accurate results.
E1
Condone increase sample size provided no nonsensical statements follow.
2. An economist suggested the rate of unemployment and the rate of wage inflation are independent. Amy sets about investigating this suggestion. She collects unemployment data and wage inflation data from a random sample of regions in the UK and decides that it is appropriate to carry out a significance test on Pearson’s product moment correlation coefficient.
Amy’s summary statistics for percentage unemployment, \(x\), and percentage wage inflation, \(y\), are shown below.
(a) Calculate Pearson’s product moment correlation coefficient for these data. [4]
(b) Carry out Amy’s test at the 5% level of significance and state whether the economist’s suggestion is reasonable. [4]
Amy also collects unemployment data and wage inflation data from a random sample of 10 regions in Spain and calculates Pearson’s product moment correlation coefficient to be \(-0{\cdot}2525\).
(c) Should this change Amy’s opinion on the economist’s suggestion above? What could she do to improve her investigation? [2]
(d) What assumption has Amy made in deciding that it is appropriate to carry out a significance test on Pearson’s product moment correlation coefficient? [1]
It suggests that the rate of unemployment and the rate of wage inflation are not independent.
E1
Only award E1 if previous three B1 awarded E0 for categorical statements
Mark scheme (c)
Scheme
Marks
Comments
Valid comment. e.g. This should cast doubt on Amy’s opinion based on her answer in (b)
E1
FT their conclusion from (b)
Valid suggestion. e.g. She could look at more countries. She could come to different conclusions for different countries. She could consider more regions within each country
E1
Mark scheme (d)
Scheme
Marks
Comments
The underlying distribution is bivariate normal. The data come from a bivariate normal distribution.
6. The University of Arizona surveyed a large number of households. One purpose of the survey was to determine if annual household income could be predicted from size of family home.
The graph of Annual household income, \(y\), versus Size of family home, \(x\), is shown below.
(a) State the limitations of using the regression line above with reference to the scatter diagram. [1]
The data for size of family homes between 2000 and 3000 square feet are shown in the diagram below.
(b) Calculate the equation of the least squares regression line to predict Annual household income from Size of family home for these data. [5]
Mark scheme (a)
Scheme
Marks
Comments
Valid comment. Eg. Data suggests a non-linear relationship. The scatter diagram seems to have three distinct sections. The regression line will give an underestimate for some values of x e.g. between 2500 and 4000 and an overestimate for others e.g. between 1500 and 2500. Three separate lines for the distinct sections would be better.
5. A bowling alley manager in the UK is concerned about falling revenues. He collects data from the United States, hoping to use what he finds to revive his business in the UK.
He finds data which seem to show correlation between margarine consumption and bowling alley revenue. He attempts to carry out some statistical analysis in order to present his findings to the board of directors. He produces the scatter diagram shown below.
The product moment correlation coefficient for these data is \(-0{\cdot}7617\). He carries out a one-tailed test at the 1% level of significance and concludes that higher margarine consumption is associated with lower revenue generated by bowling alleys.
(a) Show all the working for this test. [5]
The manager also conducts a significance test for bowling alley revenue and fish consumption per person. He produces the computer output, shown below, for the analysis of bowling alley revenue versus fish consumption per person.
# Pearson’s product-moment correlation
# data: revenue and fish
# t = 3.8303, df = 8, p-value = 0.005215
# alternative hypothesis: true correlation is not equal to 0
# sample estimates:
# correlation
# 0.8024423
(b) Comment on the correlation between bowling alley revenue and fish consumption per person and what the board of directors should do in light of the manager’s findings in part (a) and part (b). [3]
(c) Give one possible reason why the board of directors might not be happy with the manager’s analysis. [1]
Mark scheme (a)
Scheme
Marks
Comments
\(H_0:\rho=0\qquad H_1:\rho<0\)
B1
B0 for \(H_0\): correlation = 0. Population must be stated or implied.
\(n=10\)
B1
si
\(\text{TS}=-0.7617\)
B1
B1 for use of or stating as test statistic
CV = \((\pm)\) 0.7155
B1
Since \(|\text{TS}|>\ 0.7155\) (or \(\text{TS}<-0.7155\)), there is sufficient evidence to reject \(H_0\).
B1
Mark scheme (b)
Scheme
Marks
Comments
p-value \(=0.005215\) is less than 0.01
E1
Condone any significance level less than or equal to 0.1
Bowling alley revenue is (positively) correlated with fish consumption per person.
E1
Do not allow weak correlation
The correlations, despite being significant, seem irrelevant in both cases and therefore the board of directors should not implement any suggestions made by the manager based on his calculations.
E1
Award E1 for a comment that implies the board of directors should do nothing, or that the correlations are irrelevant with a reasonable course of action.
Do not allow “the bowling alley should sell more fish and less margarine”
Do not allow board of directors should do nothing because there is not correlation.
Mark scheme (c)
Scheme
Marks
Comments
Valid comment. e.g. It is unclear what each data point represents. e.g. He has carried out the tests without considering how irrelevant the findings may be. e.g. It’s unclear whether this data is a sample, so no inferences can be made. e.g. It’s not clear what the population is. e.g. The data is not bivariate normal
5. Huw investigates the opinion of the public about their local primary schools. He visits his town centre and administers a questionnaire to 120 people over the course of several days.
(a) What type of sampling is this? [1]
(b) How could Huw’s sampling be improved? [1]
Huw wishes to investigate the relationship between the population and the number of primary schools in a county. He searches the internet for data and selects a random sample of 24 counties from the UK. He produces the following scatter diagram.
The equation of the regression line for this dataset is given by
\[\text{Number of Primary Schools}=22{\cdot}7+0{\cdot}2406\times\text{County Population (in 1000s)}.\]
(c) State a limitation of this regression line. [1]
(d) Calculate an estimate for the number of primary schools in a county with a population of 889 000. [2]
(e) Statisticians often say, “Correlation does not imply causation”. Comment on this statement in this context. [1]
Mark scheme (a)
Scheme
Marks
Comments
Opportunity sampling.
B1
Mark scheme (b)
Scheme
Marks
Comments
Valid comment. eg. Go to different areas, as this is a biased sample. Many of his responders will be from the same community.
B1
Mark scheme (c)
Scheme
Marks
Comments
Appropriate statement. eg. The line is only valid for predicting the number of primary schools in counties with populations between approximately 100,000 and 2,800,000.
E1
Allow: The model predicts that a county with a population of 0 has about 23 primary schools
Mark scheme (d)
Scheme
Marks
Comments
\(\text{No. of Primary Schools}=22.7+0.2406\times889\)
M1
M0 for using 889,000
\(=237\text{ schools}\)
A1
A0 for 236.5934
Mark scheme (e)
Scheme
Marks
Comments
Appropriate comment. eg. It is likely that number of primary schools IS caused by a greater population (because local authorities are obliged to provide education).
Closing a school doesn’t automatically decrease the county population (as pupils would be absorbed into other local schools)
Increasing the county population doesn’t automatically increase the number of schools, (but they are clearly related)
E1
Do not accept a categorical statement of causation
(a) Sketch a scatter diagram of a dataset for which Spearman’s rank correlation coefficient is +1, but the product moment correlation coefficient is less than 1. [2]
Two judges were judging cheese at the UK Cheese Festival. There were 8 blue cheeses in a particular category. The rankings are shown below.
Cheese
A
B
C
D
E
F
G
H
Judge 1
1
5
8
7
6
4
3
2
Judge 2
1
3
8
5
2
4
6
7
(b) Calculate Spearman’s rank correlation coefficient for this dataset. [3]
(c) By sketching a scatter diagram of the rankings, or otherwise, comment on the extent to which the judges agree. [2]
Mark scheme (a)
Scheme
Marks
Comments
B2
B2 Increasing ranks and non linear. At least 3 points. B1 for positive correlation.
Mark scheme (b)
Scheme
Marks
Comments
\(\sum d^2=58\)
B1
\(r_s=1-\dfrac{6\times58}{8\times63}\)
M1
\(\phantom{r_s}=0.3095\ldots=\frac{13}{42}\)
A1
Mark scheme (c)
Scheme
Marks
Comments
Valid comment on agreement. e.g. Both judges agree on the best and worst cheese. e.g. The judges agree on 3 of the 8 cheeses.
E1
Valid comment on disagreement. e.g But they almost completely disagree about the others.
7. A university professor conducted some research into factors that affect job satisfaction. The four factors considered were Interesting work, Good wages, Job security and Appreciation of work done.
The professor interviewed workers at 14 different companies and asked them to rate their companies on each of the factors. The workers’ ratings were averaged to give each company a score out of 5 on each factor.
Each company was also given a score out of 100 for Job satisfaction.
The following graph shows the part of the research concerning Job Satisfaction versus Interesting work.
(a) Calculate the equation of the least squares regression line of Job satisfaction (\(y\)) on Interesting work (\(x\)), given the following summary statistics. [5]\[\textstyle\sum x=46{\cdot}2,\qquad\sum y=898,\qquad S_{xx}=3{\cdot}48\]\[S_{xy}=49{\cdot}45,\qquad S_{yy}=1437{\cdot}714,\qquad n=14\]
(b) Give two reasons why it would be inappropriate for the professor to use this equation to calculate the score for Interesting work from a Job satisfaction score of 90. [2]
The board's mark scheme printed “\(b=\dfrac{49.4511}{3.48}\)”. Changed to “\(b=\dfrac{49.45}{3.48}\)” because the question gives \(S_{xy}=49.45\), and \(49.45\div3.48=14.2(0977\ldots)\), the value on the next line.
Original: see the board's mark scheme PDF (link above).
Mark scheme (b)
Scheme
Marks
Comments
The regression line \(x\) on \(y\) would be more appropriate. OR would not be able to rearrange this equation to find \(x\) from \(y\).
E1
Gives a value outside the range. OR \(90\) is outside the range.
5. Rebecca is a farmer who is monitoring prices for products to use on her farm. She records the prices of two products made from different grains, wheat and oats, at random points in time, to investigate whether there is any correlation.
The product moment correlation coefficient for the data is 0·244. There are 12 data points, and the \(p\)-value is 0·4447.
(a) Comment on the correlation between the prices of Feed Wheat and Feed Oats. [2]
Rebecca also records the prices of two wheat products at random points in time, to investigate whether there is any correlation.
The product moment correlation coefficient for the data is 0·653. There are 12 data points.
(b) Stating your hypotheses clearly, test at the 5% level of significance whether there is any evidence of correlation between the prices of these two products. [5]
(c) Without referring to the positioning of the points on the graphs, suggest why the product moment correlation coefficient is higher for the second set of data. [1]
Mark scheme (a)
Scheme
Marks
Comments
p- value for ‘Feed Wheat’ versus ‘Feed Oats’ > 0.05
E1
Price of Feed Wheat does not seem to be correlated with price of Feed Oats
E1
Mark scheme (b)
Scheme
Marks
Comments
H0: \(\rho\) = 0 H1: \(\rho\neq 0\)
B1
H0: \(\rho\) = 0 H1: \(\rho\) > 0
TS = 0.653 CV = 0.5760 Since TS>0.5760 Reject H0. Sufficient evidence to suggest there is a correlation between the prices of feed wheat and wheat straw.
B1 B1 B1 E1
TS = 0.653 CV = 0.4973 Since TS>0.4973 Reject H0. Strong evidence to suggest there is a positive correlation between the prices of feed wheat and wheat straw.
Mark scheme (c)
Scheme
Marks
Comments
Appropriate comment implying understanding that the second graph is comparing two wheat products whereas the first is comparing different grains.
5. A baker is aware that the pH of his sourdough, \(y\), and the hydration, \(x\), affect the taste and texture of the final product. The hydration is measured in ml of water per 100 g of flour (ml / 100 g). The baker researches how the pH of his sourdough changes as the hydration changes.
The results of his research are shown in the diagram below.
(a) Describe the relationship between pH and hydration. [2]
(b) The equation of the regression line for \(y\) on \(x\) is \[y=5{\cdot}4-0{\cdot}02\,x.\]
(i) Interpret the gradient and intercept of the regression line in this context.
(ii) Estimate the pH of the sourdough when the hydration is 20 ml / 100 g. Comment on the reliability of this estimate. [4]
Mark scheme (a)
Scheme
Marks
Comments
Strong linear relationship
E1
The higher the hydration the lower the pH.
E1
Mark scheme (b)
Scheme
Marks
Comments
(i) Each additional ml of water per 100g of flour decreases the pH by 0.02 on average.
E1
Or equivalent
The intercept would imply that at zero hydration the pH would be 5.4.
E1
(ii) \(y=5.4-0.02\times20\)
\(y=5\)
B1
From use of regression line.
Any correct comment. e.g. Outside the data set Extrapolation, etc.
4. On a Welsh television game show, contestants are asked to guess the weights of a random sample of seven cows. The game show judges want to investigate whether there is positive correlation between the actual weights and the estimated weights. The results are shown below for one contestant.
Cow
A
B
C
D
E
F
G
Actual weight, kg
614
1105
718
1001
889
770
682
Estimated weight, kg
700
1500
850
1400
750
900
800
(a) Calculate Spearman’s rank correlation coefficient for this data set. [5]
(b) Stating your hypotheses clearly, determine whether or not there is evidence at the 5% significance level of a positive association between the actual weights and the weights as estimated by this contestant. [3]
(c) One of the game show judges says, “This contestant was good at guessing the weights of the cows.” Comment on this statement. [1]
Mark scheme (a)
Scheme
Marks
Comments
The ranks are
Cow
A
B
C
D
E
F
G
Actual weight
7
1
5
2
3
4
6
Estimated weight
7
1
4
2
6
3
5
B1 B1
Correct values for first row. Correct values for second row. Accept reverse ranks.
We'd like to use Google Analytics cookies to count visits and see which pages help most. They're optional, and the site works exactly the same either way. Privacy policy