Scatter Graphs - Worksheets, Questions and Revision
51 original exam-style questions - 36 pages of questions with a full mark scheme - free printable PDF.
Scatter Graphs revision guideMore GCSE Maths practiceMock exams
This topic is chapter 7 of GCSE Maths Foundation Topic Workbook 9 and chapter 12 of GCSE Maths Grade 5 Booster 3.

A scatter graph is a diagram that plots pairs of data values as points, used to show whether there is a relationship, or correlation, between two variables. It appears in GCSE statistics, where students describe positive, negative or no correlation, spot outliers, and draw a line of best fit to estimate unknown values.
- Plot each pair of data as a point on the grid, using the x-value for the first variable and the y-value for the second.
- Look at the overall pattern of the points to decide the type of correlation: positive (both increase together), negative (one increases as the other decreases), or no correlation (no pattern).
- Identify any outliers, points that do not fit the general trend, and consider whether to include them when judging the correlation.
- Draw a line of best fit through the data: a straight line that follows the trend with roughly equal numbers of points above and below it.
- Use the line of best fit to read off an estimate for a value of one variable given the other.
- Check whether your estimate is interpolation (within the range of the data) or extrapolation (outside it), since extrapolated estimates are unreliable.
Worked example
A gardener records the height of a plant (cm) against the number of days since planting. A line of best fit has been drawn through the points (10, 20) and (30, 100). Use the line of best fit to estimate the height of the plant after 15 days.
- Find the gradient of the line of best fit: gradient = (100 - 20) / (30 - 10) = 80/20 = 4.
- Write the equation relating height (y) to days (x) starting from the point (10, 20): y = 20 + 4 x (x - 10).
- Substitute x = 15 into the equation: y = 20 + 4 x (15 - 10) = 20 + 4 x 5.
- Work out the value: y = 20 + 20 = 40.
- State the final answer: the estimated height after 15 days is 40 cm. Since 15 days is within the range of days used to draw the line, this is a reliable interpolation.
4.20 Scatter Graphs
Do these points show a positive correlation, negative correlation, or no correlation? Explain in one short sentence.
Give a one-sentence description of the strength of the correlation (strong, moderate or weak) and state whether it is positive or negative.
- A) As one variable increases, the other variable also increases.
- B) As one variable increases, the other variable decreases.
- C) There is no relationship between the two variables.
- D) Both variables always stay the same value.
Describe the correlation shown by the scatter graph.
Find the equation of the line of best fit, and use it to estimate y when x = 12.
Estimate the number of tubs sold when the temperature is 15°C by drawing a rough line of best fit through the points and reading off your estimate.
Give your estimate.
Number of customers (x): 4, 6, 8, 9, 11, 13, 15, 17, 19, 22
Tips collected in GBP (y): 12, 18, 25, 30, 38, 45, 52, 60, 68, 78
Temperature in °C (x): 2, 5, 7, 9, 12, 14
Woolly hats sold (y): 42, 35, 30, 24, 18, 14
The remaining two days had a temperature of 16°C with 10 hats sold, and a temperature of 19°C with 5 hats sold.
Points: (45, 4.5), (50, 5.0), (55, 5.2), (60, 5.6), (48, 4.8), (52, 5.1), (58, 5.4), (62, 5.8).
Hours of sleep (x): 4, 5, 6, 7, 8, 9, 10, 11
Reaction time in ms (y): 340, 310, 290, 265, 105, 230, 200, 180
Pages read per day (x): 3, 5, 6, 8, 9, 11
Minutes spent reading (y): 15, 25, 30, 42, 47, 57
Temperature in °C (x): 14, 16, 18, 19, 21, 22, 24
Bottles of water sold (y): 20, 24, 28, 30, 34, 36, 40
On the remaining two days, the temperature was 26°C with 44 bottles sold, and 28°C with 48 bottles sold.
Hours on app (x): 1, 2, 3, 4, 5, 6, 7, 8
Test score out of 50 (y): 22, 25, 8, 30, 34, 38, 41, 45
Question 1
- (3, 5) plotted or coordinate given as (3, 5) caoB1
- Answer: (3, 5)
Question 2
- x = 7, y = 2 caoB1
- Answer: x = 7, y = 2
Question 3
- positive correlation (oe)B1
- Answer: Positive correlation
Question 4
- identify positive correlationM1
- brief reason, e.g. as x increases y increases or points rise left to right caoA1
- Answer: Positive correlation. As x increases, y increases.
Question 5
- identify positive correlationM1
- identify strength as strong (points lie close to a rising straight-line trend) caoA1
- Answer: Strong positive correlation.
Question 6
- correct option BB1
- Answer: B
Question 7
- negative correlation (oe)B1
- Answer: Negative correlation
Question 8
- compares how y changes as x increasesM1
- no correlationA1
- Answer: no correlation
Question 9
- gradient (25.5 - 19.5) / (21 - 9)M1
- finds the interceptM1
- y = 0.5x + 15A1
- estimate 21A1
- Answer: y = 0.5x + 15; estimate 21
Question 10
- (a) positive correlationB1
- (a) Answer: Positive correlation
- (b) negative correlationB1
- (b) Answer: Negative correlation
- (c) no correlationB1
- (c) Answer: No correlation
Question 11
- (a) all 10 points plotted correctly (allow small tolerance); award B1 for at least 7 points plotted correctlyB2
- (a) Answer: Points at (2,18), (4,25), (5,30), (7,42), (9,50), (10,55), (12,65), (14,72), (15,80), (18,90)
- (b) positive correlationB1
- (b) Answer: Positive correlation
Question 12
- £42,000 (or 42) caoB1
- Answer: £42,000
Question 13
- negative correlation (oe)B1
- Answer: Negative correlation
Question 14
- (a) positive correlationB1
- (a) Answer: Positive correlation
- (b) negative correlationB1
- (b) Answer: Negative correlation
- (c) no correlationB1
- (c) Answer: No correlation
Question 15
- £38 (or 38) caoB1
- Answer: £38
Question 16
- any correct feature, e.g. it follows the trend of the points with roughly equal numbers of points above and below it, or it passes through the mean point (oe)B1
- Answer: It should follow the trend of the data with roughly equal numbers of points above and below the line.
Question 17
- sensible line of best fit through points or description of trend and interpolation at x = 15M1
- interpolate between (14, 23) and (16, 28) or use a line giving around 25 per 15 degreesM1
- 25 tubs cao (allow answers 24 to 26 awrt)A1
- Answer: 25
Question 18
- (a) identify negative correlationM1
- (a) state strength e.g. moderate negative correlation (points generally fall as x increases) caoA1
- (a) Answer: Moderate negative correlation
- (b) interpolate between x = 4 (y = 2) and x = 6 (y = 2) or use nearby points to estimateM1
- (b) 2 books cao (allow 1 or 3 if justified)A1
- (b) Answer: 2
Question 19
- 200 pages is outside the range of the data collected (extrapolation)B1
- the trend seen for 20 to 60 pages may not continue for much larger magazines (e.g. different paper or cover may be used)B1
- Answer: The estimate is unreliable because 200 pages is far outside the data collected (20 to 60 pages), so it involves extrapolation and the same trend may not continue.
Question 20
- (a) 6 caoB1
- (a) Answer: 6 hours
- (b) 64.7 (or 64.67, or 64 2/3) awrt 64.7B1
- (b) Answer: 64.7% (1 d.p.)
- (c) (6, 64.7) ft from parts (a) and (b)B1
- (c) Answer: (6, 64.7)
Question 21
- (a) Graph AB1
- (a) Answer: Graph A
- (b) the points in Graph A lie closer to a straight line / are less scattered than in Graph BB1
- (b) Answer: The points in Graph A lie closer to a straight line, showing a stronger correlation than the scattered points in Graph B.
Question 22
- (a) point (27, 98) plotted correctlyB1
- (a) point (30, 112) plotted correctlyB1
- (a) Answer: (27, 98) and (30, 112)
- (b) positive correlationB1
- (b) Answer: Positive correlation
- (c) correct method, e.g. gradient = (110-30)/(30-10) = 4, then 30 + 4 x (20-10)M1
- (c) 70 caoA1
- (c) Answer: 70 ice creams
- (d) 38 is outside the range of the data (maximum recorded was 30), so this would be extrapolation and the trend may not continueB1
- (d) Answer: 38°C is outside the range of the data collected (up to 30°C), so using the line here is extrapolation and unreliable.
Question 23
- (a) point (8, 4500) plotted correctlyB1
- (a) point (9, 3700) plotted correctlyB1
- (a) Answer: (8, 4500) and (9, 3700)
- (b) correct method, e.g. gradient = (3000-10000)/10 = -700, then 10000 - 700 x 6M1
- (b) £5800 caoA1
- (b) Answer: £5800
- (c) 25 years is outside the range of the data collected (up to 9 years), so this is extrapolationB1
- (c) at 25 years the line would give a negative value (10000 - 700x25 = -7500), which is impossible, showing the linear trend cannot continueB1
- (c) Answer: 25 years is far outside the data range, and substituting gives a negative value (-£7500), which is impossible, so the line of best fit cannot be used here.
- (d) the value of the car decreases by about £700 for each extra year of ageB1
- (d) Answer: The value of the car decreases by approximately £700 for each additional year of age.
Question 24
- (a) correct method, e.g. gradient = (26-10)/(10-2) = 2, then 10 + 2 x (6-2)M1
- (a) 18 MPa caoA1
- (a) Answer: 18 MPa
- (b) day 0 is outside the range of the data collected (curing days from 2 to 10), so this is extrapolationB1
- (b) concrete needs time to cure, so there is no real evidence the same linear pattern applies right at the start of the processB1
- (b) Answer: Day 0 is outside the tested range of curing days (2 to 10), so this is extrapolation, and there is no evidence the linear trend holds before curing has properly begun.
Question 25
- (a) point (7, 29) plotted correctlyB1
- (a) point (8, 34) plotted correctlyB1
- (a) Answer: (7, 29) and (8, 34)
- (b) positive correlationB1
- (b) Answer: Positive correlation
Question 26
- identifies positive correlationB1
- correct statement in context, e.g. as the number of weeks increases, the height of the plant tends to increaseB1
- Answer: There is positive correlation: as the number of weeks increases, the height of the plant tends to increase.
Question 27
- (a) (5, 20) caoB1
- (a) Answer: (5, 20)
- (b) any sensible reason, e.g. the student may have been unwell on the day of the test, or the score may have been recorded incorrectly (oe)B1
- (b) Answer: The student may have been unwell or unable to prepare properly on that occasion.
Question 28
- (a) all 10 points plotted correctly (allow small tolerance); award B1 for at least 7 points plotted correctlyB2
- (a) Answer: Points at (4,12), (6,18), (8,25), (9,30), (11,38), (13,45), (15,52), (17,60), (19,68), (22,78)
- (b) positive correlationB1
- (b) Answer: Positive correlation
Question 29
- (a) point (16, 10) plotted correctlyB1
- (a) point (19, 5) plotted correctlyB1
- (a) Answer: (16, 10) and (19, 5)
- (b) negative correlationB1
- (b) Answer: Negative correlation
Question 30
- identifies negative correlationB1
- correct statement in context, e.g. as the temperature increases, the number of woolly hats sold tends to decreaseB1
- Answer: There is negative correlation: as the temperature increases, the number of woolly hats sold tends to decrease.
Question 31
- (a) positive gradient identifiedB1
- (a) Answer: Positive gradient
- (b) use interpolation or a sensible straight line through the points to find y at x = 54M1
- (b) estimate 5.15 minutes cao (allow 5.0 to 5.3 awrt)A1
- (b) Answer: Approximately 5.15 minutes
- (c) extrapolate using line of best fit to give a sensible time at x = 70 (eg about 6.2 to 6.6)M1
- (c) comment: not very reliable because 70 is outside the data range (extrapolation) caoA1
- (c) Answer: Estimated time about 6.4 minutes. Not very reliable because 70 kg is outside the range of the given data (extrapolation).
Question 32
- (a) positive correlation identifiedB1
- (a) Answer: Positive correlation, moderately strong.
- (b) use interpolation between (4,45) and (5,50) or use line of best fit to find y at x = 4.5M1
- (b) calculate midpoint: 45 + 0.5*(50-45) = 47.5 or equivalentM1
- (b) 47.5 orders cao (allow 47 to 48 awrt)A1
- (b) Answer: Approximately 47.5 orders
- (c) identify it as an outlier or notM1
- (c) reason: it is below the trend (fewer orders than expected for that spend) or lies far from the line of best fit caoA1
- (c) Answer: Yes, it is an outlier, because it lies well below the trend/line of best fit for the other points.
Question 33
- (a) positive correlation caoB1
- (a) Answer: Positive correlation
- (b) method: gradient = (change in y)/(change in x) using two given pointsM1
- (b) 4/3 or 1.333... caoA1
- (b) Answer: 4/3 or 1.333...
- (c) 8.67 cao (answer to 2 dp)B1
- (c) Answer: 8.67
Question 34
- the relationship is not linear (it curves / changes shape across the range)B1
- a single straight line would not model the curved trend well, giving poor estimates for the youngest and oldest runnersB1
- Answer: The data does not follow a linear pattern (it curves at the youngest and oldest ages), so a straight line of best fit would not fit the trend well and would give poor estimates at the extremes of age.
Question 35
- (a) correct method, e.g. gradient = (90-30)/10 = 6, then 30 + 6 x 12M1
- (a) 102 (or over 100)A1
- (a) explains estimate is unreliable, e.g. 12 hours is outside the range of the data (extrapolation) and/or a test score cannot exceed 100%B1
- (a) Answer: 102%, which is impossible since a test score cannot exceed 100%; 12 hours is also outside the range of the data collected, so this is extrapolation and unreliable.
- (b) correlation does not necessarily mean causationB1
- (b) gives another possible factor that could affect scores, e.g. natural ability, quality of revision, or prior knowledgeB1
- (b) Answer: The scatter graph only shows correlation, not causation; other factors, such as natural ability or the quality of revision, could also affect test scores.
Question 36
- (a) (8, 105) caoB1
- (a) Answer: (8, 105)
- (b) any sensible reason, e.g. the reaction time may have been recorded incorrectly, or this player may have had an unusually fast practice attempt (oe)B1
- (b) Answer: The reaction time may have been recorded incorrectly, or this player had an unusually fast response on that attempt.
Question 37
- 40 sessions is outside the range of data collected (1 to 12), so this is extrapolationB1
- the same trend/rate of weight loss may not continue that far beyond the data, e.g. there could be a natural limit to how much weight can be lost, or attending 40 sessions a month may not be realistic (oe)B1
- Answer: The estimate is unreliable because 40 sessions is far outside the data collected (1 to 12 sessions), so it involves extrapolation and the same trend may not continue.
Question 38
- (a) 7 caoB1
- (a) Answer: 7 pages
- (b) 36 (minutes) caoB1
- (b) Answer: 36 minutes
- (c) (7, 36) ft from parts (a) and (b)B1
- (c) Answer: (7, 36)
Question 39
- (a) Graph AB1
- (a) Answer: Graph A
- (b) the points in Graph A lie closer to a straight line / are less scattered than the points in Graph B (oe)B1
- (b) Answer: The points in Graph A lie closer to a straight line than the scattered points in Graph B.
Question 40
- (a) FalseB1
- (a) Answer: False
- (b) TrueB1
- (b) Answer: True
- (c) TrueB1
- (c) Answer: True
- (d) FalseB1
- (d) Answer: False
Question 41
- (a) point (26, 44) plotted correctlyB1
- (a) point (28, 48) plotted correctlyB1
- (a) Answer: (26, 44) and (28, 48)
- (b) a single straight line, drawn by eye, that follows the trend of all 9 points with a roughly equal number of points on each sideB1
- (b) Answer: A straight line following the trend of the 9 points, e.g. close to the line y = 2x - 8.
- (c) 38 (bottles), accept 36 to 40, ft from candidate's lineB1
- (c) Answer: 38 bottles (accept 36 to 40)
Question 42
- (a) positive correlation identifiedB1
- (a) Answer: Positive correlation, fairly strong but leveling off at higher x.
- (b) use the plotted points to estimate at x = 2.5: interpolate between x = 2 (y = 55) and x = 3 (y = 60), or read from a sensible line or curve through the dataM1
- (b) calculation or reading at x = 2.5, e.g. mid-value 55 + 0.5*(60-55) = 57.5M1
- (b) state an estimate between 55 and 60, the scores either side of 2.5 hours (57.5 by interpolation; accept a decimal or a rounded value)M1
- (b) an estimate from 56 to 59 inclusive (57.5 by interpolation; a smooth curve through the points gives about 58 and a straight line of best fit about 56.5)A1
- (b) Answer: Approximately 57.5% (accept any sensible estimate from 56% to 59%)
- (c) state extrapolation to x = 20 is not reliableM1
- (c) reason: 20 is far outside the data range and trend may level off or change, so prediction not trustworthy caoA1
- (c) Answer: Not accurate; predicting at x = 20 is unreliable because it is far outside the observed data and the trend appears to level off.
Question 43
- (a) (3, 8) caoB1
- (a) Answer: (3, 8)
- (b) the outlier does not follow the trend/pattern shown by the rest of the dataB1
- (b) removing it means the line is not pulled away from the trend by this single unusual result, so it better represents the general relationship for the remaining pupils (oe)B1
- (b) Answer: The outlier does not fit the pattern of the rest of the data, so removing it stops the line of best fit being distorted, giving a line that better represents the true relationship for the other 7 pupils.
- (c) 20 hours is far outside the range of data collected (1 to 8 hours), so this would be extrapolation and the trend may not continue (oe)B1
- (c) Answer: 20 hours is well outside the data collected (1 to 8 hours), so predicting from the line would be extrapolation and may not be reliable.
Question 44
- correctly interprets the negative correlation in context, e.g. as the number of adverts shown increases, the percentage of viewers who watch to the end tends to decreaseB1
- states that correlation, however strong, does not prove causationB1
- identifies a plausible alternative explanation/confounding factor, e.g. longer videos may naturally have more adverts and viewers may stop watching because of the video's length rather than the adverts themselves (oe)B1
- Answer: As the number of adverts increases, the percentage of viewers watching to the end tends to decrease, but this does not prove adverts cause viewers to leave; another factor, such as longer videos having both more adverts and lower completion for unrelated reasons, could explain the pattern.
Question 45
- (a) identify positive correlationM1
- (a) explain in context: as x increases, y tends to increase oeA1
- (a) Answer: Positive correlation: as x increases, y tends to increase.
- (b) gradient = (6-2)/(5-1) = 1M1
- (b) y = x + 1 caoA1
- (b) Answer: y = x + 1
- (c) substitute x = 8 into y = x + 1 to get y = 9M1
- (c) state extrapolation, since x = 8 is outside the original data range (1 to 5), so the estimate may be less reliable caoA1
- (c) Answer: y = 9; this is extrapolation, since x = 8 lies outside the given data range of 1 to 5, so the estimate may not be reliable.
Question 46
- (a) the line of best fit predicts y = 2 + 1 = 3 when x = 2M1
- (a) the actual value (15) is far from the predicted value (3), much further than any other point, so (2, 15) is an outlier caoA1
- (a) Answer: The line predicts y = 3 at x = 2, but the actual value is 15, a much larger difference than for any other point, so (2, 15) is an outlier.
- (b) no, outliers are usually excluded (or investigated separately) since including them would distort the line of best fit away from the general trend of the other points oeB1
- (b) Answer: No; an outlier is usually excluded (or investigated separately), since including it would pull the line of best fit away from the general trend shown by the rest of the data.
- (c) gradient = (8-2)/(6-1) = 6/5 = 1.2M1
- (c) use point (1,2) to find intercept: 2 = 1.2*1 + cM1
- (c) y = 1.2x + 0.8 caoA1
- (c) Answer: y = 1.2x + 0.8
Question 47
- (a) gradient = (9-3)/(4-1) = 2M1
- (a) use point (1,3) to find intercept: 3 = 2*1 + cM1
- (a) y = 2x + 1 caoA1
- (a) Answer: y = 2x + 1
- (b) substitute x = 3.5 into y = 2x + 1M1
- (b) 8 (tens of ice lollies) caoA1
- (b) Answer: 8 (i.e. 80 ice lollies)
- (c) set 15 = 2x + 1 and rearrangeM1
- (c) x = 7 caoA1
- (c) Answer: 7 hours
- (d) not necessarily reliable; x = 7 is outside the original data range (1 to 4 hours), so this is extrapolation and the pattern may not continue oeB1
- (d) Answer: Not necessarily reliable, since x = 7 is outside the original data range (1 to 4 hours) - this is extrapolation, and the relationship may not continue in the same way beyond the observed data.
Question 48
- (a) cannot determine whether the estimate is reliable without knowing the strength of the correlation (oe)B1
- (a) 195 mm is within the observed range of 165 mm to 210 mm, so this is interpolation; it would be more reliable if the points lie close to the line of best fit (oe)B1
- (a) Answer: 195 mm is within the observed range of 165 mm to 210 mm, so this is interpolation. Its reliability cannot be determined without knowing how closely the points follow the line of best fit.
- (b) not reliable / noB1
- (b) 260 mm is well outside the range of data collected (165 mm to 210 mm), so this is extrapolation and the relationship may not continue in the same way beyond the observed data (oe)B1
- (b) Answer: Not reliable, because 260 mm is well beyond the range of wing lengths collected (165 mm to 210 mm), so this is extrapolation and the trend may not continue.
Question 49
- (a) as x increases, y also tends to increaseM1
- (a) the points lie close to a straight line (little scatter), so the association is strong oeA1
- (a) Answer: As x increases, y tends to increase, and the points lie close to a straight line, showing a strong association.
- (b) state that correlation does not imply causationM1
- (b) identify that another factor could be responsible for both variables (a confounding variable)A1
- (b) give a plausible reason: hot weather leads to more ice cream sales and, separately, more people swimming (so more risk of drowning), rather than ice cream sales directly causing drownings oeA1
- (b) Answer: Correlation does not imply causation. A third factor, such as hot weather, could cause both more ice cream sales and more people swimming (and so more drownings), without ice cream sales directly causing drownings.
- (c) recognise that x = 3 lies outside the original data range (5 to 20), so this would be extrapolationM1
- (c) state that extrapolated estimates are less reliable, since the linear relationship may not hold outside the range of the observed data oeA1
- (c) Answer: Not reliable: x = 3 is outside the data range (5 to 20), so using the line here is extrapolation, and the relationship may not hold outside the observed range.
- (d) temperature can plausibly cause both more ice cream sales and more swimming directly, so it is a more sensible candidate for a genuine underlying cause than ice cream sales themselves oeB1
- (d) Answer: Temperature is more plausible as a genuine cause, since hot weather can directly explain both more ice cream sales and more people going swimming, whereas ice cream sales have no direct mechanism for causing drownings.
Question 50
- (a) identifies that only sunny weekend days in May to August were sampled, excluding weekdays, poor-weather days and other seasonsB1
- (a) explains a consequence, e.g. the relationship seen may not represent the true pattern across all days/seasons, since weekday attendance or attendance in other weather/seasons could behave differently, so conclusions may not generalise (oe)B1
- (a) Answer: Only sunny weekend days in May to August were sampled, so weekdays, poor weather and other seasons are missing; the relationship found may not hold for the park's attendance the rest of the year, so conclusions drawn from it may not be reliable more generally.
- (b) any sensible improvement, e.g. collect data across weekdays and weekends, throughout the year, and in a range of weather conditions (oe)B1
- (b) Answer: Collect data on a mix of weekdays and weekends, across all seasons and a range of weather conditions, not just sunny summer weekends.
Question 51
- (a) (58, 52) caoB1
- (a) Answer: (58, 52)
- (b) recognises this athlete is an outlier/anomalous point that does not follow the trend shown by the rest of the groupB1
- (b) explains the implication, e.g. a lower resting heart rate is generally linked with a higher VO2 max in this data, but this athlete has a low heart rate and a low fitness score, so some other factor must be affecting their result (oe)B1
- (b) Answer: This athlete does not follow the general trend: in the rest of the group, a lower resting heart rate goes with a higher VO2 max, but this athlete has a low heart rate and a low VO2 max score, so some other factor must be affecting their fitness score.
- (c) states that correlation, even a strong one, does not prove causationB1
- (c) identifies a plausible confounding factor, e.g. overall training/fitness level could improve both resting heart rate and VO2 max at the same time, rather than one directly causing the other (oe)B1
- (c) states that establishing causation would require further evidence, such as a controlled study, not just an observed correlation from a scatter graph (oe)B1
- (c) Answer: The claim is not proven: correlation does not establish causation. Both resting heart rate and VO2 max could be improved together by a third factor, such as overall training level, rather than a lower heart rate directly causing higher fitness. A controlled study, not just this scatter graph, would be needed to establish a genuine causal link.
Mark your answers
Check answers without an account. Some answers can be checked automatically; use the mark scheme to mark the rest.