Scatterplot worksheet with answer key
8 problems on scatterplot, ready to print for class or homework. The answer key prints on its own page, so hand out everything before it.
Name: ______________________________ Date: ______________ Period: ______
Problem 1
A school data club is deciding what to plot. For each situation, say which variable is explanatory and which is the response, and which axis each belongs on. (a) A ferry operator records, for each of 30 crossings, the wind speed in kilometers per hour and the crossing time in minutes. (b) A dealer records, for each of 18 second-hand vans, the odometer reading in thousands of kilometers and the price the van sold for. (c) A coach records each runner's squad (junior or senior) and their 5 km finishing time. (d) A physiotherapist records each patient's forearm circumference in centimeters and their grip strength in kilograms, with no plan to predict either one from the other.
Problem 2
A repair shop tests 9 phones of one model and records the phone's age in months, , and the battery's measured capacity as a percent of its capacity when new, : , , , , , , , , . (a) Which variable is explanatory? (b) Sketch the points and describe the association completely. (c) A student answers only "strong and negative". What does that answer miss, and why does it matter?
Problem 3
A county's winter contractor plots 12 storms from the past three seasons. Here is the snowfall in centimeters, ranging from 8 cm to 44 cm, and is the plow time in hours. The scatterplot is linear, positive, and strong, and the line drawn through the points passes exactly through and . (a) Find the equation of that line. (b) Predict the plow time after a 25 cm storm. (c) One 25 cm storm took 31 hours. How far from the line does it sit, and what does that position mean? (d) Interpret the slope in context. (e) The county asks for the predicted plow time after a 90 cm storm. Give the number the line produces and explain why you would not report it as a prediction.
Problem 4
A bike-share operator plots its 10 docking stations, with the number of docks and the average number of rides started per day. Eight stations lie close to the line : , , , , , , , . The other two are station P at and station Q at . (a) Show that no station in the group of eight sits more than 2 rides from that line. (b) Which of P and Q is an outlier in the direction? (c) Which of the two has high leverage? Compute (read "x-bar"), the mean number of docks over all 10 stations, to support your answer. (d) Which of P and Q is more likely to change the slope of a fitted line, and why? (e) A new station R opens at . Classify it and say what it would do to the line.
Problem 5
A bakery chain plots its 12 shops, with the floor area in square meters and the weekly transactions in hundreds. Six shops are on high streets: , , , , , . The other six are in retail parks: , , , , , . The plot the manager sees does not label which shop is which. (a) Describe the scatterplot completely. (b) Compute the mean point of each group of six and use the two mean points to state the overall drift of the plot. (c) Within each group, what is the direction of the association? (d) The manager says larger shops sell less, so the chain should stop opening large shops. Explain what is wrong with that reading. (e) What would you change about the plot?
Problem 6
A tutoring center plots 7 students, with the number of sessions attended and the points gained between two practice tests: , , , , , , . Every point lies within 1 point of the line . (a) Describe the plot completely. (b) An eighth student attended 13 sessions and gained 4 points, having been ill on test day. How far from the line does that point sit, and where does it sit in the direction? (c) What does adding it do to the direction, to the apparent strength, and to a line fitted to all 8 points? (d) Suppose instead that same student had attended 7 sessions and gained 4 points. How would the effect on the fitted line differ, and why?
Problem 7
A cold-brew producer steeps the same coffee for different lengths of time and measures the caffeine in the finished brew. With the steep time in hours and the caffeine in milligrams per 100 mL, the readings are , , , , , , , . (a) Find the change in caffeine over each 2-hour step and describe what those changes show. (b) Describe the scatterplot completely. (c) The producer's analyst draws the straight line through the first and last points. What does that line predict at 6 hours and at 8 hours, and how far off is it? (d) If a least-squares line were fitted anyway, what pattern would the residuals show? (e) The analyst uses the straight line to predict the caffeine after 24 hours. Compute that prediction and explain why it cannot be right.
Problem 8
A city council plots its 8 policing neighborhoods, with the number of bus stops and the number of reported bicycle thefts in a year: , , , , , , , . Every point lies within 1 theft of the line . (a) Name the explanatory variable and describe the association completely. (b) Use the line to predict the thefts in a neighborhood with 40 bus stops, and say whether that prediction is safe. (c) A councillor proposes removing bus stops in order to cut bicycle theft. Explain why this plot cannot support that. (d) The council also records residents: the 25-stop neighborhood has 9,000 and the 50-stop neighborhood has 19,000. Compute thefts per 1,000 residents for each and say what that does to the councillor's argument. (e) What would it take to establish that bus stops cause thefts?
From statslearn.org/worksheets/scatterplots-worksheet. Free to print and copy for classroom use.
Scatterplot: answer key
(a) Wind speed on the x-axis, crossing time on the y-axis. (b) Odometer reading on the x-axis, selling price on the y-axis. (c) Neither: squad is categorical, so no scatterplot is possible; compare the two time distributions instead. (d) Both are quantitative so a plot works, but there are no natural roles; pick the axis from the question, for example forearm circumference on the x-axis if you want to predict grip strength.
(a) Age in months. (b) A strong, negative, linear association between a phone's age and its battery capacity: every point lies within 1 percentage point of the line , and there are no outliers, clusters, or gaps. (c) It names direction and strength but omits the form (linear) and the unusual features (none), and it never names the variables in context.
(a) . (b) 26 hours. (c) hours above the line, so the line underpredicted this storm by 5 hours. (d) Each extra centimeter of snow predicts about 0.8 hour, or 48 minutes, more plowing on average. (e) The line gives 78 hours, but 90 cm sits far outside the observed 8 cm to 44 cm range, so it is an extrapolation and should not be reported as a prediction.
(a) The largest gap is 2 rides, at the 10-dock station. (b) P: it sits rides above the line, while Q sits exactly on it. (c) Q: docks, and Q's 40 docks is 20.7 above that, further out than any other station. (d) Neither moves the slope much. P has a big miss but almost no leverage at 17 docks, and Q has huge leverage but no miss, so it anchors the existing slope. (e) R has high leverage and sits 118 rides below the line, so it is influential: it would drag the right end down and flatten the slope.
(a) Two clusters (30 to 60 and 140 to 200 square meters) separated by an 80 square meter gap; the overall drift is negative but only moderate, with about , and the form is two offset linear bands rather than one, so what strength there is comes mainly from the gap between the groups. (b) Mean points and : area rises 123.33 square meters while mean transactions fall 7.17 hundred, so the drift is negative. (c) Positive inside both clusters. (d) The manager is reading a difference between two location types as an effect of floor area; within each type, larger shops do more transactions. (e) Record and plot location type, with different symbols or two separate plots, and describe each group's association separately.
(a) A strong, positive, linear association between sessions attended and points gained, with every point within 1 of and no outliers, clusters, or gaps. (b) It sits points below the line, at 13 sessions, which is 5 above and near the right edge of the data. (c) Still positive but much weaker, and the fitted line's right end is dragged down, so the slope flattens and the intercept rises: the point is influential. (d) At 7 sessions the miss is points, but the point sits at the center of the x values, so it pushes the whole line down without tilting it and the slope barely changes.
(a) Steps of mg per 100 mL: still rising, but the rate falls from 14 mg per 100 mL per hour to 1 mg per 100 mL per hour. (b) A strong, positive, non-linear (concave down) association between steep time and caffeine, with no clusters, gaps, or points off the curve. (c) The chord has slope mg per 100 mL per hour and predicts 63.71 mg at 6 hours (24.29 mg low) and 75.57 mg at 8 hours (26.43 mg low). (d) An arch: residuals negative at both ends and positive through the middle, the signature of a wrong form. (e) It gives 170.43 mg, which is impossible here: the curve has flattened near 123 mg and 24 hours is beyond the data.
(a) Bus stops is explanatory. There is a strong, positive, linear association between the number of bus stops and reported bicycle thefts, with every point within 1 theft of and no outliers, clusters, or gaps. (b) thefts, and 40 is inside the observed 12 to 58 range, so the prediction is an interpolation and safe to report. (c) The data are observational and the plot shows association only, so it cannot say what removing stops would do. (d) and thefts per 1,000 residents, so the neighborhood with more stops has the lower rate; population is a lurking variable inflating both counts. (e) A randomized experiment assigning stop removals, or failing that, comparing rates across neighborhoods matched on population, density, and bicycle ownership.
Every step is worked at statslearn.org/practice/scatterplots-practice. Free to print and copy for classroom use.
The key gives final answers only. The full worked solution for each problem is on the matching practice set, and every other sheet is on the worksheets page.