Digital SAT Math · Problem-Solving and Data Analysis

SAT Two-Variable Data: Models and Scatterplots Practice Questions

These questions cover scatterplots, lines of best fit, linear and exponential models, and interpreting what a model's slope or growth rate means.

Questions in the bank
209
Easy / Medium / Hard
67 / 93 / 49
Subtopics
20
Typical time
~1.2 min each

Updated . Questions are original BlitzSAT items written to the College Board's published skill specifications. They are not official SAT questions.

How Two-Variable Data: Models and Scatterplots questions appear on the SAT

Scatterplots with a line or curve of best fit, tables of paired data, and questions such as 'Which equation best models the data?' or 'What does 0.8 represent in the model?'

Two-Variable Data: Models and Scatterplots is part of the Problem-Solving and Data Analysis domain, which makes up about 15% of the Math section (5–7 questions per test). A calculator is allowed on every Math question, and the built-in Desmos graphing calculator is available in Bluebook.

How to approach Two-Variable Data: Models and Scatterplots questions

  1. 1Read the axis labels and scales before you do anything else.
  2. 2Estimate the slope of a fitted line from two points that lie on the line, not from individual data points.
  3. 3A constant difference between values points to a linear model. A constant ratio points to an exponential model.
  4. 4To predict, use the model's value. To compare with reality, use the actual data point. Residual = actual − predicted.

Common mistakes

  • Reading a value from the wrong axis scale.
  • Interpreting correlation as causation.
  • Computing the residual as predicted − actual.

10 SAT Two-Variable Data: Models and Scatterplots practice questions with answers

Pick an answer to check it. Each explanation covers the correct answer and why each of the other choices is wrong. The set runs from easy to hard.

0 / 10 answered
1EasyCalculate a residual from a data table
In a hypothetical materials experiment, a fitted model predicts indentation depth using d^=0.4F+1\hat d=0.4F+1, where FF is the applied force in newtons and d^\hat d is the predicted depth in millimeters. The table shows observed depths.
Applied force (newtons)Observed indentation depth (millimeters)
52.7
105.8
156.6
A residual is the observed depth minus the predicted depth. What is the residual, in millimeters, for the observation at 10 newtons?
Show answer and explanation

Answer: D

At 10 newtons, the model predicts 0.4(10)+1=50.4(10)+1=5 millimeters. The observed depth is 5.8 millimeters, so the residual is 5.8−5=0.85.8-5=0.8 millimeters.Why the other choices are wrongAThis subtracts the observed depth from the predicted depth, reversing the residual definition.BThis subtracts only the intercept, 1, from the observed depth instead of subtracting the full prediction of 5.CThis is the predicted indentation depth, not the difference between the observed and predicted depths.
2EasyPredict a response using a stated model
For a hypothetical school project, the model s=6n+14s=6n+14 predicts the total number of sheets of paper used after nn work sessions. How many sheets does the model predict will have been used after 5 work sessions?
Show answer and explanation

Answer: A

Substitute n=5n=5 into the model: s=6(5)+14=30+14=44s=6(5)+14=30+14=44. The model predicts 44 sheets.Why the other choices are wrongBThis calculates 6(5)6(5) but omits the additional 14 sheets in the model.CThis adds 5+145+14 without multiplying the number of sessions by 6.DThis reverses the roles of 6 and 14, calculating 14(5)+614(5)+6 instead of 6(5)+146(5)+14.
3EasyChoose a reasonable line for a small coordinate table
Which equation provides the most reasonable linear model for the data in the table?
xxyy
07.1
18.8
211.1
Show answer and explanation

Answer: A

The value of yy is approximately 7 when x=0x=0, and yy increases by approximately 2 for each increase of 1 in xx. The equation y=2x+7y=2x+7 predicts 7, 9, and 11, which are close to all three listed values.Why the other choices are wrongBThis equation predicts that yy decreases as xx increases, but the table shows an increasing relationship.CThis equation has a reasonable slope but predicts y=0y=0 when x=0x=0, rather than a value near 7.1.DThis equation interchanges the approximate slope and intercept. It predicts 2 and 16 at the endpoints, far from 7.1 and 11.1.
4MediumCompare predictions and observations with units
In a hypothetical environmental investigation, water depth in a collection basin is modeled by h=0.04r+0.18h=0.04r+0.18, where rr is rainfall in millimeters and hh is predicted water depth in meters. The table shows observed depths.
Rainfall (millimeters)Observed water depth (meters)
50.39
120.61
201.01
For the observation with 12 millimeters of rainfall, how does the observed water depth compare with the model's prediction? There are 100 centimeters in 1 meter.
Show answer and explanation

Answer: A

At r=12r=12, the model predicts h=0.04(12)+0.18=0.66h=0.04(12)+0.18=0.66 meter. The observed depth is 0.66−0.61=0.050.66-0.61=0.05 meter less. Converting the difference gives 0.05(100)=50.05(100)=5 centimeters less than predicted.Why the other choices are wrongBThe difference is 5 centimeters, but the observed depth of 0.61 meter is less than the predicted depth of 0.66 meter.CThe difference of 0.05 is measured in meters. Converting to centimeters requires multiplying by 100.DThis results from omitting the constant term: 0.61−0.04(12)=0.130.61-0.04(12)=0.13 meter. The complete model includes the additional 0.18 meter.
5MediumIdentify association from coordinate pairs
The table gives five coordinate pairs.
xxyy
111
210
38
45
56
Which statement best describes the association between xx and yy?
Show answer and explanation

Answer: B

The larger xx-values generally correspond to smaller yy-values, indicating a negative association. However, the changes in yy are −1-1, −2-2, −3-3, and 1 for successive increases of 1 in xx. Since these changes are not constant, the points do not lie exactly on one line. The final increase does not erase the overall downward association.Why the other choices are wrongAThis focuses on the final increase from 5 to 6 while ignoring the overall decline from the larger yy-values at smaller inputs.CAn association describes an overall tendency; it does not require every successive pair to follow the same direction.DAn exact linear relationship requires constant changes in yy for equal changes in xx. The table's changes are not constant.
6MediumDistinguish interpolation from extrapolation
A model y^=2x−1\hat y=2x-1 was fitted to the coordinate data in the table.
xxyy
−6-6−14-14
−2-2−4-4
3344
771818
The model predicts y^=−1\hat y=-1 when x=0x=0 and y^=15\hat y=15 when x=8x=8. Which statement correctly classifies these two uses of the model?
Show answer and explanation

Answer: D

Interpolation uses an input within the range of observed inputs, while extrapolation uses an input outside that range. The observed xx-values range from −6-6 to 7. Since 0 is within this range, the first prediction is an interpolation. Since 8 exceeds 7, the second is an extrapolation. The predicted yy-values do not determine these classifications.Why the other choices are wrongAAlthough both predicted outputs lie within the observed output range, classification depends on the input range. The input 8 is outside that range.BAn input need not appear as a row in the table to support interpolation. The input 0 lies between the observed inputs −6-6 and 7.CThis reverses the classifications: 0 is inside the observed input range, whereas 8 is outside it.
7MediumInterpret a fitted line slope
In a hypothetical transport planning analysis, the fitted model T=18+0.6dT=18+0.6d predicts a shuttle's travel time TT, in minutes, for a route of length dd kilometers. Which statement correctly interprets the slope of this model?
Show answer and explanation

Answer: A

The slope is 0.6 minute per kilometer, so each additional kilometer increases predicted travel time by 0.6 minute. Converting to seconds gives 0.6(60)=360.6(60)=36 seconds.Why the other choices are wrongBThe value 100 results from taking the reciprocal of 0.6 and multiplying by 60. The slope already gives minutes per kilometer and should not be inverted.CThe coefficient 0.6 is measured in minutes per kilometer, not seconds per kilometer.DThe value 18 is the model's intercept, not its slope. The predicted change per additional kilometer is 0.6 minute.
8HardInterpret a fitted line intercept
In a hypothetical library trial, teams sort identical batches of returned books. Let nn be the number of staff members on a team and TT be the batch-sorting time, in seconds. All teams observed in the trial have between 6 and 10 staff members, inclusive.An analyst defines x=n−4x=n-4 and y=T−120y=T-120 and fits the line y^=180−15x\widehat{y}=180-15x, where y^\widehat{y} is the predicted value of yy. Which statement correctly interprets the line's yy-intercept in terms of the original quantities and the observed range?
Show answer and explanation

Answer: D

The intercept gives y^=180\widehat{y}=180 when x=0x=0. Since x=n−4x=n-4, this means n=4n=4. Since y=T−120y=T-120, the predicted original time is 180+120=300180+120=300 seconds, or 5 minutes. A team of 4 is outside the observed range of 6 to 10 staff members, so this prediction is an extrapolation, not an observed result.Why the other choices are wrongAThe team size and predicted time are correct, but 4 staff members is outside the observed range. The prediction is an extrapolation.BThe intercept of 180 is a predicted value of T−120T-120, not of TT. Restoring the 120-second offset gives 300 seconds, or 5 minutes.CThe intercept occurs at x=0x=0, but x=n−4x=n-4. Thus, it corresponds to 4 staff members, not 0 staff members.
9HardIdentify limitations of a stated fitted model
In a hypothetical recreation survey, volunteers attending weekday morning sessions report their travel time xx, in minutes, and their number of center visits yy per month. The fitted model y^=12−0.2x\hat{y}=12-0.2x has small residuals for these volunteers, whose travel times range from 55 to 2525 minutes.Membership records show that every center member's travel time is also between 55 and 2525 minutes and that the mean for all members is 1818 minutes. An analyst substitutes 1818 into the model and estimates that all members average 8.48.4 visits per month.Which statement identifies the primary limitation of this estimate?
Show answer and explanation

Answer: A

Small residuals show that the line fits the surveyed volunteers well, not that it fits all members well. Weekday morning volunteers may have different attendance patterns from other members. The calculation 12−0.2(18)=8.412-0.2(18)=8.4 is valid, and all travel times are within the observed range, but applying the relationship to the full membership is not justified by those facts alone.Why the other choices are wrongBFor a linear model, averaging individual predictions gives the same result as evaluating the model at the mean input: the average prediction is 12−0.2xˉ12-0.2\bar{x}.CExtrapolation involves inputs outside the observed range, not inputs different from the mean. Every member's travel time is within the surveyed range of 55 to 2525 minutes.DA predictive association need not be causal. The problem is whether the surveyed group's association applies to all members, not whether travel time causes attendance changes.
10HardCompare linear and exponential models from a table
In a hypothetical library project, staff must review 1,0241{,}024 catalog records. The table shows the cumulative number reviewed after tt weeks.
Time, tt (weeks)Records reviewed
00
1256
2448
4700
An analyst considers either a linear model for the cumulative number reviewed or an exponential model for the number remaining. Which choice identifies a model consistent with every table value and gives that model's predicted number of additional records reviewed from the end of week 44 to the end of week 66? Predicted counts need not be whole numbers.
Show answer and explanation

Answer: C

Subtracting the reviewed counts from 1,0241{,}024 gives remaining counts of 1,0241{,}024, 768768, 576576, and 324324. These fit R(t)=1,024(3/4)tR(t)=1{,}024(3/4)^t: the remaining count is multiplied by 3/43/4 each week, and 576(3/4)2=324576(3/4)^2=324 accounts for the two-week gap. At week 66, the model predicts 324(3/4)2=182.25324(3/4)^2=182.25 records remaining. Thus, 324−182.25=141.75324-182.25=141.75 additional records are predicted to be reviewed during those two weeks. The changing weekly increases in reviewed counts rule out a linear model fitting every row.Why the other choices are wrongAThis extends the overall average rate from week 00 to week 44, which is 700/4=175700/4=175 records per week. That linear rate does not fit the intermediate counts of 256256 and 448448.BThe exponential model is appropriate, but 182.25182.25 is the predicted number remaining at week 66, not the number reviewed between weeks 44 and 66.DThis repeats the increase of 700−448=252700-448=252 records observed from week 22 to week 44. A constant two-week increase does not fit the whole table; the exponential model predicts smaller increases as fewer records remain.

These are 10 of 209 questions on this skill in the BlitzSAT question bank.

Practice the rest by difficulty, timed or untimed, with your history saved. Free, no credit card.

Practice all 209 questions

What the 209 Two-Variable Data: Models and Scatterplots questions cover

The BlitzSAT bank splits this skill into 20 subtopics. Counts are from the current bank. 59 questions (28%) are student-produced responses, where you type in your answer. 10 include a graph, table, or figure.

SubtopicQuestionsEasyMediumHard
Calculate a residual from a data table20695
Choose a reasonable line for a small coordinate table20695
Compare predictions and observations with units20695
Distinguish interpolation from extrapolation20695
Identify association from coordinate pairs20695
Identify limitations of a stated fitted model20695
Interpret a fitted line intercept20695
Interpret a fitted line slope20695
Predict a response using a stated model20695
Compare linear and exponential models from a table19694
Compare candidate linear models to plotted data1010
Compare distances from a model rather than response heights1100
Distinguish within-group and pooled association1010
Find an input for a specified fitted response1010
Identify extrapolation from plotted observation limits1100
Interpolate a model value within an observed range1100
Interpret a graph slope with units1100
Interpret the intercept of a fitted line without extending a measured range1100
Read a signed residual from a scatterplot and fitted line1100
Recognize curvature in a scatterplot1100

Other Problem-Solving and Data Analysis skills

More Math skills

Keep going