Question 56·Medium·Two-Variable Data: Models and Scatterplots
A technician records the screen size (in inches) and the mass (in grams) of six smartphone models. The data are shown in the table and appear to follow a roughly linear trend.
| (in.) | 4.7 | 5.0 | 5.5 | 6.1 | 6.5 | 6.8 |
|---|---|---|---|---|---|---|
| (g) | 135 | 145 | 155 | 170 | 182 | 190 |
Which equation best models the relationship between and for the data?
For questions asking which line best models a roughly linear data set, first estimate the slope quickly by taking the first and last data points and computing . Compare this estimate to the slopes in the answer choices to eliminate options with slopes that are clearly too small or too large. Then, using your approximate slope and one data point in , estimate the intercept and match it to the remaining choices; if needed, plug a mid-range -value into the candidate equation to see which one predicts a close to the actual value.
Hints
Think about slope first
Look at how changes when increases from the smallest to the largest value. About how many grams does the mass increase when the screen size increases by a little over 2 inches?
Compare your slope to the choices
Use the first and last points to compute an approximate slope . Then, look at the slopes in the four equations (the numbers multiplying ). Which are obviously too small or too large?
Now find the intercept
Once you know a good estimate for the slope, plug that slope and any one pair from the table into to solve for . Then match this approximate intercept to the choices.
Desmos Guide
Enter the data as a table
In Desmos, create a table and enter the -values (4.7, 5.0, 5.5, 6.1, 6.5, 6.8) in the first column and the corresponding -values (135, 145, 155, 170, 182, 190) in the second column. You should see six plotted points.
Graph each candidate line
On new lines in Desmos, type each option exactly: y = 15x + 26, y = 26x + 15, y = 40x - 60, and y = 60x - 40. All four lines will appear on the same coordinate plane along with the data points.
Visually compare fit
Look at how close each line is to the six data points overall. Focus on which line stays nearest to most of the points (not just one point). The equation whose line best follows the cluster of points is the model you should choose.
Step-by-step Explanation
Recognize you need a linear model
The question says the data follow a roughly linear trend, so we want an equation of the form , where:
- is the slope (change in mass for each 1-inch change in screen size), and
- is the -intercept.
We will estimate from the table and then find .
Estimate the slope from the data
Use the first and last data points to estimate the slope:
- First point:
- Last point:
Compute the changes:
- Change in :
- Change in :
So the slope is approximately
Look at the answer choices: their slopes are 15, 26, 40, and 60. Only one choice has a slope close to 26, so all lines with slopes far from 26 can be eliminated.
Estimate the y-intercept using the slope
Now use the approximate slope with any data point to estimate .
Using the point in :
Compute :
So
The intercept should be around 13. Compare this to the intercepts in the answer choices: 26, 15, , and . The only one near 13 is the line with intercept 15.
Check that this line fits the data well
Take the equation with slope 26 and intercept 15 and test it on another data point, say (which has actual ):
Compute:
The predicted mass g is very close to the actual g, and similar checks work for other points. Therefore, the equation that best models the relationship is .