The table represents points in a scatterplot.
A researcher believes that, except for one measurement error, the data follow an exponential model of the form , where , , and are positive constants. Which choice gives the -value that is most likely due to measurement error?
When an exponential model includes a constant added outside the exponent, do not expect the original -values to have a constant ratio. Instead, examine consecutive differences. For a model , the differences follow an exponential pattern with common multiplier . Use consistent points to build the model, then test each point for an outlier.
Hints
Look at changes in the outputs
Find the differences between consecutive -values. A vertical shift, represented by , does not prevent an exponential pattern from appearing in the differences.
Use the exponential structure
For , each first difference is multiplied by the same factor to get the next first difference.
Verify the suspicious point
Use points that appear consistent to determine the model, then compare the model's predicted value at each with the table.
Desmos Guide
Prefer the difference method first
Algebra using consecutive differences is faster here because the model includes a vertical shift. Use Desmos to verify the resulting model visually.
Enter the data
Enter the -values and -values into a Desmos table so the six data points appear on the graph.
Graph the model
Graph . Five of the six points lie on this curve; identify the plotted point that does not lie on the curve.
Step-by-step Explanation
Compare consecutive differences
The first differences between the listed -values are , , , , and . For , consecutive first differences should have a constant multiplier of .
Use the first three points to find the model
The first two differences are and , so the multiplier is . Thus .
Using the first two points gives and . Subtracting shows that , so .
Check the point at
The model is . Therefore,
The table instead gives when . The values at and match the model, so is the measurement error.