The only two models that give a fairly close approximation are A and C. Here is a table comparing their values with those for the given function E.

Year t6
(1976)
8 101214161820222426
(1996)
Expenditure E
(Billions)
$2.53.03.03.13.03.13.23.33.43.33.1
Model A2.482.542.62.662.722.782.842.92.963.023.08
Model C2.5282.6722.82.9123.0083.0883.1523.23.2323.2483.248

From the table, we can see that, for the most part, Model C gives a better approximation.

(One could actually make this claim precise by computing the "sum of squares error" for each model: square and add the errors (quantity predicted by the model minus actual quantity). Model C gives a much smaller sum-of-squares-error,)


Just close this window to return to the tutorial.
Lost the tutorial? Press here.