Question on expectations for S.ID.6 and S.ID.8

[dward]

When students fit exponential and quadratic functions to data, is the expectation that they will use a calculator to first examine a scatter plot and then use the calculator to determine a regression equation that best models the data? If so, should the analysis of the residuals plots and correlation coefficient come into play for such situations? Standard S.ID.8 indicates that students interpret the correlation coefficient of linear fit. Does this include the correlation coefficient that the TI gives with the exponential regression?

[Bill McCallum]

I think students will be using some sort of technology (not limited to calculators) for all the operations with data: scatterplots, curve-fitting, plotting residuals. They should be looking at realistic data sets, which are too large for anything else. It makes sense to me to analyze residuals while looking at a different curves of fit, in order to see how the distribution varies. Then you eventually just start looking at the curve of best fit and the correlation coefficient.

[Author]

Posts