We fear that some readers will hear in our analysis and recommendations a call
to abandon the teaching of noninferential exploratory methods of data analysis and
to eschew data from other than well-defined samples. In fact, we believe that we
should begin teaching informal methods of data analysis in the spirit of EDA to
students at a young age. Moreover, we are not recommending that the teaching of
data analysis should be grounded in, or necessarily headed toward, the technical
question of drawing formal inferences from carefully constructed samples.