Chapter PII, Problem 28RE

To determine

To discuss whether such a conclusion can be drawn from this regression and if so discuss the supporting evidence and if not then explain why.

Answer to Problem 28RE

No, it cannot be drawn.

Explanation of Solution

It is given in the question that the September $1998$ issue of the American psychologist published an article by Kraut et al. that reported on an experiment examining, the social and psychological impact on the Internet on $169$ people in $73$ household during the one to two years online. Thus, the experiment was conducted and the conclusion of the study made news headlines that those who spent more time online tended to be more depressed at the end of the experiment. The summarized regression of depression is as:

  $\begin{array}{l}{R}^2=4.6\%\\ s=0.4563\\ \text{Intercept}=0.5655\\ \text{Internet use}=0.0199\end{array}$

The news reports about this study clearly concluded that using internet causes depression. Thus, this conclusion drawn from this regression is false because depression is not a quantitative unit or variable so it cannot be measured and cannot be defined by this study.