Business Applications of Multiple Regression

CORRELATION ANALYSIS 47

• They were alcoholics or drug users who were unemployable.

• They were unemployable for other reasons.

• They were under 18 and too young to work.

It is likely that the poor performance from children from these

groups was bad enough that it drove down the likely higher perfor-

mance by children who had loving, concerned mothers who stayed

home for their children.

Even if none of these factors were present and the data were com-

pletely valid, there is another equally likely explanation for the data:

Families are more likely to make the sacrifice for the mother to stay

home when the child is having problems. ...