A psychologist once suggested this possible relationship between the IQs of children and their parents [27]:

What statistical phenomenon would explain this psychologist’s use of the coefficient 0.5, rather than 1.0? A few years later another researcher used the data in Table 9.17,

relating the mean IQ scores of children and their fathers, grouped according to the father’s occupation, to estimate the regression equation Y = α + βX + ε. Draw a scatter diagram and estimate this equation by least squares.

a. Which variable did you use as the dependent variable? Explain your reasoning.

b. Is your estimated equation consistent with the theory that (child IQ − 100) = 0.50(adult IQ − 100)?

c. What is suspicious about your estimated equation?

Transcribed Image Text:

(child IQ-100) = 0.50(adult IQ-100)