Using GSS2006 data, here is a non-linear regression equation, using hours of television watched as the dependent variable (EDUC stands for years of education, all slopes are statistically significant): TVHOURS = 3.36 + .12 (EDUC) -.011 (EDUC)² Which of the following is the most accurate description of what is going on? A) People with low levels of education watch a lot of television, then people with moderate levels of education watch less television, and then people with high levels of education watch a lot of television. B) People with low levels of education watch a little of television, then people with moderate levels of education watch more television, and then people with high levels of education watch a little of television. C) People with low levels of education watch the most television, and for each additional year of education, hours of television watched goes down by the same amount. D) People with low levels of education watch the most television, and as education increases, hours of television watched starts going down, slowly at first, and then more and more rapidly.

Question

Quizplus · Accepted Answer

The non-linear regression equation shows that as education increases, the hours of television watched decreases. However, the decrease in hours watched starts off slow and then accelerates as education levels increase. This is indicated by the negative coefficient of the non-linear term (EDUC)², which is squared and implies a curved relationship between education and TV watching. Therefore, people with low levels of education watch the most television, and as education levels increase, TV watching decreases at an increasing rate.

Using GSS2006 Data, Here Is a Non-Linear Regression Equation, Using