Question

35-43. A room of researchers are debating what variables predict attitudes towards scientists.

Researcher 1 says, "In a comparison of individuals, older people will rate scientists higher than younger people."

Researcher 2 says, "In a comparison of individuals, those with higher education will rate scientists higher than those with less education."

The following is a real regression model created using the 2016 ANES. The dependent variable, feelings towards scientists, ranges from 0-100. Party ID is coded 0=Democrat, 1=Independent, 2=Republican. Female is a dummy variable, coded 0=male and 1=female. Education is coded 0=high school or less, 1=some college, 2=college degree, and 3=graduate degree. Age is coded 0=18-29, 1=30-49, 2=50-64, 3=65+.

Coefficientsa
Model		Unstandardized Coefficients		Standardized Coefficients	t	Sig.
		B	Std. Error	Beta
1	(Constant)	77.318	.851		90.866	.000
	age 4	.102	.318	.005	.320	.749
	educ 4	2.818	.312	.149	9.025	.000
	Party ID 3 cat	-4.146	.345	-.200	-12.011	.000
	female	-.839	.643	-.022	-1.304	.192
a. Dependent Variable: FT Scientists

35. What does the coefficient for age mean? (Explain in one sentence)

36-37. Is Researcher 1's hypothesis right? How do you know? (note: to get the point, you must explain your answer.)

38. What does the coefficient for education mean? (Explain in one sentence)

39-40. Is Researcher 2's hypothesis right? How do you know?

41. What would be the predicted rating of scientists for an 25 year old, high school educated, Democrat male? Show your work.

42. What would be the predicted rating of scientists for an 25 year old, college educated, Republican male? Show your work.

43. R-squared for this model is .063. What does this mean? Explain in one sentence.

Answer 1

35)For every one block increase in age, the feelings towards scientists value is expected to increase by 0.102

36-37)Researcher 1 is not right, We can see that the age coefficient is positive which means that as age increases, the rating increases. But, this isnt sufficiently justified by the test. the pvalue=0.749 meaning, the researcher cannot conclude this significantly.

38)For every one unit increase in education, the feelings towards scientists value is expected to increase by 2.818

39-40)Yes researcher 2 is right, We can see that the educ coefficient is positive which means that as educ increases, the rating increases. This is significantly proven by the pvalue=0.000 which means the researcher can conclude significantly

41)Rating=77.318+0.102*0+2.818*0-4.146*0-0.839*0=77.318

42)Rating=77.318+0.102*0+2.818*2-4.416*2-0.839*0=74.106

43)Rsq=0.063 means that only 6.3% of the variation in the dependent variables is being explained by the predicted variables. This also means that the model is not good.