This preview has intentionally blurred sections. Sign up to view the full version.View Full Document
Unformatted text preview: Homework #1—Review of Multiple Regression—Page 1 Soc 63993, Advanced Social Statistics II Homework No. 1 Review of Multiple Regression I. The attached table is from an article in Journal for the Scientific Study of Religion, 1990, Vol.29(3):297-314, Religious Practice: A Human Capital Approach , written by Laurence Iannaccone. (Table appears on p. 304.) Read the table carefully and answer the following questions. 1. When the dependent variable is CONTRIBUTE, the unstandardized coefficient of INCOME is 9.025. What does this mean? 2. When the dependent variable is CONTRIBUTE, the significance level of INCOME is p .001. What does this mean? Which type error does it test? Write out the null and alternative hypotheses that are being tested by the T statistic. 3. Suppose the researcher believes that older people tend to attend fewer masses than do younger people. Do these results support her? 4. What is the standard error of the coefficient for MARSAME in the CONTRIBUTE equation? 5. For the CONTRIBUTE equation, test whether or not R 2 significantly differs from 0 (i.e. compute the F value, its degrees of freedom, and its significance). NOTE: You can use the Ftail function in Stata to compute the significance. So, for example, if your computed F = 2.4 with d.f. = 10, 120, the F value is significant at the .0124 level. Ergo, you would reject the null if using the .05 level of significance but not reject if you were using the .01 level. . display Ftail(10, 120, 2.4) .01236157 6. When a male and female have identical values on all other IVs, which one will be likely to attend mass more (and by how much)?...
View Full Document
- Spring '11
- Regression Analysis, unstandardized regression coefficients, Multiple Regression—Page, MARSAME, Advanced Social Statistics