• Home
  • Textbooks
  • Statistical Methods for the Social Sciences
  • Introduction to Advanced Topics

Statistical Methods for the Social Sciences

Alan Agresti, Barbara Finlay

Chapter 16

Introduction to Advanced Topics - all with Video Answers

Educators

Chapter Questions

Problem 1

Refer to Table 16.1.
a) Interpiet the estimated effects of race and gender on the hazard rate.
b) Shou how to test the effect of race. and interpret

Check back soon!
02:06

Problem 2

In studying the effect of race on job dismissals in the federal bureaucracy. C. Zwerling and H. Silver (American Saciological Revier: Vol. 57, 1992, p. 651) used event history analysis tu model the hazard rate regarding termination of employment In modeling yvoluntary terminations using a sample of size 2141, they reported $P<.001$ in significance tests for the partial effects of 1ace and age.
a) They reported an estimated effect on the hazard rale of $e^k=2.13$ for the coefficient of the dummy variable for being black. Explain how to interpret.
b) The gender effect was not signuficant, but they reported an estimated multiplicative effect on the hazard of 1.06 for the dummy variable for heing fesnale Interprel

Marc Lauzon
Marc Lauzon
Numerade Educator
01:42

Problem 3

Explain what is mean by a censored observation. and give an example in which most observations would be censored.

Lucas Finney
Lucas Finney
Numerade Educator

Problem 4

Consider the vanables $l=$ annual incomc, $E=$ attained oducational level, $Y=$ number of ycars experience in job. $M=$ motuvation. $A=$ age. $G=$ gender. and $P=$ parents' attauned educational level
a) Construct a path diagram showing your view of the likely relationships among those variables.
b) Construct the regression models you would need to fit to cstimate the path coelficients for that diagram

Check back soon!

Problem 5

Rerer to Table 9.13 and Problem 9.17. Draw a path diagram relating $B=$ buth rate, $G=$ GNP, $L=$ literacy. $T=$ television ownership, and $C=$ contraception, Specify the models you would need to fit to estimate path coefficients for your diagram.

Check back soon!

Problem 6

In Table 9.1, consider the variables ınurder rate, percentage metropolitan. percentage high school graduates, and percentage in poverty. Do not use the observation for D C
a) Construct a reahstic path diagram for these variables.
b) By fitting the appropriate models for these data, estimate the path cocfficients. and construct the final path diagram. Interpret.

Check back soon!
01:57

Problem 7

Refer to Table 9.16 in Problem 9.24. Consider the spurious causal nodel for the association between crime rate and percentage high school graduates, controlling for percentage urban. Analyze the data. and explain whether the data are consistent with this model.

Tyler Moulton
Tyler Moulton
Numerade Educator

Problem 8

Refer to Table 9.1. Using soltware as show'n by your instructor. conduce a factor analysis How many factors scem appropriate? Interpret the factors, using the estimated factor loadings.

Check back soon!
07:39

Problem 9

Refer to the previous problein Remove the observation for D.C., and repeat. Hou sensitive are the estimatcd factor luadings and your identification of factors to that onc observation?

Heather Duong
Heather Duong
Numerade Educator

Problem 10

Construct a diagram representing the following covanance structure modct, three observed rcsponse vanables are described by a single latent variable, and that latent variable is regressed on four observed predictors.

Check back soon!

Problem 11

Construct a diagram representing the following covanance structure inodel. for vanables measured for each state. The latent response variable is based on two observed indicators, violent crime rate and murder rate. The two predictor variables for that latent variable are the observed values of percentage of residents in poverty and percentage of single-parent families; these are treated as perfoculy measured

Check back soon!
01:15

Problem 12

Refer to the previous problem. Using software that your instructor introduced. fit this model to the data in Table 9.1 and interpret results

Raymond Matshanda
Raymond Matshanda
Numerade Educator

Problem 13

Construct a diagram representing a covanance structure model in which (i) in the measurement part ol the model. a single factor represents violent crime ratc and murder rate and a single factor represents percentage high school graduates, percentage in poverty, and percentage of single-parent families, and (ii) in the structural cquation part of the model, the first factor depends on the second factor as well as on the percentage of metropolitan residents

Check back soon!