Researchers at an insurance company want to study the risk factors associated wi
ID: 3268940 • Letter: R
Question
Researchers at an insurance company want to study the risk factors associated with heart disease, so they collect data from 240 people. As their dependent variable, they use a question that asks respondents if they have been hospitalized for heart problems 10 = no, 1 = yes2. They also collect information on the respondent's sex (0 = female, 1 = male), weight (in pounds), tobacco use (packs of cigarettes per week), and exercise (hours per week). Their analysis produces the following logistic regression table:
a. Why did they use logistic regression instead of multiple regression?
b. Interpret the results for each of the five variables, paying special attention to e b and statistical significance.
b
SE
Wald
Sig
eb
constant
-23.499
10.699
4.824
0.028
0.000
sex
4.523
2.552
3.142
0.076
92.149
weight
0.143
0.061
5.524
0.019
1.154
tobacco
-0.495
0.316
2.448
0.118
0.610
exercise
-0.132
0.175
0.576
0.448
0.876
b
SE
Wald
Sig
eb
constant
-23.499
10.699
4.824
0.028
0.000
sex
4.523
2.552
3.142
0.076
92.149
weight
0.143
0.061
5.524
0.019
1.154
tobacco
-0.495
0.316
2.448
0.118
0.610
exercise
-0.132
0.175
0.576
0.448
0.876
Explanation / Answer
a.
Logistic regression is used when the dependent variable is a categorical variable. Here, in this case also, dependent variable is a binary variable whose answers would be either 0 (no) or 1 (yes).
b.
The values mentioned in the b column are the coefficients. For example, value of 0.143 for weight indicates that a one unit change in weight results in 0.143 unit change in the log of the odds. SE column refers to the standard error. Wald is a chi-square value. Sig is a 2-tailed p-value which tests the null hypothesis that the coefficient is 0. Here, only weight would be considered statistically significant because it is having value less than 0.05. The final column eb is the exponential of the coefficient b which can be considered as odds ratio for predictors.
Related Questions
drjack9650@gmail.com
Navigate
Integrity-first tutoring: explanations and feedback only — we do not complete graded work. Learn more.