• 100% Satisfaction Guarantee

Math-John, PhD in Statistics
Category: General
Satisfied Customers: 1310
Experience:  I have rich teaching, tutoring and consulting experience.
62644613
Type Your Statistical Analysis Question Here...
Math-John is online now

### Resolved Question:

What are the advantages and disadvantages of logistic regression, sequential logistic regression, and stepwise logistic regression?? What criteria would be used to decide which method to use?
Submitted: 4 years ago.
Category: Statistical Analysis
Expert:  Math-John replied 4 years ago.

Hi, welcome to Just Answer. Here it is.

1. DV and error terms don't have to be normally distributed,

2. DV doesn't have to have equal variance in each group

2. No linear relationship between the IV and DV has to be assumed.

3. Can handle nonlinear effect, interaction effect and power terms

4. IV can be categorical variable and bounded.

Disadvantage: requires large sample size to achieve stable results.

Please ACCEPT. Bonus and Positive Feedbacks are greatly appreciated. To request me, begin your questions with "For John". Thanks.

John

Customer: replied 4 years ago.

Is your answer referring to logistic regression, sequential logistic regression, or stepwise logistic regression??

Also, what criteria would be used to decide which method to use?

Expert:  Math-John replied 4 years ago.

Sorry I thought you asked the pros and cons of logistic regression in general. I assume "logistic regression" means using all predictors. Please let me know if otherwise.

Logistic regression

Pros: use all predictors, will not miss important ones.

Cons: may have multicollinearity .

Stepwise logistic regression

Pros: can find a model that is parsimonious and accurate. Maybe able to find relationships that have not been tested before.

Cons: may over fit the data.

Sequential logistic regression

Pros: can test the relationship that the research is interested. Maybe able to find relationships that have not been tested before.

Cons: may miss the chance to find important relationship.

• 1. If the interest is the relationship between all predictors and dependent variables, logistic regression with all predictors is appropriate to use.
• 2. If one wants to add variables in a pre-specified order, sequential logistic regression is appropriate to use.
• 3. If one wants a parsimonious and accurate predictive model, stepwise logistic regression is appropriate to use.

Please let me know if you have any questions and accept. Thanks.

John