Solved: How to weigh 5 as the best and 0 as the worst. Multiple regression (fit model)

Thestudent · Apr 5, 2022 08:03 AM

Its many years since I used JMP, so sorry for the stupid question.

I made a survey where the answers is rated from 0 (Not included) to 5 (The best). I need to have this included in my regression model. How do I do that? Its concerning many variables which all explains preferences. Some are very important (rated 5) and some are less important (from 4 to 0)

I just need to know how I weigh them in my fit model

Mark_Bailey · Apr 5, 2022 08:23 AM

Is not the rating (0-5) the response? If so, you can use Numeric or Character data type, but use the Ordinal modeling type. Enter this column in the Y role after selecting Fit Model from the Analyze menu. This action will automatically change the Personality to Ordinal Logistic regression. You will model the log odds of the cumulative logit of the response.

Is that your question?

View solution in original post

Mark_Bailey · Apr 5, 2022 08:23 AM

Is not the rating (0-5) the response? If so, you can use Numeric or Character data type, but use the Ordinal modeling type. Enter this column in the Y role after selecting Fit Model from the Analyze menu. This action will automatically change the Personality to Ordinal Logistic regression. You will model the log odds of the cumulative logit of the response.

Is that your question?

How to weigh 5 as the best and 0 as the worst. Multiple regression (fit model)

Re: How to weigh 5 as the best and 0 as the worst. Multiple regression (fit model)

Re: How to weigh 5 as the best and 0 as the worst. Multiple regression (fit model)