Analysis of parental leave


New Member
Hello, in my bachelor thesis i have to use regression analysis, i used a questionnaire to collect the data. Dependent variable is the time that respondents have factually spent on parental leave.However, i didn't set this question clear, and now i have answers from respondents about when their leave started, and in what period did it finish. So, the questions and optional answers are:
1)Start of maternity leave - more than 8 weeks before bearing, 8 weeks, 7, 6, less than 6 weeks;
2)End of parental leave - less than 12 months, 12m to 18m, 18m to 2 years, 2 to 3 years, 3 to 4 years, and "decided not to return to job".
Is there a right way to adjust this proxies for analysis? (I was thinking of using maximal values: 12m, 18m, 2 years, 3, 4years respectively, and denote not returning to job for 5 years; however, that not sure if it's right)
Thanks for help


No cake for spunky
You can use categories as the dependent variable if your run ordinal or multinomial logistic regression. But I am not sure how you can generate categories because you are subtracting a range of values from a range. How can you tell if its more than 8 weeks how many weeks it is, or what (as an example) it means to start more than 8 weeks and end less than 12 months. How long is that?