0

I was inspired by the following question:

How to write a logit and probit regression equation?

To ask for advice on how to write a logit regression equation in a suitable format for an academic journal:

The analysis is conducted in longitudinal panel data in Stata, where I consider the impact of parental job loss on the probability of children being overweight (a binary outcome variable) in a fixed effects logistic regression (xtlogit), while controlling for region (binary urban or rural), year (2002, 2006, 2010), parental age (continuous), mothers education (4 categories), mothers marital status (4 categories) and child's age in months (continuous).

If this was a linear equation I would express it as:

Dummy dependent variable = intercept term + dummy parental job loss + dummy region + categorical year + continuous parental age + categorical maternal education + categorical maternal marital status + continuous child age.

But I would like to write this like the response by Alexis to the question linked above, can anyone advise?

Edit:

I am asking this question because although I have read the response by Alexis I did not understand it, and thus do not know how I can apply it to my own model. If I could understand how to do this here then I would know how to do it in the future for any other logit models I run.

I need to write the model as a formal equation based on the requirements of the journal I am targeting.

James Moore
  • 74
  • 9
  • 3
    Could you please say why you think you need to write the model as a formal equation? Logistic regression is well enough established as a technique that you should simply be able to specify the predictors you used in the text and then provide a table of regression coefficients, their standard errors, etc. Is there some specific reason why you need to write the model formally as an equation? – EdM Sep 13 '19 at 18:43
  • 1
    There doesn't seem to be anything new in this question: it's the same as the one you reference, only with a slight change in the names of the variables. What, then, are you looking for as an answer? You do mention it's a panel study, but your linear equation example doesn't reflect that. – whuber Sep 13 '19 at 19:11
  • I have edited above to reflect your comments. – James Moore Sep 14 '19 at 10:05
  • Are there any suggestions on the above? Thank you, James. – James Moore Sep 20 '19 at 11:27

0 Answers0