I would like to evaluate the effect of the independent variable (ratio of females to males) on the dependent variable Y in R. We often use cbind(y, N - y) as a dependent variable in GLM with binomial distribution. Similarly, can we use cbind() as independent variable?
I have written the following equation:
res
It is just like when you put an X matrix in, as in lm(y~X), so lm(y ~ cbind(x1, x2)) is equivalent to lm(y ~ x1 + x2). I think Sal Mangiafico 's point is that if you have only two gender categories your model wouldn't make sense, but if you had an OTHER option that would just be dummies for male and female.
Are "male" and "female" really different variables or are these actually values of one variable (sex, what can be either female or male)?
It would also be helpful if you'd attach "DF".
Jochen Wilhelm
Thank you for taking the time to comment on my question. A part of the DF has been uploaded. Each row shows the Y (population trend index varying between 0 and 1), number of females, and number of males in each area.
I am interested in evaluating whether the number of females or the number of males has a stronger impact on population trends. Due to the strong collinearity of the number of females and males, I could not include them as independent variables in the model. So I tried to simply convert them to the proportion of females and include them in the model, but then I would lose the information about the difference in magnitude of the original numbers. I am looking for a way to verify effect of the difference in proportion while retaining the information from the original numbers.
- Badges
- Science topic
- Similar topics
- Mathematics
- Statistics
- Sampling Design
- Sample Size