0
$\begingroup$

This question is related to How to analyze curvilinear seasonal data

I have data like following: 

subjectID, month, vnum1, age, gender
1          1       5    10      1 
2          1       6    15      2
3          2       6    18      2
...

I have one continuous variable (called vnum1) which is showing a curvilinear relation with month of the year (lower in summer months). However, I have 2 other variables also- age and gender. I have to determine if the curvilinear relation of vnum1 with month is independent of age and gender. How can I test this. Thanks for your help.

Improve this question
$\endgroup$
4
  • $\begingroup$ Age and gender will need to be predictors and then you need to test for interactions of however you are handling month with age and gender. Standard linear model stuff. In this case the data are for people who are (certainly) aging over time but (probably) constant in gender, so at least the first change needs to be considered even if it is neglected. $\endgroup$
    – Nick Cox
    Commented Nov 27, 2014 at 12:26
  • $\begingroup$ I tried to clarify in previous question also that subjects were not tested repeatedly. Different subjects were tested over 3 years. There was no age and gender selection of subjects tested at different times. Hence any relation between age or gender with month is likely to be by chance only. $\endgroup$
    – rnso
    Commented Nov 27, 2014 at 12:33
  • $\begingroup$ If no person is ever measured twice, then OK. $\endgroup$
    – Nick Cox
    Commented Nov 27, 2014 at 12:40
  • $\begingroup$ No person was ever measured twice. $\endgroup$
    – rnso
    Commented Nov 28, 2014 at 3:08

0

Reset to default

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.