0
$\begingroup$

I have a sample size of 167. In addition to DV, IV, and a moderator variable, I would like to use 6 variables. Is it too many? How do you determine acceptable numbers of covariates?

Also, fewer covariates can be a better study model or a worse model? I appreciate your opinions.

It is a hierarchical model, 1)covariates, 2)IV, moderator, 3)interaction.

Improve this question
$\endgroup$
5
  • $\begingroup$ How many parameters will the model estimate? $\endgroup$
    – Michael M
    Commented Oct 6, 2019 at 6:46
  • $\begingroup$ You could use Lasso or ridge regression to regularize complexity $\endgroup$
    – Sebastian
    Commented Oct 6, 2019 at 9:01
  • $\begingroup$ What kind of model are you running? $\endgroup$
    – Dave
    Commented Oct 6, 2019 at 14:33
  • $\begingroup$ Hierarchical linear regression with 3 steps, 1)covariates, 2) predictor, moderator, 3)interaction. Thank you! $\endgroup$
    – Yumi
    Commented Oct 6, 2019 at 16:09
  • $\begingroup$ Could you please write down the model you want to use? $\endgroup$
    – Dave
    Commented Oct 6, 2019 at 16:25

0

Reset to default