2
$\begingroup$

I have run bivariate analysis predicting cholesterol and iron from gender and age. The coefficient of gender is $-0.57$ (male=1, female=0) and the coefficient of age is $0.5$. Now I am wondering how one can interpret these coefficients?

$\endgroup$
2

1 Answer 1

2
$\begingroup$

When generating a multivariate regression model, each term's coefficient pertains to its main effect when controlling for the effect of the other coefficients. In your described case of modeling cholesterol (continuous) with respect to age (continuous) and gender(categorical), a coefficient of .5 for age means that controlling for (this is different than ignoring) gender, for every year older an individual is, their cholesterol level increases by .5.

Since your intercept is "female" for your gender variable, a coefficient of -.57 indicates that controlling for age, males have .57 lower cholesterol than females on average.
If you're having a hard time with the controlling for vs. ignoring concept, I'd recommend reading through this answer.

Edit: User has since changed and/or clarified their question

$\endgroup$
2
  • 1
    $\begingroup$ thanks for the comment. But here I have two dependent variables. And these are the coefficients of multivariate regression and not univariate regression. "Since your intercept is "female" for your gender variable, a coefficient of -.57 indicates that controlling for age, males have .57 lower cholesterol than females on average." this interpretation is for univariate analysis when I have just cholesterol as DV. $\endgroup$
    – LoveStat
    Jun 26, 2017 at 17:30
  • $\begingroup$ Ah, I've edited your original question to reflect the now clarified intention. $\endgroup$
    – Connor
    Jun 26, 2017 at 17:39

Your Answer

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

Not the answer you're looking for? Browse other questions tagged or ask your own question.