Hi,
I have done a linear model with 3 IV including female dummy on weight for height Z score..
I need help with understanding the effect of coeff value (positive value) with regards to intercept and male reference, as well as effect of a higher beta value compared to other IVs in model..
Please help..
If your dummy variable is properly scored as 0/1, then the intercept is the mean score for the category coded zero (the so-called "omitted" category), after controlling for the other variables in the equation. The coefficient for the dummy variable is the difference the two categories, i.e., the amount that would need to be added to or subtracted from the intercept in order to produce the mean for the second category. The test of significance for the dummy variable is thus a test of the difference between the two category means (again, after controlling for the other variables in the equation).
It's usually either the unweighted grandmean from focal group's intercept (effect coding) or groups lean away from overall weighted sample's mean (dummy).
Hope this helps,
Matt
Generally, dummy variables are only used as controls in a linear regression. If they are significant, it only means that they can be used to separate the sample into distinct groups. In this case MANOVA can be used. Interpreting them in linear terms of beta coefficients or intercepts is contestable and best avoided.
- Badges
- Science topic
- Similar topics
- Mathematics
- Statistics
- Regression