I need to know if I can perform multiple linear regressions between a dependent variable (predicted) that I calculated using a formula that I took from the bibliography and regressors that I measured?
well, if i get your question right, there's no reason why you should not perform MLR unless your formula already includes your independent variables. I think it might be a better idea to provide more details on the variables.
I try to predict the age of gestation in domestics animals, that I analyzed after slaughter.
I do not know the exact values of gestational age because I have no history in this regard, but I found in the bibliography a formula already established by other authors in a similar population to that in which I did the study. so I collected the results of Fetomaternal independent variables (regressors)
is what I can calculate gestational age using the predefined formula and achieve thereafter linear regressions with regressors ?
If I undestand correctly you're colling "regressors" to the independent variables, let's call them x1, x2, ..., xN, and your dependent variable, the gestational age, would be y. Since you don't have y, you calculate y from a formula y=F(x1,x2,...,xN) that you found in the literature. Then you want to use those y values in a linear regression study of y versus x1,x2,...xN. Am I correct?
If that's what you want to do, you cannot.
That would be fine. Your analysis won't event depend on you calculating GA=(X1+23.99)/0.366, since it's just a linear transformation, thus it will affect the value of the regression coefficients but not any non-scale-dependent analysis such as calculating standarized coefficients, correlations, partial or semipartial correlations, etc. So go ahead!
- Badges
- Science topic
- Similar topics
- Bioinformatics
- Bioinformatics and Computational Biology