This is often much faster (in human seconds you're making the computer do all the work again but that's usually a few hundred milliseconds) You can also get and estimate of b1-b2 and its standard error (and typically a t-value and p-value) by taking a simple (linear) transformation of your predictors and then running a new regression. Its estimate is computed as the element corresponding to the row for b1 and the column for b2 (or vice-versa) in the estimate of the covariance matrix for the parameters, which is s 2 -1 (Most multiple regression packages will give this to you often as an option) So you need the covariance between the parameter estimates. If we assume that your estimates are b1 and b2: I take it B1 and B2 are the population values. R-bloggers - blog aggregator with statistics articles generally done with R software.
0 Comments
Leave a Reply. |
AuthorWrite something about yourself. No need to be fancy, just an overview. ArchivesCategories |