When you enter a variable (X), the R-squared increases, and has a p-value. Then X has a p-value. These two p-values are exactly the same. If you enter another variable, the p-value associated with X will change, but that's because it's the value when you've controlled for another variable. Send your output, and then we can see what happened. jeremy On 12 August 2012 08:39, Maryam al-Humairah Adam <[log in to unmask]> wrote: > Hi, > > I have done a multiple regression analysis to examine some predictors of > life satisfaction. One of the results shows that no significant change of > the variance when a new predictor variable (say predictor X) was entered, > but the standardised beta coeeficient yielded significant results (e.g. > p=.04) . Can I say that predictor X has a potential predictive effect on > life staisfaction? > Thanks very much > > Mazni Mustapha > Plymouth University