The partial F corresponding to each stimulus measures whether the whole stimulus is siginificant. In the other words, the partial F value would help answer the following question: if you remove the stimulus from the model, is the added variance by removing this stimulus significant compared to the case with the stimulus in the model? If you have only two lags, the null hypothesis for the partial F is H0: beta0 = 0 and beta1 = 0.

The t value corresponding to each specific lag simply tests whether the coefficient for that lag is significantly different from zero. In the case of beta0 (or beta1), the null hypothesis is H0: beta0 = 0 (or beta1 = 0).

The relationship between partial F and t values is apparent in terms of their null hypotheses: If the partial F is not significant, the two t values would not be significant; However, if either or both t values are not significant, it does not necessarily mean partial F would be insignificant.

Hope this would clarify the diference.

Gang