In the current study on dysarthric speech, we investigate the effect of web-based treatment, and whether there is a difference between content and function words. Since the goal of the treatment is to speak louder, without raising pitch, we focus on acoustic-phonetic features related to loudness, intensity, and pitch. We analyse dysarthric read speech from eight speakers at word level. We also investigate whether there are differences between content words and function words, and whether the treatment has a different impact on these two classes of words. Linear Mixed-Effects models show that there are differences before and after treatment, that for some speakers the treatment has the desired effect, but not for all speakers, and that the effect of the treatment on words for the two categories does not seem to be different. To a large extent, our results are in line with the results of a previous study in which the same data were analyzed in a different way, i.e. by studying intelligibility scores.