Stepping carefully away from the quantitative/qualitative debate (although I believe there is a need for both sides to shake hands and come out "triangulating"), I have a question relating to your conclusion. In your paper, the conclusion is that "media influence learning." Given this, your response to the question by Jeff Oliver [20 Oct 96], "Finally, how were the post-tests delivered? Print? Screen? Audio?,"...
[quoting Tripp, 21 Oct 96.a] Irrelevant, but the test was paper and pencil. Clark's hypothesis is not about the form of the post-test.
is appropriate. However, later you change this to...
I believe that my evidence, combined with Furnham and Gunter's, provides good reason to believe that people learn factual information better from text than from audio.
In this case, I think that the type of post-test does become important. Whether media comparison studies are valid and/or of any worth is (still) being debated. Nevertheless, one of their shortfalls is that the evaluation of learning outcomes is nearly always done using paper and pencil (i.e., print). The way information is stored and retrieved may depend on it's form. As Margaret Nissen [19 Oct 96] wrote...
If I study a photograph, I retain an image. If I learn a song, I can "hear" it playing in my head. However, when I read a book, I process symbolic information in the form of words.
Perhaps your results would be different if you used an oral/audio test?
It seems to me that if the learning outcomes of multimedia are to be evaluated then we should be looking at new (multimedia?) ways of testing them.