WEEK 4 - Research on media differences

(created on Jan. 28)

Articles Read:

· Barron, A.E., & Kysilka, M.L.(1993). The Effectiveness of Audio in Computer-based Education. Journal of Research on Computing in Education 25(3), 277-289.

· Baggett, P. (1979). Structurally equivalent stories in movie and text and the effect of the medium on recall. Journal of Verbal Learning and Verbal Behavior, 18(3), 333-356.

· Beentjes, J.W.J., & van der Voort, T.H.A. (1993). Television viewing versus reading: Mental effort, retention, and inferential learning. Communication Education, 42(1), 191-205

Article Summaries:

Barron & Kysilka(1993)

This piece is to examine the effectiveness of adding digital audio to CBT.
In the experiment, all the students completed an identical CBT (computer-based training) program. However, they were randomly assigned to one of three treatment groups: text-only delivery, full-text and totally redundant audio delivery, or partial-text (bulleted) with full-audio delivery.
It is found that (a) Achievement: there was no significant difference among the three treatment groups in achievement gain; (b)Time: there was a significant different in the mean completion times with the text-only version requiring the least time on task; (c) Perception: student perceptions indicated a high degree of program acceptance across all levels of treatment. So the conclusion is “the various CBT delivery modes did not have a significant effect on overall comprehension of the tutorial content”.

Baggett (1979)

The study examined two questions. The first is whether the perception of episodic structure (exposition, complication, resolution) is medium (text versus movie) independent (Experiment 1 through 4); and the second is what is the effect of the input medium (text versus movie) on cued recall (Experiment 5). The findings were the following: a. stories told in a text and in a movie without words can be constructed so that people perceive them to have equivalent episodic structure; b. statements of structure in recall protocols of subjects who study the structurally equivalent movie and text stories are very similar; c. After a 7-day delay, recall of structural statements deteriorated faster for text subjects. The conclusion was story processing is independent to the input medium.

Beentjes & van der Voort, (1993)

It compares children’s learning from structurally equivalent television and print versions of two stories. Children were randomly assigned to four orders of presentation (to control for sequence effects), and each child was presented with both stories and both media in two different combinations (story A on TV and story B in print or story A in print and story B on TV).There are three findings. a. Mental effort in processing: when measured by means of Salomon’s retrospective measure, children invest more mental effort in processing print stories compared to television stories. However, on the reaction-time measure the reverse was found. b. Inferential learning: television stories resulted in more inferential learning than the print stories, which is in contradiction to the hypothesis derived from Salomon’s (1984) model. c. Immediate retention: Immediate retention is not affected by the medium through which the story is presented. d. Delayed retention: Viewers’ retention was superior to that of readers.

My Responses:

What do you think makes for a good research question to address with respect to learning via differing media? Why is this question better than others?

In response to this question, I will first analyze each of the three studies.
Barren et al's (1993) study suggests there is no differences among various delivery mode. However, the task is not sensitive to the two types of media.The task is CD-ROM tutorial. It is not surprising that using sound or text won’t make much different. It is highly possible whether there are pictures or not will make a difference in such kind of task. In addition, the study might need more sensitive measurements.
In Baggett's (1979) study, one of the measurements is the delayed effect of media on recall, which helped identify the differences in text and movie representation. This suggests that more sensitive measurement is necessary for such kinds of study.
Among the three studies in our required readings, I think Beentjes & van der Voort (1993)'s study is the best. In the study, the authors used multiple tasks and multiple measurements. First, the author used two different kinds of films, so as to “provides a check on the generalizability of the study’s results” (p.196). In addition, the authors adopted two different methods in the measure of mental effort and of retention. However, I am not sure the second-task procedure is an appropriate approach here. It is more likely to be used in measuring engagement. Student slower response in TV condition may suggest that they are more engaged (in things beyond the stage of making sense of the meaning/decoding) instead of spending more mental efforts in understanding the content (decoding).
Another question arouse from Beentjes et al's study is whether print version and TV version really comparable. In more general terms: Is two different media really comparable? Print versions can be very different in many ways! The level of difficulty of language, the techniques of writing…all these will bring very different experiences to the readers.
From the above analysis, it is clear that a good research on learning with different media should consider the variations of the tasks. In addition, as the different media may only lead to differences in learning processes (as was argued by Cobb (1997)) or very slight measurable differences in learning outcome, more sensitive measurements should be used to avoid drawing premature conclusions that various media has no differences.

Question for Discussion

Is two different media really comparable?

To what degree can we generalize the findings in such comparative studies since most time the stu