Fernand Gobet

ESRC Centre for Research in Development, Instruction and Training

Department of Psychology, University of Nottingham

University Park, Nottingham NG7 2RD

United Kingdom


Phone (0115) 951 5402, Fax (0115) 951 5324




ESRC Centre for Research in Development, Instruction and Training

July 1997

Technical Report No. 48





Observations on the use of visual aids (VAs) in two international conferences on psychology are reported. Results based on 85 talks show that most speakers use VAs, but that in more than half of the cases VAs were not legible. Suggestions for further research on the misuse of VAs are given.





Giving or attending talks constitutes an important portion of scientists’ life, and it is therefore desirable to know what is the impact of this mode of communication. From a practical point of view, this boils down to: How much benefit does a listener gain from a talk? To answer this question, one can use at least two general principles derived from research in cognitive psychology. First, in gathering sensory information, it is preferable to use multiple channels rather than a single channel. Second, sensory input of poor quality are difficult to process by the cognitive system and is unlikely to leave a durable trace in memory.

A large corpus of empirical evidence supports the first principle. It is known that organisms, including humans, can take advantage of whatever sensory information is provided when learning (Hilgard & Marquis, 1961). This result is also valid in educational settings, where, for example, material presented verbally and visually is better recalled than material presented in a single modality, assuming an audience able to understand the message of both the verbal and visual modalities (Romiszowski, 1988). Several theoretical explanations may be given for the raison d’être of this first principle. For example, it has been shown that information is best recalled when encoded in a redundant way, allowing its retrieval from several different cues (Anderson, 1990; Richman, Staszewski, & Simon, 1995). In addition, use of several modalities forces a deeper treatment of the information (Craik & Lockhart, 1972).

Informal observations show that psychologists apply the first principle consistently: an imposing majority of them use visual aids when giving a talk. (The empirical evidence reported below generally supports this claim.).

The second principle has also been established beyond any doubt by scientific research, and it has almost become a truism that the characteristics of sensory stimuli play an important role in the way information is perceived, as is illustrated, for example, in the use of the signal detection procedure in various psychophysical domains (Goldstein, 1989). However, it is at the same time quite common in discussions between psychologists to hear complaints about the poor use of visual aids (later VAs) in professional talks and about the fact that, in spite of numerous guidelines in the literature on the use of VAs (e.g. Bell, 1987; Stuart, 1988; Wilkinson, 1989), certain VAs are legible only for those with eagle-sharp eyes in the audience.

I took the opportunity of attending two psychology conferences to test quantitatively whether psychologists, as experts of human information processing, apply the two principles mentioned above. Specifically: (a) Do they use VAs to enhance the clarity and memorability of their presentations? (b) Do they apply consistently the guidelines of clarity and readability that can be derived from basic perceptual processes?



Observations were made during two international conferences on psychology: the Second International Conference on Memory (Padova, July 14th-19th 1996), and the "Growing Mind" Conference (Geneva, September 14th-18th 1996). For each of the talks (including plenary sessions) I attended in the two conferences, I took a measure of the legibility of the visual aids. Talks attendance was decided by my own scientific interests, which are likely to be independent from the quality of VAs. VAs consisted exclusively of slides, overheads, and blow-ups of computer screens.


Measure of legibility

Although it is typically stressed that the contents of VAs should be large (e.g. Bell, 1987; Stuart, 1988; Wilkinson, 1989), few recommendations have been made about the minimum font size allowing good readability. I used the following arbitrary (and probably lenient) criterion: the equivalent of 14 points in Geneva or New York fonts. (Note that this information was estimated in real time from my viewing point, which changed from talk to talk).

I defined a VA as hard-to-read if some critical information was written in a size equal or less than the criterion. Critical information include, among others, mathematical formulae, labels for axes and lines in figures, contents of table, and written summary of important points in talk.

Talks using VAs were classified as follows:

1. Low readability: more than 2/3 of the VAs are hard-to-read.

2. Medium readability: from 1/3 to 2/3 of the VAs are hard-to-read.

3. High readability; less than 1/3 of the VAs are hard-to-read.


As my main interest in attending these talks was to learn about scientific developments in the field, the counting method had to be fast, and therefore, somewhat approximate. This did not seem to hamper seriously the results, as most presenters were quite consistent in using either large or small slides, and as the incertitude comes mainly from the Medium readability category. As we will see, this does not influence the conclusions of this paper.



41 observations were made in the Memory Conference (5 plenary sessions, and 36 talks), and 44 in the "Growing Mind" Conference (6 plenary sessions, and 38 talks). Table 1 gives the quantitative results and show that, while most psychologists do use VAs, only few of them take the time to prepare VAs with high readability. There is a tendency for plenary sessions to be more readable, although the small number of observations makes generalization tentative. In the Memory Conference, the quality was better for plenary sessions (respectively 40%, 20%, and 40%, for low, medium, and high readability&endash;all used VAs) than for talks (61%, 27%, and 11%). In the "Growing Mind" conference, 50% of the plenary sessions did not use VAs; in the other sessions, 17%, 33% and 0% had low, medium, and high readability, respectively, as compared with 60%, 8% and 13% in talks only.




# talks attended
% talks with VAs
2nd Conference on Memory
Conference "The Growing Mind"

Table 1: Number of talks attended, percentage of talks where VAs were used, and percentage of talks having low, medium, and high readability, for two psychology conferences.



The goal of these observations was to examine how experts in human cognition apply the principles they study in their training years and help develop in their own research. Two international psychology conferences were chosen as testing ground. It was found that most of the speakers use some sort of VAs, but that, disappointingly, more than 55% of the VAs had a low readability. This result indicates that most psychologists, who should precisely be aware of the limits of human cognition in acquiring new information and of the ways to at least not worsen these limits, do not apply the principles developed in their field.

This paper has addressed only one (important) aspect of preparing good VAs, and a fortiori, of giving a good talk (see Bell, 1987; Stuart, 1988; and Wilkinson, 1989, for an extensive discussion on the use of VAs or the art of giving talks). Further research may be conducted in several directions. Cognitively, do psychologists also neglect other principles of good communication, as described by Bell (1987), Stuart (1988) and others? Sociologically, does this neglect extend to scientists of other fields? Economically, what is the proportion of good and exciting ideas that are not being properly communicated because of poor VAs, and what is its monetary cost? Clinically: does reading bad VAs have the same distracting, frustrating and irritating effects as reading a text full of spelling errors and other typos? Finally, psychoanalytically: why do psychologists, who suffer from the poor VAs used by others, keep presenting poor VAs on their own?



Anderson, J. R. (1990). Cognitive psychology and its implications. New York: Freeman.

Bell, G. (1987). The secrets of successful speaking and business presentations. Oxford: Butterworth-Heinemann.

Burnett, R. (1990). Technical communication. Belmont, CA: Wadsworth.

Craik, F. I. M. & Lockhart, R. S. (1972). Levels of processing: A framework for memory research. Journal of Verbal Learning and Verbal behavior 11, 671-684.

Goldstein, E. B. (1989). Sensation and Perception. Belmont, CA: Wadsworth.

Hilgard, E. R., & Marquis, D. G. (1961). Conditioning and learning. (Second edition, revised by G. A. Kimble). New York: Appleton-Century-Crofts.

Richman, H. B., Staszewski, J., & Simon, H. A. (1995). Simulation of expert memory with EPAM IV. Psychological Review, 102, 305-330.

Romiszowski, A. J. (1988). The selection and use of instructional media. New York: Nichols Publishing.

Stuart, C. (1988). Effective speaking. Hants: Gower.

Wilkinson, L. (1989). SYGRAPH: The system for graphics. Evanston, IL.: Systat, Inc.