The World Wide Web is rapidly becoming a popular new medium for communicating messages and ideas. This study investigates whether designing with enhancing visual elements in a Web presentation would contribute to greater understanding of the message being presented in the Web presentation compared to using text only. Significant effect for the use of enhancing visual elements was found in participants' immediate understanding of the message, aesthetic satisfaction of the site, and memory retention over time.



INTRODUCTION

The World Wide Web emerged as a new media experience during the mid 1990's. The Web exists in cyberspace "... the labyrinth of computers and computer resources inter-connected through modem high-speed digital networks" (Berghel, 1996). It has become popular world-wide because it is a technology that enables the production, distribution, and viewing of multimedia on remote networked computers.

Much research exists concerning the effects of mass media on viewers, and how the public uses mass media for their own purposes. Until recently mass media has existed as television and radio programs, film, video, and graphic art via hardcopy in the forms of magazines and newspapers. In this study they will be referred to as "traditional media" in order to distinguish them from the new medium of the Web. There is less research on the effectiveness of messages presented on the Web or on the effects of using enhancing visual elements (i.e. color, graphics, texture, text formatting, lines, icons) in a Web presentation.

With this research I have investigated whether designing with enhancing visual elements (EVE) in a Web presentation will contribute to greater understanding of the message being presented in the Web presentation compared to using text only. Because the Web is used to present messages to users, it is important to determine if this presentation process can be aided by enhancing design elements in light of what researchers in traditional mass media indicate. The following considers the value of using visual information in traditional media. The relation of visual information to information perception, retention, accuracy of understanding, memory enhancement, visual cuing, symbols, immediacy of understanding, and emotional reenforcement of message are briefly examined. Finally a Web study by Randolph (1996) is reviewed.

LITERATURE REVIEW

Graber (1989), in her study on content and meaning in TV news, discusses the value of visual information: “Research has shown that audiences report visual content more accurately than verbal content and retention rates are much higher for visual information” (p. 149). Messaris (1993) argues images can provide immediate understanding without requiring a time of initiation, and the understanding of visual conventions presented in mass-mediated TV and films seems to be derived from analogies to real-life perceptual cues. He states “ what makes images unique as a mode of communication is precisely the fact that they are not merely another form of arbitrary signification” (p. 290)and further indicates images can cross over cultural boundaries much easier than language.

The phrase “perceptual cues” suggests the “iconic” or “symbolic” nature of visual images which are very important aspects of visual perception. We are familiar with the old saying “a picture is worth a thousand words”. Think about interstate billboard advertising. Interstate billboard designs are powerful because they present significant information about the product or client in simple direct messages. Pictures have the ability to summarize information through known visual symbolic connections. They can be used in Web design to present the main content of the Web message in condensed quick-reference visual shorthand which offers visual metaphors that help cue general references in the viewer’s memory. This cuing gives the viewer help with immediate understanding of the presented information and memory storage for retrieval when the viewer wants to search the Web for this kind of information again. Icons work as even more condensed graphic images.

In their stories journalists and news reporters use visual cues that are culturally and socially relevant to help viewers and readers draw on general information from their memories to supplement the content of the stories (Graber, 1989). This stereotypical information is part of what is called schema. Wicks (1992), quoting from a study by Friske and Taylor (1984), said schema can be seen as a “cognitive structure that represents organized knowledge about a given concept or type of stimulus” (p.139) taken from previous experience. Gamson (1987) claims schema enable us to group certain information into associations or general categories of focus. We may have trouble understanding information that falls outside of these associations.

The use of icons and pictures for information enhancement and memory reinforcement can be considered in light of communication schema research on how TV pictures affect memory. As we watch television, the visual and audio information can trigger thoughts about something to us. These thoughts can come from schema in our memories. If we regularly access information through these schema sources, processing new information through schema becomes easier (Wicks, 1995). Psychological literature claims that images “possess special advantages over text producing accelerated recall over time” (Wicks, 1995).

In addition to helping people access their memories for associations, visual images can enable people to directly process information and thus make fewer errors in understanding the information (Graber, 1990). In her study on how visuals contribute to learning from the news, Graber found that visual images contributed to story comprehension and information gain by adding reality and clarification the story, and providing emotional support. Basil (1994) suggests though people process audio and written words in common ways, we probably process pictures differently. Pictures take less mental resources to process, picture processing is more automatic.

These studies on the effects of visual image in traditional media support the usefulness of visual images in giving strength to messages. Because the Web is used to communicate messages and is very capable of presenting visual information, it seems very reasonable to apply these findings to Web design.

Randolph’s (1996) research on the subject of visual information used in Web publication is closely related to this study. Randolph chose two research issues to examine the use of graphics and hypermedia in on-line publishing. He wanted to see their effects on memory in both immediate recall and longer-term retention. Randolph created an experiment using 98 college students, over two thirds of which were apparently beginner-level computer users. Students were divided into 4 groups, each group receiving one of four versions of a news story influenced by one of the following: graphics, no graphics, hypermedia links, or no hypermedia links. The Web site was presented on computers in a classroom lab. After the on-line publication was viewed, the Web browser (Netscape) was turned off and students took a quiz over the material. Students took the same quiz a week later without rereading the publication. The graphics related results showed no significant effect on immediate recall, but did show significant effect on long-term retention. The study seemed to emphasize that the students forgot less through the graphics enhancement.

Randolph’s study was reserved to reading on-line publications and observing the effects of graphics and hypermedia to memory processes. The concern of the current study is the broader framework of Web Presentation in general and whether the use of enhancing visual elements (i.e.. color, graphics, textures, text formatting, icons) enable the Web message to be more effectively communicated to the Web user. Considerations in addition to memory are user interest; aesthetic attractiveness of the site; clarification of message; emotional reenforcement; and immediacy of understanding of message. The independent variable, enhancing visual elements, is defined in this study as: visual elements of line, graphics, pictures, color, text formatting, and icons that are carefully and sensitively chosen to be incorporated into the design of the Web presentation to complement and emphasize the textual message; to give greater embellishment to the textual message. The dependent variables of memory retention, interest, emotional bonding, aesthetic satisfaction, clarification of the message, and immediacy of understanding message are the intended effects of applying enhanced visual elements to the Web design.

Memory retention refers to both immediate and long-term retention of memory of the Web message or parts of the Web presentation. Interest refers to the ability of the Web presentation to attract and keep the Web user’s attention. Emotional bonding refers to providing the Web user with some kind of emotional support for the message and the presentation. Aesthetic satisfaction refers to providing a visually (and sometimes audibly) fulfilling experience for the Web user due to the creative and purposive use of visual design elements. Clarification of message refers to using these visual elements in such a way as to order the presented information so it is not ambiguous or confusing and thus easier to assimilate. Immediacy of understanding refers to the desire to present the information in a direct way so the user can quickly and accurately perceive the intended meaning of the message.

The following hypotheses were made concerning these considerations:

This study seeks to show the benefits of using enhancing visual elements in Web presentations. The Web is a unique form of media. The fact that it enables the use of enhancing visual elements as aids in communication makes it possible to expect results in visual communication effects which are similar to those found in traditional media research.

METHODS

In order to test whether the use of enhancing visual elements in a Web presentation will present the Web message in a more beneficial way to the Web user than using black text only, this researcher used a controlled experimental method using two versions of a Web presentation, one with black text only (BTO) on a grey background (the default color of Web browsers) and the other one containing the use of enhancing visual elements (EVE). The subject of both versions was the same, selling insurance. The textural content of both was meant to be the same. The EVE version had two sections of an actual larger Web presentation which this researcher had been designing for a client, independently to this study. The (BTO) version was created for this study. It was designed with the text material from the (EVE) version. The part of the original Web presentation that was used in these two versions was the home page (i.e. a kind of introductory/directory page in Web presentation) and a page containing information about the insurance agency.

The design of both pages of the (EVE) site used a mildly textured blue background. The type was formatted in different font sizes in various combinations of two fonts, Times and Helvetica. Dark blue, blue, black and red were used for the font colors. In the first page of the (EVE) site, the type was designed around a graphic line drawing of the gable end of a house. The body of this house was divided into four sections. Each section contained the word describing one area of insurance that the local company offers (ie. home, auto, life, business). This version contained a navigation link to the second page. The second page (EVE) contained a list of 7 statements, made from the consumers’ point of view, about what they want in an insurance agency. Each statement was precluded by a cartoon picture that illustrated the statement content. The (BTO) version was all on one page. There was one horizontal line dividing the first section from the second section. All of the information in the (BTO) version was aligned flush left. The type was black. The background was white. With the exception of the name of the local agency which was in all capital letters, no attempt was made to use any text formatting that could be judged to be visually enhancing.

There were three experimental groups, one viewing the BTO version, one viewing the EVE version, and one viewing both. Participants were randomly assigned to one of these groups except for one participant who arrived right after the group was assigned. This person was conveniently placed in a group. Memory retention, clarification of message and immediacy of understanding were tested with a twelve question fill-in-the-blank quiz, each variable having four questions, taken by the participants immediately after viewing the Web Presentation. The four questions used to test memory retention were:

This set of questions was tested for reliability using Cronbach’s alpha and produced an alpha of .55.

The four questions used to test clarification of message were:

This set of questions was tested for reliability using Cronbach’s alpha and produced an alpha of .85.

The four questions used to test immediacy of understanding were:

This set of questions was tested for reliability using Cronbach’s alpha and produced an alpha of .68.

A second quiz, with just the four memory retention questions, was offered to the participants in the BTO and EVE groups one week later. The BTO/EVE group were not included in this second quiz because the researcher thought their results would have been corrupted through viewing both sites.

Interest, emotional bonding, and aesthetic satisfaction were tested with a Likert- scale- type questionnaire. The questions each had five attitude scale items to choose from, such as really enjoyed it to really didn’t enjoy it. To avoid the possibility of bias occurring in the participants responses to the questionnaire due to repeated exposure to the same range of answers, these items were worded slightly differently in each question. Interest, emotional bonding, and aesthetic satisfaction were tested with one question each. The question for interest was: Was this Web presentation interesting to visit? The question for emotional bonding was: How much did you enjoy visiting this Web presentation? The question for interest was: Was this Web presentation interesting to visit? The question for aesthetic satisfaction was: This Web site was pleasing to look at.

PROCEDURES

The participants were students in a graduate communication class at the University of Arkansas. These students were asked if they would like to volunteer to participate in a research study. Volunteers were given an informed-consent form to fill out and promised anonymity. In an effort to avoid influencing the participants’ answers by giving them prior knowledge about the experiment, the only thing the participants knew about the research study was they would be viewing a Web presentation, answering questions, and taking a quiz.

The experiment occurred in a computer lab on the University of Arkansas campus. In the lab 16 participants were randomly assigned to one of three groups, either the BTO group or the EVE group, or the BTO/EVE group by asking them to number off from one to three repeatedly until all participants had a number of either one, two, or three. One’s were assigned to the BTO group, two’s to the EVE group, and three’s to the BTO/EVE group. Slips of paper, each with one BTO, EVE, or BTO/EVE Web address typed on it and some with further instructions, were folded in half and stapled shut. The EVE and BTO/EVE address papers had instructions telling the participant to click on the words Ingram Insurance Inc. when they were through looking at the first page. Each participant was given one of these pieces of paper according to the group he or she was in. One more student arrived right after this procedure and was conveniently assigned to group two.

All of the participants viewed the Web presentations at the same time in the same room. Only minimal instructions were given concerning how to use the Web browser. Participants were asked not to discuss the Web presentations with each other and to turn off the presentations when they were done. As soon as they turned off the presentations they took the quiz and filled out the questionnaire. Participants were asked to take the same quiz one week later. Only those who were in the BTO and EVE groups were asked to take Quiz 2. Only one from both groups was absent from Quiz 2.

RESULTS

The mean scores of the first quiz and of the questionnaire taken by the participants immediately after viewing the Web sites show the groups who viewed a Web site designed with enhancing visual elements (EVE) had higher mean scores on all items except for the item of memory retention. Participants who viewed the Web site designed with black text only (BTO) seemed to remember more immediately after viewing the Web site than those viewing the EVE or BTO/EVE versions. For the variables of interest, emotional bonding, and aesthetic satisfaction, those in the BTO/EVE group had second highest mean scores.

Memory retention over time was tested in the second quiz given six days later to participants of the BTO and EVE groups. This time the EVE mean score was higher (.600) compared to BTO (.500). Mean scores are shown in Table 1.

To test H2 (Enhancing visual elements used in a Web presentation will help the Web user develop more interest in the site) an analysis of variance (ANOVA) test was used. No significant difference was found between groups. For H3 (Enhancing visual elements used in a Web presentation will help the Web user develop more emotional bonding with the message presented) an analysis of variance (ANOVA) test was used. No significant difference was found between groups for this variable either.

H4 (Enhancing visual elements used in a Web presentation will, related to the Web presentation, increase aesthetic satisfaction in the Web user) was tested with an analysis of variance (ANOVA). Aesthetic satisfaction was significant at the .05 level: f(2,22) = 4.243, p<.03). The mean score for BTO was 1.2 and the mean for EVE was 3.0. In the Bonferroni post hoc test there was significant difference found between groups: p<.03. H5 (Enhancing visual elements used in a Web presentation will help clarify the message to the Web user) was tested with an analysis of variance (ANOVA) test. There was no significant difference found between groups. H6 (Enhancing visual elements used in a Web presentation will help the Web user understand the presentation more immediately)was tested with an analysis of variance (ANOVA). Immediacy of understanding was significant at the .05 level: f(2,16) = 7.515, p<.01). The mean score for BTO was 1.0 and the mean for EVE was 1.3. In the Bonferroni post hoc test there was significant difference found between groups: p<.01.

H1 (Enhancing visual elements used in a Web presentation will help the Web user with longer-term retention of the Web message) was tested for the BTO and EVE groups with a paired samples T-Test comparing memory retention questions of Quiz 1 with memory retention questions of Quiz2, taken six days later. Significance was found at the .05 level: (t = 3.411, p<.01). Reviewing the mean scores on these two quizzes shows those viewing the BTO version had higher memory retention immediately after viewing the Web site, but when both groups took Quiz2, those viewing the EVE Web site had more memory retention than those viewing the BTO. These results support H1, enhanced visual elements used in the design of a Web site will aid in long-term memory retention. The BTO/EVE group was not compared for retention over time because the researcher did not give that group the second quiz.

DISCUSSION

With this study I have investigated whether designing with enhancing visual elements in a Web presentation (site) will contribute to a greater understanding of the message being presented compared to using text only. Two versions of a Web site were used in this experiment. One version was designed with black text only (BTO) and the other with enhancing visual elements (EVE). Sixteen participants were randomly divided into one of three groups, and one late participant was conveniently placed in a group. One group viewed the BTO version. One group viewed the EVE version. The third group viewed both (BTO/EVE).

The first hypothesis, enhancing visual elements used in a Web presentation will help the Web user with longer-term retention of the Web message is supported in this study with the EVE group remembering more information than the BTO group on the second quiz. This finding is consistent with Randolph’s (1996) Web study. He found that the use of graphics increases retention of information over time. His study did not find any significance for use of graphics to help immediate recall. This result is similar to the findings in this study in which mean scores for immediate memory retention were higher for the BTO group than the EVE group.

No significant statistical support was found for the second hypothesis, that enhancing visual elements used in a Web presentation will help the Web user develop more interest in the site. Neither was any found for the third hypothesis, enhancing visual elements used in a Web presentation will help the Web user develop more emotional bonding with the message presented. However, the EVE group participants did score highest for these variables with the BTO/EVE scores second. These findings suggest that if the sample was larger these hypotheses might have shown significance at the .05 level.

The fourth hypothesis was that enhancing visual elements used in a Web presentation will increase aesthetic satisfaction in the Web user. There was a significant difference between the responses of the BTO group and the EVE group for the variable of aesthetic satisfaction. The mean score for the EVE group was highest, the BTO/EVE score second, both higher than the BTO group. This hypothesis was supported.

No support was found for the fifth hypothesis, that enhancing visual elements used in a Web presentation will help clarify the message to the Web user. However, as with H2 and H3, the mean score for the groups was highest with EVE and BTO/EVE. In fact they were tied. This also suggests a larger sample might have resulted in a significant difference.

The sixth hypothesis, enhancing visual elements used in a Web presentation will help the Web user understand the presentation more immediately, was supported. The test showed a significant difference between how immediately the BTO and the EVE group participants understood the Web message. The mean scores for these groups indicate that the EVE participants more immediately understood the Web message.

It is interesting to observe that the BTO/Eve group mean score was lowest on immediate memory retention and on immediacy of understanding. They also tied for second with the EVE group on clarification of message. In other words, the people in the BTO/EVE group retained less immediate understanding of the Web sites they viewed; didn’t perceive the meaning of the Web sites very quickly at all; and apparently the message of the sites was not very clear to them. The reason this is worth noting is these participants had the opportunity to see two sites conveying the same information. Their BTO encounter was supplemented with the EVE version.

As the researcher, my first response to the low scores on the first quiz was to decide to not include the BTO/EVE group in the second quiz. I felt at the time the answers to the first quiz were corrupted by the fact that the participants of the BTO/EVE group were asked to view both sites. Actually the point of having the third group was to see how they would differ from the other two groups through having them view both sites. It would have been better to have them take the second quiz also.

So why did they score so poorly? Here are some possibilities. Perhaps they were not given enough time. They were given their addresses last of the three groups, not for any particular reason. There wasn’t much time lag between groups receiving their addresses, but even a little bit of difference could have mattered since they had twice as many sites to look up. Maybe they were confused by the difference in the design of the two sites. They might have spent most of their time comparing the two sites rather than noticing the message content of the sites. This might suggest to Web designers to be careful and keep a regular integration of enhanced visual elements and text and not switch back and forth between main areas of only text and other areas of enhanced visual elements.

It is assumed the effective use of visual elements in Web presentations is governed by careful sensitive visual design choices. Just arbitrarily adding visual elements into a Web presentation is not being suggested in this study. Visual elements chosen for the Presentation should be carefully subordinated to the Web presentation message (Berghel & Willis, work in progress, A priori guidelines for Web design ). Pictures will probably work most successfully in concert with text information. It is likely few images can work as stand-alone presenters of meaning in a Web site. This concept is echoed in television research by Basil (1994) when he indicates visual cues aid learning when they contribute to the audio information and there are enough mental resources left from processing the audio information. (Basil supposes written and audio information are processed similarly.)

With this study I found support for the hypotheses that enhancing visual elements used in a Web presentation will, related to the Web presentation, increase aesthetic satisfaction in the Web user, enhancing visual elements used in a Web presentation will help the Web user with longer-term retention of the Web message, and enhancing visual elements used in a Web presentation will help the Web user understand the presentation more immediately. Overall the mean scores indicate some support for H2, H3, and H5 even though no statistical significance was found for these hypotheses. The scores were higher for interest, emotional bonding, and clarify the message for the EVE groups than for the BTO groups. These higher mean scores for the EVE groups suggest further studies could be done with a larger sample of participants. It seems likely more of the hypotheses presented in this current study could be supported.

These results can provide useful information to Web message presenters and Web designers to aid them in preparing their Web presentations. Future research could study specifically how various enhancing visual elements effect the Web message (i.e. how does graphics effect the message; how does the inclusion of sound effect the Web message; how would the use of video files effect the Web message; how is navigation effected by enhancing visual elements).

Randolph’s study (1996) emphasized the communication value of graphics in Web design as opposed to thinking of graphics as mere decoration. The results of this study seem to support this conclusion as well only extending the conclusion to include a broader range of design elements than just graphics. These results echo the studies in traditional media effects which have also pointed out that visual elements are very important contributors to communicating messages through media. This current experiment along with Randolph’s study extends to Web design research the importance of carefully considering how enhanced visual elements can be designed into a Web presentation in order to more effectively communicate the desired Web message.

REFERENCES

Basil, M. D. (1994). Multiple resource theory 1. Application to television viewing. Communication Research, 21 (2), 177-207.

Berghel, H. L. & Willis, C. D. (work in progress). A priori guidelines for Web design.

Fiske, S. T., & Taylor, S. E. (1984). Social cognition. Reading, MA: Addison-Wesley.

Gamson, W. A. (1987). The 1987 distinguished lecture: A constructionist approach to mass media and public opinion. Symbolic Interaction, 11, (2), 161-174.

Graber, D. A. (1989). Content and meaning. What’ it all about? American Behavioral Scientist, 33, (2), 144-152.

Graber, D. A. (1990). Seeing is remembering: How visuals contribute to learning from television news. Journal of Communication, 40, 134-155.

Messaris, P. (1993). Visual “literacy”: A theoretical synthesis. Communication Theory, 3, (4), 277-294.

Randolph, G. B. (1998). Effect of hypermedia and graphics on recall and retention in on-line publishing. Communication Research Reports, 15, (1), 1998, 64-70.

Wicks, R. H. (1992). Schema theory and measurement in mass communication research: Theoretical and methodological issues in news processing. Communication Yearbook,15, 115-145.

Wicks, R. H. (1995). Remembering the news: Effects of medium and message discrepancy on news recall over time. Journalism and Mass Communication Quarterly, 72 (3), 666-681.