Text vs. Graphical Links in an E-Commerce Website

Conclusions

Impact for practitioners

Our findings showed that text links in an e-commerce environment are faster and produce less errors. On the other hand, the users subjective satisfaction survey tells us that the graphical and text links were easier to use than the the text only links. Designers of an e-commerce site might view this experiment and want to use text only links, but they must be careful. Yes, text only is faster for the user and produces less errors, however the majority of the users found the graphics + text links to be easier to use. An e-commerce website is a business and one of the golden rules of business is "The customer is always right". Therefore graphics + text links should be used in a site that is structured similarly to the one we tested. Differently structured sites would require further experimentation.

Suggestions for future researchers

This experiment is really a basis for future studies. We had limitations that did not allow us to expand the scope of this experiment as we would have liked. Future researchers should build on this experiment and explore the many issues that accompany it.

Our experiment did not explore the very important issue dealing with the number of links. It would be interesting to know what impact a more "crowded" web page would have on our results. Future researchers may find that when there are more links text only is a better option than text and graphics. Maybe the top level of the web site should be text only links and as we get deeper add graphics to the links.

We also did not take into account the load time for the graphics. There was no difference in load time for the pages with text only and those with graphics. As the world moves faster and faster toward high-speed internet this may not be an issue, however not all users have high speed connections today. The time it would take to load graphics on a 56K modem may be enough to change the subject satisfaction ratings that scored text only links below graphics+text links.

The graphics only links did not score well in any category. Future researchers may want to explore the thought of customers in which english is not their first language. Our study lumped all the people together and did not ask if english was their first language. If this study was run again and the results were compared within non-english speaking customers(english is not thier first language), the graphics only links would have done much better. But as a general rule based on our findings, graphics only links are not advisable.

Refine the theory or develop a new one

For the standard e-commerce customer, Text only links produce a much lower error rate and a faster traversal than the graphics only or the graphics + text links. However, users find the graphics + text links slightly easier to navigate with than the text only links, both of which provide much easier navigations than the graphics only links.

Other Suggestions

When using graphics only links in a page there should be a border or some indication of what the entire link is. We found that most subjects were confused by our graphical links and felt they had to click on a certain portion of the link. Also, graphical links need to be a little smaller than the ones we used. Most subjects felt the graphics were too big.

Department of Computer Science: Direct questions and comments to the student editorial team