The influence of the meaning of pictures and words on Web page recognition performance

Firms spend high sums trying to make their "home" page as memorable as possible to attract repeat visits. For this purpose, fancy pictures and words are used to catch the attention of visitors. Interestingly, the effectiveness of all of this effort is nearly completely unknown. This study investigated how picture and word selections affected the recognition success rates of the sites visited by subjects. The clear finding in cognitive psychology is that meaningful pictures (Bower et al., 1975) and words (Wanner, 1968) are remembered with much higher performance than non-meaningful ones. In corporate Web sites, therefore, attention can be paid to the meaningfulness of both pictures and words in attempting to make them more memorable. Building on previous research by cognitive researchers that suggests higher memory retention for meaningful pictures and words versus meaningless ones, we hypothesized that Web pages where the salient picture and words are both business-meaning related (BMBM) will be much easier to remember than those with only dictionary meaning or no meaning at all, and that those with dictionary meaning (DMDM) will be easier to remember than those with no meaning. An experiment with 41 participants was performed to test these hypotheses. All possible combinations of pictures and text with no meaning, dictionary meaning, and business meaning form a 3 × 3 matrix. Support was found for two of the three hypotheses; the recognition success rate of the combination BMBM was statistically much better than any of the other groups (with the exception of the group DMDM for which it was about the same). Rather than appealing to fancy pictures and words, this study suggests that companies should aim at having consistent business-related meaningful salient pictures and words in their main Web pages. Future research should manipulate the time delay between stimulus and measurement (e.g.; hours, days, etc.) perhaps to reduce the mean scores and increase variability, as well as to investigate interactions of the meaningfulness of the salient elements with other graphics and page elements.