Beautification

Overall, visual inspection of Table 23 and Table 24 suggested that type of quality changes made were mostly the addition of elements, and rarely deleted items; and the changes (total minus quality changes) made were mostly relocation and addition of elements, followed by change of text in labels and change of element type. The deletion of elements was mostly considered as extra changes and only a few were of quality (as reflected in Table 23 and Table 24). However, interestingly, the proportion of addition of elements was relatively higher in the high formality design, which also reflects the low number of other types of changes made. In other words, it was likely that subjects would add elements to a design with high formality than making other types of changes.

By counterbalancing the order of presentation of the five conditions (presentation of designs with different levels of formality – refer to Figure 1 in the methods section), order effects was controlled for, and it was expected that overall, order effects would not influence results.

Nonetheless, although order effects was controlled for, according to Heiman (2000), order effects may reduce internal validity; therefore this was also examined as it may have played a role in producing a (weaker/stronger) trend of changes made across levels of formality. However, as there was only a small sample (N = 30), and that it was not the purpose of the study to test for order effect, no statistical tests were performed to search for significant order effects. Instead, graphs were used for visual inspection. Data were grouped and examined according to the order of conditions presented (e.g. 54321, 12345, 43215, 23451…etc), one multi-line graph for total changes (Figure 24a); quality changes (Figure 25a); and expected changes (Figure 26a). Visual inspection of the Figure 24a, 25a and 26a suggested that generally, although with different orders of conditions presented, there was still a linear trend found in the number of total, quality and expected changes made across levels of formality – also identified in the trends test in one-way repeated measures ANOVA. There was little to some order effect – this was more noticeable in orders 54321 and 12345. In order 54321, added with the practice effect over the five conditions, such order levered the main effect of formality – less changes made in the first condition (high formality), but as formality decreased, the number of changes increased – also seemed to be affected by practice effect. However, order 12345 worked against the main trend – in other words, with practice effect present, and at its strongest during the fifth condition (high formality), the main effect of formality on the number of changes made was weakened by the opposing force – practice effect. Figure 24b, 25b and 26b illustrates the contrast between the curves of orders 54321 and 12345.

The weaker or stronger trends shown in Figures 24a, 24b, 25a, 25b, 26a and 26b, must be interpreted with extreme caution as there were only four subjects (n=4) in each order group. Overall, as expected, order effect did not influence the results significantly as linear trends were present regardless of the order of presentation of conditions, which further highlights the general significant linear trend.

Figure 24a. Mean total changes made across levels of formality according to the order of conditions presented.

Figure 24b. Mean total changes made across levels of formality according to the presentation of conditions in the 54321 (n=4) and 12345 directions (n=4).

Figure 25a. Mean quality changes made across levels of formality according to the order of conditions presented.

Figure 25b. Mean quality changes made across levels of formality according to the presentation of conditions in the 54321 (n=4) and 12345 (n=4) directions.

Figure 26a. Mean expected changes made across levels of formality according to the order of conditions presented.

Figure 26b. Mean expected changes made across levels of formality according to the presentation of conditions in the 54321 (n=4) and 12345 (n=4) directions.

Subjects’ rankings of the five designs with different levels of formality were examined (see Appendix P). “Overall enjoyment” was defined as the subjects’ ranking of the five designs in the order from the “most-liked” design with a rank of 1 (i.e. the design they enjoyed working on the most in comparison to other the other designs presented) to the “least-liked” design with a rank of 5 (i.e. the design they enjoyed working on the least in comparison to the other designs presented). Figure 27 shows the overall ranking, from the most liked (lower scores) to the least liked (higher scores), when working on each design in comparison to other designs in terms of enjoyment.

Figure 27. A bar graph showing mean rank and standard deviation, in terms of preference, according to the overall enjoyment, in working on each design with a different level of formality.

Overall, the high formality design was ranked most frequently as the “most liked” design (by seventeen out of thirty, or 56.7% of participant). Most participants (twenty out of thirty, or 66.7%) ranked the low formality design presented on the Tablet PC as the least liked design. Interestingly, however, fourteen out of thirty (46.7%) participants ranked the low formality design presented on paper as the second least liked. In other words, when given low formality designs, participants liked (enjoyed) working on the design presented on paper more than working on the design presented on the Tablet PC. This was reflected in participants’ preference for design mediums in the experiment: thirteen out of thirty (43.3%) participants preferred using paper and pen, while fifteen out of thirty (50%) participants preferred using the Tablet PC, and two participants (6.6%) indicating no preference of one medium over another medium.