Design Method of Desktop Widget based on Users' Information Perception

Design Method of Desktop Widget based on Users' Information Perception

RESEARCH QUESTION 

Desktop Widgets were created to provide useful information to users without overly distracting them from their main task.  However, absorbing such peripheral information can potentially have a significantly negative impact on a user's performance of their main task.  This study explores methods on how to best design Desktop Widgets so that they can convey as much information as possible while having a minimal effect on a user's performance of their main task.

Figure 1. An example of a typical desktop widget. 

Source: http://www.flickr.com/photos/yahoo_presse/3192543894

METHODOLOGY

In order to appropriately design a Desktop Widget, it is necessary to determine the kind of cognitive processes users undergo when they absorb peripheral information while performing their main task.  Experimental methodologies from the field of cognitive psychology were used in the following two experiments. 

1. Test subjects performed a main task that required them to mentally carry out mathematical addition while simultaneously drawing upon their short-term memory.  At the same time, the amount of peripheral information they absorbed from a widget from both text and images respectively was measured in terms of two indices:depth of information and presence or absence of keywords, which also served as an indirect indicator of a user's level of interest in the information presented. 

2. Multiple Resource Theory explains that people draw upon different cognitive resources when performing verbal and spatial tasks.  To verify the relation between the nature of tasks performed and user performance, test subjects were asked to perform 4 combinations of tasks.  Task performance was evaluated to see whether using both verbal and spatial cognitive resources instead of only one would improve test scores. 

EXPERIMENTAL METHODS AND RESULTS

The bulk of my work focused on the second experiment, and it is these results that are presented here. 

Figure 2. Combination of verbal and spatial tasks. 

Participants were asked to answer a series of questions according to the following combinations.  For instance, in combination A, participants were presented with main and secondary tasks of a verbal nature.  In this case, they would only have to draw on their verbal cognitive resources.  Different combinations were designed to engage one or both types of resources to determine its effect on task performance. 

Figure 3. A screenshot of one of the stages in the experiment. 

Subjects were presented with text on the left side of the screen (their main task, of a verbal nature), and 7 pictures were simultaneously displayed by the right margin (this spatially presented information constituted their secondary task).  This corresponds to Combination B as listed in Figure 2. 

Figure 4. Comparison of task performance. 

In Combinations A and C, only a single cognitive resource was engaged during task performance (verbal and spatial cognitive resources respectively).  However, in Combinations B and D, when the tasks were varied to engage both verbal and cognitive resources simultaneously, subjects showed an improvement in task performance. 

Figure 5. Task performance comparison according to the number of types of cognitive resources used. 

Mean scores for Combinations A and C were summed up to give an overall indicating of task performance when only one cognitive resource was engaged.  Summed scores for Combinations B and D give an indication of task performance when both verbal and spatial cognitive resources were used during the task.  The results show that using both verbal and spatial cognitive resources simultaneously gives rise to better task performance. 

The possible usage and implication of these results were then applied to a Google search result.

Figure 6. Google search result for the search entry "speaker" in the current format. 

The word ' speaker' was entered into the search engine and the result is shown on the left. Note that the ads (peripheral information) are currently displayed in a text format, in addition to the text results of the main search. 

Figure 7. New proposed format for the Google search result for the search entry "speaker". 

According to the research results, the Google search page could be revamped such that the ads (peripheral information) are presented in a pictorial form.  This would ensure that users can draw on different cognitive resources to process verbal (text) and spatial (image) information.  Users would be able to absorb the maximal amount of peripheral information without detracting from the performance of their main task, which in this case is the processing of the search result. 

CONCLUSION

In the first experiment, results show that users were only able to absorb peripheral information in which they were interested in, and that conveying complex information was better achieved by using text and not images.

More importantly, the results of the second experiment suggest that user performance is indeed significantly higher when tasks of a different nature are performed, thereby engaging both verbal and spatial cognitive resources, as compared to when tasks of a similar nature are performed.  This has important implications for desktop widget, search engine or website design.  To convey the maximal amount of peripheral or secondary information without decreasing a user's performance of his main task, as well as to maximize information retention from both the main and secondary task, peripheral information has to be presented in a way that is different from the main task, i.e. secondary information should be of a graphical (spatial) nature if one's main task is presented in a text format and vice versa. 

PUBLICATIONS

• Jaeyoon Jung, Yoshihiro Kawahara, Kazuhiko Yokosawa, Tohru Asami,  Design Method of Desktop Widget based on Users' Information Perception, Information Processing Society of Japan, 6z-2, March 2010 (pdf in Japanese)

• Jaeyoon Jung, Yoshihiro Kawahara, Kazuhiko Yokosawa, Tohru Asami, The Impact of Desktop Widget on Task Performance, The Institute of Electronics, Information and Communication Engineers, B-15-7, September 2009 (pdf in Japanese)