You are here:

The icon response phenomena: Examining creativity and icon association errors in computer and Web-based research

, University of Pittsburgh, United States

University of Pittsburgh . Awarded


This study examined the relationship between creativity and icon association in terms of how icons appear on computer and web-based electronic reference databases. One hundred senior students from a high school were administered the Remote Associatives Test to determine their creativity level. The sample was stratified into three creativity levels (low, middle, and high). Later, students were given two Associative Icon Response Surveys (AIRS), two weeks apart. The first treatment (AIRS1) consisted of ten-isolated icon displays (icons without screen context). The second treatment (AIRS2) showed the same icons as AIRS1; however, the icons appeared in screen context. The subjects were instructed to associate with each icon for one minute.

After the treatments were given to the sample, (n = 45), the responses and means were calculated for each group. Both treatments were quantitatively and content analyzed. AIRS1 was designed to measure ideational fluency in regards to “ number” of icon associations a respondent can make for one minute. Whereas, AIRS2 examine analogical reasoning in reference to the “correctness” of icon associations. The “correctness” of icon associations was based upon the manufacturer's intended meaning.

The treatments' data was analyzed individually using a one-way ANOVA for both the quantitative and content analyses. The AIRS1 was analyzed quantitatively. The results showed that creativity significantly contributed (71.79%) to the students' ability to generate associations from isolated icon displays. Secondly, AIRS2 was content analyzed. The results showed that students' creativity significantly contributed (62.94%) to the arrival of the correct icon's meaning when observing icons within context. On average, both analyses' findings were corresponding to the students' creativity group.

As secondary analyses, the treatments' data were crossed using the other's primary analysis (i.e. AIRS1 x content analysis, AIRS2 x quantitative analysis). It was revealed that the absence or presence of screen context plays a major role in icon association.

Overall, the study provided groundwork for instructional designers of research databases aimed toward school students. From these findings, manufacturers may be able to improve their interface design to be more student-friendly. In addition, the results will alert librarians to individual differences contributing to their students' success when using on-line databases.


Ent, V.I. The icon response phenomena: Examining creativity and icon association errors in computer and Web-based research. Ph.D. thesis, University of Pittsburgh. Retrieved April 23, 2021 from .

This record was imported from ProQuest on October 23, 2013. [Original Record]

Citation reproduced with permission of ProQuest LLC.

For copies of dissertations and theses: (800) 521-0600/(734) 761-4700 or