Influence of Familiarity towards Assessment of Website Interface
Bachelor's Thesis ▫️Role: Sole researcher▫️Project language: German▫️Methods: A/B testing
Bachelor's Thesis ▫️Role: Sole researcher▫️Project language: German▫️Methods: A/B testing
If users are familiar with the text-based content on a website, does this positively influence their evaluation of the user interface? This Bachelor’s thesis examined the effect of familiarity regarding a website’s text-based content on evaluations of aesthetics, fluency (ease of processing), and usability.
Interfaces with familiar text-based content would receive higher ratings for aesthetics, fluency, and usability.
The topic “Classical Music” was selected. Therefore, the study used examples from the field of classical music, such as classical composers and musical instruments.
Two website interfaces with the same layout were developed. The only difference was the content: one with familiar or more known composers and instruments and one contains unfamiliar or less known composers and instruments. This familiarity was decided after pre-study, where participants evaluated their familiarity with the chosen names of composers and musical instruments.
Study design: Between-subjects design with two groups (website with familiar and unfamiliar contents), 50 participants each group (100 total)
Download song lyrics
Play sound sample of a musical instrument
using UEQ+ questionnaires
Aesthetics
Expected usability*
Perceived usability
Fluency
*Expected usability: Measures participants' usability evaluation from the landing page only, before interacting with the actual website
t-tests for independent samples
Correlations
Mean ratings between familiar and unfamiliar cohorts in terms of aesthetics, easiness of tasks, expected usability, fluency, and perceived usability
Mean ratings between familiar and unfamiliar cohorts in terms of fluency and trust
Investigating the effect of brand recognition on e.g. e-commerce websites could be interesting, whether those featuring well-known brands or products are perceived as more trustworthy than featuring unknown ones, even with the same layout.
It might be interesting to investigate the duration of the task completion to see if familiarity has an impact on this.
Gorilla.sc was not well-suited for some browsers or operating systems because it did not record completely on some of them. The use of LimeSurvey could be better for remote studies.
More tasks could be added in the study, so that participants could spend more time on the website.