The phenomenon of banner blindness explains that users can mentally ignore online advertisements (ads). However, eye-tracking studies have shown that users still fixate on ads, and even without direct gaze, ads still fall within a user's peripheral vision, which may negatively overload cognition. It is therefore unknown how blind, banner blindness, truly is, and what other effect ads may have on user's information seeking. To address this gap, a withinsubjects design experiment was conducted with 37 participants who performed search tasks from the TREC 2017 Common Core News Collection, where 3 search tasks contained various types of ads, and one search task had no ads. Although our results showed that on average, participants retrieved similar amounts of relevant documents regardless of whether ads were present or absent, participants took significantly longer achieving this performance when ads were present. Furthermore, when ads were absent, participants reported less frustration, and not only believed they learned more, but a post-task recall test showed that participants actually did learn up to 38% more. Consequently, our findings suggest that banner blindness is more costly than just mere annoyance, and that the influence of ads on user's information retrieval recall may extend current theories of visual crowding. CCS CONCEPTS• Information systems → Users and interactive retrieval; • Human-centered computing → HCI design and evaluation methods.

No abstract

No abstract

When too much visual stimuli is present, the phenomenon of clutter is known to degrade an individual's perception across a variety of domains, ranging from completing search tasks incorrectly, to decreasing reading speed when letters are too close together. However, research is lacking as to whether the negative effects of clutter impact learning when too many words are visible at any one given time. Furthermore, colour has been implicated in affecting clutter. Thus, the present study created a recognition experiment whereby 42 participants had to learn target words that were presented in black or red font and positioned amongst no clutter, clutter words (distractor words surrounded the target), and clutter non-words (sequence of random letters surrounded the target). Results found that words learned in isolation were identified faster and significantly more accurately than words learned in both forms of clutter. Although red target words did not eliminate the negative effects of clutter, red words did show a trend towards higher accuracy of recognition compared to black words. These results would appear to be explained by existing clutter theories that state the limits of attentional resources and short-term memory cannot process excess visual stimuli. These findings have real-world implications for establishing optimal reading formats to improve learning.

Perceptual Speed (PS) is a cognitive ability that is known to affect multiple factors in Information Retrieval (IR) such as a user's search performance and subjective experience. However PS tests are difficult to administer which limits the design of user-adaptive systems that can automatically infer PS to appropriately accommodate low PS users. Consequently, this paper evaluated whether PS can be automatically classified from search behaviour using several machine learning models trained on features extracted from TREC Common Core search task logs. Our results are encouraging: given a user's interactions from one query, a Decision Tree was able to predict a user's PS as low or high with 86% accuracy. Additionally, we identified different behavioural components for specific PS tests, implying that each PS test measures different aspects of a person's cognitive ability. These findings motivate further work for how best to design search systems that can adapt to individual differences. CCS CONCEPTS • Human-centered computing → Human computer interaction (HCI); • Computing methodologies → Machine learning; • Information systems → Users and interactive retrieval.