Intuitive Design

A common goal in interface design is to provide users with an effortless interaction experience. It is assumed that when interactions are intuitive they can be completed using few, if any, cognitive resources and they require very little instruction for successful use. Functionally speaking, the benefits of an intuitive interaction come about when the userís mental representation of the interaction is congruent with the actual use of the interaction. From a traditional cognitive perspective, an intuitive interaction is little more than a familiar interaction.


Nevertheless, a major challenge in this area of research lies in determining what features of a design will be familiar, or intuitive, to the user. Although it is possible to ask users to identify intuitive aspects of a design, this approach may not reveal the true sources of intuitive interactions. The cognitive processes supporting intuition occur automatically and are based on implicit representations of previous interactions. There are, therefore, many cases in which the user would not be able to accurately describe what makes an interaction intuitive. Still, Still, and Grgic (2015; Interacting with Computers) identified one such case.

This research line will continue in two related areas. One aim is to identify familiar elements in designs using techniques that are sensitive to familiarity and provide objective measures of familiarity. The second aim is to examine the extent to which measures of familiarity and intuition reflect common processes.

Presented Information near the subjective threshold of awareness

As computing devices become more pervasive, it is increasingly important to understand how to maintain calm and lightweight interactions. One approach is to minimize the amount of cognitive resources required to interpret information; this can be done by presenting information below the userís subjective threshold of awareness. This approach is being tested in a variety of ubiquitous computing environments including intelligent systems and automobile interfaces. My initial work in this area targeted the development of five practical recommendations for designers who want to present information below the threshold of awareness: 1) Support claims of awareness with divergent results and multiple tasks, 2) Measure awareness immediately, 3) Use several stimulus intensities, 4) Determine the importance of awareness by examining outcomes, 5) Examine resource requirements (Still & Still, accepted with revision; IJHCI). In continuing this research, I plan to leverage my previous experience investigating repetition blindness, masked orthographic priming, and semantic priming; all of which utilize stimuli presented near the threshold of subjective awareness. In addition to using traditional behavioral measures like response time and error rate, I have used and plan to continue using non-invasive physiological measures like galvanic skin response and pupillometry (Geller, Still, & Morris, 2016; Memory & Cognition) to enhance this line of research.

Electronically-Mediated Communication

Electronically-mediated communications (e.g., email, instant messaging, text messaging, social network postings) are increasing exponentially. The pervasive use of mobile devices is driving this trend. According to FCC estimates in 2009, approximately 270 million individuals in the United States subscribed to wireless services. In addition, the accepted use of electronically-mediated communication has gone beyond that of personal communication, as it is now used in professional settings and in advertising. Therefore, it has become necessary for us to understand how these messages are processed.

 

TXTMSG Abbreviations

The question of how exactly abbreviations are processed is relatively new to word recognition research. Nevertheless, there appears to be sufficient evidence to suggest that abbreviations are processed somewhat like words. For instance, exposure to FBI produces similar ERPs to those produced by words (e.g., N400; Lazslo & Federmeier, 2007) and abbreviations can prime related words (e.g., ABSbrakes; Brysbaert, Speybroeck, & Vanderelst, 2009). These findings indicate that abbreviations access semantic information and may have their own entries in the mental lexicon. My research examines whether or not emotions are also accessed during abbreviation recognition. For example, does I H8IT elicit an emotional response similar to one produced by I hate it? (Materials used in this line of research will be available here at a later date) In a series of experiments I have found that “emotional” acronyms and abbreviations do not elicit strong emotional responses. The implication of this finding is that abbreviated forms of communication may not convey their intended meaning.

 

Still, M. L., Morris, A. L., & Jones, K. (2011, November). IDK if you realy H8IT unless you spell it out. Poster presented at the Annual Meeting of the Psychonomic Society, Seattle, WA. Abstract

 

Still, M.L. & Morris, A.L. (2012, November). Alternative methods for measuring the frequency of TXTMSG Abbreviations. Poster presented at the Annual Meeting of the Psychonomic Society, Minneapolis, MN. Abstract

 

Emoticons

If acronyms and abbreviations fail to evoke emotion, perhaps emoticons do. Another aspect of my research is designed to investigate this possibility. Some researchers have suggested that emoticons serve as nonverbal cues relating the sender‘s emotional state to the receiver. Dresner and Herring (2010) suggest that emoticons can be more than general indicators of emotion; they can be part of the speech act providing information to the receiver that can help clarify the intent of the message. I have two primary questions regarding emoticons. First, do they elicit emotion in the receiver? Second, do emoticons provide non-emotional information? If so, what kind of information do they provide?

Practice Orientated Research

Over the last three years I worked with industry partners, which resulted in use-inspired basic research. For example, I collaborated with McAfee (now Intel Security) and two Human Factors and one Cognitive colleague at San Jose State University. The goal of the project was to find ways to improve cybersecurity by better conveying the risk associated with downloading an application. We manipulated the type of app being downloaded, the risk associated with the app (high, medium, and low), the type of risk associated with the application, and whether the app rating seemed to be coming from another user or seemed to be generated by a computer program. Completing the project required coordination between two laboratories and coordination with McAfee. Our general finding was that users were more likely to trust security ratings that appeared to be generated by a program over those provided by another user (Schuster, Still, Still, Lim, Feria, & Rohrer, 2015; Human Aspects of Information Security, Privacy, & Trust).