Eye-tracking Construal Levels: How Temporal Distance Affects Attention

Introduction We all make plans for the future, perhaps to join in a local festival next week, or participate in an online course starting next month. However, we don’t necessarily stick to the future plans. People’s intentions change with time—we can all recount times when we plan to do something, but then decide against it when the time comes. This presents challenges when designing systems that support future behaviors, such as event-based social network systems [e.g. 9], personal tracking devices with goal-setting features, [e.g. 10] and calendar systems [e.g. 11].

A recent study argues that intertemporal change of intentions could be systematic rather than random, and explains the underlying mechanism that makes people change their behavioral intention over time [10]. Drawing upon Construal Level Theory (CLT), which posits that people’s construal (a mental representation, what is salient in regards to an object or event) differs by the temporal distance toward a behavior [11], the study finds that when people make a plan for an event in the far future, the extent that the behavior outcome is expected as desirable (positive attitude toward the behavior) is important to the behavioral intention to perform the behavior; however, it becomes less salient as planned date approaches [10]. Thus, the study argues that when an event is a month ahead (e.g., deciding whether or not to attend a yoga class) people sign up for the event focusing on why to attend the

Minhyang (Mia) Suh Human Centered Design & Engineering, University of Washington Seattle, WA 98105, USA miasuh@uw.edu

event (e.g., “yoga makes me relaxed”—a high-level construal); however, many people drop out a few days before the event because their focus changes to how to perform the behavior (e.g., “The class time is not convenient for my schedule”—a low-level construal) [10].

The study demonstrates a need for tailored strategies in regards to the temporal distance from a target event [10]. For example, a designer should focus on supporting the why of the event (e.g., emphasizing importance of performing the behavior) when one aims to encourage people to RSVP for a distal event; however, the focus should be changed to support the how of the event, presenting it as more feasible (e.g., highlighting the ease of performing the behavior) when the event date approaches. On the other hand, if one aims to predict the attendance of an event more accurately, the focus on the how of the event may be more useful because it may discourage people who expect that their attendance is not actually feasible in the future (e.g., reminding them of their conflicted schedule).

[10] assumes that people’s attention is a part of an intertemporal change of intentions toward a behavior, although the findings rely on self-reported ‘perceived’ construal, rather than attention itself, which is a behavioral attribute. Other researchers have noted the limitations of self-reported data to explain behaviors: One may not be honest when answering a question; or one may lack the introspective ability to provide an accurate response to a question [5]. Moreover, people may have different understandings on the intervals or ratings in the questions [9]. These concerns may reduce reliability of the theoretical premise of CLT and hamper its broader applicability in practical settings. A

question arises: aside from perception, do people also differently assign their attention on visual information according to the temporal distance from the target behavior?

I used the eye-tracking methodology to understand how attention plays a role in temporal change of intentions to perform a behavior. Eye-tracking measures an individual’s eye-movements, allowing us to understand where and how one is looking, and to identify important factors in visual-based information processing [3]. Prior work often regards the fixation duration as the indicator of attention people award toward an object [2], and this is related to their attitudes or behavior [9]. The longer the fixation time that people place on an object, the more likely it is that they prefer the object compared to others [9], and will recall it better [5]. This suggests that beyond construals, if people also have different attention according to the temporal distance toward a behavior, we should also be able to observe the different fixation time using eye-tracking. We hypothesize:

H1. When people read about an event in the far future, they will have greater fixation duration on the ‘WHY’ aspects of the event, compared to an event in the near future.

Methods I conducted a within-subject lab experimental study with an eye-tracker (Tobii Pro) while introducing to participants two different social events in the local area (one planned for a week later, the other a month later). The stimuli was modelled upon the event-based application [12], consisting of the event title, time, location, and descriptions of why and how to participate in the event. Each participant (N = 13; 8 were female, m age = 21.6, range = 18 - 30 years old) read both

stimuli, and I measured how the participants’ visual attentions (fixation duration and frequency) differed by two events at different temporal distances.

Preliminary Result To analyze the data, I conducted a paired t-test and found that H1 is supported. The fixation duration between the near future (M=11.22) and the far future (M=15.63) was significantly different (t(13)=-2.38, p<.05) Figures 1 and 2 are heat maps of the result. This finding indicates that people’s attention on visual information is differently assigned depending on the temporal distance from the target behavior.

Discussion This study demonstrates an intertemporal predisposition on visual attentions, differing by the temporal distance toward an object. When considering social events in the distal future, people place attention on ‘why’ to join the event, however, this attention significantly reduces if the same event is planned in the near future. It confirms CLT’s theoretical premise by showing the same effect at the attention level, objectively measured by eye-movements. It also suggests that practitioners should note the differently assigned attention with time, and use this knowledge when creating persuasive strategies for interface design or messages.

Future Research & Conclusion Building on this study, I plan to investigate how to effectively utilize different attentions at different times in system designs. Prior works suggest that when the temporal distance from an object is matched with arguments inducing a certain level of construals, people are more likely to be persuaded [4,6,8]. Researchers speculate that it is because people may feel ‘right’ [1,7], however, this has not been explored in the context of

event-based social network applications. Furthermore, no prior work has examined ‘the matching effect’ at the levels of attention. Thus, I plan another eye-tracking study focusing on the matching strategy of construals. While continuing my research endeavors, I would appreciate feedback from other graduate students at InfoSocial 2017, and I would like to share my experience of using eye-tracking methodology to assist others and to strengthen and extend a theory. References 1. Jennifer L Aaker and Angela Y Lee. 2006.

Understanding regulatory fit. Journal of Marketing Research 43, 1: 15–19.

2. Sophie C Boerman, Eva A Van Reijmersdal, and Peter C Neijens. 2015. Using eye tracking to understand the effects of brand placement disclosure types in television programs. Journal of Advertising 44, 3: 196–207.

3. Paul Dourish. 2014. Ways of Knowing in HCI. https://doi.org/10.1007/978-1-4939-0378-8

4. José Mauro da Costa Hernandez, Scott A. Wright, and Filipe Ferminiano Rodrigues. 2014. Attributes Versus Benefits: The Role of Construal Levels and Appeal Type on the Persuasiveness of Marketing Messages. Journal of Advertising 44, 3: 243–253. https://doi.org/10.1080/00913367.2014.967425

5. R J Jacob and Keith S Karn. 2003. Eye tracking in human-computer interaction and usability research: Ready to deliver the promises. Mind 2, 3: 4.

6. EunHa Jeong and SooCheong C. Shawn Jang. 2015. Healthy menu promotions: A match between dining value and temporal distance. International Journal of Hospitality Management 45: 1–13. https://doi.org/10.1016/j.ijhm.2014.11.001

7. Sherri Jean Katz and Sahara Byrne. 2013. Construal Level Theory of Mobile Persuasion. Media Psychology 16, January: 245–271. https://doi.org/10.1080/15213269.2013.798853

8. Hakkyun Kim, Akshay R Rao, and Angela Y Lee. 2009. It’s time to vote: The effect of matching message orientation and temporal frame on

Heat Maps of the Result

Figure 1. Aggregated Fixation Duration for the event in the Far Future

Figure 2. Aggregated Fixation Duration for the event in the Near Future: When the event is planned in the near future, aggregated fixation duration on description of ‘Why to join the event’ is significantly smaller compared to those of in the far future (Figure 1.)

political persuasion. Journal of Consumer Research 35, 6: 877–889.

9. Franziska Marquart, Jörg Matthes, and Elisabeth Rapp. 2016. Selective Exposure in the Context of Political Advertising: A Behavioral Approach Using Eye-Tracking Methodology. International Journal of Communication 10: 20.

10. Minhyang Mia Suh and Gary Hsieh. 2016. Designing for Future Behaviors: Understanding the Effect of Temporal Distance on Planned Behaviors. In Proceedings of the 2016 CHI Conference on Human Factors in Computing Systems, 1084–1096.

11. Yaacov Trope and Nira Liberman. 2003. Temporal construal. Psychological review 110, 3: 403–421. https://doi.org/10.1037/0033-295X.110.3.403

12. Meeup. Retrieved from meetup.com 13. Fitbit. 14. Google Calendar.