You are on page 1of 6

Towards Stress-less User Interfaces: 10 Design Heuristics based on the Psychophysiology of Stress

Neema Moraveji Calming Technology Lab Stanford University 450 Serra Mall Stanford, CA 94305 USA neema@stanford.edu Charlton Soesanto Calming Technology Lab Stanford University 450 Serra Mall Stanford, CA 94305 USA csoesant@stanford.edu

Abstract
In this paper we identify 10 design heuristics used to minimize the number of stressors in an interface. They are meant to complement other interface design heuristics. The heuristics are based on characteristics of stressors drawn from decades of empirical study. These include both evolutionary (survival) and psychosocial sources of stressors. The list is followed by the results of an exploratory heuristic evaluation conducted on four different mobile Twitter clients. It shows how the heuristics complement Nielsens usability heuristics by enabling designers to evaluate and differentiate interfaces along two dimensions: usability and potential for stress. The results of the paper are useful for designers and scholars interested in systems that induce, or mitigate the risk of inducing, negative psychophysiological state change. Current methods of evaluating interfaces on this dimension require user interviews and physiological monitoring.

Author Keywords
Stress; heuristic evaluation; calming technology.
Copyright is held by the author/owner(s). CHI12, May 510, 2012, Austin, Texas, USA. ACM 978-1-4503-1016-1/12/05.

ACM Classification Keywords


H.5.2. [Information interfaces and presentation]: Evaluation/methodology.

Introduction and Motivation


Neither usability nor user experience encapsulate an important side effect of using many products: whether or not proper use of the product still induces stress. Stressors, as this paper shows, can exist despite, or in addition to, usability or user experience issues. E.g., a usable email client can enable one to process mail efficiently but induce stress because users now feel they must process massive amounts of mail. The stress derived from even effective use of a product can carry over and affect other areas of life. Stress is not purely an experiential descriptor. Stress (specifically, distress, as distinguished from eustress or good stress [14]) accumulates over time and can have an impact on cognitive, physiological, and affective state. Instead of using user experience as an organizing principle, this paper motivates the design and evaluation of interfaces using principles from research on stress given its long history of scholarly study [14] (reviewed in [7]) and demonstrated effects on the brain [12], long-term recall [13], cardiovascular system [6], and cognitive performance [6]. We draw upon the vast literature of stress to dissect interfaces into components that can be independently evaluated to identify likely stressors. The resulting heuristics may be useful to designers as they continue to design user interfaces.

prevention, NH6: recognition rather than recall, NH7: flexibility and efficiency of use, NH8: aesthetic and minimalist design, NH9: help users recognize, diagnose, and recover from errors, NH10: help and documentation. The heuristics help ensure designs meets basic user expectations and reduce erroneous user input, confusion, or mistakes. We have not seen in the literature a set of design heuristics meant to mitigate known stressors in particular. The sympathetic nervous system activates the stress response when faced with external or (more commonly) internal or self-imposed stressors; the parasympathetic response counterbalances it. It is the chronic triggering, or ineffective release, of the stress response that causes damage. Because stress can be so subjective, it seems daunting to discern which user interfaces may be stressinducing. However, studies have identified some patterns that help predict which elements will likely cause stress. These stressor characteristics (SCs) have been summarized in a review by Lupien, et al [7] and are presented below: SC1: Feels unpredictable, uncertain, or unfamiliar in an undesirable manner [7]. SC2: Evokes the perception of losing/lost control [4]. SC3: Has potential to cause harm or loss to ones self or associated objects, living things, or property [2]. SC4: Is perceived as judgment or social evaluative threat including threats to ones identity or selfesteem [1].

Related Work
Arguably the most widely used interface design heuristics are Nielsens heuristics [11] (NH). They are: NH1: visibility of system status, NH2: match between system and the real world, NH3: user control and freedom, NH4: consistency and standards, NH5: error

Heuristics
To devise heuristics based on stressor characteristics, we reviewed empirical studies where administrators induced or observed stress. We grouped the studies together and ranked them according to frequency and then created heuristics that counteracted them. The following 10 stress-less design heuristics aim to reduce the likelihood of an interface containing a known stressor based on the stressor characteristics above. They complement and sometimes overlap Nielsens; the goal is not to completely differentiate from those, but to focus particularly on evaluating stressors in user interfaces. The design heuristics mitigate known stressors and affect users physiological state, which then affect users state of mind. Each heuristic is listed with the stressor characteristic it mitigates. 1. Reveal Ability to Control Interruptions (SC1, SC2) Unpredictable interruptions compromise the control users have on their own focus. A branch of HCI deals primarily with addressing the problem of interruption [5]. This heuristic asks whether or not an interface reveals settings to block, control, or temporarily disable interruptions during slideshows, presentations, phone calls, and other important attention-heavy tasks. This may include the practice of letting the user specify, Dont show me this again, or Would you like reminders about this in the future? (see Figure 1) 2. Reduce Feelings of Being Overwhelmed (SC2, SC4) Large datasets are very common in applications that have many users and social elements or use datasets on the web. In these instances, the possibility of introducing stressors exist by making users feel that they cannot control the amount of information or that

they will never be finished using the application (see Figure 5a). Ones self-identity is threatened in these cases because a user may feel he or she isnt engaging enough with the application, isnt keeping up with other users, or has not added sufficient input. One can also feel overwhelmed from too many features or options.

Figure 1. TweetDeck, a desktop Twitter client which allows for simultaneous Twitter feeds, allows users to select what types of notifications to receive and from whom to receive them, minimizing unwanted interruptions.

3. Acknowledge Human Interpretations of Time Passing (SC1, SC2) The progress bar indicates system progress to the user from the systems perspective. However, humans do not perceive time purely linearly; time is experienced as slower the longer one waits (reviewed by Harrison, et al. [3]). This can induce stress by creating unpredictability or a lack of control over when one can use the system. One way to address this is to arrange longer processes to complete early on. Another technique is to distract users during waiting periods, with diverting stimuli.

Figure 2. The use of humor to acknowledge system error, mitigating stress that could arise from losing data or state.

4. Use Appropriate Tone and Emotion (SC1, SC4) Users engage computers with overlearned social behaviors [10] such as politeness and reciprocity. When this fact is not acknowledged by the system, social stressors can be introduced where users are surprised by inappropriate tone or emotion. To mitigate this, designers can introduce human tone and conversational emotion when appropriate. Some examples include apologetic or funny acknowledgments of system or user errors (see Figure 2). Polite requests (rather than demands) are another technique. 5. Provide Positive Feedback to User Input and Events (SC1, SC4) Negative feedback (e.g. for invalid input, unavailable features, etc.) can induce stress by threatening ones self-esteem or by triggering a feeling that users cannot provide the machine with the information it needs. Such feedback violates expected norms of conversational interaction with social agents [10], further inducing stress by being unpredictable. A current norm among application designers is to give feedback on operator errors, not successes. By simplifying tasks and acknowledging successes, user interfaces can build confidence and resilience to stress in their users (e.g. Thanks for filling out the form and You successfully updated the application). By pointing out common mistakes, users can realize they are part of a group with similar experiences and may not feel threatened (e.g. 74% of people had errors filling out this form - but not you! or 18 other members made the same simple mistake this hour no worries!). 6. Encourage Prosocial Interaction (SC4) Social stressors include instances where users imagine how they appear to others and that image violates their

desires or expectations [1]. Increasingly, applications with social components have greater potential for stress as users manage self-presentation. Simple ways of communicating prosocial interactions (e.g. Likes Retweets, and Mentions, +1) can help mitigate the effect of such stressors. Figure 3 shows an example of a Twitter application that reduces barriers to mentioning others on Twitter, spelling and capitalizing Twitter handles correctly. 7. Relieve Time Pressure (SC1, SC2, SC4) The manner in which time pressure induces stress is well documented [9] but has not yet been applied to interactive systems, where instances of time pressure are not infrequent. In contrast to heuristic 3, which addresses the experience of time while waiting for the system, this heuristic addresses the temporal context the user experiences during interaction. Users may feel a lack of control when they are pressured for time or even be worried about how they appear in a competitive sense if the time they took to complete a task feels too long. Unnecessary time pressures (see Figure 4) should be eliminated in interactive systems. 8. Choose Naturally Calming Elements (SC1) Involuntary attention or fascination at natural settings can cause shifts toward a more positively-toned emotional state, positive changes in physiological activity levels [15]. These beg the question of how to integrate natural elements into virtual environments and interactions (e.g. sounds, images, and animation rhythms). Examples of known practices include soothing error tones, naturalistic animations, and desktop wallpapers taken from the natural world.

Figure 3. The Mobile Twitter application allows users both to reduce recall and increase prosocial interaction by autocompleting Twitter handles using the proper case.

Figure 4. This dialog does not sufficiently acknowledge the Restart Later action by giving reason for why it is disabled. It also provides a countdown for an action that has no reasonable alternative. The resulting lack of control can induce stress.

9. Acknowledge Reasonable User Actions (SC1, SC2) On any screen or dialog, there exist a number of reasonable actions that users may want to take. Even if those actions are disallowed by the system, they should be acknowledged or guided in some way. When a user expects to take an action that is not available, the stress response can be triggered and grow if not addressed. E.g., in Figure 4, the system should at least acknowledge why the Restart Later button is disabled. 10. Demystify the Interface (SC1, SC3) Faced with myriad choices, unsure of what the results of ones actions will be, a user can feel stressed. Further, asking for help may threaten a users selfesteem. To complement Nielsens heuristic of Help and Documentation, this heuristic asks designers to evaluate whether or not they have demystified the features (see Figure 5) to the users before users ask for help or documentation.

Twittelator (top left) has unimplemented features (violating NH5) but also allows natural, soothing sounds (SH8) and an interruption preference section (SH1). NatsuLion (bottom left) did not follow the drag-down to refresh design paradigm (violating NH4), intermittently displayed authorization failed while loading new tweets (violating NH9). It also had no settings to control push notifications (violating SH1).

A Stress-less Heuristic Evaluation


This section contains the results of a brief heuristic evaluation of four different mobile Twitter clients using the stress-less heuristics (SHs) identified above and Nielsens usability heuristics. Through evaluation, each client was found to have features that placed it into a different quadrant based on a two dimensions: usability and potential for stressors (see Figure 6). By examining several features of the mobile Twitter clients, one can easily distinguish usable and stressless as important interface characteristics. SimplyTweet (top right) has a simple design (NH8) and humanizes user feedback options with prompts (SH4). It also groups tweets together to reduce noise (SH2).

Figure 5. (a, top) The user can feel overwhelmed by the number of new messages. (b, bottom) A playful means of demystifying features.

Figure 6. Heuristic evaluation using usability and stress-less heuristics produced a two-dimensional space of Twitter clients.

Twitterrific has a minimalist, simple design (NH8). However, as new Tweets are available (80 in our case), the user is brought to the least recent one. This feature gives users the sense that they must read through all the tweets before they are done and create a feeling of being overwhelmed (SH2).

This brief evaluation demonstrates that usability and stress do not always directly correlate and that stressless heuristics can catch seemingly small features that can cause an adverse stress response from the user.

[4] Henry, J. P., & Grim, C. E. (1990). Psychosocial mechanisms of primary hypertension. Journal of Hypertension, 8, 783793. [5] Iqbal, S., Horvitz, E. (2007). Disruption and Recovery of Computing Tasks: Field Study, Analysis, and Directions, Proc. of CHI 2007, San Jose, CA. [6] Lupien, S., et al. (2007). The effects of stress and stress hormones on Human Cognition: Implications for the field of brain and cognition. Brain & Cognition, 65(3). [7] Lupien, S.J., Maheu, F., et al (2007). The effects of stress and stress hormones on human cognition: Implications for the field of brain and cognition. Brain and Cognition. [8] Mason, J. W. (1968). A review of psychoendocrine research on the pituitaryadrenal cortical system. Psychosomatic Medicine, 30, 576 607. [9] Maule, J., Hockey, G. R. J. (1993). State, Stress, and Time Pressure. In Time pressure and stress in human judgment and decision-making. Plennum Press. [10] Nass, C., Steuer, J., Tauber, E. (1994). Computers are social actors. Proc. of ACM CHI, Boston, MA. [11] Nielsen, J. (1994b). Heuristic evaluation. In Nielsen, J., and Mack, R.L. (Eds.), Usability Inspection Methods, John Wiley & Sons, New York, NY [12] Rose, R. M. (1980). Endocrine responses to stressful psychological events. Psychiatric Clinics of North America, 3. [13] Sapolsky, R. (1996). Why Stress is Bad for Your Brain. Science. Vol. 273: 749-750. [14] Selye, H. (1975). Confusion and Controversy in the Stress Field. Journal of Human Stress, 1, Issue 244. [15] Ulrich, R., et al. (1991). Stress recovery during exposure to natural and urban environments. Journal of Environmental Psychology.

Conclusion
The ubiquity of user interfaces in modern society means the HCI community must acknowledge that even usable interfaces can have adverse effects on users. Thus, HCI researchers are no longer concerned solely with how efficient or even enjoyable the product is to use. Products can also cause stress. The heuristics in this paper provide designers and researchers with the ability to evaluate interface components on the basis of stress before conducting a user study.

Future Work
These heuristics are based on prior studies of stressor characteristics. However, they should be supported with physiological or at least self-report measures of stress taken during the use of different user interfaces before and after the heuristics are applied. The heuristics can then be expounded upon in greater depth, backed by multiple examples that guide their application.

References
[1] Dickerson, S., Margaret, K. (2004). Acute Stressors and Cortisol Responses: A Theoretical Integration and Synthesis of Laboratory Research. Psych. Bulletin. [2] Dienstbier, R. A. (1989). Arousal and physiological toughness: Implications for mental and physical health. Psychological Review, 96, 84 100. [3] Harrison, C., et al. (2007). Rethinking the Progress Bar. Proc. of UIST, Newport, RI.

You might also like