How accurate are designers in inferring the thoughts and feelings of users?

In design, designers need to understand users accurately. Understanding of users has been considered to lead to the development of products and services that better match user needs and goals. Therefore, a relevant question is how accurately designers can infer the users’ thoughts and feelings.

Surprisingly, previous research on an interpersonal accuracy task among friends (Ickes & Hodges, 2013), strangers (Ickes et al., 1990), and psychotherapists-clients (Marangoni et al., 1995) have reported 20–30 percent accuracy of people inferring the thoughts and feelings of others. A recent study (Chang-Arana, Piispanen et al., 2020) uncovered that when designers inferred design-related contents, they were able to reach approximately 50 percent accuracy, which was higher than inferring mental contents irrelevant to the design. This difference suggests that interpersonal accuracy is a skill that could be trained. We present recent research results related to the interpersonal accuracy of designers, suggest design methods to support the interpersonal accuracy of designers and raise ethical considerations regarding user understanding.

User understanding is good business

Successful businesses and projects have been associated with practicing design-thinking or human-centred design (Brown, 2019; Kramer et al., 2016). For instance, based on a McKinsey (2018) study of 300 publicly listed companies, the revenue growth of top design performers was almost double that of their industry counterparts, and shareholder return growth was 70 percent higher than their industry peers (Sheppard et al., 2018). In a study by Adobe and Forrester (2016) greater market share was reported by 41 percent of design-led organizations and competitive advantage by 46 percent of design-led companies compared to competitors that do not have a design approach. As an approach to design, design thinking highlights the role of empathy. However, there is no widely accepted or consistent definition of empathy in design (Chang-Arana, Piispanen et al., 2020). This presents a challenge to the design field: if there is no clarity on what empathy means nor how to measure it, how can the designers and professionals involved in design be successful in their work?

Not just empathy, but more precisely empathic understanding and the accuracy of it

Recent doctoral dissertation of Chang-Arana (2023) tackles this problem by incorporating into design the concept of mutual understanding, and interpersonal accuracy (Hall et al., 2016) as a collection of methods to measure it. Aligned with Surma-aho and Hölttä-Otto’s call for better operationalization of empathy (2022), two interpersonal accuracy methodologies, empathic accuracy (EA; Ickes et al., 1990; Ickes & Hodges, 2013; Ickes, 2016) and emotion recognition accuracy (ERA; Bänziger, 2016) were adapted for design and music performance contexts (Chang-Arana, 2023). Although, the latter context exceeds this article’s focus.

Designers may intend certain uses for and anticipate attitudes toward their creations. Yet, end-users, having their own needs, interests, and circumstances, may not share such design expectations (Hassenzahl, 2018). Through interpersonal accuracy, it is possible to identify and assess how accurately designers infer users’ needs, expressed as thoughts and feelings. We think users’ thoughts and feelings are a door towards better-tuned design interventions, and hence, better design outcomes.

Ways to measure understanding between users and designers

Empathic accuracy

Empathic accuracy (Ickes, 2016) is a set of performance-based methods to measure mind-reading accuracy between an interacting dyad (e.g., designer and user, Chang-Arana, Piispanen et al., 2020). A conversation between a dyad is videoed. A member of the dyad acts as the target of inference. They rewatch the interaction and annotate thoughts and feelings remembered, and the exact recording time when these occurred. The other member of the dyad acts as a decoder. They rewatch the interaction, only that it is stopped at the time points annotated by the target. The decoder must infer as accurately as possible the target’s mental contents at those exact time stops. The accuracy of these inferences is rated by a group of naïve raters. The higher the score, the higher the EA (see Ickes, 2016, for further details on EA methods).

Emotional recognition accuracy

Emotional recognition accuracy (ERA) is the ability to correctly identify someone else’s non-verbal emotional expressions (Bäzinger, 2016; Chang-Arana, 2023). Methods of ERA may be more varied than the standard EA methods. Yet, some methodological commonalities can be established. A typical ERA study will ask participants to infer what is the emotional state of a target based on a predetermined set of alternatives. Perceivers will be presented with different stimuli, such as video recordings, photographs, gestures, or audio-recorded vocal expressions. The accuracy will be determined by the self-reported intention of the target or by the emotional intention chosen by the researchers, and the answers provided by the participants.

We are quite inaccurate

So far, only EA has been adapted to the design context. When applying EA in design, designers obtained approximately 50 percent accuracy when inferring the mental contents of users in different design contexts (Chang-Arana, Surma-aho et al., 2020). Although a 50 percent accuracy may seem low, it is considerably higher than past findings in other research contexts. Previous EA studies have measured the interpersonal accuracy of male strangers (EA = 24 %) vs. male friends (EA = 36 %; Stinson & Ickes, 1992), and psychology students’ EA towards psychotherapy clients (23–24 %; Marangoni et al., 1995).

Yet, a closer look into designers’ EA shows higher accuracy when inferring mental contents relevant to design vs. other mental contents (Chang-Arana, Surma-aho et al., 2020). The difference may be attributed to the design context. Designers approach users intending to understand design-relevant experiences. Their attentional resources, tailored interviews and other methods, and demonstrations of objects or actions, all support designers’ user-understanding. Still, designers’ EA could be higher. Are there ways to achieve such an increase?

How can we get more accurate via various design methods?

Interpersonal differences impact interpersonal accuracy (Li & Hölttä-Otto, 2023). For example, coming from different national cultures lowers the accuracy significantly compared to a user-designer pair that comes from the same national background (Li & Hölttä-Otto 2020; Li et al., 2021). This can be mitigated via travel, but other actions can be taken to improve the accuracy (Li & Hölttä-Otto 2023; Li et al., 2021). Chang-Arana (2023) provides two general recommendations for improving understanding between designer-user dyads. First, for a long time, psychology has focused on investigating perspective-taking or people’s capacity to infer someone else’s mental contents (Batson et al., 1997). More recently, however, psychologists are exploring perspective-getting (Damen, Pollmann et al., 2021; Damen, van Amelsvoort et al., 2021): a higher understanding of someone’s mental events can be achieved by asking directly from them rather than inferring such contents. For example, if a user describes a chair as “uncomfortable”, a perspective-getting approach would encourage the designer to actively inquire what “uncomfortable” means to this designer in this context until both agree on its meaning. Second, utilizing accessories and other resources during an interaction to exemplify. Ambiguity in the communication between designers and users may be reduced through props. Considering these general recommendations, we will next briefly review some tools utilized by designers that are thought to increase the understanding of a user. We will comment on these methods using psychology literature on empathy, where we find a longer research tradition and fuller understanding of empathy than in design (Chang-Arana et al., 2022). Yet, we acknowledge that empathy in psychology is also a problematic concept (Cuff et al., 2016).

Designers may use wearable simulators (Kullman, 2016) such as goggles, restrictive bodysuits, and other types of equipment which may temporarily limit the movements and physical possibilities of a designer (Chang-Arana et al., 2022). These tools may help bridge differences in state influences (e.g., ethnic background, age, shared experiences). By state, we mean an entity’s “relative stability of its basic components or elements. Although the components or elements are essentially qualitatively stable, it is possible for them also to be dynamic, as in a hyperactive state or a state of flux” (APA, 2023, par. 1).

Other methods may require less bodily engagement and more intellectual and imaginative capacities of designers. Such is the case with design tools such as personas (Miaskiewicz & Kozar, 2011), storytelling (Madsen & Nielsen, 2010), and empathy maps (Both & Baggereor, n.d). These tools may engage more of the designers’ top-down control. That is, willingly taking the perspective of users. This capacity is often referred to as cognitive empathy or the cognitive process which allows adopting someone else’s psychological perspective or point of view (Shamay-Tsoory, 2011).

Related to taking the perspective of users, in design there is a clear understanding that, even when sharing similar backgrounds (Taffe, 2015), designers are limited in their capacity to understand users. That is, designers are limited by their empathic horizon (Chang-Arana et al., 2022; McDonagh et al., 2018). Our studies of empathic accuracy in design allowed us to quantify the limited understanding designers have of users (as per our operationalization of empathic accuracy). Relatedly, health sciences are encouraging their professionals to “recognize when they are in the position of power with their patients or students and attempt to be open, self-aware, egoless, and flexible.” (Foronda, 2020, p. 10). Similarly, the concept of the empathic horizon (Heylighen & Dong, 2019; McDonagh et al., 2018) could remind designers of disparities in understanding and power over users.

On a wider level, the system intelligence approach to empathic design emphasizes the importance of the whole, recognizing the interconnectedness, interdependence, and systematic feedback (Saarinen & Hämäläinen, 2004) present in the interaction between designers, users, and their context. Taking a systems intelligence approach to design, a designer acknowledges that improvements in micro-behaviors such as eye contact, expressions of gratefulness for others and reciprocity can create positive systemic growth (Jumisko-Pyykkö et al., 2021, p. 199). Tactics such as approaching people with warmth and acceptance and having a positive look on the future can support more empathic design (Jumisko-Pyykkö et al., 2021).    

Designers make use of objects and props to help improve the outcome of interactions (e.g., a contextual interview or a prototype session). This facilitates understanding what, otherwise, could be unclear user needs. We think this practice may be consistent with perspective-getting, or “actively inquiring about other peoples’ mental states” (Damen, Pollmann et al., 2021, p. 2).   

Ethical considerations about interpersonal accuracy in design

Before concluding, we consider some ethical considerations on understanding accurately a user (Chang-Arana et al., 2022). Designers have highlighted the pragmatic use of “empathy”. For example, Kouprie and Sleeswijk Visser (2009, pp. 437–438) claimed that empathy in design allows designers to “get closer to the lives and experiences of (putative, potential or future) users, in order to increase the likelihood that the product or service designed meets the user’s needs”. Designers recognize that empathic design without a serious pondering of its professional, social, and ethical implications (Devecchi & Guerrini, 2017; Hess et al., 2017; Hess et al., 2021; Walther et al., 2017) may risk viewing people as “simply the economic bottom line of designers” (Walther et al., 2017, p. 132). Thus, we encourage designers to engage with users while carefully considering their interventions’ ethical consequences. 

In short

Empathy is deemed necessary for successful design outcomes. However, its definition is not clear. Interpersonal accuracy is an alternative methodological approach to understanding empathy in design. We are quite inaccurate when inferring what someone else is thinking or feeling, and this is true for designers as well. Yet, we think that psychology offers designers conceptual and methodological tools to increase interpersonal accuracy. Perspective-getting may be a useful concept for designers trying to improve user understanding. Using accessories and other resources during an interaction may reduce ambiguity in the communication between designers and users. Equally important is to practice design while being aware of its ethical implications for the well-being of society and the sustainability of our environment.

Conclusion

An important takeaway for design professionals is that when understanding users of your products and services, quite likely, you are going to get it (mostly) wrong. Acknowledge you are not going to get it right. Move beyond specifications or features of the product and engage in perspective-getting, a more active understanding of a user’s thoughts and feelings. For educators and design students the insights of this article may be helpful in design skills development and expanding the current toolkit of design methods. Lastly, while exploring user-understanding (and across the whole design process), keep in sight the ethical consequence of your design practices.