EPELI: a novel virtual reality task for the assessment of goal-directed behavior in real-life contexts

Psychological Research https://doi.org/10.1007/s00426-022-01770-z ORIGINAL ARTICLE EPELI: a novel virtual reality task for the assessment of goal‑directed behavior in real‑life contexts Erik Seesjärvi1,2 · Jasmin Puhakka1 · Eeva T. Aronen3,4 Matthias Kliegel6,8 · Matti Laine9 · Juha Salmi1,10,11 · Alexandra Hering5 · Sascha Zuber6,7 · Liya Merzon10 · Received: 6 June 2022 / Accepted: 6 November 2022 © The Author(s) 2022 Abstract A recently developed virtual reality task, EPELI (Executive Performance in Everyday LIving), quantifies goal-directed behavior in naturalistic conditions. Participants navigate a virtual apartment, performing household chores given by a virtual character. EPELI aims to tap attention, executive function, and prospective memory. To ensure its applicability to further research and clinical work and to study its relationship to relevant background factors, we examined several key properties of EPELI in 77 typically developing 9–13-year-old children. These included EPELI’s internal consistency, age and gender differences, sensitivity to gaming experience, head-mounted display (HMD) type, and verbal recall ability, as well as its relationships with parent-rated everyday executive problems. Of the eight EPELI measures, the following six showed acceptable internal consistency: task and navigation efficacy, number of correctly performed tasks and overall actions, time monitoring, and controller movement. Some measures were associated with age, gender, or verbal encoding ability. Moreover, EPELI performance was associated with parent-rated everyday executive problems. There were no significant associations of gaming background, task familiarity, or HMD type with the EPELI measures. These results attest to the reliability and ecological validity of this new virtual reality tool for the assessment of attention, executive functions, and prospective memory in children. Introduction In cognitive psychology and neuropsychology, there is a long history of using experimental paradigms that contain a relatively narrow set of stimuli and allow for only a limited range of behavioral responses from the experimentee (Hatfield, 2002). The use of such paradigms and measures has unarguably yielded a wealth of information on * Erik Seesjärvi 1 Department of Psychology and Logopedics, University of Helsinki, P.O. Box 4, 00014 Helsinki, Finland 2 Child Neurology, University of Helsinki and Helsinki University Hospital, Helsinki, Finland 3 Child Psychiatry, University of Helsinki and Helsinki University Hospital, Helsinki, Finland 4 Laboratory of Developmental Psychopathology, New Children’s Hospital, Pediatric Research Center, Helsinki, Finland 5 human cognition in both healthy participants and clinical groups. However, some critics have questioned whether this approach alone is sufficient when the subject of the study is something as complex as human goal-directed behavior taking place in complex everyday situations (e.g., Bronfenbrenner, 1977; Brunswik, 1943; Gibson, 1970; Kingstone et al., 2008; Neisser, 1976). In particular, there has been a call to develop neuropsychological tests with increased ecological 6 Swiss National Centre of Competence in Research LIVES, Geneva, Switzerland 7 Institute on Aging and Lifelong Health, University of Victoria, Victoria, Canada 8 Department of Psychology, University of Geneva, Geneva, Switzerland 9 Department of Psychology, Åbo Akademi University, Turku, Finland 10 Department of Neuroscience and Biomedical Engineering, Aalto University, Espoo, Finland 11 MAGICS, Aalto University, Espoo, Finland Department of Developmental Psychology, Tilburg University, Tilburg, The Netherlands 13 Vol.:(0123456789) Psychological Research validity, that is tasks which are more closely related to everyday tasks (Franzen & Wilhelm, 1996). There are two general types of ecological validity of a test measure to consider here, namely verisimilitude and veridicality. Verisimilitude refers to how accurately the test properties resemble corresponding situation in the real world, whereas veridicality refers to the extent to which results on a test reflect or predict the skills or task performances in everyday life (Franzen & Wilhelm, 1996). There are different approaches to assess and improve the verisimilitude and veridicality of a test (Chaytor & Schmitter-Edgecombe, 2003; Chaytor et al., 2006; Parsons, 2015). In some studies, cognitive functions have been measured “in the wild”, by instructing the participants to perform specific cognitively demanding tasks, for instance, in a shopping street (Garden et al., 2001; Shallice & Burgess, 1991). As these tasks take place in real-life environments, they have high verisimilitude (Spooner & Pachana, 2006). However, it is difficult to control for the complex environmental effects in reallife conditions, and the precise measurement of behavior becomes challenging too. Therefore, this type of conditions may suffer from limited reliability and participants safety might also be compromised (Logie et al., 2011). Furthermore, taking the testing to actual real-world environments can be time-consuming and impractical. Another, more convenient option is to include features to a computer task that will increase its ecological validity. For instance, in the Virtual Week task, participants are asked to perform a set of everyday tasks in a board game (Rendell & Craik, 2000). This type of tasks can be considered to share higher resemblance with real-life situations than conventional paper-andpencil tasks or experimental cognitive psychology tasks with restricted stimuli, but it is debated whether such tasks measure the same cognitive processes that are required in reallife situations (Parsons, 2015). In computerized tasks, the interaction with the environment is typically quite artificial and time scale of task performance is unrealistic. Moreover, even though a structured task has benefits raising from clearly defined measures, it can be criticized that functionled tasks aiming at high ecological validity (e.g., the Virtual Week) may not generalize to multiple different situations. This is the aim with conventional construct-laden neuropsychological tasks, such as Wechsler’s Intelligence Scale for Children, that are designed to tap several core cognitive functions reflecting general abilities (Wechsler & Kodama, 1949). All in all, there are multiple difficult choices that the experimenter needs to make to select an appropriate task, which will influence how well the outcome measure can be defined and how accurately does the respective measure reflect something being actually done in everyday life (Iverson et al., 2008). Virtual reality (VR) provides means to improve the ecological validity of a psychological test without 13 compromising the experimental control considerably. VR refers to using digitally generated, artificial environments to recreate real-world activities to participants. Wh (...truncated)