Emotions are integral to learning and decision-making. Every part of a virtual environment can influence a user’s emotions. In this project we explored how game characters can be designed to evoke positive, negative, an neutral emotional states in players. I helped conduct research on how facial expressions, colors, shapes, and dimensionality (2D vs 3D) of game characters affect players’ emotional states.


Contributions: Research design, Psychological design, Data collection, Data analysis, and Scientific writing.
– Led a mixed-methods research study observing players’ emotions when engaged in Virtual Reality and 2D gameplay.
– Helped conduct multiple studies on the effect of facial expression, shape, color, and dimensionality of game characters on players’ emotions.
– Published a chapter summarizing factors in emotional design in the Handbook of Game-based learning, and co-authored multiple articles in academic journals.
Related Publications:
Learning & Instruction: Emotional Design for Digital Games for Learning: The Affective Quality of Expression, Color, Shape, and Dimensionality of Game Characters. Access article
Cognitive Development: Activating Adolescents’ “Hot” Executive Functions in a Digital Game to Train Cognitive Skills: The Effects of Age and Prior Abilities. Access article
American Educational Research Association 2018: Exploring the Emotional Effect of Immersive Virtual Reality Versus 2-D Screen-Based Game Characters. Access article
Handbook of Game-based learning: Emerging Design Factors in Game-Based Learning: Emotional Design, Musical Score, and Game Mechanics Design. Access chapter

Background & Goal

The goal of this project was to study how design of game characters can influence players’ emotional states. This was studied in the context of game-based executive functions training game. Executive functions (EF) are core cognitive process that regulate attention and memory. EF are important because they influence academic outcomes and social well-being. The project goals were as follows:
Design and evaluate game characters for evoking distinct emotional states in players during gameplay.
Observe how colors and shapes of game characters influence emotional states.
Test if 3D game characters in virtual spaces evoke stronger emotional response than 2D game characters in a similar context.


All you can E.T. in virtual reality (AYCETVR) is an executive function training game developed by the Consortium for Research and Evaluation of Advanced Technologies in Education (CREATE). AYCETVR trains players’ shifting skills, which is a EF sub-skill related to switching between mental sets. It is the ability to adopt new information while disregarding old conflicting information. In this game, players are immersed in a 3D virtual space and use VR motion controllers to feed aliens. Feeding rules are shown on the food canons and tell players the correct food items for different aliens. For example, feed cupcakes to blue aliens and feed milkshakes to orange aliens. These rules change frequently, for example: feed milkshake to blue aliens and feed cupcakes to orange aliens. To follow these rules, players have to take actions by switching between distinct mental sets. This trains their shifting skills.


I helped conduct two experiments to study how colors, shape, and dimensionality (2D vs 3D) of game characters affect players’ emotional states. The first study compared a ‘Hot’ vs ‘Cool’ versions of an EF training game. Game characters were designed based on previous research on shape and colors of characters. The ‘Hot’ version characters had emotionally activating design, and the ‘Cool’ version characters had emotionally neutral design. The second study compared a 2D vs VR version of the same EF training game. In this study, participants emotional states were observed throughout gameplay to see which version elicited more activating emotions.

Study 1

In study 1, we explored how facial expressions, colors, and shape of game characters affect players’ emotions. We conducted a quantitative study with 233 adolescent participants. The goal was to compare the ‘Hot’ vs ‘Cool’ versions, and see if emotionally activating gameplay can improve EF training outcomes. A between subjects pre-post design with random assignment was used. The users completed pretest measures including EF tasks and surveys, then played either the ‘Hot’ or the ‘Cool’ version of the game for 20 minutes, and finally completed posttest measures. We compared gamelog, EF task, and survey data of the treatment group with the control group.

Study 2

This was a mixed-methods study comparing the effect of 2D vs VR gameplay on players’ emotions. We designed think-aloud protocols and surveys to capture players’ emotion activation. Users were first trained in thinking aloud. Once users were comfortable with expressing their emotional states during gameplay, they played both versions of the game. Half of the participants played VR version first, and the other half played the 2D version first. This counterbalance design helped avoid order effects. We prompted participants to keep thinking-aloud during both session of gameplay. The number of times players elicited emotional utterances was used as a measure of emotional activation. A custom positive negative affect scale (PANAS) survey was completed by all participants at the end of the study.


Results from the study 1 were as follows:
1. We found a main effect of treatment on EF scores. Users who played the hot version scored higher on the posttest DCCS task.
2. There was a two-way interaction between age and treatment. Older adolescents who played the ‘Hot’ version performed better than those who played the ‘Cool’ version.
3. There was also a three way interaction between age, treatment and prior EF. This indicated that players’ prior EF skills and their age determined if ‘Hot’ or ‘Cool’ would be more beneficial for them.

Results from study 2 were as follows:
1. Users’ reported higher emotion activation for VR gameplay than 2D gameplay.
2. Users’ had more emotion utterances when playing the VR version than the 2D version.
3. The recency of VR gameplay affected users’ report of emotion activation. Users’ who played the VR version right before completing the PANAS survey reported higher emotion activation than those who played the 2D version closer to completing the PANAS survey.