Avatar Customization and Prosocial Behavior in Games

ISEF Category: Behavioral and Social Sciences

Subcategory: Social Psychology  ·  Difficulty: Advanced  ·  Setup: Home Setup  ·  Time: 1 to 2 Months

The Hook

A tiny change in how you look on screen can change how other players treat you. In cooperative games, avatars can act like social masks, and people may help, trust, or ignore teammates based on those cues. You can turn that into a real research question by comparing chat tone and helping behavior across different avatar styles.

What Is It?

Avatar customization means the visible choices you make for a game character, like race cues, gender cues, age cues, clothing, or style. Those cues can shape first impressions fast, before anyone knows the real person behind the screen. In social psychology, that matters because people often respond to appearance first and identity second.

Prosocial behavior means actions that help other people, such as sharing items, reviving teammates, or sending encouraging messages. Chat sentiment means the emotional tone of the words players use. Think of the avatar like a school costume day outfit, the same person can get treated differently depending on what other people think they see.

Why This Is a Good Topic

This topic gives you a clean mix of human behavior, text analysis, and game data. You can measure something real, like help offers or positive chat, without a wet lab, and you can test one change at a time. It also connects to online safety, team trust, and bias, so your question has a clear real-world angle. A strong version teaches you how to code data, compare groups, and explain behavior with evidence.

Research Questions

• How does avatar gender affect the rate of helping actions in cooperative matches?
• What is the effect of avatar race cues on positive chat sentiment toward teammates?
• Does avatar age appearance change how often players offer support after a teammate asks for help?
• To what extent do customized avatars change the number of cooperative actions compared with default avatars?
• Which avatar traits predict more encouraging messages after a win or a loss?
• How does the presence of a visible avatar change prosocial behavior across solo queue and premade teams?

Basic Materials

• Computer with internet access.
• Free account in a cooperative game that saves match logs or replay history.
• Spreadsheet software, such as Google Sheets or Excel.
• Python in Google Colab or a local install.
• A coding sheet for avatar traits, helping actions, and chat sentiment.
• Screen capture or export tool for saving logs and avatar screens.

Advanced Materials

• De-identified match-log archive from a university lab or public dataset.
• R with tidyverse, lme4, and ggplot2.
• Python with pandas, spaCy, and scikit-learn.
• Qualitative coding software such as Taguette.
• Access to IRB-approved survey software, such as Qualtrics or REDCap, for follow-up measures.

Software & Tools

• Python: Cleans log files, scores chat sentiment, and counts prosocial actions.
• R: Tests group differences and fits models that control for player and match factors.
• Google Colab: Lets you run Python notebooks in a browser without local setup.
• Taguette: Helps you label chat messages with clear coding categories.

Experiment Steps

1. Define the avatar traits you will test, and keep the rest of the match setting steady.
2. Choose one primary prosocial outcome, and decide exactly how you will spot it in logs.
3. Build a coding plan for chat tone, so every message gets the same rules.
4. Set the controls you will track, such as match size, player skill, and game mode.
5. Decide how you will compare groups, handle missing data, and check whether one loud player skews results.
6. Plan how you will combine behavior counts and sentiment scores into one final claim.

Common Pitfalls

• Mixing race, gender, and age changes in the same test, which hides which avatar cue actually mattered.
• Counting every friendly message as prosocial, which inflates your behavior score with normal small talk.
• Ignoring player skill or rank, which can make strong players look like they caused the effect.
• Feeding raw chat text into sentiment software without cleaning slang, emojis, and usernames, which distorts the score.
• Pooling different game modes or team sizes, which adds noise and can erase a real pattern.

What Makes This Competitive

A stronger project separates avatar effects from player skill, match size, and game mode. It also treats behavior and language as two linked outcomes, not one. If you code messages carefully, test the pattern across several avatar traits, and check whether the effect survives a stricter model, your work starts to look like real social science. A clear comparison across subgroups or a new angle on team trust can push it farther.

Project Variations

• Compare default avatars with customized avatars in casual matches, then see whether helping rises in one mode more than the other.
• Test one trait at a time, such as gender cues or age cues, so you can see which cue drives chat tone most.
• Swap in a different cooperative game with public replay logs and check whether the same avatar effect shows up again.

Learn More

• PubMed: Search review articles on avatar cues, social identity, and prosocial behavior in online groups.
• PubMed Central: Read full-text papers on deindividuation, gaming behavior, and sentiment analysis.
• APA Dictionary of Psychology: Look up plain-language definitions for prosocial behavior, stereotype, and social identity.
• MIT OpenCourseWare: Find free lectures on research methods, statistics, and study design in the psychology course catalog.
• PLOS ONE: Search open-access studies on online behavior, cooperation, and text analysis.