TL;DR
This paper investigates how VR avatar gender and narrative context influence creative decision-making in film education. By placing students in avatars of different genders and asking them to film a dance sequence, researchers found that "gender-swapping" can actually heighten a user's sense of presence and shift their cinematic style—effectively allowing male students to adopt a "female gaze" and vice versa.
What Problem Does This Paper Address?
While Virtual Reality (VR) is increasingly used to teach technical skills like cinematography and lighting, we know very little about how the virtual environment itself—specifically the user's avatar—shapes their creative "voice."
In film theory, the "male gaze" often refers to a perspective that objectifies subjects and focuses on action/physique, while the "female gaze" tends to emphasize emotional intimacy and reciprocity. The authors sought to understand if the Proteus Effect (the phenomenon where a user's behavior shifts to match their avatar's characteristics) could be used as a pedagogical tool. Can changing a student's virtual gender help them recognize and overcome their own narrative biases?
Methods
The researchers developed Film DanceStudio, a VR environment built in Unity where users can control cameras and lighting to film a 45-second dance sequence. They employed a $2 \times 2 \times 2$ factorial experimental design with 48 participants:
- User Gender: Male or Female.
- Avatar Gender: Male or Female.
- Narrative Actor Gender: A male-led or female-led dance sequence (inspired by the film Black Swan).
Participants underwent a 15-minute tutorial followed by a 12-minute filming task. The resulting films were then evaluated by two industry experts (one male, one female) who coded the shots for "Male Gaze" (high contrast, wide action shots, focus on body parts) vs. "Female Gaze" (soft lighting, close-ups on facial expressions, focus on emotional interaction).
Datasets
The study did not use pre-existing datasets. Instead, it generated a primary dataset consisting of:
- Quantitative Survey Data: Igroup Presence Questionnaire (IPQ) and Standardized Embodiment Questionnaire (SEQ) scores from 48 participants.
- Qualitative Interview Data: Transcripts from semi-structured post-task interviews.
- Video Content: 48 unique 45-second film clips created by the participants, categorized by expert raters.
Key Insights and Ideas
- Productive Discomfort: Contrary to the common belief that "matching" avatars increase immersion, this study found that gender-incongruent avatars (e.g., a male user in a female avatar) led to a stronger sense of presence and self-awareness. Users reported "watching every movement" more closely because the body felt different from their own, leading to more deliberate creative choices.
- The Avatar as a Creative Lens: The Proteus Effect was clearly visible in the artistic output. Male students using female avatars were significantly more likely to adopt "female gaze" techniques, focusing on emotional nuances and soft lighting.
- Narrative Can Override Identity: The "narrative lead" (who initiates the action in the dance) also played a major role. Even when users were in gender-matched avatars, a narrative led by the opposite gender could shift their filming style, suggesting that story context is a powerful tool for breaking habitual creative patterns.
- Perspective Playground: The authors propose that VR shouldn't just be a technical simulator but a "perspective playground" where students can systematically swap identities and narrative roles to dismantle unconscious biases in filmmaking.
Key Results
- Presence & Awareness: Statistical analysis (ART ANOVA) showed a significant interaction between user gender and avatar gender on self-presence ($p = 0.049$). Users in non-matching avatars reported higher levels of "Body Sensation" and "Awareness."
- Cinematic Style Shifts:
- Group A (Male User/Male Avatar/Male Narrative): Produced works almost exclusively classified as "Male Gaze."
- Group B (Male User/Male Avatar/Female Narrative): Despite the male identity, more than half of the works shifted toward a "Female Gaze."
- Group H (Female User/Female Avatar/Female Narrative): Primarily produced "Female Gaze" content.
- Expert Validation: Content analysis achieved substantial inter-coder agreement ($\kappa = 0.78$), confirming that the shifts in lighting and framing were objectively observable and not just subjective feelings of the participants.
Limitations and Future Work
- Sample Size: With only 48 participants (6 per experimental cell), the statistical power is relatively low ($0.25$), meaning some subtle effects might have been missed.
- Binary Focus: The study focused strictly on cisgender binary identities. The authors acknowledge that future work should include non-binary and androgynous avatars to explore more fluid gaze dynamics.
- Demographics: The participants were primarily students from a single academic institution and of East Asian descent, which may limit the generalizability of the findings to other cultural contexts.
- Future Directions: The researchers suggest moving toward collaborative VR environments where multiple users with different avatar/narrative configurations must negotiate creative decisions together.
Summary updated