190 likes | 416 Views
Pilot work on Transformed Social Interaction and Learning. Behavioral Flags Transformed Conformity NonZeroSum Proximity Why Potential new studies/LIFE projects Data analysis strategies Publication Venues. Collaborative Virtual Environments. Transformed Social Interaction (TSI). Actual
E N D
Pilot work on Transformed Social Interaction and Learning • Behavioral Flags • Transformed Conformity • NonZeroSum Proximity • Why • Potential new studies/LIFE projects • Data analysis strategies • Publication Venues
Transformed Social Interaction (TSI) Actual Behavior Strategic Filter Transformed Behavior
Three dimensions of TSI • Transforming Self Representation • Transforming Social-Sensory Abilities • Transforming Social Context
Behavioral Flags, Study 1 • Classroom situation • Subject is the teacher, instructed to use head movements (eye gaze) to engage the students • Students are agents with idling behaviors • Students turn translucent in proportion to out of teacher’s field of view • Can teacher use the flags, do they help gaze • Varied • Presence of Flags or not • Teacher speaking or not (30 second stories about key words) • DVs • Gaze
Percentage of time student became invisible (was out of the FOV for more than five seconds)
Behavioral Flags, Study 2 • Classroom situation • Subject becomes a student sitting either in the center of the room or on the very end of the table (periphery of teacher) • Other 8 students are agents with idling behaviors • Teacher reads 4 minute “fever” passage • Varied • Whether subjects received teacher using head movements from Study 1 subjects who received Flags or not • Teacher speaking or not (30 second stories about key words) • DVs • Learning • Subjective response • Return gaze to teacher (not yet analyzed)
Results(# multiple choice questions correct, out of 10) Nothing significant
Results(mean score on 12 traits about teacher— “informed, sincere, likable, etc.) Significant interactions with flags and Seat, flags and gender
Todo • Use head gaze inview as CV • Use regression • Split out questions of recall and inference • Change “flag name” • Partial out effects of time perceived, naturalness • Look at bimodal effect of gender, seat, and flag • Journal of learning sciences, Journal of Ed Psych • Show how it generalizes away from VR, focus on the theory of it • Look at joint attention literature, common ground • Can we use physio to isolate arousal versus other explanations (is it receiving gaze versus knowing the teacher is aware of you, separate out theory of mind) • Look at other types of flags (student turns translucent when she looks away from you) • Use gaze from another student to perhaps increase the arousal of the student • Use dynamic nonverbal cues of the teacher to automatically respond to differential performance of the students, automatically capture teacher’s responses to good and bad performing students in realtime, and record and play back those behaviors to another set of students. • Compare actual head movements to randomly generated spots (true random follow the dot) • Make the flag moving the seat farther away versus closer • Use phsyio measures • Look at picard’s opera glass, happy sad glasses used in class • Selective attention literature on kids and television (what has them look at the screen?), fred rogers, looking and learning, norman fell • Metaknowledge of the gaze of the teacher, do the other students know about it? • Add some history such that opacity decreases quicker if recent gaze has been high, what type of gaze, not amount is conducive to learning. • Head movements (three subjects, one teacher, two students, parse learners by teachers movements
Transformed Conformity • Classroom situation • override peer behaviors with optimal learning behaviors • Networked CVE • Pilot study – does conformity to learning behavior occur? • Agent teacher with lesson on pharm companies • Student always sits in same place, 15 other students in the room • Conditions • Conformity (between, n=30 in each) • Negative (not looking at teacher, looking at watch, shaking head in disgust, watching distracting events) • Positive (looking at teacher, nodding, taking notes, not looking at distracting events) • Empty • Distracting Events ( within, car outside window, four times over 8 minutes) • DVs • Learning • Room Recall • Eye gaze • Amount of head movement • Looking at teacher • Looking at distracting events
Agent student Agent student Agent student Agent student window Agent student Agent student Agent student Agent student Subject Agent student Agent student Agent student Agent student Agent student Agent student Agent student Teacher
Results • No differences in learning by condition (recall of passage facts) • Interaction between gender and condition in terms of gaze • Significant correlation between yaw and learning
Todo • Put the learning data, focus on gaze itself, not the conditions, how does VR inform the learning sciences community • Middle grounds for how the results inform learning sciences, etc, vary theoretical to empirical • Don’t focus on memory, because it is not the best way to teach, also memory is not correlated • Measure how good a teacher they are later on after getting the vr lesson (transfer), “going beyond the information given”
NZSP • Classroom situation • Subject is the student • One ‘sweet spot’ for learning • Networked CVE, everyone gets the sweet spot • Pilot study – where is the sweetspot? • Angle: see data from Behavioral Flags study, distance is constant, angle varies • Distance Varied • Distance (short far, within subject, n=48) • Population of room (full empty, between subject, 24/24) • DVs • Learning • Gaze (no differences)
NZSP results:Learning S2 S1 Teacher Sig main effect, marginal interaction