12.07.2015 Views

Deliverable 4.4 - INSEAD CALT

Deliverable 4.4 - INSEAD CALT

Deliverable 4.4 - INSEAD CALT

SHOW MORE
SHOW LESS

You also want an ePaper? Increase the reach of your titles

YUMPU automatically turns print PDFs into web optimized ePapers that Google loves.

AtGentive IST-4-027529-STP - Attentive Agents for Collaborative Learners(before the beginning of the pilot) defined by their teachers as low performers might hadused the question mark and unhappy buttons more that children who had been definedas high performers. The results shown in figure 3 indicate that this was actuallyhappening. Therefore, the students who were asking for help indeed needed the extrahelp.Average # of Motivational Interv inPaper task by Team LevelAverage # of Cognitive Interv inPaper task by Team LevelAvg. # of M interv.54.543.532.521.510.50low medium highTeam levelAvg. # of C-Interv.2.521.510.50low medium highTeam levelFigure 3 – Average number of interventions in the Paper task by team (children) level. Itappears that the majority of Cognitive and Motivational interventions were given tochildren defined as low or medium performers by their teachers.The students understood the manner of asking for help and were able to indicate to thesystem that they needed more help. However the system was unable to bring theirperformance up to the standards of the successful students. This observation opened thefurther question of how much help the system was able to supply to children of differentlevels of ability, we discuss this aspect in section 4.It is interesting to note that interventions seem to have been particularly effective instimulating the children to be more active in the forum by asking more questions to theexpert. We found a positive correlation between all types of interventions provided duringthe forum task and the number of questions asked to the expert (see table 4).Gamma, ( ase), (p-value)N_MC_Forum recoded(0-12->0,13-218->1) 0.89, (0.12), (0.004)N_C_Forum recoded(0->0,1-11->1) 0.83, (0.18), (0.02)N_M_Forum recoded(0->0, 1-5->1) 0.86, (0.12), (0.003)Table 4 – Chi square Association between the number of questions asked to the expertand the number of interventions sent during the forum activityThe results obtained, without recoding, using a Spearman’s correlation (table 4') stillreports the positive correlation between the forum activity and the Meta-Cognitiveinterventions, they don't however report the correlation with Cognitive and Motivationalinterventions. This confirms the observation made in the analysis of figure 2: there is aclear boundary between teams that didn't use the feedback/help buttons (which elicitedthose type of interventions) and children who did.Del <strong>4.4</strong>: AtGentive Final Evaluation Report – Appendix C page 7

Hooray! Your file is uploaded and ready to be published.

Saved successfully!

Ooh no, something went wrong!