12.07.2015 Views

Deliverable 4.4 - INSEAD CALT

Deliverable 4.4 - INSEAD CALT

Deliverable 4.4 - INSEAD CALT

SHOW MORE
SHOW LESS

Create successful ePaper yourself

Turn your PDF publications into a flip-book with our unique Google optimized e-Paper software.

AtGentive IST-4-027529-STP - Attentive Agents for Collaborative LearnersHowever, when we considered the same type of interactions limiting the sample to theExperimental group (table 6), we found that the number of questions asked to the expertis correlated with the paper quality (p=0.0300).Independent VariableQuestions_asked_recod(0->0; 1,2,..,7->1) 0.0300intro recod(0,..9->0,10->1) 0.9400Good_goal 0.5900cc recod(0,..5->0,6,..10->1) 0.2400Chi Square test (P-value)Table 6 - Each independent variable vs Quality of Paper for the Experimental Sample.Significant interaction between the number of questions asked to the expert and thepaper quality. Chi-square test.This effect is not found if the sample is limited to the Control group (see table 7): wefound no significant interaction between the quality of the final paper and other parts ofthe learning process in the Control group.Independent VariableQuestions_asked_recod(0->0; 1,2,..,7->1) 0.2100intro recod(0,..9->0,10->1) 0.7200Good_goal 0.8400cc recod(0,..5->0,6,..10->1) 0.4700Chi Square Test (P-value)Table 7 - Each independent variable vs Quality of Paper for the Control Sample. Nosignificant interaction found for the Control Group as regards to paper quality. ChiSquare.This data reflects the fact that the interaction between the questions asked and thequality of paper applies only to children in the Experimental group. From theanalysis shown in table 2 we know that such interaction is negative (meaning that askingmore questions may actually lower the quality of paper).In order to gain a better understanding of the interaction between the quality of the paperand the questions asked to the expert, we have first looked at how the average numberof questions asked, with respect to a specific paper quality varied between theExperimental and Control group. Figure 4 shows that not only children in the Controlgroup asked significantly less questions on average, but also that the trend to a positiveinteraction between the number of questions asked and paper quality present in thecontrol group, reverses in the Experimental group.Del <strong>4.4</strong>: AtGentive Final Evaluation Report – Appendix C page 9

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

Saved successfully!

Ooh no, something went wrong!