Critique
From Course talk:CPSC522/Improving the accuracy of Affect Prediction in an Intelligent Tutoring System
Hi Abed,
Nice work! I It fits in perfectly from where you left in the previous assignment. You also did a good job presenting your work today. It made understanding your page a lot easier after seeing you present.
Here are a few comments/suggestions:
- The Statistical analysis section could be improved by providing the reader a bit more intuition behind what is happening. For instance, you could briefly explain the ANOVA table for curiosity by showing how we interpret the values.
- Rather than just having a single level for boredom and curiosity, breaking them down into smaller categories like “Extremely bored” and “Moderately bored” might give us a more accurate analysis. You could probably try that it out in the future.
Keep up the good work.
Best Regards,
Adnan
Hi Adnan, Thank you for your feedback and your suggestions. I would try to update the page to the best of my ability
- I agree that the analysis section can be updated a bit, specially with giving intuition behind what is happening. I would try to add a section there
- I forgot to mention in the page that some amount of thresholding was done in the original paper which lead to the dataset being binary. The actual dataset contains data in a 5 point likert scale. This is definitely a future work prospect.
Thanks Abed
MDAbedRahman (talk)