Miscell announcements & gyan

Hi all,

Just to make sure, no section missed out on these announcements and occasional gyan:

1. If *any* slide is shown in the classroom, even for a few seconds, pls look up and read up the corresponding portion from the textbook (ensure you know the definitions at least). These will be fair game for MCQ in the exam.

2. Pls practice class exercises on SPSS at home. Make sure you know how to perform any particular data analysis procedure covered in the class. Ask around if you are having trouble navigating the spss menu. Additionally, the textbook gives detailed step by step spss directions, something worth looking up if you get stuck on spss anywhere.

If you're unsure about something, feel free to ask.

3. Optionally, see if you can perform the same procedures on MSexcel. Gives great confidence and certainty when a proc can be flawlessly programmed into and executed on excel. Again, this is purely optional and won't be important from an exam viewpoint.

4. While I mentioned that ANOVA is a subset of linear regression with dummy variables, what I may not have mentioned/stressed ion all sections is that ANOVA is important for another reason - it does *not* require/make any assumptions about a linear functional form.

So in the classroom example, if mobile usage depended on gander and rateplan, i.e.
usage = f(gender, rateplan)
Then using a regression with dummy variables would necessarily asume that the f(.) is linear. 2-way ANOVA avoids making this assumption.

Sudhir