Early morning project mailbag

Got this email:

Dear Professor Voleti,

The original instruction regarding surveys was that it has to be filled by people outside of ISB, so we did not approach any ISB students. But we have heard that many ISB students have filled the survey due to requests from other groups. Will these numbers be added to the total count? This will be a disadvantage to those groups that diligently followed the instructions.

Also, is the 8% grade for reaching 90 surveys per group? Or would you scale the grade based on the max. surveys per group?

Please let us know.

Thanks,
S

My response:

Hi S

The within-ISB survey responses will be discounted to a large degree. We will do an IP address check for the same. Am counting on IT to come through for the same. Yes, higher the # of completed responses from a group, the better it is. There are other factors also – such as respondent selection –

(i) to what degree was the survey complete? The counter counts as complete and well-filled out? So someone somewhere serially checking all the leftmost buttons for the matrix questions, or leaving them out altogether etc would not as much weight as respondents who have filled out the entire survey.

(ii) to what degree was the respondent within the target population of interest for xyz corp? So someone who’s too young to obtain a driver’s license or someone whose first car was a Merc wouldn’t qualify, for example.

Also, let me mention that it is very difficult to get zero% in Phase II, much more difficult than to get 8%. So the range of variation across groups wouldn’t be 0-8 but more like 4-8 (perhaps 5-8), I would think.

Hope that clarifies.

Sudhir