Harmon Foods Stats Case

[Daniel Zarovy]

Hey Guys,

I've attached the write up that Sho and I worked on for the Harmon Foods case.  Please review it.  Also, Dustin, there's a note in there for you about the ROI of Consumer Packs. 

I can print it and turn it in tomorrow.  In addition to the write up, I was planning to print a page from Excel with the results of our (well...Dustin's) regression analysis.

Also, Claire's search for a rich banker reminded me of a remake of Single Ladies by Columbia Business School.  Good luck Claire. :)

Later,
Daniel

[Dustin Goot]

Guys,

Beautifully calculated, and your intuition on #3 is right on. One tiny issue: the coefficients I had calculated did not include the 90%, 5% and 5% weighting. I just added that to my raw data and suggested that you could plug it into the regression after I left if you wanted. Sorry for the confusion. It turns out that the 90/5/5 weighting has almost exactly the same fit (> .91 R-squared), so I went ahead and substituted the correct coefficients.

Another thing to note, just to keep in mind for the final, etc. To calculate confidence interval of the entire equation (not just a single coefficient), you have to multiple standard error by SQRT(1 + (1/N)). Also, we should be using the T-value here since we only have 48 observations. (It doesn't make a huge difference in calculation, but it's a more correct approach.)

Can someone print the revised assignment and turn in? The last tab in the spreadsheet has the new numbers. I have a feeling I might oversleep and/or skip stats once again.

--D