zaslow: Ralph, the spreadsheet is great. Here are a couple of questions:
1. Row 15 Professions. How should we distinguish between "corporate" and "professional"? For example, interview subject 14 is an engineer who works for a Fortune 500 company.
- The professions are traditionally law and medicine. An engineer working for a corporation would be 'corporate'.
2. Row 26 Driving Intensity. Errands and carting the kids around make up a lot of some subjects driving. Does that come under "work"? "Recreation?" Do we need a third category?
- Yeah...we probably do need a third one. In fact, we probably need to rethink that whole section.
3. Row 33 Not sure what you meant by "average monthly maximum range"
- I was trying to capture the difference between people who go on long trips all the time and those who only do it occasionally. In both cases, range might be the same. So, we need to capture some sort of local maxima...I figured monthly might be good enough.
Attached is the spreadsheet for interview cluster analysis. I filled in the first six interviews that Meg and I did as a test run and adjusted the categories accordingly. I'm sure that Ruth and Chris will need to add things here and there.
General Guidelines and Notes:
- Each category has a scale (usually 1-3) or qualitative breakdown. The qualitative breakdown lists capitalize the shorthand for each choice. Thus, under the region category "Mid-Atlantic" would be listed as MA and under the City Density category "Mid-sized city" would be listed as M
- Some of the terms could probably be improved. 'Saviness' could use a different synonym. Feel free to adjust these.
- When you 'check-out' the file to edit, put a note on the Wiki. When you check it back in erase the note.
- The interviews I looked at don't really capture some of the driving behavior variables very well. particularly in the case of peak usage or peak type . Or, perhaps peak isn't the right way to operationalize the construct. We may need to change these categories or go back to our subjects if possible.
Once all the data is entered I'll perform a rudimentary cluster analysis (both statistical and intuitive) to see where the areas of greatest variance are.
Let me know if you have any questions.
Thanks, Ralph
Interview Spreadsheet-1.xls
cluster analysis spreadsheet.xls
Use this revised spreadsheet for uploading added data (last revision 8/11 by Ralph)
clusteranalysis[1-27].xls This is the latest version with interviews 1-27 completed (8/12 by chrisw).
clusteranalysis(1-30)(1).xls Interview 28 data added (8/12 by ruth). Interview 29 data added (8/19 by ruth). Interview 30 data added (8/29 by ruth)
I
