[nc-whois] Updated baskets and statistics.
Please find attached new statistics which take into account Tony Harris' and Ram Mohan's work on fixing the baskets, plus some additional work I did tonight, in order to get the "raw gems" (unqualified '#'; see my earlier message) under control (some of which didn't turn out to be that valuable). The results are in the usual Excel sheet format; I've put them into zip files by chapter of the final report. The baskets used have been extended; Kristy has put up a Word document with the revised baskets some two or three hours ago. You'll have to download that, too. What should you do to these results? In short: Look for anything interesting you notice, and write it up. If you believe that the number of baskets is too large for any question, please come up with fewer baskets, by _merging_ existing baskets. An example of how this can be done is in the preliminary report's evaluation of question 16. (In that case, it's not baskets, but answers to multiple-choice questions. Still, the principle remains the same.) If you come up with such revised baskets, PLEASE DO NOT SIMPLY ADD UP THE NUMBERS, except in order to get a quick estimate of the approximate order of magnitude of the results - the categories are generally not mutually exclusive. For the actual report, please send me a message telling me what baskets you want to have merged. Once I know that, I can quickly produce new statistics using the revised baskets. Please note that decreasing the number of baskets is not a bad thing at all: You may lose some very fine-grained information, but at the same time, the results you have will become more robust, since baskets which are close to each other, and may have been interpreted in different ways by different members of the task force, will most likely put together into one category. (In fact, there are several questions with extremely large numbers of baskets. In these cases, I'd _very_ _strongly_ suggest that you try to come up with simplifications.) Also, please note that I have the entire basketing information in a database, so even complex searches are easy. If you want to know what questionnaires were put into a specific category in order to understand what is meant by that category, please don't hesitate to send me a message. If you have any more advanced ideas to massage the data, also don't hesitate to write. I'll respond as quickly as possible to such inquiries. As an alterantive, you can also search the Excel sheet contained in "all.zip" (attached, too), which contains _all_ the basketing information currently available. Asking me to query the database will be the faster and easier way to get to the information, though. ;-) Regards, and good night, -- Thomas Roessler <roessler@does-not-exist.org> iii-uniformity-centralization.zip |