Barbie

The YAPC::NA 2010 Surveys have now closed. I'm now reviewing the data to produce the emails for speakers, the website pages for results for public consumption and the feedback documents for organsiers. If all goes well expect to see some postings tomorrow.

In addition, I've been refining the scripts that I use to perform the analysis of the results. One aspect that has been missing for the past few years is the output that provides the raw data. I'm please to say that this has now been added. This now means that when I upload the webpages, I can now provide all the data used to perform the analysis for the webpages. Note that the feedback questions and all the talk evaluations are omitted from this data.

This also means I can at last provide all the data from previous years too. As I need to prepare for YAPC::Europe this week, expect to see the historical data uploaded sometime during August. The data will be provided in JSON format, consisting of 3 parts; 'survey' provides the structure of sections and questions for the survey, 'results' is the analysis used to create the webpages and 'data' is the raw records entered by respondees. The 'data' records consists of 'name', 'value' and 'code' fields. 'name' is the question id that maps into both the 'survey' and 'results' data structures, with the 'value' being the value given by the respondee.

The 'code' field is used to allow corrolation analysis between answers of a single respondee. Note that the corrolation code is only be available in later surveys, and is only applied to questions which avoid respondees from being uniquely identified. The intention for this field is to allow those that want to look into more detailed analysis, to see whether one answer is related another. For example, does the number of events attended have any relation to the involvement in the Perl community?

I'm not expecting a rush of mashups, but hopefully it will be useful to future organisers, where if they get asked a specific questions about attendees, that isn't provided in the basic results on the website, they will have the raw data available to them to try and answer them.

Expect more details tomorrow.