268
edits
2021 Survey Results: Difference between revisions
→Anonymization of Survey Data to Preserve Privacy
|
5d. In the second copy of the survey data, delete the columns containing language codes, country codes and names, and all demographic data except gender, but preserve the gender information, so that analysis of responses by gender will be possible.<br>
6. In both spreadsheets, delete the 'Response ID', 'Date submitted', 'Last page', 'Start language', 'Seed', 'May we share (anonymously, of course) your comments with the OpenStreetMap community?', 'Optional Demographic: Are you a... [Other]' columns from both spreadsheets to make the possibility of merging of the datasets less likely.<br>
7. Run script to scramble randomly the responses by gender in order to prevent merging with the first spreadsheet and thereby deanonymizing the data.
==Plan for Presentation of Survey Results==
Please note: this is the plan for data release, but it is subject to change if difficulties or problems are encountered in processing the survey data. Respecting privacy and anonymity of respondents is our highest priority, followed closely by presentation of the data and analysis of the data as quickly as we can.
Feb 15 -<br>
1. Release participation numbers by country.<br>
2. Release list of participating countries.
Feb 18 -<br>
1. Release comments permitted by assent, in English, with translations of comments in other languages into English, after removing any identifying information (some respondents gave their names)
Feb 20 -<br>
1. Obtain normalization data and normalize survey results.<br>
2. Anonymize the survey results. This will result in two sets of spreadsheets, one set with gender data, and one set without gender data.
Feb 22 -<br>
1. Post raw anonymized survey data in .ods format to wiki.osmfoundation.org (remove comments from the spreadsheet, group any variables with n<20)<br>
2. Release normalized and anonymized data based on OSMstats data on mapping activity by country in .ods format.<br>
3. Release summary statistics and graphs based on both raw and normalized data via wiki.osmfoundation.org, blog, OSMWeekly.
Feb 24 - Release cross-sectional P-tests (probability tests):<br>
* male vs. female+nb+other for F1-F5
* mappers vs. non-mappers for F1-F5
* OSMF members vs. non-members for F1-F5
* cross-regional for F1-F5:
** EU vs. Rest of World
** Germany vs. Rest of EU
** Germany vs. Rest of World
** U.S. vs. Rest of World
** U.S. vs. EU
** France vs. Rest of World
** UK vs. Rest of World
** Russia vs. Rest of World
** Ukraine vs. Rest of World
** Asia vs. Rest of World
** Brazil vs. Rest of World
** Latin America vs. Rest of World
** Africa vs. Rest of World
** Other Europe vs. Rest of World
** Other No Am/Caribbean vs. Rest of World
** India vs. Rest of Asia
Mar 4 – GeoMob podcast interview with Ed Freyfogle on the survey.
Mar 10-11 - Series of three BBB video conferences open to OSM community 8 hours apart to answer questions about data, at noon UTC, 2000 UTC, and 0400 March 11 UTC (these correspond to 0700, 1500, and 2300 EST).
| |||