A total of 50,386 images were uploaded to Commons for the six reported events, and 3,456 (7%) images uploaded during these events have already been included in Wikimedia project pages.
The average photo upload initiative uploads just over 2,000 images to Commons. The average amount of images placed in Wikimedia projects was 885. Events that had fewer than 5,000 image uploads had a higher image rate, but we need more data to evaluate that.
We collected photo upload counts for all six photo upload initiatives. A total of 50,386 photos were uploaded to Wikimedia Commons from these events. Upload counts ranged from 87 to 33,086. The average event uploaded 2,206 photos.[2] Regarding unique image use in Wikimedia projects, a total of 3,456 (7%) unique images are used in Wikimedia projects from these events.Unique image use ranged from 21 to 1824. The average number of unique images used from a photo upload initiative is 318 images[3] (see Graph 2).
A very small percentage of images from photo upload initiatives are awarded quality status on Wikimedia Commons.
Out of 50,386 photos uploaded for the reported photo upload initiatives, 156 were rated as Quality, Valued, and/or Featured Image (see Graph 2). Out of those 156:
152 were rated as Quality Images
1 was rated as Valued Images
3 were rated as Featured pictures
Some photos are included in multiple categories.
Graph 2: Increasing quality. As highlighted in the bar graph above, the biggest impact in terms of increasing quality, regarding the six reported photo upload events, is that 50,386 photos were added to Commons, of which 3,456 (7%) were used in project pages, 152 were rated as "Quality Images," 1 as a "Valued Image," and 1 was a "Featured Picture"
The average photo upload initiative spends just over $0.60 cents USD for each photo uploaded.
Three implementations (50%) reported budget data, and we were able to determine the cost of each image based on those budgets. The cost ranged from $0.06 USD to $1.70 USD. The average was $0.64 USD (see Graph 3).
Graph 3: Dollars to photos/media added. As illustrated in the box plot, photo upload events demonstrated a rather narrow range of variation in the input dollars to the amount of content added; from as little as $0.06 US per photo/media added to $1.70 US per photo/media added. The median reported input budget for the three events reported was $0.64 US per photo.
The number of photos uploaded ranged from 87 to 33,087; more research needs to be done in order to determine how the number of participants affects the rate of uploaded photos.
Graph 4: Participants to photos. As illustrated in the graph, the number of photos uploaded ranged from 87 to 33,086 uploaded by 4 to 365 contributors, of which, 6% to 24% photos had been used (as of November 4, 2013). In the graph, the number of participants is illustrated by bubble size and label. It appears that, the number of participants did not necessarily relate to the use rate of uploaded photos, but that the use rate was higher for events in which the number of uploads did not exceed 5,000 photos. More data are needed to confirm and/or further explore this observation.
References
↑Note: Although "content production" is a direct product of the program event itself and technically a program output rather than outcome most of the program leaders who participated in the logic modeling session felt this direct product was the target outcome for their programming. To honor this community perspective, we include it as an outcome along with quality improvement and retention of "active " editors
↑Image use appeared higher for those events for which more than a month or two had passed. Specifically, looking at the three events which had reached more than a one or two month follow-up point, the use rate of uploaded images was 19%.
Recruitment and retention of active editors
As of reporting time, we were able to evaluate recruitment and retention of active users for five of the photo upload events for which we had usernames. A few things to remember with this section:
An active retained user is considered one that makes 5 or more edits a month. We examined both the "pages created" (a new page gets created when someone uploads a file) and "edit count" using Wikimetrics to measure user activity on Wikimedia Commons[1].
A "survived" editor is defined as one that made at least one edit in a given window of time after the event ended.
User retention was measured for both survival and active user retention. Retention was assessed for the first three months after the event (Quarter 1) and the second set of three months (Quarter 2: months four, five, and six) following the event. Here we considered users as surviving if they had made a single contribution, and "active" if they had averaged [2]5 or more edits a month, during the follow-up window.
Six months after the end of these photo upload events, 0% of new users were retained as active contributors to Wikimedia Commons.
91 new users participated in the five events for which usernames were available for tracking. 44 (48.3%) of these uploaded five or more files and 45 (49.5%) made five or more edits during their upload event time. With regard to retention (see Graph 5):
Three-month follow-up (First quarter)
Survival (Any contributions)
11 (12.1%) of new users had made at least one edit and uploaded at least one new file to Wikimedia Commons.
Active (Average 5+ contributions a month)
4 (4.4%) of new users had made 15+ edits and uploaded at least 15 files during the first quarter of retention follow-up.
Six-month follow-up (Second quarter)
Survival (Any contributions)
2 (2.3%) of the 86 new users eligible for 6 month follow-up had made at least one edit and uploaded at least one new file to Wikimedia Commons.
Active (Average 5+ contributions a month)
0 (0.0%) of the 86 new users eligible for 6 month follow-up had made 15+ edits and uploaded at least 15 files during the first quarter of retention follow-up.
Graph 5: New user retention.
This bar graph shows the number users who actively contributed to the event and the number of new users retained for periods of three months at three and six months following the events. A total of 91 new users participated in the five events for which usernames were available.
As with Wiki Loves Monuments, existing editor retention is higher than new user retention for photo upload events, and participants demonstrate a higher rate of “active” participation following these events as compared to Wiki Loves Monuments. Six months after the events, 73.6% of existing user participants were actively editing on Wikimedia Commons and 64.2% were actively uploading files.
77 existing users participated in the five events for which usernames were available. 52 (67.5%) of these uploaded five or more files and 60 (77.9%) made five or more edits during the events. With regard to retention (see Graph 6):
Three-month follow-up (First quarter)
Survival (Any contributions)
67 (87.0%) of existing users had made at least one edit and 61 (79.2%) had uploaded at least one new file to Wikimedia Commons.
Active (Average 5+ contributions a month)
58 (75.3%) of existing users had succeeded at making 15+ edits during the first quarter of retention follow-up.
51 (66.2%) of existing users had uploaded 15+ files during the first quarter of retention follow-up.
Six-month follow-up (Second quarter)
Survival (Any contributions)
44 (83%) of the 53 existing users eligible for 6 month follow-up had made at least one edit and 39 (73.6%) had uploaded at least one new file to Wikimedia Commons.
Active (Average 5+ contributions a month)
39 (73.6%) of the 53 existing users eligible for 6 month follow-up had succeeded at making 15+ edits during the second quarter of retention follow-up.
34 (64.2%) of the 53 existing users eligible for 6 month follow-up had uploaded an average of 15+ files during the second quarter of retention follow-up.
Graph 6: Existing user retention, 2012
This bar graph shows the number users who actively contributed to the event and the number of new users retained for periods of three months at three and six months following the events. A total of 77 existing users participated in the five events for which usernames were available.
References
↑Pages created and edits were assessed for namespace 6 for files on Commons.
↑by counting those who had made 15 or more edits for the three month quarter. However, users could make 15 or more edits during any period of time, not five or more each month necessarily. This solution allows us to even much of the month-to-month variation of users to get at those users who are more or less staying active on the project. For example, someone who made five edits in the first month, three edits in the second month, and six edits in the third month following the event would be counted as active for the full three-month quarter. At the same time, someone who made all their contributions in the first month following the contest, and nothing further, would also be counted as surviving or active for the quarter.
