diff --git a/literature/notes b/literature/notes index 9914cd4a0e37ab4561f8ab62fb9c4999689ad7a7..f608010a66accfcceb0d389a94e673b97e8fe96f 100644 --- a/literature/notes +++ b/literature/notes @@ -2275,3 +2275,138 @@ works, but also amateur videos using their music in the background, or musicians tribute to a band by playing live along with the commercial recording as a backing track" "This kind of content fingerprinting, being both easy and oblivious to nuance, encourages these kinds of shotgun tactics." //compare blog post on upload filters + +=================================================================== +\cite{HalKitRied2011} + +purpose of reverts: +"fix mistakes, repair vandalism, and help enforce +policy." + +effect of reverts on newcomers and more experienced editors; +demotivate editors; reduce contributions vs higher quality edits +effect of reverts according to tenure of the reverting editors; + reverts by anonymous editors not very impactful + reverts by more experienced users deminish contributions significantly +"Essentially, +editors reverted by anonymous editors recover to the same +average level of activity within a couple of weeks, but those +reverted by named editors do not recover for at least one +month (if ever)." +"we suspected that the long- +term effect on reverted editors could have been due to edi- +tors being demotivated enough to leave Wikipedia entirely." + +contribution: "ours is the first study we are aware of to quantify the +impact of reverts on editor behavior." +Findings in a nutshell: +"(1) reverts do have a negative +impact on editor contribution and survival, especially for +newcomers; and (2) when editors do continue to contribute +after a revert, the quality of their contributions increases." + +What can findings be used for: +"design of intelligent tools for supporting reverts that enhance +their beneficial effects while minimizing costs." + +4 Research questions: +RQ1: How does being reverted affect the quantity +of editor work? +RQ2: How does being reverted affect the quality of +editor work? +RQ3: How does being reverted affect communica- +tion? +RQ4: How does experience moderate the effects of +reverts on contribution? + +editing is easy for everyone: also for malicious/biased users +so reverting is also easy -> reduces cost for fixing damage + +total percentage of reverts has increased over time + +"tasks that may pro- +duce high group value can increase individual motivation;" +"Getting reverted may make individuals feel that their con- +tributions are not valued by the group and are not leading to +positive group outcomes, resulting in demotivating effects." + +"being reverted could be part of the learning +process for editors." + +Wikipedia’s Bold, Revert, Discuss cycle + +more policies and guidelines -> newcomers are reverted at higher rates + +Data sample: 400 000 revisions; January 2010 data dump +200 000 samples of not reverted revisions used as a control group +no two revisions performed by the same editor + +Quantity measures: +Revisions/day +Words added/day: number of non-stop words added to articles per day + +Quality measures: +Reverts/revision: "the proportion of an editor’s revisions +that have been reverted in a given timespan." +PWR/word: "the average number +of revisions that words added by an editor persist. Higher +quality contributions should, on average, last longer." + +Boldness measures: +Words changed per revision +Establishment of words removed: the average PWR of +words that an edit removes +users may appear to make higher quality contributions by making "safer" edits + +Productivity measures: +combines quality and quantity to estimate impact +PWR/day: + +all these measures focused on articles; don't consider improvements on policy pages, etc; + +Measuring communication: +- about articles: article talk page -> Article Talk revisions/day +- personal communication between editors: user talk page -> User Talk revisions/day + +"editors who are reverted are less likely +to get reverted in the future." + +"After a revert, old-timer editors do experience a tempo- +rary reduction to their article activity, but they return to the +level of activity of their not-reverted counterparts within two +weeks of being reverted. For newbie editors, the difference +in the activity delta is both stronger and longer-lasting. Re- +verted newbies take more than four weeks to return to the +activity levels of not-reverted newbies." + +but less experienced editors learn more from reverts and have bigger improvements in their article quality + +"activity +of editors reverted by newbies will recover within two weeks, +the activity of editors reverted by old-timers did not recover +in the four observed weeks." + +"Another interpretation is that old-timers +have an enhanced ability to identify unconstructive editors +and chase them away" // wistful thinking + +interestingly: +"Under our measure of pro- +ductivity (PWR/day), the net effect of reverts on Wikipedia +is positive: on average, an editor who is reverted produces +more persistent words per day – even if we include those +editors who withdraw from Wikipedia in the calculation!" +"Our research suggests that overall reverting activity in +Wikipedia is healthy and valuable," +"there are +specific cases in which reverting activity might be managed +better" + +"the reverting editor should be +encouraged to provide clear feedback to help the reverted +editor grow as a member of the community." // yeah. I bet this never happens +"Newcomers should be reached out +to actively to help them become socialized into Wikipedia." +"more curmudgeonly old-timers should be +kept away from newcomers until they have gained some ex- +perience in the system."