Wikipedia:WikiProject Wikidemia/feedback to contributors

Central question: What is the effect of feedback?

Research outline:

  • Describe existing feedback mechanisms.
  • Develop ways of measuring characteristics (quantity, valence, detail, person-specific or work-specific) of feedback.
  • Develop ways of measuring contributor motivation (number and length of edits per unit of time).
  • Interventions to manipulate feedback:
Pick a sample of contributors.
Randomly divide into N groups.
One group is the pure control.

(The following are example interventions...)

In one group, we take one or more recently-edited articles by those contributors and place, on those articles' pages, notes of the form: "Jane Doe recently contributed feedback to this article. Please give her feedback on her talk page.
In one group, we take one or more recently-edited articles by those contributors and place, on those articles' pages, notes of the form: "Jane Doe recently contributed feedback to this article. If you think she made a positive contribution, please give her feedback on her talk page.
In one group, the researcher or confederates provide positive feedback (eg, "I liked your recent edits to the article on Bob Dylan.")
In one group, the researcher or confederates provide negative feedback (eg, "I didn't like your recent edits to the article on the Rolling Stones.")
These last two treatments are problematic because 1) they can't be organically replicated after the conclusion of the study, 2) the confederates might be detected by the contributors in these groups.


An alternative approach for implementing this type of treatment could potentially try to enlist the help of members of the welcoming committee to randomize the order in which they welcome new editors.