Page MenuHomePhabricator

Article recommendation user testing protocol
Closed, ResolvedPublic

Description

Protocol for T112030. Various models for this type of user study exist. I intend to base my protocol on the ResQue model described here: http://ceur-ws.org/Vol-612/paper3.pdf (see Appendix A), adapted to a semi-structured interview format, which will also include questions focused on learning the existing practices of content translators.

Event Timeline

Capt_Swing claimed this task.
Capt_Swing raised the priority of this task from to Needs Triage.
Capt_Swing updated the task description. (Show Details)
Capt_Swing moved this task to In Progress on the Design-Research board.

he's a draft of the protocol. I'll refine it after the first session tomorrow: https://meta.wikimedia.org/wiki/Research:Evaluating_article_recommendations_for_content_translation/Interview_protocol

Looks good. Just a couple of small comments:

  • When asking about the background, you mention "article translation" (in contrast with other kinds of translation like UI messages). That is great, but we may want to go one step further and ask also a bout some types of "article translation": using translation to create a new article, or using translation to add contents to an existing article. I think the distinction is relevant since article suggestion (and also Content Translation) is currently focused on existing articles that do not exist.
  • It may be interesting to ask about the collaboration aspects as a motivator. For example, checking if the user has done translations as part of editathons or some kind of on-line/off-line campaign.