Change Details

This task [and its subtasks] is about the work involved with instrumenting both the existing and revised [1] adding/modifying links and citations workflows on mobile, in VE. == Open questions == - [ ] How will we be able to interact with the data our instrumentation/event logging is gathering? - [ ] Will we need to oversample Edit events? == Background == We know what proportion of edits using the visual editor involve using common editing features. [2] We do not //yet// know how many contributors take which actions //within// these workflows. We are pursuing instrumentation to help us answer questions related to this second part. Questions like... - What % of contributors who start to add/modify a link and citation are successful in doing so? How do these proportions compare across the existing and revised workflows? -- This kind of information could be helpful in identifying opportunities to make adjustments that will help these contributors find their way (read: be successful in) adding/modifying links/citations. - Do "Edit cards" encourage more people to engage more deeply with adding/modifying links and citations? -- Asked another way: "What percentage of contributors who are shown the edit card (revised) or context item (existing) end up engaging with it?" - What % of edit sessions, in which contributors successfully add/modify a link, result in them attempting to save their edits?" How do the revised and end existing flows compare in this way? - Do "Edit cards" cause more people to add/modify links and citations more efficiently? This work sits within our larger effort to improve the mobile edit completion rate. Work that depends on us being able to know what contributors actually do after starting an editing session. == Instrumentation needs == The instrumentation specification is defined in [this doc](https://docs.google.com/document/d/1n0MKZWiNm2ZviBHiImnNRnUjSp5e5JqfXorYtIsuP-0/edit?pli=1#) (publicly commentable). ###Resources - [Schema:VisualEditorFeatureUse](https://meta.wikimedia.org/wiki/Schema:VisualEditorFeatureUse) - [Wireframes: linking w/ contextual inputs](https://wikimedia.invisionapp.com/freehand/document/s1duRw6io) - [Mobile context items hypotheses](https://phabricator.wikimedia.org/T221247) - [Desktop VE feature use](https://docs.google.com/spreadsheets/d/1FZsIZk7ty9xB92TQRBnxdmXpt3vEcWCjxZy0YPt85Nc/edit#gid=0) --- ###References 1. **Revised workflows**: https://phabricator.wikimedia.org/T221247 2. **Feature use data**: https://www.mediawiki.org/wiki/VisualEditor_on_mobile_report

This task [and its subtasks] is about the work involved with instrumenting both the existing and revised [1] adding/modifying links and citations workflows on mobile, in VE. == Open questions == - [ ] How will we be able to interact with the data our instrumentation/event logging is gathering? - [ ] Will we need to oversample Edit events? == Background == We know what proportion of edits using the visual editor involve using common editing features. [2] We do not //yet// know how many contributors take which actions //within// these workflows. We are pursuing instrumentation to help us answer questions related to this second part. Questions like... - What % of contributors who start to add/modify a link and citation are successful in doing so? How do these proportions compare across the existing and revised workflows? -- This kind of information could be helpful in identifying opportunities to make adjustments that will help these contributors find their way (read: be successful in) adding/modifying links/citations. - Do "Edit cards" encourage more people to engage more deeply with adding/modifying links and citations? -- Asked another way: "What percentage of contributors who are shown the edit card (revised) or context item (existing) end up engaging with it?" - What % of edit sessions, in which contributors successfully add/modify a link, result in them attempting to save their edits?" How do the revised and end existing flows compare in this way? - Do "Edit cards" cause more people to add/modify links and citations more efficiently? This work sits within our larger effort to improve the mobile edit completion rate. Work that depends on us being able to know what contributors actually do after starting an editing session. == "Done == |**Status**| **Task** | **Owners** | | Define research questions instrumentation needs to answer |@ppelberg | | | Define metrics to be derived from instrumentation| @ppelberg + @Neil_P._Quinn_WMF | | | Define data instrumentation will need to generate | @ppelberg + @Neil_P._Quinn_WMF | | | Confirm instrumentation definition | @dlynch + @Neil_P._Quinn_WMF | | | Define schema and queries and other attributes of the logging @dlynch + @Neil_P._Quinn_WMF | | | Create queries that will generate results | @Neil_P._Quinn_WMF | | | Make legal aware that schema privacy review will be needed | @JTannerWMF | | | Email "privacy@" with link to Phabricator ticket for review | @JTannerWMF | | | Finalize schema | @dlynch + @Neil_P._Quinn_WMF | | | Implement instrumentation | @dlynch | | | Test instrumentation | @dlynch + @Neil_P._Quinn_WMF + @ppelberg | | | Update data dictionary | @Neil_P._Quinn_WMF | | | Activate production level logging | @dlynch | == Instrumentation needs == The instrumentation specification is defined in [this doc](https://docs.google.com/document/d/1n0MKZWiNm2ZviBHiImnNRnUjSp5e5JqfXorYtIsuP-0/edit?pli=1#) (publicly commentable). ###Resources - [Schema:VisualEditorFeatureUse](https://meta.wikimedia.org/wiki/Schema:VisualEditorFeatureUse) - [Wireframes: linking w/ contextual inputs](https://wikimedia.invisionapp.com/freehand/document/s1duRw6io) - [Mobile context items hypotheses](https://phabricator.wikimedia.org/T221247) - [Desktop VE feature use](https://docs.google.com/spreadsheets/d/1FZsIZk7ty9xB92TQRBnxdmXpt3vEcWCjxZy0YPt85Nc/edit#gid=0) --- ###References 1. **Revised workflows**: https://phabricator.wikimedia.org/T221247 2. **Feature use data**: https://www.mediawiki.org/wiki/VisualEditor_on_mobile_report