Initiatives

Numeric Exploration

Understanding reader, editor and wiki-specific phenomena through quantitative analysis

NY Public Library
photo by Thomas Hawk, CC BY-SA 2.0, via Flickr

Overview

In addition to it's qualitative and experimental work, the Wikimedia Foundation Design Strategy team supports product development through quantitative methods. From tracking key characteristics of the 732 wiki projects worldwide to testing/validating hypotheses about the factors that drive momentum in the Free Knowledge Movement, these sorts of studies use numeric data to provide insights that enable the Product teams to make more informed decisions.

Studies

  1. Momentum: Contribution, Content, & Reading in Wiki Growth

    This study test some key hypotheses: that building awareness of the brand drives more people to read our content; that by growing readership the Movement itself grows as readers are converted into content contributors; that as the community grows so does the amount and diversity of content offered; and that more, and more diverse, content attracts more readers.

  2. Wikimedia Commons User Interviews and Data

    A project that combines usage data and in-depth interviews with Commons users to understand the current state of the platform.

  3. Templates and Trust-o-meters: Towards a widely deployable indicator of trust in Wikipedia

    This work identifies three key challenges: 1) empirically determining which metrics from community approaches most impact reader trust; 2) validating indicator placements and designs; and 3) demonstrating that such indicators can both lower trust and increase perceived trust in the system when appropriate.

  4. Long Tail Topic & Influence Ontology

    This work provides a method for visualizing and comparing topical coverage from one wiki project to another for the purpose of better understanding where knowledge gaps and subject matter expertise exist across the Free Knowledge Movement.

  5. Trusted: Signals, Inferences & Indicators

    The goal of this work is to develop systems for automatically detecting and characterizing the editorial debates behind Wikipedia articles for the purpose of surfacing indications of trustworthiness to the reader. This project is a collaboration between the Wikimedia Foundation and the CMU Human Computer Interaction Institute.

  6. Wiki Comparison

    This initiative, lead by Neil Shah-Quinn, established a baseline for comparing key characteristics of 732 wiki projects (number of editors, amount of content, readers, readers by platform, etc).