Wikimedia Meta-Wiki

Community Insights/Community Insights 2020 Report/Endnotes

< Community Insights | Community Insights 2020 Report
2020 Community Insights Report

Below are methodological endnotes corresponding to in-text references.

Brand Awareness

edit
  1. n = 4249 for all contributors responding to the question, 83.7% responding "yes" (95% CI: 82.6%, 84.8%). Respondents who finished less than 50% of the survey were less likely (81.1%) to indicate "yes" (n = 2531, χ2 = 31.306, p = 0.000), though this effect size was quite small (𝜑 = 0.088). When compared to respondents on other continents, a Pearson χ2 test found awareness of the Wikimedia Foundation was higher among those in Northern America (n = 166, χ2 = 12.481, p = 0.000) and lower for those in South America (n = 155, χ2 = 13.175, p = 0.000) and Eastern Europe (n = 221, χ2 = 6.357, p = 0.012). A Fisher’s exact test found Foundation awareness was also lower in Central America (n = 26, p = 0.002).
  2. n = 4434 for all contributors responding to this question in 2018.
  3. “How well does this vision represent your values…” (n = 3975, average score 3.31, 95% CI: 3.29, 3.34. Those who finished less than 50% of the survey had a lower average score of 3.28, n = 2197, U = 1,193,970.5, p = 0.005), though this effect size was very small (r = 0.001). “The Wikimedia Foundation has the right programs and strategies…” (n = 2284, average score of 3.76, 95% CI: 3.72, 3.80). “I am confident in the Wikimedia Foundation’s ability to support the global movement” (n = 3021, average score of 3.68, 95% CI: 3.65, 3.72. Those who finished less than 50% of the survey had a lower average score of 3.64, n = 1459, U = 690,065, p = 0.002), though this effect size was very small (r = 0.001).
  4. Foundation Support is calculated by averaging responses to the three index questions (or those index questions responded to), standardized to a scale of 1 to 5. A Foundation Support score (FSS) was calculated for 4080 contributors who answered any combination of the three index questions (average FSS score was 3.92, 95% CI: 3.89, 3.94). For the following statistical tests, results of a t-test are reported for clarity even though assumptions of normality and outliers are violated. A Mann-Whitney U-test is used in each case to confirm findings. When compared to other contributors, an independent samples t-test found average FSS to be higher for those living in Africa (4.36, n = 21, t = -2.778, p = 0.006; mean rank of 851.94 compared to 618.11, U = 15164, z = 2.754, p = 0.006), Southern Asia (4.12, n = 66, t = -2.347, p = 0.019; mean rank of 741.79 compared to 617.39, U = 29552.5, z = 2.190, p = 0.028), and Northern Europe (4.03, n = 131, t = -2.012, p = 0.044; mean rank of 731.48 compared to 612.60, U = 63657, z = 3.084, p = 0.002). A similar test found average FSS to be lower among contributors living in Eastern Asia (3.64, n = 147, t = -4.172, p = 0.000, not assuming equal variances; mean rank of 490.59 compared to 635.89, U = 52717, z = -4.277, p = 0.000).

Global Advocacy

edit
  1. n = 2101 for all contributors responding to the question, 95% CI: 64.6%, 68.6%. Respondents who finished less than 50% of the survey were less likely (61.8%) to indicate policy knowledge (n = 584, χ2 = 8.224, p = 0.004). When compared to other respondents, a Pearson χ2 test found contributors living in Western Asia (n = 37, χ2 = 9.734, p = 0.002), Southern Asia (n = 56, χ2 = 13.916, p = 0.000), and South America (n = 152, χ2 = 11.178, p = 0.001), newcomers (n = 236, χ2 = 9.110, p = 0.003), on-wiki admins (n = 172, χ2 = 12.560, p = 0.000), and movement organizers (n = 493, χ2 = 94.016, p = 0.000, not weighted by editing activity) were more likely to indicate public policy knowledge in at least one area. Those in Northern America (n = 153, χ2 = 5.368, p = 0.021) and Western Europe (n = 327, χ2 = 28.784, p = 0.000) were less likely to indicate knowledge of public policy.
  2. n = 2831 for all contributors responding to the question, 95% CI: 15%, 17.2%. Respondents who finished less than 50% of the survey were less likely (14.0%) to indicate participation in policy advocacy (n = 1025, χ2 = 4.362, p = 0.037), though this effect size was quite small (𝜑 = 0.039).
  3. n = 2493 for all contributors responding to the question. 74.8% indicated that access was important (95% CI: 73%, 77%), 43% that free expression was most important (95% CI: 42%, 45%; respondents who finished less than 50% of the survey were less likely (40%) to say this was important (n = 729, χ2 = 4.320, p = 0.038), though this effect size was quite small (𝜑 = 0.042)), 29% that copyright was important (95% CI: 27%, 31%), 28% that privacy was important (95% CI: 26%, 29%), and 9.7% that intermediary liability was (95% CI: 8.5%, 11%; respondents who finished less than 50% of the survey were less likely (7.7%) to say this was important (n = 729, χ2 = 4.650, p = 0.031), though this effect size was quite small (𝜑 = 0.043)). A Fisher’s exact test found those in Africa more likely to value knowledge access than other contributors (n = 20, p = 0.007). Compared to other contributors, a Pearson χ2 test found those in Southern Asia (n = 63, χ2 = 3.924, p = 0.048) and South-eastern Asia (n = 50, χ2 = 6.078, p = 0.014) more likely to value access. Contributors in Eastern Asia (n = 144, χ2 = 39.660, p = 0.000) and on-wiki administrators (n = 185, χ2 = 4.340, p = 0.037) were less likely to value access. Contributors in Africa (n = 21, χ2 = 5.737, p = 0.017), Southern Asia (n = 63, χ2 = 7.070, p = 0.008), and South-eastern Asia (n = 50, χ2 = 9.087, p = 0.003) were less likely to value free expression, but those in Eastern Asia (n = 143, χ2 = 8.725, p = 0.003) and Southern Europe (n = 185, χ2 = 7.097, p = 0.008) to be less likely to do so. Contributors in South-eastern Asia were more likely to value privacy (n = 51, χ2 = 3.853, p = 0.050).

Platform Evolution

edit
  1. n = 179 for all volunteer developers reporting the country in which they live. When compared to all contributors, a Pearson χ2 test found volunteer developers more likely to live in Asia (n = 54, χ2 = 15.827, p = 0.000), less likely to live in Europe (n = 74, χ2 = 16.303, p = 0.000), and more likely to be under the age of 25 (n = 43, χ2 = 14.864, p = 0.000).
  2. n = 41 for volunteer developers who began contributing in this capacity in 2018 or 2019. A Pearson χ2 test found newer developers to be more likely to be under 25 (χ2 = 6.686, p = 0.010), to live in Asia (χ2 = 8.571, p = 0.003), and less likely to live in Europe (χ2 = 7.466, p = 0.006) than more tenured developers.
  3. n = 157 for volunteer developers responding to “MediaWiki and related software currently have a safe, healthy, and sustainable codebase” (3.52, 95% CI: 3.37, 3.68). n = 160 for “The code for MediaWiki and related software can be applied to other Wikimedia projects when needed” (3.83, 95% CI: 3.68, 3.97). n = 161 for "The code for MediaWiki and related software is appropriate for Wikimedia and other projects" (3.81, 95% CI: 3.67, 3.95). n = 141 for "The code development processes for MediaWiki and related software make it easy for me to contribute high-quality code" (3.31, 95% CI: 3.10, 3.51). n = 130 for “The MediaWiki deployment tooling for Wikimedia projects is easy to use” (3.24, 95% CI: 3.03, 3.45). n = 127 for “The MediaWiki deployment tooling for Wikimedia projects is effective” (3.51, 95% CI: 3.32, 3.71). For the following statistical tests, results of a t-test are reported for clarity even though assumptions of normality and outliers are violated. A Mann-Whitney U-test is used in each case to confirm findings. When compared to other contributors, an independent samples t-test found volunteer developers living in Europe to be less confident in the health and safety of the codebase (3.33, n = 63, t = 2.049, p = 0.042; mean rank of 65.67 compared to 78.63, U = 2121, z = -1.964, p = 0.050). Similar tests found newer developers more likely to agree that MediaWiki code could be applied to other projects (4.10, n = 31, t = -2.011, p = 0.046; mean rank of 84.15 compared to 64.57, U = 2112.5, z = 2.661, p = 0.008).
  4. Developers indicated whether they experienced each barrier never (1), some of the time (2) most of the time (3), or all of the time (4). n = 111 for volunteer developers responding to "Difficulty setting up development environment" (average score 2.27, 95% CI: 2.11, 2.43). n = 130 responding to “Unclear or missing documentation about how to contribute software” (2.48, 95% CI: 2.33, 2.63).  n = 109 responding to “Slow code review process” (2.51, 95% CI: 2.34, 2.68). n = 130 for "Difficulty finding a person or documentation to get questions answered" (2.46, 95% CI: 2.30, 2.62). n = 126 for “Members of the technical community are not helpful or not friendly" (1.68, 95% CI: 1.54, 1.83). For the following statistical tests, results of a t-test are reported for clarity even though assumptions of normality and outliers are violated. A Mann-Whitney U-test is used in each case to confirm findings. When compared to other contributors, an independent samples t-test found volunteer developers living in Asia less likely to experience slow code review (2.14, n = 35, t = 2.973, p = 0.004; mean rank of 39.34 compared to 56.51, U = 747, z = -2.975, p = 0.003). They were also more likely to report an unfriendly technical community as a barrier to contributing code (1.93, n = 40, t = -2.630, p = 0.010; mean rank of 71.13 compared to 52.70, U = 2025, z = 3.081, p = 0.002).
  5. n = 1353 for contributors responding to “I know how and where to report a technical problem” (3.04, 95% CI: 2.97, 3.10).  n = 1492 for “I have the right buttons or tools to do what I need to do” (3.69, 95% CI: 3.65, 3.74).  n = 1506 for “The software works the way I expect it to” (3.73, 95% CI: 3.69, 3.77). n = 1536 for “The software is easy to use” (3.66, 95% CI: 3.62, 3.71).
  6. For the following statistical tests, results of a t-test are reported for clarity even though assumptions of normality and outliers are violated. A Mann-Whitney U-test is used in each case to confirm findings. When compared to other contributors, an independent samples t-test found contributors living in South America were more likely to report that software is easy to use (3.85, n = 147, t = -2.337, p = 0.020; mean rank of 558.73 compared to 500.88, U = 61966, z = 2.282, p = 0.022). Newcomers were more likely to say the software worked as they expected (3.91, n = 206, t = -3.083, p = 0.002, not assuming equal variances; mean rank of 555.65 compared to 482.58, U = 81839, z = 3.419, p = 0.001) and that they had the right buttons or tools (3.85, n = 218, t = -2.422, p = 0.016; mean rank of 525.87 compared to 479.13, U = 80080.5, z = 2.271, p = 0.023). Those in Southern Asia (3.60, n = 41, t = -4.023, p = 0.000, not assuming equal variances; mean rank of 597.02 compared to 435.21, U = 15171.5, z = 3.300, p = 0.001), Western Asia (3.81, n = 28, t = -3.433, p = 0.001; mean rank of 620.78 compared to 435.79, U = 12205.5, z = 3.321, p = 0.001), very highly active editors (3.35, n = 966, t = -4.632, p = 0.000, not assuming equal variances; mean rank of 1051.09 compared to 941.01, U = 548291.5, z = 4.426, p = 0.000, not weighted by editing activity), movement organizers (3.54, n = 128, t = -4.991, p = 0.000; mean rank of 625.01 compared to 461.39, U = 41265.5, z = 4.968, p = 0.000), and administrators (3.62, n = 162, t = -6.687, p = 0.000; mean rank of 625.51 compared to 463.75, U = 34001, z = 4.465, p = 0.000) were all more likely to say they knew how to report technical problems.
  7. For the following statistical tests, results of a t-test are reported for clarity even though assumptions of normality and outliers are violated. A Mann-Whitney U-test is used in each case to confirm findings. When compared to other contributors, those in Eastern Asia were less likely to agree that the software is easy to use (3.40, n = 118, t = 3.107, p = 0.002, not assuming equal variances; mean rank of 429.41 compared to 516.21, U = 36567, z = -3.049, p = 0.002), that they had the right buttons or tools (3.40, n = 119, t = 3.500, p = 0.001, not assuming equal variances; mean rank of 397.56 compared to 498.65, U = 34109.5, z = -3.764, p = 0.000), and that they knew where to report technical problems (2.70, n = 105, t = 2.750, p = 0.007, not assuming equal variances; mean rank of 362.92 compared to 448.14, U = 26915, z = -3.016, p = 0.003). Contributors who identified as women were less likely to agree that the software worked as they expected (3.52, n = 139, t = 3.098, p = 0.002, not assuming equal variances; mean rank of 413.91 compared to 496.75, U = 37277, z = -3.161, p = 0.002), as were contributors not fluent in English (3.65, n = 381, t = -2.311, p = 0.021, not assuming equal variances; mean rank of 467.51 compared to 515.91, U = 108269, z = 2.618, p = 0.009). Those not fluent in English were also less likely to know where to report bugs (2.87, n = 336, t = -3.079, p = 0.002; mean rank of 400.13 compared to 465.38, U = 91109, z = 3.471, p = 0.001).
  8. n = 1017 for contributors responding to “When the Wikimedia Foundation creates software, they collaborate well with people in my project or community” (3.31, 95% CI: 3.25, 3.37). n = 1118 for those responding to "The Wikimedia Foundation prioritizes the software that is most needed for Wikipedia and Wikimedia projects" (3.55, 95% CI: 3.50, 3.60). For the following statistical tests, results of a t-test are reported for clarity even though assumptions of normality and outliers are violated. A Mann-Whitney U-test is used in each case to confirm findings. When compared to other contributors, The most active (tier 4) editors were less likely to agree that the Foundation collaborates well (3.11, n = 800, t = 4.401, p = 0.000; mean rank of 756.28 compared to 850.37, U = 411208.5, z = -3.719, p = 0.000; not weighted by editing activity level), as were those in Western Europe (3.17, n = 199, t = 2.639, p = 0.009, not assuming equal variances; mean rank of 280.93 compared to 319.56, U = 26771, z = -2.277, p = 0.023), and Northern Europe (2.89, n = 73, t = 4.096, p = 0.000; mean rank of 223.21 compared to 319.41, U = 9538, z = -3.810, p = 0.000). Newcomers (3.58, n = 139, t = -3.812, p = 0.000; mean rank of 361.06 compared to 312.77, U = 35097.5, z = 2.721, p = 0.006), contributors under the age of 25 (3.60, n = 154, t = -4.045, p = 0.000; mean rank of 356.41 compared to 291.62, U = 34557.5, z = 3.833, p = 0.000), and those living in Southern Asia (3.78, n = 45, t = -4.211, p = 0.000, not assuming equal variances; mean rank of 446.67 compared to 305.19, U = 12935, z = 4.498, p = 0.000), South-eastern Asia (3.77, n = 35, t = -2.850, p = 0.004; mean rank of 387.36 compared to 308.85, U = 9359, z = 2.288, p = 0.022), and South America (3.69, n = 99, t = -4.139, p = 0.000; mean rank of 354.40 compared to 305.75, U = 25112, z = 2.417, p = 0.016) were all more likely to agree that the Foundation collaborates well.
  9. For the following statistical tests, results of a t-test are reported for clarity even though assumptions of normality and outliers are violated. A Mann-Whitney U-test is used in each case to confirm findings. When compared to other contributors, the most active editors were less likely to agree that the Foundation prioritizes the right software (3.41, n = 835, t = 3.255, p = 0.000; mean rank of 824.72 compared to 893.32, U = 339612, z = -3.055, p = 0.002; not weighted by editing activity level), as were contributors in Northern Europe (3.23, n = 76, t = 3.374, p = 0.001; mean rank of 300.21 compared to 358.63, U = 13735.5, z = -2.102, p = 0.036). Youth (3.69, n = 166, t = -2.519, p = 0.012; mean rank of 374.68 compared to 335.72, U = 39703, z = 2.166, p = 0.030), and those living in Latin America (3.75, n = 134, t = -2.796, p = 0.005; mean rank of 393.38 compared to 347.19, U = 37730.5, z = 2.367, p = 0.018), Sub-Saharan Africa (4.34, n = 7, t = -2.294, p = 0.022; mean rank of 562.50 compared to 352.72, U = 3354, z = 2.701, p = 0.007), and Southern Asia (3.78, n = 46, t = -2.100, p = 0.041, not assuming equal variances; mean rank of 438.90 compared to 350.90, U = 12293, z = 2.450, p = 0.014) were all more likely than others to agree that the Foundation prioritizes the right software.
  10. n = 1289 for contributors responding to “The Wikimedia Foundation produces high-quality software” (3.67, 95% CI: 3.62, 3.72). For the following statistical tests, results of a t-test are reported for clarity even though assumptions of normality and outliers are violated. A Mann-Whitney U-test is used in each case to confirm findings. When compared to other contributors, The most active (tier 4) editors were less likely to agree that the Foundation produces quality software (3.52, n = 915, t = 3.658, p = 0.000; mean rank of 907.41 compared to 995.14, U = 411208.5, z = -3.719, p = 0.000; not weighted by editing activity level). Those who live in Western Europe were more likely to agree (3.78, n = 272, t = -2.126, p = 0.034, not assuming equal variances; mean rank of 446.05 compared to 403.78, U = 63998.5, z = 2.283, p = 0.022).

Thriving Movement

edit
  1. n = 1570 for all indicating both tenure (year when began contributing) and country or territory of residence. A Fisher’s exact test found new contributors more likely to live in Africa than tenured contributors (p = 0.007). A Pearson χ2 test also found new contributors more likely to live in Asia (χ2 = 46.400, p = 0.000) and Latin America or the Caribbean (χ2 = 5.619, p = 0.018), and less likely to live in Europe (χ2 = 35.299, p = 0.000) and Northern America (χ2 = 5.153, p = 0.023) than more tenured developers. Even though Oceania includes Melanesia, Micronesia, and Polynesia, the only respondents from this region lived in Australia and New Zealand.
  2. n = 1691 for all indicating gender identity. 86.9% identified as male (95% CI: 85.3%, 88.5%), and 11.6% identified as women (95% CI: 10.1%, 13.1%). A Pearson χ2 test found that newer contributors were more likely to identify as women (χ2 = 31.350, p = 0.000).
  3. The source for a global estimate of metropolitan area populace is Moreno-Monroy A.I., Schiavana M., Veneri P. (2020). Metropolitan areas in the world. Delineation and population trends. Journal of Urban Economics, published online ahead of print, https://www.sciencedirect.com/science/article/pii/S0094119020300139. n = 2382 for contributors responding to a question about rural/urban location, and 34.52% said they live in the country or in a small city or town (95%CI: 32.26, 36.79). A Pearson χ2 test found new contributors more likely to live outside of a metropolitan area (χ2 = 4.632, p = 0.031).
  4. n = 1545 for contributors indicating their level of education by number of years. 16 years of formal education was used as a reasonable equivalence for a 4-year college degree. A Pearson χ2 test found new contributors less likely to have 16 or more years of formal education (χ2 = 34.171, p = 0.000). Estimate of worldwide rates of post-secondary degree completion is sourced from Barro R., Jong-Wha L. (2013). A new data set of educational attainment in the world, 1950-2010. Journal of Development Economics, 104, 184-198.
  5. n = 244 for new contributors responding to “easy to learn editing policies” with an average score of 3.44 (95% CI: 3.31, 3.58). n = 253 for “easy to learn the editing tools” with an average score of 3.66, (95% CI: 3.53, 3.79). n = 243 for “I feel like I understand the rules” with an average score of 3.78 (95% CI: 3.67, 3.89). n = 251 for “learning to edit has been easy” with an average score of 3.85 (95% CI: 3.73, 3.97). n = 222 for “other editors have been kind” with an average score of 3.88 (95% CI: 3.77, 3.99). n = 228 for “other editors have been helpful” with an average score of 3.79 (95% CI: 3.67, 3.91). The Newcomer Empowerment Score is calculated by averaging responses to the six index questions (or those index questions responded to), standardized to a scale of 1 to 5. A score was calculated for 256 new contributors who answered any combination of the six index questions (average score was 3.74, 95% CI: 3.65, 3.83).
  6. For the following statistical tests, results of a t-test are reported for clarity even though assumptions of normality and outliers are violated. A Mann-Whitney U-test is used in each case to confirm findings. Newcomers who were youth had a higher average empowerment index than other newcomers (4.00, n = 66, t = -3.608, p = 0.000; mean rank of 104.46 compared to 79.45, U = 4235.5, z = 3.089, p = 0.002). They were also more likely to agree that learning to edit was easy (4.07, n = 66, t = -2.250, p = 0.026; mean rank of 95.52 compared to 80.38, U = 3775.5, z = 2.243, p = 0.025), that they understood content editing rules (4.05, n = 66, t = -3.911, p = 0.000, not assuming equal variances; mean rank of 101.50 compared to 76.00, U = 4088, z = 3.664, p = 0.000), and that learning editing policies (3.75, n = 62, t = -2.444, p = 0.015; mean rank of 96.51 compared to 76.00, U = 3640.5, z = 2.757, p = 0.006) and tools (4.06, n = 66, t = -4.103, p = 0.000, not assuming equal variances; mean rank of 104.57 compared to 76.96, U = 4260, z = 3.794, p = 0.000) was easy. Contributors in South America were more likely to agree that learning to edit had been easy (4.16, n = 38, t = -2.778, p = 0.007, not assuming equal variances; mean rank of 119.12 compared to 95.49, U = 3785.5, z = 2.589, p = 0.010). Newcomers who knew of opportunities to be mentored by a more experienced editor (54.7% of newcomers, 95% CI: 47.8%, 61.7%) had a higher overall empowerment index than other newcomers (4.00, n = 111, t = -5.989, p = 0.000, not assuming equal variances; mean rank of 103.60 compared to 64.65, U = 5242, z = 5.177, p = 0.000). Newcomers who had ever attended an in-person movement event were more likely to understand rules for content (4.11, n = 28, t = -2.080, p = 0.039; mean rank of 125.98 compared to 97.61, U = 2723, z = 2.582, p = 0.010) and find other editors helpful (4.22, n = 30, t = -2.816, p = 0.005; mean rank of 113.73 compared to 90.21, U = 2606, z = 2.196, p = 0.028) than newcomers who had not attended an event. Editors living in Africa were more likely to find other editors helpful (4.53, n = 8, t = -2.282, p = 0.023; mean rank of 130.50 compared to 88.11, U = 1008, z = 2.394, p = 0.017) and kind (4.56, n = 6, t = -2.050, p = 0.042; mean rank of 123.00 compared to 84.65, U = 717, z = 2.050, p = 0.040).
  7. For the following statistical tests, results of a t-test are reported for clarity even though assumptions of normality and outliers are violated. A Mann-Whitney U-test is used in each case to confirm findings. Newcomers who identified as women (3.50, n = 56, t = 2.071, p = 0.042, not assuming equal variances; mean rank of 85.57 compared to 107.71, U = 2931.5, z = -2.288, p = 0.022), live in Eastern Asia (3.36, n = 35, t = 3.009, p = 0.003; mean rank of 72.70 compared to 107.69, U = 1838, z = -3.151, p = 0.002), or who are not fluent in English (3.48, n = 64, t = -3.339, p = 0.001; mean rank of 82.13 compared to 114.00, U = 5621, z = 3.385, p = 0.001) had lower empowerment index scores than others. Those not fluent in English were less likely to agree that learning to edit generally had been easy (3.59, n = 63, t = -2.621, p = 0.010, not assuming equal variances; mean rank of 85.62 compared to 110.01, U = 5136, z = 2.996, p = 0.003), as well as learning tools (3.40, n = 64, t = -2.662, p = 0.008; mean rank of 88.38 compared to 109.79, U = 5103, z = 2.547, p = 0.011) and policies (3.51, n = 61, t = -2.676, p = 0.008; mean rank of 83.15 compared to 106.92, U = 4879, z = 2.774, p = 0.006), and that they understood rules for editing content (3.45, n = 59, t = -3.243, p = 0.020, not assuming equal variances; mean rank of 77.93 compared to 107.60, U = 4899.5, z = 3.672, p = 0.000). Contributors in Eastern Asia also indicated greater difficulty learning editing tools (3.30, n = 35, t = 2.094, p = 0.042, not assuming equal variances; mean rank of 83.64 compared to 103.83, U = 2199, z = -2.023, p = 0.043) and policies (2.96, n = 33, t = 2.818, p = 0.005; mean rank of 76.00 compared to 102.32, U = 1904, z = -2.577, p = 0.010). Those identifying as women were less likely to agree that other editors had been helpful (3.58, n = 55, t = 2.045, p = 0.042, not assuming equal variances; mean rank of 71.91 compared to 95.63, U = 2227, z = -2.845, p = 0.004) and that they understood the rules for editing content (3.37, n = 54, t = 2.872, p = 0.005, not assuming equal variances; mean rank of 80.50 compared to 103.40, U = 2622, z = -2.757, p = 0.006) than other contributors.
  8. The Belonging index is an averaged score of disagreement or agreement with two statements: (1) I feel like I belong in the Wikimedia movement; and (2) I feel respected in the Wikimedia movement. A Belonging index score was calculated for 1939 people responding to these questions, with an average score of 3.56 (95% CI: 3.52, 3.60). Those who finished less than 50% of the survey had a lower average score of 3.41, n = 241, U = 134425.5, p = 0.003), though this effect size was very small (r = 0.002). The Engagement index is an averaged score of disagreement or agreement with five statements: (1) I would recommend Wikimedia projects as a great place to contribute; (2) The vision of the Wikimedia movement is to freely share in the sum of all knowledge. This vision inspires me to contribute more to Wikimedia projects than I would somewhere else; (3) I take pride in contributing to the Wikimedia projects; (4) I often think about quitting as a contributor to Wikimedia projects (scale reversed); and (5) I expect to be contributing to Wikimedia projects two years from now. An Engagement Index score was calculated for 2411 people responding to these questions, with an average score of 4.01 (95% CI: 3.98, 4.04). Those who finished less than 50% of the survey had a lower average score of 3.91, n = 606, U = 356857, p = 0.003), though this effect size was very small (r = 0.002).
  9. For the following statistical tests, results of a t-test are reported for clarity even though assumptions of normality and outliers are violated. A Mann-Whitney U-test is used in each case to confirm findings. Contributors to the Japanese language Wikipedia (2.86, n = 56, t = 5.890, p = 0.000; mean rank of 444.58 compared to 720.81, U = 22131.5, z = -4.947, p = 0.000), the Dutch language Wikipedia (3.32, n = 50, t = 1.897, p = 0.058; mean rank of 576.23 compared to 714.59, U = 23298.5, z = -2.215, p = 0.027), and the simplified Chinese language Wikipedia (3.33, n = 55, t = 1.901, p = 0.057; mean rank of 572.06 compared to 714.41, U = 21530.5, z = -2.198, p = 0.028) all had lower average Belonging Index scores, as did those living in Eastern Asia (3.28, n = 132, t = 3.919, p = 0.000, not assuming equal variances; mean rank of 454.73 compared to 583.95, U = 43915.5, z = -4.022, p = 0.000).  Contributors to the Japanese language Wikipedia (3.62, n = 81, t = 4.646, p = 0.000; mean rank of 649.82 compared to 922.02, U = 47033, z = -4.468, p = 0.000), the Dutch language Wikipedia (3.76, n = 60, t = 2.514, p = 0.012; mean rank of 682.25 compared to 916.95, U = 32837.5, z = -3.128, p = 0.002), the German language Wikipedia (3.76, n = 136, t = 3.515, p = 0.001, not assuming equal variances; mean rank of 762.82 compared to 919.82, U = 76498.5, z = -3.025, p = 0.002), the Russian language Wikipedia (3.80, n = 104, t = 2.800, p = 0.005; mean rank of 745.72 compared to 918.47, U = 59070.5, z = -2.956, p = 0.003), those living in Eastern Asia (3.88, n = 147, t = 2.839, p = 0.005; mean rank of 512.58 compared to 625.30, U = 55421, z = -3.362, p = 0.001), and those not fluent in English (3.95, n = 491, t = -3.244, p = 0.001; mean rank of 655.09 compared to 574.84, U = 186300, z = 3.580, p = 0.000) all had lower average Engagement Index scores.
  10. For the following statistical tests, results of a t-test are reported for clarity even though assumptions of normality and outliers are violated. A Mann-Whitney U-test is used in each case to confirm findings. Contributors who had attended movement events (3.85, n = 395, t = -7.061, p = 0.000, not assuming equal variances; mean rank of 713.41 compared to 583.77, U = 138483, z = 5.206, p = 0.000), movement organizers (3.96, n = 162, t = -6.132, p = 0.000, not assuming equal variances; mean rank of 923.64 compared to 694.53, U = 86490.5, z = 45.5121, p = 0.000), contributors living in Southern Asia (3.90, n = 62, t = -2.830, p = 0.006, not assuming equal variances; mean rank of 735.27 compared to 565.71, U = 27895.5, z = 3.249, p = 0.001), in Africa (4.10, n = 19 t = -2.447, p = 0.015; mean rank of 750.75 compared to 568.95, U = 11876, z = 2.254, p = 0.024), and on-wiki administrators (3.83, n = 185, t = -3.906, p = 0.000; mean rank of 752.10 compared to 604.98, U = 50843, z = 3.496, p = 0.000) each had higher average Belonging Index scores than others. Contributors who had attended movement events (4.19, n = 406, t = -4.937, p = 0.000; mean rank of 768.29 compared to 629.19, U = 154413, z = 5.162, p = 0.000), movement organizers (4.27, n = 168, t = -4.535, p = 0.000; mean rank of 1135.06 compared to 897.03, U = 111555.5, z = 4.485, p = 0.000), contributors to WikiData (4.24, n = 79, t = -2.701, p = 0.007; mean rank of 1137.33 compared to 903.17, U = 63175, z = 3.326, p = 0.001), to Commons (4.10, n = 301, t = -2.179, p = 0.029; mean rank of 997.67 compared to 900.58, U = 168176.5, z = 2.398, p = 0.016), to Spanish Wikipedia (4.11, n = 233, t = -2.009, p = 0.045; mean rank of 1001.59 compared to 893.36, U = 182616.5, z = 2.793, p = 0.005), and those living in Africa (4.58, n = 23, t = -3.389, p = 0.001; mean rank of 896.35 compared to 613.93, U = 17717, z = 3.526, p = 0.000) each had higher average Engagement Index scores than others.
  11. n = 1637 for all indicating both role and country or territory of residence. n = 233 for movement organizers providing this information. A Pearson χ2 test found movement organizers more likely to live in Southern Asia (10.3% compared to 3.2% of others, χ2 = 16.287, p = 0.000), and Western Asia (χ2 = 5.619, p = 0.000), as well as to identify as women (χ2 = 15.304, p = 0.000). Fisher’s exact tests found organizers more likely to live in South-eastern Asia (8.7% compared to 2.8% of others, p = 0.002) and Western Asia (7.1% compared to 2.3% of others, p = 0.004) than other contributors. A Pearson χ2 test found organizers less likely to live in Eastern Asia (3.2% compared to 9.1% of others, χ2 = 5.295, p = 0.021) and Europe (47% compared to 58% of others, χ2 = 5.619, p = 0.018).
  12. n = 166 for administrators indicating where they live. A Pearson χ2 test found administrators more likely to live in Southern Asia (6.6% compared to 3.5%, χ2 = 3.914, p = 0.048) and less likely to live in Latin America or the Caribbean (4.2% compared to 13%, χ2 = 11.073, p = 0.001). A Fisher’s exact test found administrators more likely to live in Oceania (2.4% compared to 0.7%, p = 0.047).
  13. n = 163 for administrators responding to "When I am presented with a challenge in my on-wiki role, I usually know how to address it" (4.06, 95% CI: 3.95, 4.16). n = 160 for those responding to "...usually able to address it in the way I think is most appropriate" (4.05, 95% CI: 3.96, 4.14). n = 160 for those responding to "...usually have the resources I need to address it" (3.93, 95% CI: 3.82, 4.04). The Administrator Empowerment Score is an averaged score of disagreement or agreement with these three statements. The average Index score was 4.00 (95% CI: 3.92, 4.08), calculated for 164 administrators. For the following statistical tests, results of a t-test are reported for clarity even though assumptions of normality and outliers are violated. A Mann-Whitney U-test is used in each case to confirm findings. Administrators who live in South America indicated higher empowerment (4.49, n = 7, t = -2.393, p = 0.018; mean rank of 54.50 compared to 27.58, U = 157.5, z = 2.868, exact p = 0.001) while those not fluent in English indicated lower levels of empowerment (3.74, n = 34, t = -2.600, p = 0.013, not assuming equal variances; mean rank of 15.35 compared to 31.90, U = 371.5, z = 3.005, p = 0.003). n = 160 for administrators responding to "Other volunteers usually provide the help and support I need to execute my on-wiki role" (3.84, 95% CI: 3.70, 3.97). n = 94 for those responding to "If I contact them, the Wikimedia Foundation usually provides the help and support…" (3.16, 95% CI: 2.92, 3.40). For the following statistical test, results of a t-test are reported for clarity even though assumptions of normality and outliers are violated. A Mann-Whitney U-test is used to confirm findings. Administrators living in Southern Europe were more likely to indicate that other volunteers supported them in their role (4.14, n = 16, t = -2.207, p = 0.029; mean rank of 53.00 compared to 27.28, U = 250, z = 3.517, exact p = 0.000).
  14. n = 363 for organizers responding to "I know whom to ask for help with my Wikimedia program or partnership" (3.84, 95% CI: 3.74, 3.94). n = 369 for those responding to "I have people who can help me with my..." (3.99, 95% CI: 3.90, 4.07). n = 335 for those responding to "I have access to the right tools or software to implement my..." (3.58, 95% CI: 3.48, 3.68). n = 347 for those responding to "I have access to documentation and training materials to help me implement my..." (3.61, 95% CI: 3.50, 3.71). n = 338 for those responding to "I have access to communications materials and tools that help me reach the people I want to engage with my..." (3.59, 95% CI: 3.48, 3.69). n = 339 for those responding to "I have access to funding that I need to implement my..." (3.05, 95% CI: 2.92, 3.18). n = 335 for those responding to "I have access to the non-monetary resources (e.g., event space) that I need for implementing my..." (3.38, 95% CI: 3.25, 3.50). The Organizer Empowerment Index is an averaged score of disagreement or agreement with six three statements. The average Index score was 3.58 (95% CI: 3.50, 3.67). For the following statistical tests, results of a t-test are reported for clarity even though assumptions of normality and outliers are violated. A Mann-Whitney U-test is used in each case to confirm findings. Organizers who live in Africa indicated higher empowerment (3.86, n = 29, t = -2.122, p = 0.034; mean rank of 228.16 compared to 179.10, U = 6181.5, z = 2.408, p = 0.016), as did those in Central America (4.25, n = 6, t = -2.170, p = 0.031; mean rank of 267.58 compared to 181.59, U = 1584.5, z = 1.985,p = 0.047). Organizers in Eastern Asia had lower empowerment scores on average (2.97, n = 20, t = 3.666, p = 0.000; mean rank of 124.78 compared to 186.38, U = 2285.5, z = -2.545, p = 0.011).
  15. n = 342 for organizers responding to "The Wikimedia Foundation provides the help and support I need to plan and execute my projects(s)" (3.26, 95% CI: 3.14, 3.38). n = 357 for those responding to "My Wikimedia community or affiliate provides the help and support I need to plan and execute my projects(s)" (3.71, 95% CI: 3.60, 3.81). For the following statistical tests, results of a t-test are reported for clarity even though assumptions of normality and outliers are violated. A Mann-Whitney U-test is used in each case to confirm findings. Organizers who live in Sub-Saharan Africa were more likely to agree that the Foundation provided needed support (4.11, n = 19, t = -5.506, p = 0.000, not assuming equal variances; mean rank of 240.66 compared to 163.09, U = 4382.5, z = 3.552, p = 0.000), as were those in South-eastern Asia (4.05, n = 21, t = -5.308, p = 0.000, not assuming equal variances; mean rank of 235.05 compared to 162.97, U = 4705, z = 3.459, p = 0.001) and who identified as women (3.51, n = 91, t = -2.366, p = 0.019; mean rank of 186.95 compared to 160.93, U = 12826.5, z = 2.286, p = 0.022). Those in Western Europe (2.94, n = 47, t = 2.164, p = 0.031; mean rank of 138.23 compared to 172.29, U = 5369, z = -2.342, p = 0.019) and Northern America (2.91, n = 44, t = 2.260, p = 0.024; mean rank of 135.88 compared to 172.30, U = 534988.569, z = -2.436, p = 0.015) were less likely to agree that WMF provided needed support.
  16. n = 1818 for all contributors responding to "In the last 12 months, how often have you felt unsafe or uncomfortable contributing to Wikimedia projects online?" 35.3% said they had this experience at least once (95% CI: 33.1%, 37.5%). A Pearson χ2 test also found women contributors more likely to have this experience (n = 173, χ2 = 12.157, p = 0.000), as well as contributors to the Chinese (n = 58, χ2 = 14.197, p = 0.000) and Portuguese (n = 88, χ2 = 10.058, p = 0.002) language Wikipedias, those who live in Eastern Asia (n = 127, χ2 = 9.595, p = 0.002) or South America (n = 151, χ2 = 5.260, p = 0.022), and contributors not fluent in English (n = 412, χ2 = 6.744, p = 0.009). A Fisher's exact test found women contributors were significantly more likely than others to say they had felt unsafe among contributors to the German (n = 7, 0.034) and Dutch (n = 7, p = 0.013) language Wikipedias. A Pearson χ2 test also found women contributors more likely to have this experience (n = 39, χ2 = 15.617, p = 0.000) among contributors living in Western Europe.
  17. n = 581 for all contributors responding to the question: "When you felt unsafe or uncomfortable contributing to Wikimedia projects online, was it ever due to the way others respond to any of the following personal characteristics?" 16.1% selected "your level of education" (95% CI: 13.1%, 19.1%); 12.4% selected "your age" (95% CI: 9.7%, 15.1%); 9.8% selected "your race or ethnicity" (95% CI: 7.4%, 12.2%).
  18. n = 1919 for all contributors responding to "In the last 12 months, have you been harassed, bullied, or attacked in any Wikimedia online or offline space?" 14.2% said they had this experience at least once (95% CI: 12.6%, 15.8%). No statistically significant difference was found by gender for experiencing harassment generally, but the original question included an "I'm not sure" option (excluded from final analysis)-- including this option in the analysis showed women contributors to be much more likely to say they weren't sure (9.2% compared to 5.3% of contributors not identifying as women, n = 195, χ2 = 4.798, p = 0.028). Among WikiData contributors, a Fisher's Exact test found women users more likely to say they had experienced harassment (n = 6, p = 0.008). Regardless of gender, a Pearson χ2 test found contributors to the English (n = 134, χ2 = 12.794, p = 0.000) and Chinese (n = 58, χ2 = 4.872, p = 0.027) language Wikipedias were more likely to say they had been harrassed, as were movement organizers (n = 157, χ2 = 7.754, p = 0.005; this effect was even larger when analyzing organizers' experience without weighting for editing activity: 29.3%, n = 467, χ2 = 42.269, p = 0.000) and those living in Northern America (n = 156, χ2 = 19.223, p = 0.000). Those living in Europe were less likely to say so (n = 905, χ2 = 5.568, p = 0.018). n = 1950 for all contributors responding to "In the past 12 months, how often have you been sexually harassed on Wikimedia projects, such as sexual remarks made toward you or someone touching your body inappropriately?" 1.9% said they had this experience at least once (95% CI: 1.3%, 2.5%). A Fisher's Exact test found women contributors more likely to have experienced sexual harassment in an online Wikimedia space (n = 185, p = 0.024).
Retrieved from "https://meta.wikimedia.org/w/index.php?title=Community_Insights/Community_Insights_2020_Report/Endnotes&oldid=20459034"