Talk:Community Insights/Community Insights 2020 Report/Thriving Movement

Hello RMaung (WMF) ,

I would like to call your attention to something that doesn't seem to be correct in the data presented. The graphic compare new editors (apparently joined during 2018) with the established ones. This is like comparing apples and oranges, and consequently leads to non substantiated conclusions. For instance, they state 22% of the new editors are woman against less than 10% of established editors, that data is used to conclude that now there is considerably more women joining, so gender policies are a success. But for all that we know, it could always have been like that, only retention is quite low. I actually suspect that nowadays, unfortunately, it may be below the old 10% figure. For the results to make sense, new editors in 2018 should be compared to new editors in 2017, 2016 and so on, and then check what differences appear, and not with the group of established editors, which do not tell absolutely nothing of meaningful. Hypothetically, imagine you had 30% women joining in 2017, but with a very low retention. This would mean that there actually was a reduction in women joining and staying, not an increase, as is claimed in the results. The same for all other graphic, including the ones related to the Global South. When it is well known that there is quite a severe problem with retention both in gender and editors from the Global South, simply stating the new editors that joining, with no clue at all about their retention, and comparing those numbers with editors that already passed the proof of retention, do not seem to produce any meaningful results, and even less to support conclusions like (my bold) "People who identify as women are also underrepresented among Wikimedia contributors, and here we also see evidence of small gains in gender representation among newcomers" (small gains compared to what exactly?), " In 2019, newcomers were more likely to live in Latin America or the Caribbean and especially in Africa and Asia, and less likely to live in Northern America and Europe, than more tenured contributors". (yes, this may be true, but it doesn't tell anything meaningful at all, since for what we know, it may always have been like that).

I may be missing something, but it do look like a number of the conclusions are not supported by the data presented. Can you please clarify that? Thanks, --- Darwin ^Ahoy! 00:23, 25 February 2021 (UTC)

Hi DarwIn, We are careful in this report not to make conclusions about the success of gender or other policies (I'm not sure what in the report you are referring to here)-- as I agree, the issue of retention is a big one. What we do claim is that newer editor retention might be an important key in increasing the geographic and gender diversity of Wikimedians as a whole, as people newer to the movement are more likely to be women and live in underrepresented geographies. An explanation for what we see when we compare newcomers to tenured contributors could be either that (1) yes, contributors are growing more diverse or (2) we struggle to retain women and contributors from underrepresented geographies. We'll know more about what is going on as we compare these data points over time, as we will do in the forthcoming report on the 2020 survey data (coming April 2021). I disagree that comparing newcomers and tenured contributors, even across a single point in time, is meaningless, however. It provides an important starting point as we think about the importance of both attracting and retaining editors with diverse lived experiences. Thanks, RMaung (WMF) (talk) 19:51, 25 February 2021 (UTC)

@RMaung (WMF): Where have you got that "more likely to be women" thing? The data presented there says that only 22% of newcomers are likely to be women, that don't support at all what you say.--- Darwin ^Ahoy! 12:38, 4 March 2021 (UTC)

@DarwIn: Newcomers are more likely (22%) to be women than more tenured contributors (9.8%) are. RMaung (WMF) (talk) 16:49, 4 March 2021 (UTC)

@RMaung (WMF): However, the report puts the context not in comparison with the existing community in the same time frame, but with former realities. Examples: "small gains in geographic representativeness"; "here is slightly better representation among newcomers who have joined the movement in the last two years; "we also see evidence of small gains in gender representation among newcomers"(emphasis mine). In fact, you don't see any "gains" at all there, nor is anything better than 2 years before, since you're not comparing the same data. Do you understand what the problem with that report is? --- Darwin ^Ahoy! 17:09, 4 March 2021 (UTC)

@DarwIn: This year's report does not make comparisons over time. We define "newcomer" in this report as someone who started contributing in the last two years, and those who have been contributing longer as "tenured." We are comparing newcomer diversity to tenured editor diversity to explore where we might be seeing gains in diversity among newcomers, and/or a loss of diversity among tenured contributors. RMaung (WMF) (talk) 17:39, 4 March 2021 (UTC)

@RMaung (WMF): Then you can't use terms like "gain", since you may not be gaining absolutely nothing when comparing newcomers with tenured editors, since there is the possibility - or even probability - that the large majority of those newcomers leaving the projects in the short term - as actually seems to be the case. You can't also say that you have now better results than 2 years ago, when the results from 2 years ago are nowhere to be seen. Can you adjust the report, accordingly, removing or fixing those qualifiers?--- Darwin ^Ahoy! 20:32, 4 March 2021 (UTC)

@RMaung (WMF): Any news on this? --- Darwin ^Ahoy! 14:52, 25 March 2021 (UTC)