N=1, but I personally really hate it when surveys are grouped into pages and I have no idea how many questions are going to be on it. I often click off in that case or at least delay in responding (and often forget to come back).
Note that just knowing the number of pages isn’t enough, as I’ve often seen surveys where the number of questions or time/question skyrockets on one of the later pages.
If it’s just grouping to make it easy to skip, something like sections with different background colors might be a good option?
To the Marketing Team, I want an explanation on how this question got into the survey and why data on the sex gender identity is relevant.
Edit from Fri Oct 10 12:17:51 PM CEST 2025:
This comment got flagged as being inappropriate and hateful, and (rightfully) hidden. Thanks to Moderation Team for offering me the chance to rephrase and explain my previous questions.
I am fully an ally with trans people, some close friends are transitioning right now, other are still considering their options in the current political climate. I was unaware of the possibility to misunderstand my perception of this question in the Nix Community Survey, I do get that more context would have been necessary.
With the current shifts from democracies to authoritarian regimes in the States and Europe aswell, this kind of data may be misused. There is no way of skipping this questions, it’s required and the selection may not be undone. The survey being anonymous as stated in the first post, is questionable. We all know how the web works, how anonymity may not be provided anymore.
I am glad that Nix community is inclusive. I see the point in wanting to have data on such aspects of us all. But: There is one question on residence, one on age, and two on gender identity. None on income, political orientation, health, etc.
With this context, I just do not see the point. I stand by my wish to have answers on these two questions.
Simply offering an inclusive gender identity question acknowledges and respects the existence of nonbinary, transgender, and gender-diverse people which are a significant portion of the user base relative to the general population. In general, it is helpful to understand the composition of the user base. Over time and with enough data, we can see how internal or external factors change this composition. This can help governing bodies make informed decisions about the project. We can use it for mentorship programs or diversity initiatives. We can use it in collaboration with the Foundation Board to submit for grants.
None of the questions are mandatory. You can skip answering any of them. You can also pick Prefer not to say as well for this particular question.
These are the extra metadata fields I see in the response data:
Response ID - (from the limesurvey docs) each response is given a unique id. Looks like it increments from 1 for every new user that starts the survey.
Date submitted - In the data I have, for submitted surveys this is always 1980-01-01 00:00:00. Otherwise it is empty for surveys that were not submitted.
Last page - (from the limesurvey docs) lists the page number where your respondent decided to quit the survey. To check whether he filled it out or not, see also the “completed” column
Thanks for the reply and providing some context on my initial questions. I know such data points are good for a lot of reasons, also malicious one’s. What about the discrepancy of survey questions on diversity? Why the highlight only on trans – especially right now?
Getting grants approved more easily should not be a motive.
Sorry, I was wrong on that. I apologise.
Good. But that’s not the point. Declaring a “survey is anonymous” promises an environment that leads no trail back to the participant. As far as I understand the infra repo, nginx access logs are not disabled on that host.
Please, let’s only collect data we know we need and we know we can protect.
Also, Marketing Team should consider a privacy policy for the LimeSurvey.
Getting grants approved more easily should not be a motive.
I agree that it should not be the only or even the main motive, but that is not what I read the reply as saying. There were many reasons given.
I expect that these things happen on a best-effort community-led basis, and I’d assume ignorance before malicious intent about the access logs. I happen to agree this shouldn’t be the default behavior for surveys conducted officially and put together a PR to try and tackle this more productively. Specific things are always easier to act on; in an extremely unofficial capacity, thanks for investigating.
Correct and they were just given as examples. The point is that having data–and even more so a history of data–can help inform decisions in the future. I do not have any specific use for much of the data right now. In fact, only the recently submitted questions by teams have immediate use. However, we must be forward looking and ask ourselves, “What do we want to do in the future regarding the project? What kind of initiatives would benefit the project? What kind of data do we need to make those decisions? How do we get that data?” Figuring this out is currently best effort for the team but even if we do not have perfect prompts and get perfect data, getting something is better than nothing.
I really need to find a subject matter expert in this field.
Any Survey Researchers, Data Analysts, Market Research Analysts or Social Scientists out there? Anyone? Please help
Regarding this, it is not the Marketing Team’s responsibility nor do we have any authority on the matter. Hopefully, you can come to some resolution with the Infrastructure Team in your PR.
I’ve done those things, but I can’t see how to help because I think you’re doing fine!
Sensitivity about what questions are asked is normal (and arguably good) but it’s a worse problem for you than usual because your stakeholders hold such a very broad range of political opinions and none of them gets an exra vote by being your boss. I can’t think of anything you can do about that that you’re not already doing.
I appreciate the kind words @Mapybara. There are many areas where the Marketing Team needs help but I will focus on surveys here.
@Arsleust, @fricklerhandwerk, and many others did a tremendous amount of collaboration and deliberation to improve the surveys in the previous years. However, without a subject matter expert driving this forward right now, it falls upon the Marketing Team, and ultimately me, to do what we think is best. I have a wide breadth of knowledge in STEM fields but very limited knowledge when it comes to determining how to best craft surveys. I want an expert that knows the things we do not know. Specifically here are some pointed thoughts I have about the people who would be well suited to do this job and some lingering questions I have:
Someone who is passionate about surveys, metrics, data and the science behind it all. The reason being that we need someone who is informed by recent research, prior art, and best practices.
Someone who will be in ongoing communication with the various teams and governing bodies of the project to ask what kind of information they need. Then they would be able to craft that into meaningful prompts for future surveys.
I think there is a place for smaller, more focused surveys (besides the NixCon survey). They would also be the one to figure this out.
Data collection is only one half. Then comes data processing and conveying that data in a meaningful way. Are we getting the most out of the data we have?
Are we asking the right questions?
For the questions we do ask, are we asking them in the right way?
What questions are we missing?
What things have we not even considered because we are not experts in this field?
I do not expect a single individual to take all this on. It is a preposterous amount of work. I sincerely believe we need a sub-team to marketing that can focus on research, data collection, and aggregation. I have some leads on where to find professionals that may be willing to volunteer their time to help with web design and branding. I have no such leads for this effort.
This is what survey professionals usually concentrate on, in my experience. I don’t think it needs any special skills except normal communication skills. It does need time though.
As for data analysis, I have a PhD in stastical inference but my views are a bit unorthodox. Anyway, for various reasons, I always recommend not to do any statistical inference unless you absolutely have to (and I can’t think why you would have to). Just present simple summaries of the raw results in as pretty a way as you can.
(There are a few technicalities to even that, such as choosing axes, but they’re not difficult and I bet you’re already on top of them.)
Forgot to update and close the loop. The PR to disable access logs got merged in around 10 days ago!
As a bit of a different question @djacu should this thread still be stickied? If I’m not misunderstanding something, the survey is still open for about a month and I don’t seem to see this thread pinned to the top of discourse anymore.
I only discovered about the survey now (and answered right away). In previous years, I think I probably heard about it ealier through my Discourse summary emails, but this year, perhaps because of all the discussions regarding SC elections, I didn’t (and I only eventually accidentally stumbled upon it while looking more in depth at Discourse than usual).
I can only upvote the suggestion to pin the thread as a banner on Discourse. I don’t know if that’s something that is usually done, but that’s definitely doable (and we have done it on the Coq/Rocq Discourse when we did our own survey several years ago).
I see that such a banner is already on the NixOS website, which is very good, but I don’t access the main page of the NixOS website very often myself (as I don’t need to). If the banner was also on NixOS Search, I would have seen it many weeks ago, I think.
