Results From The Survey, 2011Published 10 years, 6 months past
On Tuesday — and I fully acknowledge the fact that it’s taken me until now to blog this is emblematic — A List Apart published the results of the fifth annual A List Apart Survey for People Who Make Websites. This includes anonymized data sets for the bulk of the survey, as well as standalone data sets for postcodes and a few of the answer sets for questions that allowed “Other” as an option. (Note that these last were shuffled-then-sorted, and were not filtered for potentially objectionable content. They are what they are.)
If you really want the TL;DR version, the results are largely the same as they’ve been in the past. The gender ratio, for example, is still in the vicinity of 5-to-1 male-to-female, with half a percent answering Other (a new option in the 2011 survey). Most respondents are in the age range 19-44 and live in the United States. And so on. That might sound like I’m bored by the results, but their very consistency even as the number of respondents has dropped over five years fascinates me.
It did take quite a while to publish the results. I feel personally very bad about the delay, because I run the numbers and it just took me a long time to get them run. Partly, I admit, I put it off because some of the numbers in previous years were a royal pain to generate, thanks in part to the way the data is formatted and in part because of the fine slicing that was done. This was finally addressed through various means, and now the report is done. I can’t thank Sara Wachter-Boettcher enough for her keen editing eye and firm strategic oversight, not to mention writing all the commentary text to accompany the charts. If not for her, the report might still not be done. And of course without the unwavering support and dedication of Jeffrey Zeldman, the survey might not have existed at all.
So we’ve done this five times, and the results are consistent. What now? There is much to discuss, and the answers aren’t yet clear; but I do know that this project brings me more professional pride than almost anything I’ve ever done. It tells us a lot about ourselves — and in a profession that is often characterized by single-person “web teams” and distributed offices, one which may never have a certification process or other form of registry, that’s something valuable. Thank you for helping us see ourselves a little bit more clearly.
I wish I had the stats chops to tease apart the questions like:
* How different is the perception of gender bias between men/women?
* Same thing for race?
* Controlling for education, years of experience, how do salaries break down by age/race/gender?
…etc. My wife the economist could probably do this, but I don’t think I can afford her.
We did some of that in previous years, Jemaleddin, and I could probably come up with the numbers for at least the first of those to post here, and probably the first two.
The third is a LOT harder, and much depends on your starting assumptions. For example, in last year’s report we had a chapter titled “Evidence of Bias” where we fine-sliced a couple of cohorts (e.g., ”Perception of gender bias by gender, US Master’s degree with 5+ years in the field”) but we were mostly guessing about what would be illuminating cohorts. Those are definitely the areas where we’d love to see people produce numbers on the cohorts that specifically interest them. I would be beyond ecstatic to link to other findings along those lines.
Following up: here are those first two pieces of data, Jemaleddin. Apologies for the presentation; my CSS isn’t anything like what Jason designed for the actual reports.
Percentages are based on 15,481 respondents (98.6% of all respondents)
Percentages are based on 15,450 respondents (98.4% of all respondents)
Wow – thank you, Eric! I didn’t expect my idle curiosity to turn into that! That’s a lot to think about!
No problem, Jemaleddin. For a slightly less detailed version of those charts as well as several related graphs, see chapter 6, “Perceptions of Bias”.