Reader survey results

As promised, here are the results of our reader survey. Thanks to everyone who completed it! This is a long post, and it’s mostly for our own reference. But if you’re curious, read on!

For comparison, the results of our last reader survey two years ago are here.

tl;dr version: Our readership looks about the same as two years ago, except that the gender balance is improving. And readers still like what we’re doing.

We got 426 respondents, most of whom answered every question (the few non-responses were randomly scattered throughout). This isn’t a random sample of any well-defined population. Don’t take these numbers as gospel. It’s probably best to think of them as a rough snapshot of our regular readers.

I had a quick look at some of the crosstabs. Not to test hypotheses, because I didn’t have any, but just to get a sense of the patterns of covariation. If there are any other crosstabs you want me to look at, ask me in the comments and I’ll see what I can do.

58% of respondents are male, 41% female, with one transgender respondent. That’s much less male-biased than last time (when it was 75% male), which is good. No idea to what extent the shift towards a more gender-balanced readership reflects things we’ve done, vs. factors external to us. Though it’s hard to believe the gender balance of ecology as a whole could have changed much in two years, so presumably that particular external factor didn’t drive this shift.

39% of respondents are grad students (about 6/7 of whom are Ph.D. students), 23% postdocs, 25% faculty. The faculty are split evenly between those who’ve been faculty for <6 years, and more experienced faculty. Which surprises me, I’d have thought it’d be a substantial majority of new faculty. Small numbers of respondents are in other occupations. All these numbers are similar to last time. As you’d expect, the gender balance of respondents varies with occupation: student respondents are only slightly male biased, while postdoc respondents are 60% male and faculty and other professional ecologist respondents are roughly 2/3 male. I remain surprised that postdocs are such a large fraction of our readers. But this may just reflect the fact that I’m in Canada where postdocs are relatively rare–maybe I’m forgetting that there are lots of postdocs in the world? And I remain a bit surprised that our grad student readership skews so heavily towards Ph.D. students. Ph.D. students in ecology can’t possibly outnumber M.Sc. students 6 to 1, can they? Three speculative hypotheses: (i) M.Sc. students are less likely than Ph.D. students to be considering a career in academia, and so are less likely to read an academia-focused blog like ours? (ii) M.Sc. students are more likely than Ph.D. students to be working on applied/conservation/management issues, which we don’t write about very much? (iii) M.Sc. students focus more narrowly on their own projects, while Ph.D. students spend more time reading broadly about things not directly relevant to their own work?

53% of respondents are from the US, 9% Canada, 6% UK, 18% non-UK Europe, the rest from elsewhere. That’s roughly in line with where our pageviews come from, so the  respondents are geographically representative of our total readership. The main change from last time is that our UK readership is up relative to other countries. Interestingly, it looks like gender and geography covary. Our US readership is slightly female biased, with readership from the rest of the world being strongly male biased.

Lots of variance in how long respondents have been reading Dynamic Ecology. 17% started reading my posts on Oikos Blog and followed me when I left there. 19% started reading in 2012, the year we started, 39% started reading in 2013, and 25% started reading earlier this year. These numbers are consistent with the ongoing but slowing growth in our pageviews. Consistent with the improvement in our gender balance over time, respondents who followed me over from Oikos Blog skew very heavily male, and respondents who’ve been reading since 2012 skew quite heavily male. In contrast, respondents who started reading us this year skew very slightly female (which is interesting in light of the overall male bias of ecology as a whole). So you need to explain the male bias of my Oikos Blog readership in order to explain the male bias of our readership in the previous survey.

As you’d expect, most respondents are pretty regular readers, sufficiently so that the majority want to be notified immediately when a new post goes up. 10% of respondents read all our posts, 55% read most of them, with 33% reading some of them. 57% of respondents learn about new posts primarily via RSS feed or email subscription. 31% just visit the homepage. Interesting that only 6% primarily find out about new posts by following us on Twitter. We have over 2,200 followers for our robot that tweets new posts, but apparently most of those followers aren’t getting much out of following us.

70% of respondents read other ecology blogs too. Trying to decide if that’s surprisingly low. 55% read other academia blogs, 57% read other non-ecology science blogs, and 52% read other sorts of blogs. 7% of respondents are worryingly single-minded in their attentions to us and don’t read any other blogs. 🙂 In retrospect, it would’ve been interesting to ask people how many blogs they read regularly. My guess is that most respondents read other blogs besides Dynamic Ecology, but only a few others.

There’s wide variance in how often people read the comments. 22% read the comments on most or all of the posts they read, 35% read them on some of the posts they read, 38% read them on few of the posts they read, and 5% are really missing out because they never read the comments. Gender, occupation, and geography don’t covary with how often folks read the comments.

In line with the previous survey, only 30% of respondents have ever commented on Dynamic Ecology, with ~2/3 of those having commented only once or twice. Basically, there’s a very small number of people who comment regularly (hi Jeff! hi Terry! hi Margaret!), and a much larger number of people (though still a small proportion of the readership) who comment if the topic of the post really interests them. The few respondents who have commented more than once or twice are almost entirely male; respondents who rarely or never comment are slightly female-biased compared to the full dataset. Interestingly, there’s no covariation between what fraction of our posts folks read, and how often they comment.

That ~70% of readers never comment would be worrisome if it reflected something about Dynamic Ecology that was off-putting to potential commenters. But that doesn’t seem to be the case. We asked respondents to identify the most important or common reason why they don’t comment. 42% said they don’t feel they have anything to add, 24% said they never comment on blogs or never write on the internet, and 19% said not sure/never thought about it/no reason. Only 5% said they don’t want to get into an argument or worry about being criticized, which is a reassuringly small fraction (plus, it’s not clear what fraction worry about that for reasons that are specific to Dynamic Ecology, as opposed to avoiding commenting on all blogs for that reason). So while we love the commenters we have and would love to have even more, I’m not sure there’s much we can do to address respondents’ main reasons for not commenting. (As an aside, that 24% of respondents have a policy of never writing anything on the internet is a sobering reminder that the people you read and interact with online are a non-random fraction of all people). I was reassured that less than 0.5% of respondents said that they don’t comment because they’d rather discuss the posts on Twitter or Facebook. This confirms my impression that social media isn’t cannibalizing blog comments, at least not ours. There wasn’t any covariation between gender and main reason for not commenting, which surprised me a little as I had thought women might be more likely to say that they never write things on the internet. A couple of respondents did use the “other” option to indicate that they avoid commenting on the internet because of hostility to women.

Most readers either like our substantive discussions and critiques the best (34%), or have no one favorite type of post (34%). That’s actually a pretty big shift from last time, when 61% of respondents picked our substantive discussions and critiques as their favorite type of post and only 24% said they don’t have a favorite type. 11% like our advice posts best, and 9% like the linkfests the best. As you might expect, it’s mostly students and postdocs who like the advice posts and linkfests the best.

42% of readers think we should just keep doing what we’re doing, and the rest don’t agree on what changes they’d like to see. That’s far more votes for the status quo than last time, when only 11% of respondents told us not to change a thing. The single most popular suggested change (picked by 26% of respondents) was for us to start doing posts highlighting and briefly commenting on new ecology papers. I confess I don’t really like writing such posts. In part because I don’t read such posts myself on other blogs. I don’t use blogs to find new ecology papers to read, and if I want a summary of a paper I just read the abstract. And in part because I used to write such posts on Oikos Blog, and they were by far our least-popular sort of post. Perhaps we’ll experiment with posts in which we very briefly highlight papers we thought were really cool, just in case some readers want to use our recommendations as a way of filtering the literature. The next most popular suggested change was more discussion of applied/management/conservation issues (19%). Brian’s the only one of us who has much experience to draw on there, so if we’re to go down that road it’s up to Brian or the guest posters. Speaking of guest posters, there was some support for more of them, whether as more invited guest posts (14%), inviting comments or guest posts from people who disagree with our posts (17%), or adding one or more new bloggers to the team (9%). For the folks who’d like to see more voices on the blog: we agree! But it turns out that, to a good first approximation, every ecologist who wants to write even one blog post already has their own blog. We invite more guest posts than we publish; we’ve found that many folks who are eager to do it in principle struggle in practice to carve out enough time to follow through. We’ll keep trying. Perhaps we’ll consider doing more email interviews, because that way the folks being interviewed don’t have to go to the trouble of writing a post. Although that’s a niche that BioDiverse Perspectives, The Molecular Ecologist, and others already occupy.

We’re getting better in the eyes of readers, or at least holding steady. Respondents were about evenly split among those who think we’ve gotten better (31%), stayed the same (30%), or not sure/can’t tell (36%). Only 1% said we’ve gotten worse. As you’d expect, “not sure/can’t tell” was a common response from those who only started reading us this year, and a rare response from those who’ve been reading us since 2012 or earlier. Of course, anyone who thinks we’re getting worse might well stop reading. But that’s the limitation of any poll of the readership–it doesn’t give you any data on what non-readers think of you. There was no covariation between gender and overall opinion of the direction of the blog.

The feedback was very positive overall. Not surprising, obviously, since anyone who doesn’t like us presumably wouldn’t read us. But still, positive feedback is always nice to hear. 🙂 Most of the comments said “keep it up!” or words to that effect. Following are summaries of the lengthier comments on issues not addressed in the survey:

• Several comments about how the blog helps students, ecologists in small or developing countries, non-ecologists, and people outside academia keep up with what ecologists are thinking and talking about. (I just hope nobody assumes that Meg, Brian, and I are representative of all ecologists!)
• A few comments appreciating our statistical posts
• A few comments appreciating the substantive discussions and debates we host and the respect with which we treat commenters, including those who disagree with the post. We even got one really nice comment to this effect from someone who says they disagree with a lot our posts, particularly on statistical issues. Conversely, one person said that some posts within the past year had been overly confrontational, leading to overly-confrontational comments.
• Several comments appreciating the breadth of topics covered. Conversely, one commenter suggested we stick to our niche and not get more applied, as there are other blogs and outlets for that.
• A couple of commenters wanted to see us engage on Twitter. (Sorry, but Brian and I leave that to Meg, who likes it and is good at it. Brian and I on Twitter would just be sad-making for all concerned.)
• A couple of comments had specific suggestions as to the sort of diversity we should seek if we were to add someone to the team–a grad student or postdoc, someone from outside North America, someone who is not white.
• One comment that Meg’s posts on equality in science, particularly the one on sexual harassment and assault in the field, had been eye-opening for many members of the commenter’s department.
• One request for video posts or podcasts. (Sorry, doubt it’ll happen. Too much work for us.)
• A request for more discussion of how senior scientists can pair up with and help junior scientists.
• One comment worrying that the proportion of career advice/academia/just for fun posts has been too high lately
• One commenter really likes our eye-friendly design (Thank you WordPress 2011 default!)
• A comment appreciating my posts on the IDH and how they turned into a published paper
• A comment appreciating my philosophy of science posts
• One person who finds the Friday links too much to read all at once. (Not sure if this comment refers to the amount of stuff we link to, or to the linkfest posts themselves. Either way, I suggest just skimming for any links that especially interest you; I think that’s what most readers do.)
• One person who said that my posts tend to be too long, and would be even better if shortened. (This person is right.)
• One comment appreciating advice on topics that graduate advisers don’t often address
• And finally, +1000 Internet Points to the person who requested “more prize giveaways”. Ask and ye shall receive! 🙂 As long as what you’re asking for is Internet Points. 🙂