Some years ago, I was a graduate student in economics, and one of my advisors taught me that well-done tables, all at the end, will allow a reader to capture the entire narrative of your paper quickly.
On the other hand, if I’m actually reading the paper from start to finish, tables at the end mean constant flipping back and forth.
On the other other hand, when I’m going back to papers later, it’s much easier to find the results I’m looking for if the tables and figures are together in the back.
When I review economics papers (with tables at the back), I end up keeping two PDF files of the paper open at once, one open at the text and the other open at the tables and figures.
As you can see, the vast majority of respondents prefer tables interspersed throughout the paper. Now, it’s clear from responses to the tweet that not all respondents were economists. I suspect that most respondents were people who read economics papers, so if your goal is communicating with readers, this may be a useful metric. If your goal is to impress academic economists on hiring committees, who may or may not be well represented on Twitter (I don’t have a strong prior), then it may not be so helpful.
Discussants to the tweet highlighted the points I make above, that reading on a device is easier with tables interspersed, and that tables at the back make for easier skimming.
But I’d note that if you’re going to do tables at the back, do them right: Provide clear titles and notes such that the tables really can stand alone. As one economist noted, “Tables at the end are okay with decent notes, horrible without.”
If you have a website, a few people suggested having two versions, which is an interesting idea (and more work than I will credibly do).
So, writers, pick your poison. You’re likely to annoy at least a few readers either way!