Professional cosmology is dominated by white men. Stumble into one of our workshops and this fact will be immediately obvious to you: most cosmologists look like more or less like me. This leads to all sorts of issues that my field, like other parts of the physical sciences and the tech industry, are having to come to terms with. I went to an eye-opening session at the annual meeting of the American Astronomical Society a few years back, on “Women in Astronomy.” That was the first time that I heard the expression “diversity is excellence,” and it has stayed with me. The context at the time was faculty hiring committees: I heard the story of an astronomy department that hires their faculty by seeking the person who is the most different from the group that is already in place. As you can imagine, this is not standard hiring practice, but as they had realized, it’s an excellent way of making sure that your lunchtime conversations for the next 30 years are as interesting as possible! Diversity is excellence: as a group, the more varied experiences you can collect, the better understanding you can collectively achieve.
The primary goal of observational cosmology right now is measuring the rate of expansion of the Universe as accurately as possible. It turns out that our Universe is expanding faster and faster, rather than slowing down before recollapsing. No-one knows why this is, but to test some of the possible explanations, we’re trying to make the best measurement we can. In the early noughties, we had several low precision measurements that all seemed to agree with each other about the existence of the acceleration, and that predicted a “concordance cosmology” with a very simple form. In that picture, Einstein’s law of gravity has a small correction that gives rise to the accelerating expansion without changing any of the predictions for how stars and planets move. That’s good, but this simple model would leave us with this extra term that would just have to be as big as it is and that’s that, which would be curious. Instead, if the Universe was full of some mysterious new kind of “dark energy” that somehow was causing the acceleration, we’d expect to see a slightly different set of measurements - and so it really matters exactly what numbers we get.
The tricky part is the following. Because we already saw the first few measurements agree with each other, we cosmologists somehow now expect our new measurements to also agree with those first ones, and hence confirm what we already know. There’s a danger that we might stop checking and re-checking our analyses (which are typically quite complicated, with lots of room for mistakes) once we’ve got “the right answer” - and potentially miss out on making a new measurement that disagrees, interestingly, with what we already saw.
A solution to this problem of “unconscious experimenter bias” is to not let yourself see the final measurement (nor compare it to any others) until you and your big science collaboration have agreed that all the checks that could be done, have been done. This procedure is called “blinding.” It’s a lot like evaluating job applications with the applicants’ names cut off, or judging musical performances with the performers playing behind a curtain: you are forced to be more careful, and to focus on the things that matter rather than the things that don’t. In my collaboration we do our analysis blind, and we are a lot more careful now than we used to be, I can tell you!
While there is some healthy skepticism about blinding being the solution to all our problems (it really isn’t), there is something important about being in the blinded state of not knowing what the answer is, where there is so much more possibility, so much more to learn - and it’s humbling to be continually reminded that we need to put ourselves in this state, in order to overcome our prejudices and as a result, do things right.