As I understand it, "decision fatigue" is a particular explanation for the observation in people's own lives and the observation of other people that decision-m

As I understand it, "decision fatigue" is a particular explanation for the observation in people's own lives and the observation of other people that decision-making seems to get worse after long stretches of decision-making. If I've observed, for instance, that I make more bad decisions at work later in the day, no one is necessarily saying that I am fooling myself (although to be honest I would prefer to see rigorous studies even of that claim, since judging the value of one's own thoughts and decisions is pretty well-known to suffer from many biases). In this case I'm willing to accept that decision-making does tend to get worse after long stretches, but I'm more concerned with how to verify this particular explanation for why decision-making gets worse, and for that I think it's vitally important to do controlled experiments. One's own experience almost certainly has no explanatory power for why decision-making is getting worse. It's certainly a reasonable guess that my decision-making is getting worse because I've had to make many decisions today. But there are also other reasonable guesses. Maybe it's my diet, and I had more energy immediately after breakfast and lunch but less energy later in the afternoon. Maybe sunlight has an effect. Maybe simply knowing that I get to go home from work soon makes me less invested in decisions at the end of the day. A less reasonable guess would be something like "when the afternoon train starts running my decision-making gets worse." It still naively fits my observations, but we can probably agree that a controlled test would likely show this explanation to be incorrect (but who knows, maybe the train is loud and distracts me, or makes me think about my upcoming commute home!). These are all things that can probably be controlled for, and it wouldn't be too shocking to me if the number of decisions made in a day is not in fact a significant explanation for why the later decisions are poorer. I haven't even seen the study that was preemptively dismissed, but I'd like to.