Panic! At the lab

Ever have a day where nothing went right? We had a series of unfortunate events so we couldn’t actually do the research we wanted to do. It’s not a huge deal, but what was impressive was the amount of things that went wrong, each thing compounded and let us continue the study up until the point where we were about to start. Only then did we find out we had to stop. There was no danger and the issue was figured out after the fact, but it was an impressive feat of events.

The word (phrase?) of the day is conditional probability. Conditional probability (looks like I didn’t specifically cover this!) is when one event is relying on another event (roughly). An example would probably make life easier, what’s the chances that you miss the bus, given that you slept through your alarm? What we’re saying is what are the chances you miss the bus and what are the chances you slept through your alarm? In math form it looks like this:

As we increase the number of dependent events, the probability of it occurring becomes lower and lower. That’s an important point because while a chain of events can be very long and still occur, it’s still rare. For example, what’s the probability that you get heads every time you flip a coin for 20 times? I’m not going to do the math, but it’s low. Now that we’ve thoroughly have made this point, let’s talk about the series of events that occurred (vaguely).

We had an experiment and in order to do that there was a form that needs to be filled out to verify certain things. The person doing this had forgotten to do it prior to the start, so we went ahead and started prep. We had been coordinating with someone else, who happened to be out that day unexpectedly. We went through the full steps for prep for the experiment and were about to start. Then the person remembered the form because they couldn’t find it to verify they were eligible. Then the person doing the form decided they were not eligible because of a known issue that we’ve worked out with the person who was out for that day.

See the chain of events that had to occur to make it that far? The funny thing is this is the condensed version, so not even the full list of stuff that had to happen to get to where we ended. Afterward we got ahold of the person we originally coordinated with and verified that it was okay so we’ve rescheduled the experiment and will continue. However, the fact that we made it that far was super impressive to me because we had to have so many different things to go wrong.

While it did end the experiment, it wasn’t a big issue. We work out all this stuff in advance and it wasn’t the fault of anyone in particular, it was just a string of bad luck. Our team typically does all this screening beforehand so we knew the person was eligible, we just couldn’t verify it in time for the experiment since it was very late in the day and the people who could help do this were either busy with their own things or were out.

In the end it was just a good instance of conditional probability in real life. Also the fact that outliers exist and even a very low probability event still has a chance of occuring.

If only that low probability event occurring was me winning the lottery…