I have typically been highly sensitive to decision making in small sample sizes. A single day’s outcome doesn’t matter much in a year’s worth of results.

Parts of Popper’s philosophy of falsification were thoroughly engrained in me during college, and my common exposure to research, statistics, and significance leaves me very wary of the single case of surprises. Anything to avoid a Type 1 error.

However, I think there is an area of study where a single aberrant finding can mean much more than it would in chemistry and physics. Undesirable behaviors.

I don’t yet have this thought fully fledged in my mind.

I just know that I have experienced, now multiple times, the significance of a single event when investigating problems on dairies. A single observed expression of a bad behavior; undue aggression, cutting corners, carelessness etc. has later been shown to be only one in a long list of transgressions. It has been just the “tip of the iceberg”.

Moving forward, I think I should consider longer those be single actions, and not chalk them up as anomalies.