(7.1.3.1) Repetitive scientific experiments and one-time decision making

In science, we chose an approach to increase the reliability of the hypothesis through repeated experiments. However, repeated experiments are not always possible. Let's say you are suffering from a college freshman and wondering, 'It is correct to learn programming language X'. If you do a scientific approach, you will repeat learning a few times and repeat what you do not learn a few times and compare which is better. But you can not do such an experiment alone. You can only learn or do not learn one time.

If you approach scientifically, you will be doing the following experiment, for example.

• For the sake of comparison, collect 100 students and let 50 people learn language X, leave 50 people to leave
• Effects of learning are quantitatively measured by annual income in five years
• The two groups are randomly divided and statistically test whether there is a significant difference in the annual income in 5 years

But this experiment does not meet your personal needs. It is because five years have passed while waiting for scientific knowledge, you are already not a first grader in the university. *8

Decision making is often an unrepeatable one-time event. Often there is a lack of knowledge that will be helpful at the timing when decision making is urged. Waiting for knowledge to come in that case may be a "bad decision" that misses the timing of decision making. In other words, the correctness of decision-making has properties different from mathematics and science.

*8: Also, suppose that an experiment had been conducted five years ago and the results showed that learning language X would lead to a statistically significant increase in annual income. Still, this scientific finding does not guarantee that what you personally learn now will lead to an increase in your annual income.

This page is auto-translated from /nishio/(7.1.3.1) Repetitive scientific experiments and one-time decision making Repetitive scientific experiments and one-time decision making) using DeepL. If you looks something interesting but the auto-translated English is not good enough to understand it, feel free to let me know at @nishio_en. I'm very happy to spread my thought to non-Japanese readers.