Turn crew work into research evidence

STS-42 studied subjects including balance, eye motion and other physiological responses to spaceflight. Roberta Bondar’s medical background was relevant to this research setting, but a professional credential is not itself a scientific result. The connection between a question and a conclusion runs through the experiment and its measurements.

For an original teaching example, suppose researchers ask whether performance on a tracking task changes during a mission. They define the task, instructions, instrument settings and the time at which each measurement is taken. Keeping these features explicit helps distinguish a changed response from a changed test.

The numbers that follow are invented to teach comparison. They are not STS-42 data and should not be interpreted as a diagnosis or a finding about any astronaut.

Imagine one participant takes 12 seconds to complete the same task before a flight and 15 seconds during it. The difference is 3 seconds. Relative to the 12-second starting value, the increase is 3 / 12 = 0.25, or 25 percent. A percentage requires a stated reference value.

That arithmetic describes the observation, not its cause. Practice, sleep, workload, instrument changes or the timing of the test could also affect performance. A carefully planned comparison tries to keep relevant conditions consistent and records unavoidable differences.

Repeating a test can help reveal variation, but repeated measurements of one person are not the same as independent measurements from many people. Ten trials by one participant give information about that participant’s repeated performance; they do not create ten independent crew members.

Calculate change with a stated baseline

In a fictional data package, keep the task version, units, time, instrument setup, participant code and any recorded departure from the protocol with the values. If a setting changed during a session, preserve that fact instead of silently merging the observations as equivalent.

The researcher’s later analysis depends on this work. A successful mission task might produce a well-documented measurement even when the original scientific prediction is not supported. Collecting data and confirming a hypothesis are separate outcomes.

The IML-1 photograph provides a useful reminder of the people, equipment and procedures behind a dataset. It cannot reveal the 25-percent change in our invented example: that value belongs only to this teaching exercise. The archive record preserves the actual historical scene.