Getting it right matters! Covid‐19 pandemic analogies to everyday life in medical sciences

The compass of biomedicine

“…For all our history, we as humans have had the need to know where we are and where we need to go -this is true geographically as well as it is in science: Experiments need to be planned cautiously, 19 and their results interpreted competently, focusing on what really matters. 20,21 When conducting experiments in physiological research, the compass of "blotting" has been an increasing default starting point for experiments, conducted on single cells, 22 organoids, 23 and animals 24 alike.…”

Perception

“…On an even higher level, a tertiary concept of perception is explored when scientists look at the public view on science 17 or the perception of statistical results. 18 Interestingly, science communication in biomedical research often focuses on making statistical results, which are not intuitively understood, more comprehensible, and, consequently, rendering science-based decision making more plausible. Understanding probabilities, however, remains "a thing," as much within the scientific community as outside of it: As Rao et al found in statistically trained advanced graduate students, training researchers in traditional statistics that emphasizes a 0.05 significance level for hypothesis tests more often than not create an incorrect categorical perception between so-called statistically significant and non-significant p-values.…”

“…On an even higher level, a tertiary concept of perception is explored when scientists look at the public view on science 17 or the perception of statistical results 18 . Interestingly, science communication in biomedical research often focuses on making statistical results, which are not intuitively understood, more comprehensible, and, consequently, rendering science‐based decision making more plausible.…”

The Art of Planning—How many samples are enough?

“…There are differences between disciplines, and it is common for some scientists to use alpha levels ( P ‐value <alpha) of 0.01 or below or to differentiate between different levels of certainty within one publication 17‐21 . A more detailed guide to finding the correct levels for alpha‐ and beta‐errors and important caveats is available 22 …”

“…[17][18][19][20][21] A more detailed guide to finding the correct levels for alpha-and beta-errors and important caveats is available. 22 Next up is the question of how big we expect the measured effect to be. Quite intuitively, there is an inverse relationship between the effect size and the number of experiments needed to reach desired levels of statistical certainty.…”