Seems like the problem lies with level of scrutiny applied to the research behind this device. If it is something that reads things based on skin pigmentation, wouldn’t you want to test different pigments?
Not sure this speaks to racism though as the article alleges. Research Bias and ignorance perhaps. Hopefully we can come up with calibrations for skin color here?
Hm, in that case, where does the research bias come from? What's the selection procedure for test subjects? Why does under-represent dark-skinned people?
"Racism" is a word with a wide range of meanings, from deliberate discrimination based on a consciously held world view at one end of the scale to subtle unconscious prejudice at the other. By the latter meaning, racism by definition involved in the causality chain anywhere that a racial group is statistically underrepresented.
All that being said, as a layperson, I find it perfectly possible that pulse oximeters are simply inherently more difficult to calibrate on dark skinned people. The article didn't outright say there was a systemic error in a particular direction, only that accuracy was a problem. It may be that a best-effort attempt was indeed made on an appropriate range of demographics, and this is simply a limitation of the technology. It may also be that this could be fixed with sufficient research, but racism means it's less of a priority.
Correct but I kind of stick to the traditional use of the word racism (which is usually intentional, with malice involved, often ignorance).
If we start to call what’s likely unconscious bias racism, then what word do we reserve for the truly vile conscious bias that people stubbornly hold on to? It’s a genuine question.
There are also other failure modes like too cold fingers for example.