4 ms·
How do you know if the measurements you're taking at the end of the process are valid to interpret, if you don't know the inputs? I'm not willing to trust the
by kepler1 6y ago
How do you know if the measurements you're taking at the end of the process are valid to interpret, if you don't know the inputs?
I'm not willing to trust the one (or act on it in a way that distorts what could be an otherwise fair/unbiased process) without knowing the other.
My problem with this approach is that people are too willing to stop halfway and claim that the outcome looks biased, without seeing what the inputs were. And cause (or seek to cause) massive change to people's lives based on that.
- tsimionescu 6y agoI absolutely agree, you need to try to measure boy side of the equation, not only the outcomes.