3 ms·
Reading the paper (http://whitman.syr.edu/faculty-and-research/research/pdfs/BQ_Rev2_MS_May5.pdf http://whitman.syr.edu/faculty-and-research/research/pdfs/BQ...
by panic 10y ago
Reading the paper (http://whitman.syr.edu/faculty-and-research/research/pdfs/BQ_Rev2_MS_May5.pdf http://whitman.syr.edu/faculty-and-research/research/pdfs/BQ...), it looks like each customer took the same amount of work to "check out" (moving five sliders to the correct position). The whole reason to divide a single line across multiple cashiers is to keep one slow customer from slowing down an entire line, so it's weird they didn't model this in the experiment itself.
- cortesoft 10y agoYeah, and also one of the keys gains is a lower deviation, which is a benefit separate from faster average times.