In surveys of more than 20K shiftworkers, we asked them about their shift length preferences for 24/7 coverage. We did it in a way to avoid directly mentioning specific shift lengths as shown in the graph below:

We found that asking specific shift lengths did not give reliable answers since it was based on perceptions rather than actual schedules. When we showed the shiftworkers actual schedule examples using different shift lengths, here is how they responded:

Ratings definitions:

     5 = I like it a lot

     4 = It has potential to be good

     3 = neutral or not sure

     2 = It doesn't look promising

     1 = I don't like it