I am reviewing an article so I cannot give a lot of detail.
But suppose you are comparing a number of measures (some quantitative, some dichotmous) over multiple years. The authors took the mean of 2017, 2018 and 2019 and compared that to 2020 using either Mann Whitney or Wilcoxon. Their research questions were about whether 2020 was different than earlier years (they supposed it was different and that the reason was the pandemic). They have data on the same months for each of the years.
I'm thinking that a better analysis would be a spline regression with year as the independent variable, with a knot perhaps imposed between 2019 and 2020. But I'm not sure if such a complex method is needed or if their method is OK.
