Estimate time to run statistical analysis.
Statistical work is dominated by the parts that are not statistics. Cleaning and preparation routinely take more time than the analysis itself, which is why they are entered directly rather than assumed away, and the cleaning share is reported so you can see whether your estimate is realistic — under about 30 percent usually means you have not accounted for missing data, recoding and merges. The analysis term multiplies variables by tests each, because the count of models is what drives the hours, and each test is priced with interpretation time included.
Analysis time
Analysis hours = variables x tests per variable x minutes per test / 60
Project total
Total = cleaning hours + analysis hours + write-up hours
Because real data arrives with inconsistent coding, missing values, duplicates and merge keys that do not align. Published estimates commonly put it at 50 to 80 percent of analytical project time.
No — plan the tests your hypotheses require. Running many exploratory tests inflates both the time and the false-positive rate, and unplanned multiple comparisons need correction to remain defensible.