It's generally acknowledged that IQ measured by well known IQ tests like the WAIS are fairly consistent over time. You may score a few points lower or higher, but it won't change much. Training can help a bit, but only so much. Health issues, mental problems, stress, tiredness, drugs or alcohol can all influence your score, mostly in a negative way. So I suppose that influences like these are filtered out of this research. They are interesting to research, to see what their influence is, but to study the consistency in test results over time, they should be left out.