On Wed, 18 Sep 2002 12:43:31 +0000, R.H.Alle wrote: >>The problem is that the operator has changed at a specific moment in >>time. Therefore, all data before were acquired by one operator, all data >>afterwards by another. So, if the first measurement of a patient is done >>by operator A, the second one probably is too, but the next one (and all >>subsequent ones) is done by operator B. I think blocking won't help me >>for this; or maybe I misunderstood the idea... > I see -- I neglected the element of time. I don't work with temporal > data, so I'm completely ignorant about how blocking might work in this > case, if at all. Sorry. Anyone else care to comment?
If you would redesign the research, it might work. But I am working on data already collected (over a time period of four years). The time between measurements of one subject is about half a year, so in a new experiment, if you could measure them a few times in half an hour, that would be no problem. The effects caused by time (the ageing) are on quite a different scale. Koen . . ================================================================= Instructions for joining and leaving this list, remarks about the problem of INAPPROPRIATE MESSAGES, and archives are available at: . http://jse.stat.ncsu.edu/ . =================================================================
