The higher the accuracy, the more detailed the process of testing, the more accountable the result data. The more data you have the more backing results you have the more the output cant be questioned. There’s much that Bill has done that there is a hard time questioning simply due to the data from multiple aspects and multiple steps in any test set. Lots of checks and balance in any one test set.
Yes, but the checks and balances don't necessarily apply to the human running the test. And Im not saying this is the case. However, say i build the most elaborate and expensive test bed, and then someone secretly offers me money to report different results. Say you trust my methods because you see my test bed, and Im following all the procedures that have been decided proper. But then I go in to Excel and just change a few of the numbers out of some sort of bias. It's simply a possibility you can't controll from an individual, which is why there are testing groups that do this sort of thing. Labs which have people not only double checking the equipment, but double checking each other as well. Until reviews are handled at places like that, theres NO accounting for the accurasy of the review because people can skew results, no matter if they test the block 5 times or 50. And thats why there will always be questions about reviews done by individuals.
I don’t buy this. pH, Bill, joeC, myself have all seen how terribly hard it is to control anything with testing. Environment temperature is one aspect thats almost KILLER hard to regulate. Since temperature can change at any height in the room by a number of degree's over an hour or so and really not be noticeable. I have done extended plots of temp reading of ambient temps in what I considered "controlled" env's. A room with no HVAC/no fans/and just on window that was closed with the storm closed outside of that. Temp in the room with no electronics running at the ceiling was 5 DegC warmer than the floor, and total room temp change over a 12 hour period was 4C over the entire height range of the room. With many variances in-between. The funny thing was... I couldn’t feel the temp change myself. But there’s even more changes when running electronics in a room. You get pockets of warmer air, air circulation from the fans going over a radiator, etc..
Agreed. It's nearly impossible to control ambient outside of a lab. My house actually heats up over time with the computer running, which is why I allow for a lot of time between tests, which is why the review has taken me 3 months. I agree that its still not what I would want exact. Ambient is very important, although by measuring the difference between water temperature and CPU temperature, a factor similar to c/w, ambient shouldnt have too much of an effect, should it? Ambient during my testing was always 25c, as measured by the DDoc, plus or minus a degree or two. I dont think it should skew dT too much?
I am not claiming anything. I am telling you they aren’t based on the facts you have presented. Your system has an unknown error in all the readings. Figure out, and state what the error is then the readings have some merit and at least are based in reality.
Ive known for a long time where the errors can come from, but some of them are well beyond my control as a single reviewer.
|