Crick
Gold Member
- May 10, 2014
- 28,776
- 5,566
I'm going to collect a hundred observations. I'm going to read then from some instrument and I am going to write them down on a piece of paper. I am then going to type them into a text file on a computer. A calibration sheet on the instrument tells me that I must correct its output by applying the biases in a given graph. I copy the text into an Excel spreadsheet and use the spreadsheet's functions to apply the given calibration adjustments. Then I will email them to you where you will review the numbers to verify the calibration was correctly applied and then find the statistical characteristics of the final numbers. You will take their average, their standard deviation, their mode, their median and anything else you feel like calculating. Then you will put the original observations along with your statistical results into an email and send them back to me. I will use all of that to prepare a written report with text and graphics. This will be submitted to some publishing house where it will eventually be distributed to subscribers both electronically and via printed copy.
What is wrong with that final data? It's certainly not raw. It's been processed and massaged and twisted and turned every which way. Is it any good?
The opinions you've expressed about models and data processing are unrealistic, unjustifiably biased and extraordinarily simpleminded.
What is wrong with that final data? It's certainly not raw. It's been processed and massaged and twisted and turned every which way. Is it any good?
The opinions you've expressed about models and data processing are unrealistic, unjustifiably biased and extraordinarily simpleminded.