I have to make an analysis of data collected from a lab experiment and I'm trying to find the best method and statistical approach to tackle it.

To summarize the test: It was done over an approximate 24hr period (8hrs every day) logging/collecting data every 6 seconds. The data collected was power output from a laser.

My original plan was to do a simple population analysis to observe the deviation from the expected (10W). It turns out that the output power has a lower value during the start time before it stabilizes and I'm wondering how I can better approach. I was thinking of just not including the first 3-5 minutes of each daily test run but then I have missing data, no good.

As you can see from the attached data, data points on rows 1, 4451, 7501, 9948 are new starting points for each day. Any help or suggestions on how to approach this will be appreciated.

See attachment.

Attachment(s)

PowerOutSN2825.sav   313 KB 1 version

I am by no means an expert, but when I saw the data it seemed to resemble a process control study (in SPSS, see here for example).

SPCHART  /XS=PowerW BY TstDay /RULES=All
  /ID=Timesec  /CAPSIGMA=SBAR /SIGMAS=3  /MINSAMPLE=2.

I'm not saying this is exactly the syntax you would run. I don't know what rules would matter; that would be your decision.

Thanks you much appreciated, I think this is exactly what I needed.