I am running a quality analysis. I have a can of some liquid material, and I analyze it's characteristics.
At time 0, 30 samples from this can were taken, and a few measures were taken.
at time 1, other 30 samples were taken, and the same measures were taken.
The researcher wants to know if there is a significant difference between the first group (time 0) and the second (time 1). He wants to see affect of time. I only have 2 time points, and the sample is not paired, the second sample is different, but from same big can.
Can I assume independence ? I mean, both samples are from same big can, but they are independent. A paired t-test is out of the question. Can I use a normal t-test ?
(under assumptions of normality and equality of variance of course)
I just have some questions before I say, "independant t-test"
1. What kind of statistical software are you using or are you familiar with? (e.g. Matlab, SPSS, R, SAS, Excel, etc...)
2. What kind of "measures" are you taking?
(a) Are the 30 measures taken on the same "dependant" variable? (e.g. concentration, temperature, viscoscity)
(b) Are the 30 measures ordinal, continuous, or nominal?