anova using raw data - confused... - (Apr/26/2012 )

Hi,

I am trying to perform statistics for my MTT assay but I am a bit confused as how to proceed. For simplicity sake I will present some example data.

Biological replicate 1 (done in triplicates):

Control: Absorbance: 500, 520, 510
Dose 1: Absorbance : 300, 290, 270
Dose 2: Absorbance: 200, 180, 160
Dose 3: Absorbance: 150, 130, 120


Biological replicate 2 (done in triplicates):

Control: Absorbance: 700, 720, 710
Dose 1: Absorbance : 400, 420, 430
Dose 2: Absorbance: 280, 260, 240
Dose 3: Absorbance: 210, 230, 220


Biological replicate 3 (done in triplicates):

Control: Absorbance: 400, 420, 410
Dose 1: Absorbance : 240, 230, 240
Dose 2: Absorbance: 160, 150, 155
Dose 3: Absorbance: 120, 110, 125



So if I was to take the means from each biological replicate and make the doses relative to the control, I would get almost the same % relativity between the doses. So for biological replicates 1, 2 and 3, the % relative to control for dose 1, dose 2 and dose 3 will be 60%, 40% and 30%, respectively, for all the 3 biological replicates even though there is significant variation between the absorbance of the biological replicates for the same sample (ex. control has absorbances of approx 500, 700 and 400 between the three biological replicates).

I am wondering how I would go about doing ANOVA on data such as this. Is it ok to perform anova on the percent data or should it be done on the raw absorbances?

Thanks

You are doing the analysis the right way. You first obtain means from technical replicates, and then use the 3 means from 3 biological repeats to do the ANOVA analysis. If the means are very close from each repeat, that tells the deviation is very small and the result is highly reproducible. congratulations!