I would be very grateful if someone could just doublecheck my choice of data analysis. This is for a dissertation, and I do not have the possibility to check this with my supervisor right now.

So here's the deal:

Basically, the study looks at the influence that subliminal conditioning (positive/negative) has on reaction time with varying number of pairings (low number of pairing, medium number of pairing, high number of pairings).

The experiment consists of three blocks. The blocks are short, medium, or long. In each block, participant are subliminally primed with a positive stimulus in half of the trials, and with a negative stimulus in the other half. They are then meant to react to a change in color following that priming. We are measuring the reaction time and suspect that they will react faster if the prime was positive. We also suspect that these effects will be stronger with a higher number of pairings (e.g. they will react even faster after a positive prime in the long block)

This is a very basic breakdown, but should convey the message.

So my idea was to run a 2 x 3 ANOVA, where 2= prime: positve vs negative and 3 = length of block - short, medium, long.

I have the data in the following structure: Variable 1: Positive_Block1, Negative_Block1, Positive_Block2 and so on. The outcome, reaction time, is continous, and is the mean reaction time for each participant after each prime in each block.

I was going to look at the main effect of the length of the block, the main effect of prime and the interaction between those two.

So my question:

Is the 2x3 ANOVA design correct?

Also, is it correct to enter this in a repeated measures design in SPSS? (GLM -> Repeated Measures ? ) Or do I have to do it in a mixed model (I'd very much prefer repeated).