I often use repeated measures designs with two factorial IVs, and a continuous DV, for example:
IV1: Face Emotion (2 levels - happy / angry)
IV2: Face Trust (2 levels - high / low)
DV: Mean Reaction Time (ms)
As well as being interested in the main effects and interaction between these repeated measures variables, I often also want to know how the interaction between the two variables is influenced by a participant's score on a measure of individual difference. For example, the participant's score for Anxiety.
In my designs, one measure of Anxiety is taken for each participant, then each participant experiences a series of faces that provide multiple exposures to faces representing the four combinations of the levels of my two repeated measures IVs.
In the past I have used a median split on my continuous variable when using this kind of design, and have then run a 2*2*2 mixed factorial ANOVA. However, I am keen to move away from categorizing variables like anxiety, and instead toward inserting them into the analysis as continuous predictors.
My understanding is that to add a continuous predictor to a repeated measures analysis in SPSS, I cannot simply add the variable to the covariate box in the SPSS dialogue, because I am not interested in its effects as a covariate, and the interaction terms that appear when you do that are misleading?
Can anyone suggest to me a) the type of analysis they would use to address this kind of design, b) which software package they would use, and any guidance / texts / workshops they would recommend to learn the correct procedure?
Thanks so much everyone!
IV1: Face Emotion (2 levels - happy / angry)
IV2: Face Trust (2 levels - high / low)
DV: Mean Reaction Time (ms)
As well as being interested in the main effects and interaction between these repeated measures variables, I often also want to know how the interaction between the two variables is influenced by a participant's score on a measure of individual difference. For example, the participant's score for Anxiety.
In my designs, one measure of Anxiety is taken for each participant, then each participant experiences a series of faces that provide multiple exposures to faces representing the four combinations of the levels of my two repeated measures IVs.
In the past I have used a median split on my continuous variable when using this kind of design, and have then run a 2*2*2 mixed factorial ANOVA. However, I am keen to move away from categorizing variables like anxiety, and instead toward inserting them into the analysis as continuous predictors.
My understanding is that to add a continuous predictor to a repeated measures analysis in SPSS, I cannot simply add the variable to the covariate box in the SPSS dialogue, because I am not interested in its effects as a covariate, and the interaction terms that appear when you do that are misleading?
Can anyone suggest to me a) the type of analysis they would use to address this kind of design, b) which software package they would use, and any guidance / texts / workshops they would recommend to learn the correct procedure?
Thanks so much everyone!
