for my research project, I ran an experiment in which subjects had to tell true and fabricated stories. Now I am having difficulties to find an appropriate way to analyze my data in a comprehensive way, partly because I am uncertain of how to classify my variables. In brief, each subject had to tell a true and a fabricated story at two different timepoints (T1, T2). Some of the subjects were also given helpful information about telling a convincing story before T2, but after T1 (= variable: information). Hence, my design contains two within-subjects variables with two levels each (timepoint: T1, T2 and truth status: true versus fabricated). The variable information (three levels: no information given (0), information about verbal hints (1), information about non-verbal hints (2)) I would like to treat as a Covariate.
First question: Leaving the variable information as a covariate aside, I assume a two-way repeated measures ANOVA (timepoint x veracity) is suitable. However, I am not sure whether I should treat timepoint and veracity as two independent variables, or whether I should collapse them to one single Time Variable with four levels (i.e.: T1, true (1); T1, false (2); T2, true (3); T2, false (4))?
Second question: Bringing in the variable information as a covariate, I am unsure how to classify the variable information correctly (and hence, which statistical guides to follow). That is, I understand that for time-varying covariates I would perform a SPSS mixed procedure, but I am not certain if the variable information is actually time-varying, when the definition for time-varying holds that the variable's values change across the repeated trials? For instance, while the value does change between T1 and T2 when information= 1 or 2, it does not change between T1 and T2 when information= 0 (since then no information at all is provided, irrespective of timepoint).
Thank you very much for your help in advance!