I have problems in undertanding how to correctly run a mixed linear model.

I have done an experiment to compare three groups (1 between-subject factor); for each subject, 2 sensory stimuli are given (factor1), each stimulus is given in 3 conditions (factor2), for each condition 3 different spatial positions are used on a screen (factor3). Subjects have to answer which condition they are presented.

Thus, I have 3 way crossed factorial design (2x3x3).

Each subject gives N responses for each combination of factors. Then average accuracy rates (ACC) and average reactions times are registered for each combination (RT).

Since I have missing data in some conditions I want to use linear mixed model.

For the moment I am interested in understanding how accuracy rates change according to my main effects as well as interations between my within-subject factors and between-subject factor by accounting for age, gender and RT.

I am new to mixed models.

I have inserted (in spss) my main effects as well as interactions between grouping factor and my within subject factors treating them as fixed factors. I treated covariates as fixed factors as well.

As far as I have understood I can account for differences among subjects by including a random intercept to the model.

I would like to know whether my model is correct in this way or if I have to further model covariance structure of residuals (using for instance a /repeated statement in SPSS format).

Any comment will be appreciated!



  • $\begingroup$ If you are just asking for SPSS code, that would be off topic here (but presumably on topic on Stack Overflow). But you may have a statistical question (in which case, CV is appropriate). Please edit to clarify what you need. If you want your question migrated to Stack Overflow, just flag it for the moderators. You can also ask a statistical question here, and a related coding question there. $\endgroup$ – gung May 19 '16 at 14:13
  • $\begingroup$ Hi gung, my basic question is on the correctness of the model I built. Thank for your suggestion, I modified my question accordingly. I will post a separate question on spss syntax on Stack Overflow. Best $\endgroup$ – Alessandro C. May 19 '16 at 14:33

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Browse other questions tagged or ask your own question.