Which statistical test to analyse? I am trying to find out how inflammatory markers change in response to a mechanical device change. I have daily readings for the 5 days before the event and for the 5 days after the event. I am expecting a rise in inflammatory markers in the pre readings followed by a fall after the device change. I also need to find out how the change differs in two subgroups (sample sizes of the subgroups are n=21 and n=88). The aim is to find out if the change is significant, to consider these markers as predictors of the device failing and hence needing to be changed.  
I would like to know what test would be appropriate to analyse this. I need to keep things simple as I and the people reading the results are non-statisticians.
 A: This is just a quick summary, and is in no way exhaustive: You can use the analyse/mixed model/linear tab for an LMER and the analyse/mixed model/generalized linear for a GLMM if you have binary/ordinal/etc. response variables. You then have to specify the subjects by dragging the appropriate variables into the subject box hierarchically. (E.g. people within classrooms within schools). In the fixed effects tab, you can specify your IVs by dragging them into the appropriate box. To add interactions, select all relevant variables and drag them together into the respective box (for a two-way interaction, select two variables and drag them into the "two-way" box.) You can add random effects the same way. If you only need a random intercept, tick "include intercept" in the random effects box, and set "subject combination" to whatever your subject variable is in the drop-down menu. To add random slopes, drag and drop your IVs the same way you did for random effects. In the "estimation" tab you can specify what pairwise contrasts you need in the case of categorial IVs, and what you want to save (z-scores, residuals, etc.). There are also some options related to different convergence criteria you can adjust if your model ends up not converging, but are better left alone at the outset. In addition, there are some additional settings you can use if your dataset is small, or if you need to use robust estimation, due to violations of model assumptions. Also, if you end up using the GLMM option, be sure to disable the "model viewer" in the edit/options menu, as it makes the GLMM output extremely cumbersome to navigate amd actually hides some important information.
