# Analysing a balanced mixed models design?

I am designing an experiment and I'm in need of some help.

Participants are presented with two videos on a website. In one period, a feature (a table of contents for the video) is enabled, while in the other period the feature is disabled. Cognitive load (the DV) is measured during both videos.

I know that it's sensible to counterbalance the order of treatments. However, since the participant must watch two videos, the video order needs to be counterbalanced also (even though it is not of interest). Therefore I need 4 groups to represent all of these combinations:

F=Feature
N=No Feature
V1=Video 1
V2=Video 2

Group   Period 1   Period 2
1         F&v1       N&v2
2         N&v1       F&v2
3         F&v2       N&v1
4         N&v2       F&v1


According to this website about counterbalancing (last section) I can use two between subject variables "order" and "combination" to represent the above table of group combinations. So my data looks like this:

Order   Combination cogload1    cogload2
1       1           5.00        3.00
2       1           4.00        6.00
1       2           5.00        7.00
2       2           8.00        5.00


This is where I get stuck. Given that feature/video are encoded in order/combination now, how can I determine the main effect of feature or the interaction between feature and video?

Also, I'm still getting to grips with general linear modelling, interactions and interpreting the results from SPSS. Does anyone have any resources that they could recommend?

-