Question about using ANCOVA and reporting in a paper

#1
Hi All,

I have a 3 (gait speed) x 8 (trials) within subjects experimental design and have submitted the manuscript to a journal. I was given revisions to do on it and re-submitted and have received revisions again hence my question...

I have 14 subjects that walked for 35 minutes on a treadmill-prior to walking on the TM they performed a standing balance assessment (to identify the amount of sway beneath their feet while on a force plate) - this is trial 1 (baseline). I then had them perform the same assessment every 5 minutes (total of 7 more while they had been walking). This same protocol was performed at 2 more walking speeds (hence, the 3 gait speeds). My original data analysis used a mixed model GLM with repeated measures (8-including the baseline value) with planned contrasts. On my first return of the paper, one of the reviewer's wanted me to perform a 2-way ANOVA with repeated measures and Bonferroni post hocs vs. planned and they did not like the inclusion of baseline values as one of the repeated measures. I performed the Bonferroni's but continued to use the mixed model as I was told by several statisticians and through my own reading and referenced a paper....
On my second return, same reviewer still wants the 2-way ANOVA and.... does not want me to use the baseline values in my analysis only the trials taken during the walking session and this is where my question lies..... I want to know the difference between baseline and each walking trial as well as the differences between each trial (hence the repeated measure). In doing some more research, it appears as though I may want to have my baseline values (which are constant (although vary from subject to subject) as a covariate but am not entirely sure as the readings I have come across seem to use these when there is a between subject factor which... I do not have....

The second question I have (as I have run it this way but not convinced I am right) is how do you interpret the interaction of the covariate and report it in a paper???

Please help. I have spent way too much time spinning my wheels on this and my deadline is fast approaching....

Thanks so much, sorry for the long post but wanted to be clear.

KTATC