I just got rejected a scientific study mainly because the statistical analysis that I performed I collected data through motion analysis from all the goals for the 2018/2019 season in the English Premier League. I analysed the movements using a movement classification system called: “the Bloomfield Movement Classification”. This mainly includes movements (sprint, decelerations, turns, etc), intensities of the movements (high, medium, low).
I performed Motion analysis of for the scorer, assistant, closest defender to the scorer (defender of scorer) and closest defender to the assistant (defender of assistant). Analysis was limited to the last 6 movements of each player. (note that each of this players is going to be involved in some goals but not in others)
After collecting this data, to analyse difference between players (assistant, scorer, defender of assistant and defender of scorer), group of players (attackers vs defenders), movements (Linear actions, decelerations, turns, etc) and intensities (Low, Medium, High) these were analysed through percentage confidence intervals set at 95%. When confidence intervals did not overlap we concluded that there was statistical significance between groups. For sample size of more than 30, normal approximation interval was used. For sample size of less than 30 (including 0) Wilson interval was utilized as recommended by Brown, et al. (2001).
This are the notes made by reviewers regarding the statistical analysis:
What was the rationale for using percentage confident intervals for the analysis as opposed to null hypothesis testing, particularly for the group analysis?
why did you not use another proportion test? Or another non-parametric test?
Based on the comments of the reviewers I understand that I need to change the statistical analysis and I assume they would be suggesting Chi-Square? Anyhow I'm unsure on how to apply this with my data as I’m struggling to do a contingency table with this due to its complexity.
I attach some of the figures and tables sent to the Journal for better understanding.