The sample used for this analysis includes subjects with two baselines and one post injury 0-5 days after an injury. Subjects whose total symptom score varied by more than 10 in their two baselines were excluded. This resulted in 92 total subjects, 42 of which had a post injury total symptom score of more than 10 more than the minimum of the baseline total scores and were therefore put in the concussion group, and 50 of which did not and were put in the non concussion group.

Using RC values calculated by Bruce Maxwell for a past paper, new columns were added for whether there was a reliable change for each composite score from the first baseline, the second baseline, the average of the baselines, and the best of the baseline scores.

An additional column was added for a count of how many of the composite scores each of the individuals had a reliable change. Subjects with at least two composite scores with a reliable change from the baselines to the post injury were labeled as having a concussion.

A t-test was used to compare the mean number of reliable changes in the concussion group between baseline 1 and the best baseline. The results at the 80% confidence level (t= -0.53205, p-value = 0.5961) do not provide enough evidence to suggest that there is a difference in the mean count using the first baseline and the best baseline using the 80% RC. The results at the 70% confidence level (t = -0.61437 and p-value = 0.5407) do not provide enough evidence to suggest that there is a difference in the mean count using the first baseline and the best baseline using the 70% RC.

A t-test was used to compare percent of subjects in the concussion group that were labeled as having a concussion. The results at the 80% confidence level (t = 1.9039, p-value = 0.153) do not provide enough evidence to suggest that there is a difference between the two percentages. The results at the 70% confidence level (t = 1.8301, df = 3, p-value = 0.1646) do not provide enough evidence to suggest that there is a difference between the two percentages.

**Labels:**