To evaluate TMM similarity we used UCINET (Borgatti et al., 1992),
following theprocedure developed by Mathieu et al. (2000).
This network-analysis program provides a similarity measure based on Pearson’s correlations.
As each member evaluated six
pairs of sentences for each TMMs dimension, the first step was to make team-matrices
for each dimension at each moment of data collection. Next, we used UCINET to
calculate the team similarity index for each matrix, which ranged from 21 (complete
disagreement) to 1 (complete agreement/sharedness). Similarity values were also
displayed in a matrix. The TMM similarity for each dimension of each team was then
calculated based on the average of the similarity index for each dimension. TMMs were
measured at time 1 and 3.