Skip to main content
. 2013 Feb 13;8(2):e55944. doi: 10.1371/journal.pone.0055944

Table 1. Result of linear regression of the student’s academic progress over one year (y) on four independent variables measured in January 2011.

DF SS MS F-statistic p-value
g 1 9011.83 9011.83 12.7903 0.000467***
x acquaintance 1 1877.14 1877.14 2.6642 0.104686
x friend 1 4481.44 4481.44 6.3604 0.012692*
x best friend 1 2785.15 2785.15 3.9529 0.048574*
Error 153 107801. 704.58
Total 157 125956.

y  =  −14.124 + 0.204g + 0.127x acquaintance + 0.230x friend + 0.147x best friend

Independent variables are: g: student’s academic ranking, and x i: difference in ranking between each student’s self-reported neighbors’ average GPA and his/her own GPA, where i is either acquaintance, friend or best friend. A highly significant effect of g was due to the fact that top-ranked students did not change their academic rankings much after one year.