Our results suggest relatively large effect of the various social networks; the presence of
reciprocal friends in class has a positive and significant effect on test scores in English, math, and
Hebrew: adding one such friend in class raises the student's average test scores by 0.095 standard
deviation of the test scores distribution. This relatively large effect is however limited to the first circle
of such friends, while any extensions of this social network have no effect on the academic
performance of students. The presence of followers also has significant positive effects on student
outcomes. Non-reciprocal friends in class have the opposite (negative) effect on a student’s learning
outcomes. We also find that these effects have interesting patterns of heterogeneity by gender, ability,
and student age. In addition, we present in the paper other effects of these various types of social
networks on other non-cognitive behavioral measures, including social and overall satisfaction in
school, time allocated for studying at home, and whether one exhibits violent behavior. For example,
we find that the presence of more reciprocal friends and followers in the classroom reduces fear and
intimidation from school violence and bullying, and improves social satisfaction and school
satisfaction.