Skip to main content
. Author manuscript; available in PMC: 2021 Sep 14.
Published in final edited form as: Criminology. 2019 Sep 13;58(1):35–69. doi: 10.1111/1745-9125.12227

Figure 3. Fitted Probablities of Losing a Friendship Nomination Received in Previous Wave, by Grade Size.

Figure 3.

NOTES: PROSPER sample limited to observations of in-home survey participants meeting the following criteria: attending participating school district, participated in the in-school survey, and valid data on suspension (N=2,373 observations of 766 students). Results of binomial generalized estimating equation model with controls. The model excludes 219 observations that did not receive a nomination last year. Results are combined across 20 multiply imputed datasets. Percentiles are approximate.