The Relationship Between Student Sense of Belonging and Suspension Rates

The most common form of discipline employed in American schools is suspension (Skiba & Knesting, 2002). However, a significant amount of contemporary data reveals that suspensions lack positive outcomes, and do not solve behavior issues. Research denotes the same students are suspended over and over for minor offenses including disruptive behavior, insubordination, and refusal to complete work (Skiba & Knesting, 2002). The purpose of this research was to determine the extent to which student sense of belonging accounts for the variability in out-of-school suspension rates over and above that attributed to race. To achieve this purpose, a descriptive, correlation study was used. The sample population for this study included all eighth students from four middle schools in one southwest Virginia school district. The unit of analysis for this study was the student. The investigator conducted a district-wide survey of all eighth-grade students, to collect belongingness scores using Goodenow's (1993) Psychological Sense of School Membership (PMMS) in November and December 2012. The investigator sought and received permission to access archival discipline data for participating students for the 2010-11, 2011-12 and 2012-13 school years. Possible relationships between belongingness and suspension, suspension frequency, and reasons for suspensions were initially anticipated to be examined through Analysis of Covariance (ANCOVA), above and beyond that associated to race. However, due to problems with methodology, an Analysis of Covariance was not viable; therefore an Analysis of Variance (ANOVA) and Pearson correlations were employed.

One hundred forty one students out of 693 participated, sixteen non-White students participated, none of which had been suspended. The low response rate, especially from suspended minority students, resulted in the revision of study's research questions to include belonging's relationship to all exclusionary discipline practices. Due to methodological limitations, and a low response rate, the study showed no statistically significant relationship between the two.

A number of implications for practice and future research are discussed, including recommendations for alternative research designs, larger sample sizes, and other methodologies to expand response rate and statistical significance.