Fidelity in Implementing School-Based Restorative Justice Conferences

Abstract

Many schools are attempting to reduce their reliance on suspension and exclusionary discipline to manage student behavior. One promising alternative involves restorative justice (RJ). However, many programs and interventions are being described as “restorative,” without clarity regarding key components of a RJ approach, and few studies have measured program fidelity. This paper presents a measure of fidelity in implementing RJ conferences, which was used to rate RJ 105 conferences conducted in response to truancy, chronic disruption, and incidents with and without victims. These were implemented in the context of a larger effort that also included a whole-school intervention. Conferences showed very high fidelity in the behavior of conference facilitators, high fidelity in the interaction among conference participants, in contributions of participants to the action plan, and in consensus about the action plan. Student behavior was more varied, as was the extent to which the group offered forgiveness. Emphasizing consequences was associated both with more student remorse and with more student disengagement during conferences, both of which were associated with forgiving the student, but in opposite directions. This highlights a tension in RJ between holding misbehavior accountable while also showing students respect and helping them to repair harms they have caused.

Keywords:

• Fidelity
• restorative justice
• school discipline

Acknowledgment

This study was funded by grant number 2014-CK-BX-0025, awarded by the National Institute of Justice (NIJ), Office of Justice Programs, US Department of Justice, through the Comprehensive School Safety Initiative. Points of view expressed are solely those of the authors, and do not necessarily represent the official positions or policies of NIJ, the U.S. Department of Justice, the Urban Institute, or Child Trends, or Fors Marsh.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Notes

1 Gregory et al (2014, unpublished, cited in Acosta, 2016) developed an observation instrument that rates Structure and Support – based on Baumrind's (1966) conception of authoritative parenting – and Student Voice, in order to rate RJ circles and conferences, most of which were not in response to misbehavior (Acosta et al, 2016, p. 416; Augustine et al, 2018, pp. 12 and 40). It therefore approaches RJ fidelity differently than the other instruments.

2 The formal criterion for chronic absenteeism is 10% absent (18 days), at which point a court petition may be filed, but often is not. Schools often referred students for RJ conferencing who appeared headed toward chronic absenteeism before that criterion was met.

3 Three additional schools participated early in the project (Blackstone Valley Preparatory School, and Westerly Middle and High Schools), and then dropped out. 56 additional cases were referred at these schools, and 41 conferences were held; none were observed.

4 Technical problems led to missing or incomplete data for several additional cases that are not included in the present sample.

5 86 observations used these response options. 21 earlier observations used a pilot instrument with the same questions (although student disengagement was added after the pilot), but slightly different response options (not at all; a little bit; yes). These response options were mapped onto the same 0 to 3 range as 0, 1.5, 3.

6 Several additional items did not relate to other items as expected and are not shown. (A) Two items were expected to be negatively correlated with impartiality, but were instead positively correlated positively with impartiality: “The facilitator seems aligned with the student,” and “… seemed to represent the schools or other agencies.” (B) “The facilitator largely directed the outcome(s) of the conference” was intended as negative and to be reverse coded, but it was rated quite highly (M=2.50) and was positively correlated with the Facilitator Management items.

7 Observers maintained some distance and were not reliable reporters of the contents of the action plans. YRP's conference data could not be merged with the observation data because of complex confidentiality provisions.

Additional information

Notes on contributors

Akiva Liberman

Akiva Liberman is a Senior Research Scholar at Child Trends and an Senior Research Fellow at the Urban Institute. His research has focused on juvenile justice and adolescent development, including juvenile justice reform, school discipline, and youth violence prevention. He also conducts mixed methods program evaluation, including for Pay for Success-funded programs. He has served as a program officer at the National Institute on Drug Abuse and National Institute of Justice. He holds a Ph.D. in Social Psychology from New York University.