Study 1
Augimeri and colleagues (2007) examined the immediate, short- and long-term effectiveness of the SNAP® Under 12 Outreach Project (SNAP® ORP), using a randomized controlled trial in Toronto, Ontario. The sample consisted of 32 children under age 12 (24 boys and 8 girls) who had been referred to the SNAP® ORP. To be eligible to participate in the SNAP® ORP and the study, children had to have police contact within 6 months of referral or a T–score on the Delinquency scale of the Child Behavior Checklist (CBCL) of 70 or greater (indicating behavior problems more serious than 98 percent of peers of the same age and sex). The majority of the study sample was eligible under both requirements: 81 percent had a T–score of 70 or more, and 78 percent had police contact. More than one third of the children were referred for more than one type of offending behavior. The primary reasons for referral included theft, fighting, severe defiance at home, vandalism, assault, arson, trespassing, and public mischief.
The original study design was to match 32 children, case-by-case in 16 pairs, on age, sex, and severity of delinquency (T–score) on the CBCL. One member of each pair was randomly assigned to the experimental group, which received SNAP® ORP, and the other to the control group, which received a nonclinical activity/recreation program called the Cool Runners Club. However, a few modifications had to be made to the original design for several reasons. In the end, the two groups of children were analyzed as independent samples.
The experimental group (n = 16) was 25 percent female, with an average age of 8.7 years. The control group (n = 14) was 29 percent female, with an average age of 9.1 years. There were no significant differences between the groups, except on parental marital status. The experimental group was more likely to have married or cohabiting parents. The average CBCL Delinquency score was 8.9 for both groups, and the average CBCL Aggression score was also very similar (18.8 for the experimental group and 19.4 for the control group).
Measures of children’s delinquent and aggressive behaviors were collected by having parents rate behavioral items on the CBCL. Parents rated items as not applicable or not true, somewhat applicable or somewhat true, or very often applicable or often true. In addition, records of official criminal involvement were collected. A national criminal record search was performed that captured all offenses committed between each child’s 12th and 18th birthdays. Criminal records were coded to indicate the presence or absence of a finding of guilty for one or more charges, the total number of convictions, and offense type (i.e., property versus person).
Data was collected at five time periods over 18 months: Time 1 (pretreatment); Time 2 (posttreatment, at least 3 months after Time 1); Time 3 (3 months after Time 2); Time 4 (6 months after Time 3); and Time 5 (6 months after Time 4). The change in the experimental group score between Time 1 and each subsequent time period was compared with the change in the control group using a t–test.
The study did not include a no-treatment control group. At the end of Time 2, the experimental and control groups switched service modalities. Over the period between Time 2 and Time 3, the experimental group received the Cool Runners Club and the control group received the 3-month SNAP® ORP intervention. This was done to examine whether the children in the experimental group were able to maintain treatment gains made if they entered a short-term recreation program posttreatment, and to examine whether the control group would experience the same treatment benefits that the experimental group children received from the SNAP® ORP intervention after receiving a recreational type program. In practice, however, the SNAP® ORP treatment received by the control group was not as intensive as that received by the experimental group because of clinical staffing changes at the program. The purest test of the effect of the SNAP® ORP is obtained by the comparison of results at Time 1 and Time 2. The test of the effects of the SNAP® ORP after Time 2 is conservative, because it is essentially a comparison of more and less intensive SNAP® ORP treatment.
Study 2
Lipman and colleagues (2008) assessed the impact of SNAP® ORP implemented in Hamilton, Ontario. To be eligible for the program, boys had to be 6 to 11 years old, live in Hamilton, and have had police contact or be considered at risk of police contact. Boys accepted into the program had reported police contact or risk of police contact because of elevated scores (T–score>69) for offending behaviors on the CBCL or the Teacher’s Report Form (TRF). Sessions began in February 2002 and ran three times a year (winter, spring, and fall). Because of an extensive waiting list, boys and families waiting at least 6 months formed the comparison group (beginning in April 2005).
The SNAP® ORP treatment group included 223 boys, with an average age of 9.8 years. The comparison group included 116 boys with an average age of 9.4 years. At baseline, SNAP® ORP boys were significantly older than the boys in the comparison group, had more police contacts, had higher rates of any child welfare involvement, and higher scores on the CBCL rule-breaking scale. There were no other significant differences between the groups.
Offending behavior was measured by the rule-breaking, aggressive, and conduct scales on the CBCL and TRF. Social competence was measured by the total problems scale on the CBCL and TRF. Total competence, which represents engagement in community activities, social skills, and school, was measured with the CBCL. Adaptive functioning, which includes emotional problems and learning, was measured with the TRF.
Baseline data was collected during the first telephone and face-to-face interviews. Follow-up data collection occurred at 6, 12, 24, 36, and 48 months after the core component of the program began. Comparison data was collected at pretreatment and 6 months posttreatment while the comparison group was still on the waiting list. This study reported only on the baseline and 6-month follow-up results. Pre–post data on the CBCL was available for 132 boys in the treatment group and 77 boys in the comparison group. Pre–post data measured by the TRF was available for 102 treatment boys and 67 comparison boys.
Two sample t–tests and chi-square analyses were used for descriptive statistics and precomparisons between the treatment and comparison groups. Changes in parent and teacher outcomes were analyzed using repeated measures analyses of variance (ANOVA). Age, measures of child welfare, and time between pre–post assessments (days) were included as covariates.
