Study descriptions
| Author (year), country | Intervention group | Control group | Weaknesses | Strengths |
|---|---|---|---|---|
| *p<0.05. | ||||
| Chang (1995), US19 | Behavior Pre: 21.9% Post: 44.3% p<0.001 Knowledge Seat belt (4/4), p<0.02 Car seat (1/4), p<0.001 Practice (2/4), p<0.001 | Behavior Pre: 22.1% Post: 23.8% p = NS Knowledge not tested | 1. Blind observation not reported 2. No randomization 3. No report intensity/dose 4. Random knowledge pre and post—? Same children | 1. Matched for baseline seat belt use, schools size, education philosophy 2. Minimum essential program activities 3. Control group received intervention post testing |
| Renaud (1989), Canada18 | Three intervention groups (attitude, behavior, attitude/behavior) v control Attitude mean difference 1. 0.60* 2. 0.72* 3. 0.52* Behavior mean difference 1. 0.33 2. −0.57 3. 0.94* Observation mean difference 1. 0.83 2. 2.48 3. 2.22 | 1. Attitude and behavior tested immediately following session 2. Blind observation not reported 3. Post test only | 1. Children randomly assigned 2. Trained observer 3. Measure validated 4. Control for gender and school in analysis | |
| Bowman (1987), Australia12 | Pre: 60.6% Post: 75.0% p<0.009 | Pre: 59.9% Post: 60.3% p = 0.93 | 1. Intensity/dose not reported 2. Pre/post observation? Same children | 1. Schools randomized 2. Trained observer 3. Interrater reliability 4. Matched for baseline 5. Intervention materials provided |
| Rothengatter (1984), Netherlands20 | Knowledge (parent trainer) 26.80–28.71 (Assistant trainer) 27.65–31.82 | Knowledge 27.33–29.32 | 1. Intensity/dose not reported 2. No randomization 3. Observation tool not reported 4. Blind observation not reported 5. No group statistical test differences reported | 1. Program pilot tested 2. Trained interventionist 3. Minimum dose 4. Control group received intervention post testing |
| Luria (2000), US17 | Crossing street m = 1.90, p = 0.29 Calling 911 m = 4.15, p = 0.41 Stranger danger m = 1.33, p = 0.57 | Crossing street m = 1.40, p = 0.29 Calling 911 m = 3.66, p = 0.41 Stranger danger m = 1.65, p = 0.57 | 1. Community instructors—training not reported 2. Blind observation not reported 3. Newly developed instrument—not tested 4. Intensity/dose not reported | 1. Schools randomized |
| Richards (1991), US13 | Knowledge m = 2.58, p<0.05 Seat belt use 60–75%, NS | Knowledge m = 1.28, p<0.05 Seat belt use 60–75%, NS | 1. Blind observation not reported 2. Intensity not reported 3. Self report by preschoolers | 1. Program pilot tested 2. Instrument tested 3. Teachers trained |
| Thomson (1992), UK15 | Demonstration Pre: 14% Post: 37% t = 2.41, p<0.05 Post 2: 37% t = 2.83, p<0.05 Road behavior Pre: 10% Post: 35% t = 2.17, p<0.05 Post 2: 34% t = 2.36, p<0.05 | Demonstration Pre: 4% Post: 12% Post 2: 12% Road behavior Pre: 4% Post: 12% Post 2: 12% | 1. Children randomly selected but no report of random assignment | 1. Trained interventionist 2. Observation assessment—interrater reliability 3. Randomly selected 4. Control for gender 5. Consistent intervention—dose monitored |
| Liller (1998), US14 | Knowledge OR 2.2–144.5, p<0.05 | 1. Post test only 2. School sited selected by School Board 3. No randomization 4. Blind observation not reported | 1. Teachers prepared 2. Instrument pilot tested | |
| Thomson (1998), UK16 | Road skill Pre: 15% Post: 43% t = −4.95, p = 0.001 Post 2 Pre: 15% Post: 35% t = −4.11, p = 0.001 | Road skill Pre: 16% Post: 13% Post 2 Pre: 16% Post: 16% | 1. Matched sample for gender, school 2. Trained parent volunteers 3. Intervention previously tested 4. Interrater reliability established 5. Blind observation reported 6. Consistency of dose monitored | |