Bicycle Rack Initiative: Method

EVALUATION DESIGN

The pilot project was evaluated using a pre-test post-test nonequivalent control group design (Figure 1). Randomization of schools to the intervention and control groups was not possible because the installation of the bicycle racks on school property was at the discretion of each school. Schools in agreement with the installation of bicycle racks were placed in the intervention group.

An equivalent number of schools that chose not to have the bike rack installation during the evaluation period were included in the control group. Control schools were matched as close as possible to the intervention schools in terms of rurality (number of students bused) and school population.

The change in the proportion of students who rode bicycles to school before and after the installation of the bicycle racks was the outcome measure of interest. This was calculated using the number of bicycles on the school property and school attendance at selected observation points.

STUDY POPULATION

All secondary schools involved in this evaluation were from the Limestone District School Board. Four schools agreed to the bicycle rack installation and four schools agreed to be in the control group (Table 1).

DATA COLLECTION

The characteristics of both the intervention and control schools were collected.  The intervention schools and the control schools were randomly assigned over the data collection periods to specific data collection days and times. Each school was visited 3 times pre-intervention, 3 times immediately post intervention (spring), and 3 times post intervention (fall). The number of  bicycles on the school's property was recorded at each visit. Weather conditions were also recorded for each data collection period.  See Appendix B for the evaluation plan and data collection forms.

DATA MANAGEMENT AND ANALYSIS

All data were entered into SPSS Version 15 for Windows.  Over the three observations days, the weighted proportion of students who cycled to school (number of bicycles/school) was determined for each school. The mean change in the proportion of students who cycled to school from pre- intervention to spring post intervention was compared between the intervention group and the control group using the Mann-Whitney U test. Similarly, the mean change in the proportion of students who cycled to school from the pre-intervention to fall post-intervention time periods was compared between the control group and the intervention group using the Mann-Whitney U test. The alpha level was set at .05; meaning there was a 5% chance of falsely declaring an effect if there was no actual effect. Our hypothesis was directional (i.e. there would be an increase in cycling participation at intervention schools); therefore, the one-tailed p value from the Mann-Whitney U tests was evaluated against the set alpha level (11).

ETHICAL CONSIDERATIONS

This project was reviewed in accordance with KFL&A Public Health's policy for scientific review. The project also received approval from the Limestone District School Board and the Algonquin & Lakeshore Catholic District School Board.