Reducing Vehicle Accidents at a Residential Agency

by Michele F. Klein, Bethany L. McNamara
New England Center for Children
and
Richard M. Foxx
Pennsylvania State Univesity

Ensuring the safety of children when driving is always a top priority. The Massachusetts Department of Traffic Public Safety (2004) reported that there were a total of 4,178 vehicle accidents in Massachusetts in 2003, and that 434 of those resulted in fatalities. The Massachusetts Public Safety Department also reported that motor vehicle accidents were the leading cause of unintentional deaths of children under 15. Agencies that transport children on a regular basis have a vested interest in developing training procedures designed to decrease the likelihood of vehicle accidents.

Behavior analytic techniques have much to offer in the area of reducing on-site work injuries, and in increasing driver safety (Alvero & Austin, 2004; Huybers, Van Houten, & Malenfant, 2004). For example, Ludwig and Geller (2001) used behavioral interventions to increase the safety belt use of pizza delivery drivers. The present study evaluates an antecedent intervention strategy on motor vehicle (van) accidents at a residential treatment facility. An obstacle course training was developed to deliver in-vivo instruction and feedback on specific driving skills commonly associated with accidents. Skills were identified through a review of past van accidents, and the driver errors that caused them. Although these accidents had not resulted in any serious injuries, damage to vehicles had occurred that required repair and it was considered desirable to decrease the likelihood of future serious injury.

Method

Participants, Settings and Materials
The participants were 350 staff members of the New England Center for Children (NECC). NECC is a private, not for profit program that provides educational and clinical services to children and adults with autism and related disabilities. Twenty-four teams of staff were represented. Each team consisted of 4 to 18 people. As part of their work responsibilities, staff frequently transported residents in NECC’s vans.
Participants were divided into three groups based on the total number of van accidents that occurred during an 8-month baseline period. Table 1 presents the number of teams, participants, and accidents for each group during baseline. All 350 staff members had successfully completed the NECC driver-training program within their first two weeks of employment. This program represented the baseline phase and consisted of a review of driver safety rules and expectations, followed by a written post-test, and participation in commentary driving (driving the van in real traffic conditions for ten minutes with an instructor present while the participant commented on everything he/she saw and did).

Design
A multiple-baseline design across groups was used to assess the effects of obstacle course training on the rate of vehicle accidents. The dependent variable was the total number of van accidents per month, and the intervention was obstacle course driver training.

Settings and Materials
A review of all accidents during the baseline phase revealed that accidents fell into five main categories; backing up, forward turning, tight turning, driving forward, and parking. The obstacle course was constructed such that at least one exemplar from each accident category was included (see Figure 1). The obstacle course was set up in a parking lot and traffic cones and flagged poles were used to direct the trainee Agency vans (12-passenger Ford Econoline^® vans) were used during training.

Figure 1. Depicts the layout of the obstacle course.

Response Definitions
Backing up was defined as driving the van in reverse between two lines of blue cones. The cones were set up so that the participant was required to turn slightly to the right while driving in reverse. Forward turning was defined as driving forward through a series of cones that turned to the right. The cones became narrower as the participant continued to drive forward. Tight turn was defined as driving forward through two flagged poles leading to an execution of a sharp right turn. Straight forward was defined as driving forward through a series of two rows of cones. Parking was defined as turning left to navigate between four flags placed in each corner of the parking space.

Procedure
The Trainers/Observers were ten driver trainers and one driver instructor. Driver trainers had completed a driving course and the driver instructor had completed a 10-hour Liberty Mutual certified driving course. One or two observers were present to score participant driving competency during each training session. Each participant had two opportunities to maneuver through the obstacle course without hitting any cones. Prior to driving the course, driver trainers/instructors explained the path of the obstacle course to the participant. Participants were informed that once they began to drive through the course, no further instructions would be provided. Participants were alone in the van while the observer rated driving as pass/fail on each of the five categories. Pass was scored when the participant negotiated an element of the course without making contact with either cones or flags. Fail was scored if contact was made with flags or cones.

If a participant failed during the initial trial, a second trial was run. Failure to successfully maneuver the course after two trials resulted in a driving instructor/trainer providing corrective feedback to the participant. Feedback was given by a driver trainer/instructor either in the van or with the driver trainer/instructor walking beside the van while the participant negotiated the course. Following corrective feedback the participant was required to pass the training without instructor feedback. On average, participants took eight minutes to complete the obstacle course training.

Interobserver Agreement
During 59% of all trials, two observers recorded the participants’ competency in each category. Overall agreement was determined by dividing agreements by the total number of agreements + disagreements and multiplying by 100. Inter-observer agreement was 100% across participants and categories.

Continue reading this article (page 2)...