We placed recycling receptacles in two locations in academic buildings and studied recycling behavior within an ABA multiple baseline design. During baseline, recycling receptacles were placed in a central location. During the intervention, receptacles were moved into classrooms where beverages were primarily consumed. Baseline conditions were then reinstated. The percentage of cans recycled daily increased during intervention and returned to near-baseline levels during withdrawal. The percentage of cans discarded daily in the trash decreased during the intervention and increased to near-baseline levels during withdrawal. Implications of this study include making recycling more convenient in institutional settings.DESCRIPTORS: aluminum can recycling, stimulus control, response covariationBehavioral research has identified numerous interventions that increase recycling behaviors. Austin, Hatfield, Grindle, and Bailey (1993) used signs and education to increase paper recycling in two college departments. By placing signs over the trash containers and recycling containers, they increased recyclables from 51% in baseline to 84% in the experimental condition. Williams (1991) reported that less than half of the students in residence halls recycled their daily newspapers. He suggested that a majority of students would recycle if drop-off facilities were convenient for them. Brothers, Krantz, and McClannahan (1994) found that when recycling containers were placed in a central location of an office building, 28% of recyclable paper was recycled. When the containers were moved closer to the workers, recycling increased to 88% of all The authors acknowledge the helpfulness of Denise Martz and Stan Aeschleman for their comments on earlier drafts of this manuscript. Portions of this research were presented at the annual meeting of the Association for Behavior Analysis, Washington, DC, 1995. Correspondence concerning this article should be addressed to Timothy D. Ludwig, Appalachian State University, 114 Smith-Wright Hall, Boone, North Carolina 28608 (E-mail: ludwigtd@appstate.edu).paper, and similar results were maintained for up to 7 months afterward.The present study provided a systematic replication of the study by Brothers et al. (1994) applied to student recycling behaviors in university academic buildings. By making recycling more convenient, it was predicted that patrons of two academic buildings would increase their aluminum can recycling. METHOD Participants and SettingsParticipants in the study included patrons of two academic buildings at a southeastern university over a 6-month period. Building A was a rectangular structure with three floors containing nine classrooms (no classrooms on the first floor, four on the second floor, and five on the third floor). Building B was a three-story building that contained 25 classrooms. To make the two buildings comparable, two perpendicular halls containing 10 classrooms on the second floor of Building B were used in the study. Trash containers were located inside each cl...

Safety belt use, turn signal use, and intersection stopping were observed at 3 pizza delivery locations per driver's license plate numbers. After baseline observations, employees at 1 store participated in goal setting targeting complete stops. Employees at the other store were assigned a goal. Over 4 weeks, the group's percentages of complete intersection stopping were posted. Both intervention groups significantly increased their complete intersection stops during the intervention phase. The participative goal-setting group also showed significant increases in turn signal and safety belt use (nontargeted behaviors) concurrent with their increases in intersection stopping (targeted behaviors). Drivers decreased their turn signal and safety belt use concurrent with the assigned goal condition targeting complete stops.

This paper addresses issues and research needs in the domain of behavior modification for injury control. Although much of the discussion focuses on traffic safety, the concepts and principles are applicable to all areas of injury control (e.g. on the job and in the home). Field research that has increased safety belt use is reviewed briefly to illustrate a tripartite classification of injury control factors (i.e. environmental, individual, or behavioral variables), and to introduce a heuristic framework for categorizing and evaluating behavior change strategies. A multiple intervention level hierarchy depicts a progressive segmentation of the target population as more effective (and costly) interventions are implemented; and a taxonomy of 24 behavior change techniques includes a scoring system for predicting short and long term effects of intervention programs. It is presumed that more risk-prone individuals require higher-level interventions, which are those that provide specific response information and extrinsic controls, while also eliciting active participant involvement, social support, and perceptions of autonomy. Although extrinsic controls increase the immediate impact of an intervention program, these techniques may jeopardize response maintenance when the program is withdrawn. ARTICLEInjury is physical harm to a person caused by exchanges with environmental energy; the mechanicalenergy exchanges in falls, shootings and motor vehicle crashes are the most common causes of severe and fatal injuries (Robertson, 1983). In the United States, injury is the prime cause of lost years of productive life (Waller, 1987). In fact, the number of years lost annually to injuries of Americans exceed those lost from cancer by 2.4 million years and from heart disease by 2.0 million years of life (Injury in America, 1985). The factors contributing to an injury can be categorized as person, behavior and environment variables. Figure 1 illustrates these three factors wiui examples from the domains of occupational and vehicle safety. Geller et al. (1989) termed this figure "The Safety Triad', and developed this model as a foundation for a training program to prevent occupational injuries.The person factors are the most difficult to identify as causes of injury, since these factors are numerous and most are not directly observable. For example, an individual's prior training or experience, or his/her attitude may have influenced the injury producing incident. Alternatively, a person's particular skills, abilities, intelligence or physical condition (e.g. as affected by lack of sleep or substance abuse) could have an impact. Furthermore, a risky lifestyle or personality (e.g. Donovan et al., 1983;Wilson and Jonah, 1988) increases one's propensity to emit injury-prone behaviors.The environmental conditions that are potential contributors to injury are relatively easy to identify and modify because they are observable and finite. The presence of equipment hazards fall under this category, as well as the availability ...

This clarification of the major factors emerging in the measurement of safety cultures should impact the industry through a more accurate description, measurement, and tracking of safety cultures to reduce loss due to injury.

A practical intervention program, targeting the safety belt use of pizza deliverers at two stores, increased significantly the use of both safety belts (143% above baseline) and turn signals (25% above baseline). Control subjects (i.e., pizza deliverers at a third no-intervention store and patrons driving to the pizza stores) showed no changes in belt or turn signal use over the course of 7-month study. The intervention program was staggered across two pizza stores and consisted of a group meeting wherein employees discussed the value of safety belts, received feedback regarding their low safety belt use, offered suggestions for increasing their belt use, and made a personal commitment to buckle up by signing buckle-up promise cards. Subsequently, employee-designed buckle-up reminder signs were placed in the pizza stores. By linking license plate numbers to individual driving records, we examined certain aspects of driving history as moderators of pre- and postintervention belt use. Although baseline belt use was significantly lower for drivers with one or more driving demerits or accidents in the previous 5 years, after the intervention these risk groups increased their belt use significantly and at the same rate as drivers with no demerits or accidents. Whereas baseline belt use was similar for younger (under 25) and older (25 or older) drivers, younger drivers were markedly more influenced by the intervention than were older drivers. Individual variation in belt use during baseline, intervention, and follow-up phases indicated that some drivers require more effective and costly intervention programs to motivate their safe driving practices.

A foodservice distributor in the southeastern United States implemented a voice assisted selecting tool to reduce selector errors by providing immediate feedback when errors occurred. An AB design with a nonequivalent comparison group was used to examine the effects of the voice technology on 132 selectors whose mispicks and shorts were collected over 6 weeks of baseline and 8 weeks of the intervention phase. Selector errors were reduced from 2.44 errors per 1,000 cases picked to 0.94 errors per 1,000 cases when voice technology was implemented. Further analysis indicated that the immediate feedback provided by voice had a greater impact on employees who were making the most errors during baseline.

No abstract

The purpose of this study was to evaluate a program designed to improve the work quality of employees at a food distribution warehouse. Participants (n = 23) were responsible for selecting food items to be distributed to restaurants. Selector errors were targeted using an adapted incentive program featuring tiered goals and individualized normative feedback. A pre-existing incentive program, wherein selectors forfeited money for each error, was adapted to allow the selectors the opportunity to earn back their disincentive money upon attainment of weekly goals that were increasingly demanding. The initial goal for the first two weeks was set at 2.0 errors per 1000 cases and every two weeks thereafter the new goal was reduced by 0.1 errors per 1000 cases until the final goal of 1.5 errors per 1000 cases. Weekly graphic feedback included public posting of individual errors rates. The mean errors per 1000 cases for the experimental group decreased 10% from 3.16 during baseline to 2.54 during the intervention. The largest decreases in errors was observed in the mid-level performers (with baseline error rates between 2 and 3.5 errors per 1000 cases). A cost-benefit analysis revealed potential savings of $9,799.50.