Association for Behavior Analysis International

This symposium includes four talks that span laboratory and applied organizational behavior management investigations. Two of the presentations describe human operant investigations of feedback sequencing in simulated work environments. Both studies evaluate claims of efficacy and preference for the feedback sandwich, a highly promoted method of delivering feedback in organizations. The remaining presentations were conducted in applied settings addressing real world employee performance problems. The third presentation evaluates the effects of self-monitoring involving interlocking behavioral contingencies on work performance and sales. The final presentation will share findings from a study investigating the effects of task clarification and/or feedback on the customer service skills of telephone operators. The symposium will conclude with considerations and future research directions from a discussant.

Feedback is an effective method for increasing performance in a wide range of settings, although questions remain regarding the most effective characteristics of feedback. The purpose of the present study was to investigate (1) the efficacy of and preference for feedback sequence (positive-corrective-positive, positive-positive-corrective, corrective-positive-positive, no feedback), and (2) the influence of feedback timing on efficacy and preference. Undergraduate participants performed four office tasks each associated with a feedback sequence, presented in a counterbalanced fashion. Study 1 evaluated the effects of feedback sequence on performance when feedback was delivered immediately after the session. The corrective-positive-positive sequence resulted in the greatest increase from baseline mean for two participants. The positive-corrective-positive sequence was most efficacious for one participant. An efficacious feedback sequence was not identified for the fourth participant. The corrective-positive-positive, positive-positive-corrective, and no feedback conditions were preferred by two of four participants. Study 2 evaluated the effects of feedback sequence on performance when feedback was delivered before the session. Preliminary data for two participants suggest the corrective-positive-positive and no feedback conditions resulted in the greatest increase in performance from baseline mean. Participants indicated preference for the no feedback, positive-corrective-positive, and positive-positive-corrective sequences. Data collection for the remaining two participants is underway.

The sandwich method of feedback is frequently used in organizational settings as a way to deliver information on performance. Previous research has suggested that the sandwich method improves performance, but not any more than descriptive feedback. However, participants prefer descriptive feedback to the sandwich method. The purpose of this study was to extend previous research by comparing several components of the sandwich feedback method to descriptive feedback. In addition, a choice phase was conducted to examine whether participants prefer one type of feedback over another. This was done across four participants in a multiple baseline design with a built in reversal design, measuring a response that consisted of entering information into a computerized data entry system. Results show that participant responding varied across interventions, although 3 of 4 participants selected the descriptive feedback condition in the choice phase. Interobserver agreement was 100% across all sessions, and treatment integrity was 99.12% across 48.9% of sessions.

The interlocking contingencies that exist between a customer and sales representatives may contribute to buying decisions. The purpose of this study is to redesign the sales call verbal process of 5 sales representatives to shape the verbal behaviors most likely to result in a sale. Verbal behaviors related to sales were identified by sales representatives in a focus group meeting and were used to develop a self-observation check sheet. The self-monitoring phase was associated with an average increase of 12 percentage points over baseline. These changes were associated with 65% more sales during the intervention period for the experimental group compared with 16% more sales for the comparison group. The interlocking contingencies between certain sales representative verbal behaviors, consisting of Contact (asks for customers contact information including email address) and Deposit (states how much of a deposit can get the order started) and certain customer verbal behaviors consisting of Phone (provided phone number) and Want (did customer say they want to buy a building) seemed to occasion an increase in customer buying behavior. For each sales representative, this increase in sales translates into $1,126,528 more in annualized revenue and $4,989 more in annualized sales commission.

The director of the Continuing Education department at a private university targeted four sets of employee telephone operator customer service skills for improvement. Each set of skills was a composite of several important customer service behaviors according to Telephone Doctor, a popular online customer service training provider. Composite skills included Greetings (appropriate greeting statement, department and university identification, and employee identification, all in the correct order), Communication Skills (requesting a customers name in a particular way, using the customers name to personalize calls), Closing (thanking the customer, offering additional help, and an appropriate farewell statement), and [another set of skills to be determined]. Two employees whose natural telephone performance with real customers was recorded on a daily basis for the duration of the study participated. We employed a modified multiple baseline design with counter-balancing measures to control for potential order effects following sequential exposure to the two independent variables: graphic feedback delivered by a supervisor, and the provision of task clarification. That is, graphic feedback followed by task clarification was implemented for the first two component skills and task clarification followed by graphic feedback for the final two skills. Preliminary results suggest that feedback was effective.