Literature review on behavior analysis

Breaking the Silence: Using a Token Economy to Reinforce Classroom Participation

Kurt A. Boniecki Stacy Moore University of Central Arkansas

We propose a procedure for increasing student participation, par- ticularly in large classes. The procedure establishes a token econ- omy in which students earn tokens for participation and then exchange those tokens for extra credit. We evaluated the effective- ness of the procedure by recording the degree of participation in an introductory psychology class before, during, and after implemen- tation of the token economy. Results revealed that the amount of di- rected and nondirected participation increased during the token economy and returned to baseline after removal of the token econ- omy. Furthermore, students responded faster to questions from the instructor during the token economy than during baseline, and this decrease in response latency continued even after removal of the to- ken economy.

A considerable literature attests to the importance of ac- tive learning in which students engage and process course material rather than passively receive it (e.g., Benjamin, 1991; Bligh, 2000; Bonwell & Eison, 1991). One way instruc- tors can facilitate active learning is to challenge the class pe- riodically with relevant questions and encourage students to offer questions and comments. However, instructors may avoid this form of classroom interaction because of a phe- nomenon we call “the silence,” the uncomfortable time fol- lowing the instructor’s question when no one responds. The silence is a particular problem in large classes in which stu- dents feel relatively anonymous and are reluctant to partici- pate (McKeachie, 2002). Instructors can use a variety of techniques to combat the silence, such as waiting out the si- lence (Kendall, 1994), calling on students by name (Gurung, 2002), or initiating small group discussions (McKeachie, 2002). In this article, we present another method for break- ing the silence that is effective and easy to use, particularly in large classes.

Our method relies on extra credit to reinforce participation. Other faculty have used extra credit as an incentive to improve exam performance (Junn, 1995; Nation & Bourgeois, 1978), read journal articles (Carkenord, 1994), seek writing assis- tance (Oley, 1992), demonstrate critical thinking (Junn, 1994), improve behavior modification projects (Barton, 1982), and avoid procrastination (Lloyd & Zylla, 1981; Powers, Edwards, & Hoehle, 1973). Our method creates a to- ken economy in which students earn tokens for participation. Immediately following participation, the instructor presents a token to the student. At the end of class, students exchange their tokens for extra credit toward their course grades.

Hodge and Nelson (1991) also used reinforcement to shape classroom participation. In their study, the instructor wrote students’ initials on the board and placed plus marks next to the initials of students who exhibited the desired amount of participation. Although similar to our method, Hodge and Nelson’s procedure differs from ours in several ways. For instance, their procedure is feasible only in small classes, whereas our method is relatively easy to use in classes of almost any size. Indeed, the first author has successfully used our method in classrooms that seat as many as 200 stu- dents. Also, Hodge and Nelson evaluated the effectiveness of their technique based on students’ self-reported participa- tion. In contrast, we evaluated the effectiveness of our method more objectively by having a research assistant ob- serve the degree of student participation prior, during, and af- ter the token economy.

Method

Participants

Sixty-three undergraduate students enrolled in an intro- ductory psychology course at the University of Central Ar- kansas participated in the study.

Procedure

The class met 75 min twice weekly for 16 weeks. We con- ducted the study over the final 11 class meetings of the term. During each of these 11 class meetings, the instructor period- ically directed relevant questions to the class, and students who wanted to answer the questions raised their hands. The instructor then called on students in the order in which they raised their hands until a student answered the question cor- rectly. If no one raised a hand within 60 sec following a ques- tion, the instructor announced the answer and continued with the lecture.

The first 4 of the 11 class meetings served as the baseline period. During this time, students did not receive any explicit reward for answering a question correctly. Over the next 4 class meetings, the instructor implemented the token econ- omy. The instructor announced that the first person to an- swer a question correctly would receive a token. The tokens were wooden checker pieces purchased from a local hobby store. The pieces were heavy enough to throw, but light

224 Teaching of Psychology

enough not to cause injury if they missed their target. At the end of each class meeting, students could exchange each to- ken for one point added to their next exam grade. Each exam point was worth 0.25% of the course grade. If students did not turn in their tokens at the end of the class meeting, those tokens were void, and students could not exchange them for extra credit in the future. This rule ensured that the instruc- tor had to keep a supply of tokens for only one class meeting and avoided claims of lost tokens. During the final 3 class meetings, the instructor discontinued the token economy and informed students that they could no longer earn tokens for correct answers. As required by our university’s institu- tional review board, the instructor also provided students who had not earned extra credit during the token economy with alternative extra credit opportunities during the re- moval period. After the removal period, the instructor fully debriefed students about the study.

During each of the final 11 class meetings, a research assis- tant sat in the last row of the classroom where she had an un- obstructed view of all students and posed as a student in the class (e.g., by pretending to take notes). The research assis- tant recorded the amount of directed participation (number of students who raised their hands in response to a question from the instructor), latency to participation (amount of time following each question until the first hand was raised), and amount of nondirected participation (number of times any student spontaneously asked the instructor a question or en- gaged the instructor in discussion). The research assistant measured latency using a hand-operated digital stopwatch, which she kept hidden at all times.

Results

The instructor asked 16 questions during baseline, 14 dur- ing the token economy, and 16 during removal. Overall, the instructor asked a mean of 4.18 questions per class meeting. Only once did no student raise a hand following a question from the instructor. We recorded and analyzed this question, which occurred during baseline, as zero directed participa- tion, but removed it from the analysis of latency to participa- tion. Table 1 presents a summary of all three dependent measures across the three phases.

Directed Participation

We analyzed amount of directed participation using fo- cused chi-square tests. We adjusted the expected frequencies to control for the different number of questions across the three phases. Compared to baseline, significantly more stu- dents raised their hands in response to the instructor’s ques- tions during the token economy, χ2(1, N = 77) = 11.85, p < .001. Furthermore, students raised significantly fewer hands during removal than during the token economy, χ2(1, N = 77) = 11.85, p < .001, but the number of hands raised during removal was not significantly different from baseline, χ2(1, N = 52) = 0.00.

Latency to Participation

We conducted a one-way ANOVA of the latency data. Each question from the instructor, rather than each student in the class, constituted the unit of analysis. The ANOVA re- vealed a significant difference between the mean latencies of the three phases, F(2, 42) = 8.23, p = .001, η = .53. Tukey’s honestly significant difference (HSD) test indicated that stu- dents raised their hands significantly faster during the token economy than during baseline (p = .001). However, Tukey’s HSD tests showed that latency to participation during re- moval was not significantly slower than during the token economy (p > .20), but was significantly faster than during baseline (p = .05).

Nondirected Participation

We analyzed amount of nondirected participation using focused chi-square tests. We adjusted the expected frequen- cies to control for the different number of class meetings across the three phases. Compared to baseline, students spontaneously participated significantly more during the to- ken economy, χ2(1, N = 125) = 19.21, p < .001. However, during removal students spontaneously participated signifi- cantly less than during the token economy, χ2(1, N = 120) = 11.56, p < .001. Furthermore, nondirected participation did not significantly differ between baseline and removal, χ2(1, N = 71) = 0.38, p > .44.

Discussion

As we hoped, the amount of directed and nondirected par- ticipation dramatically increased following the implementa- tion of the token economy. Students were more than twice as likely to raise their hands following a question during the to- ken economy than during baseline. Likewise, students were more than twice as likely to ask questions and to make com- ments spontaneously during the token economy than during baseline, even though the instructor did not directly rein- force this form of participation with tokens. Thus, in general, students appeared more willing to contribute to the class dur- ing the token economy. Once the instructor removed the to-

Vol. 30 No. 3, 2003 225

Table 1. Means for the Dependent Measures Across the Three Phases

Dependent Measure Baseline Token

Economy Removal

Directed participation/question 1.63a 3.64b 1.63a

Latency to participation/questiona 6.16a 0.56b 2.93b

Nondirected participation/class period 9.50a 21.75b 11.00a

Note. Values within a row not sharing a subscript are significantly different (p ≤ .05). aTime latencies are reported in seconds.

ken economy, both directed and nondirected participation fell back to baseline levels, but not below them. This result suggests that the token economy did not reduce students’ in- trinsic motivation to participate.

We also were impressed by the shorter amount of time it took students to respond to a question during the token econ- omy compared to baseline. During baseline, an average of 6 sec passed before a student raised a hand, but during the token economy, this latency dropped to less than 1 sec. A person may question whether a student can formulate a thoughtful answer in less than 1 sec. Although we collected no data to address this concern directly, the instructor and research assistant no- ticed little change in the quality of students’ responses across the phases of the study. Furthermore, we believe that, during the token economy, students often raised their hands not be- cause they had an answer, but because they wanted to be the first to answer. Reder (1987) showed that students can quickly assess whether they know an answer before actually recalling the answer from memory. Indeed, during the token economy, many students took a few seconds to formulate their response after being called on by the instructor. In contrast, during re- moval, when there was no competition for tokens, students ap- peared to wait until they formulated an answer before raising their hands—nearly 3 sec, on average, after the instructor asked the question. However, this latency during removal was still half the latency of the baseline phase, which suggests that the token economy may have a lasting effect on the speed of participation.

The contingency between the presence of the token economy and the amount and speed of participation strongly suggests that the tokens were responsible for in- creasing participation. However, we are aware that the de- sign of this study does not allow a definitive causal conclusion. A comparable control group and random as- signment would have provided a stricter test of the token economy’s effectiveness, but these methodological luxuries were not possible. Thus, alternative explanations abound. For example, the instructor covered different topics across the three periods—developmental psychology during base- line, personality and psychological disorders during the to- ken economy, and therapies and social psychology during removal. Perhaps the topics covered during the token econ- omy facilitated more participation than the topics covered during baseline and removal. Nonetheless, we have confi- dence in the token economy for two reasons beyond these results. First, a large body of research attests to the effec- tiveness of token economies and other operant techniques to modify human behavior (Glynn, 1990; Kazdin, 1982; Miltenberger, 1997). Second, the instructor in this study (the first author) has used the token economy effectively across the entire terms of several courses.

In all the classes in which the instructor has used the token economy, only one student has complained of being unable to earn tokens. One way of avoiding this complaint is to pro- vide alternative extra credit opportunities, although too many opportunities may reduce the token economy’s effec- tiveness. Another way is to set a maximum limit on the num- ber of tokens that can be earned. The “faster” students reach the limit early, thereby increasing the chance of other stu- dents earning tokens.

We believe the token economy procedure is a simple and effective means of breaking the silence, especially in large classes. In addition, the procedure serves as an excellent demonstration of operant conditioning and the utility of to- ken economies. Indeed, during the removal period, while the instructor described token economies, one student spontaneously noted that the instructor had used a token economy to increase students’ participation. We believe this sudden connection promotes an “a-ha” experience for the class and a deeper understanding of the material. Fur- thermore, the first author has noticed an increase in stu- dent attendance, enthusiasm, and preparation when he has used the token economy. Students have commented that they enjoy the procedure because it makes class more excit- ing and interactive.

Finally, the token economy system described in this study is flexible and easily adapted to an instructor’s teaching style. We understand that some instructors do not like to use extra credit in their courses. However, instead of extra credit to- ward the students’ course grades, tokens could be worth credit toward “purchasing” desirable options, such as drop- ping a quiz or being excused from the final exam (see Komaki, 1975). Alternatively, instructors could replace tokens with other easily delivered rewards, such as candy. As long as stu- dents perceive a contingency between some positive rein- forcer and their participation, instructors may develop variations to suit their teaching style.

References

Barton, E. J. (1982). Facilitating student veracity: Instructor applica- tion of behavioral technology to self modification projects. Teaching of Psychology, 9, 99–101.

Benjamin, L. T., Jr. (1991). Personalization and active learning in the large introductory psychology class. Teaching of Psychology, 18, 68–74.

Bligh, D. A. (2000). What’s the use of lectures? San Francisco: Jossey-Bass.

Bonwell, C. C., & Eison, J. A. (1991). Active learning: Creating excite- ment in the classroom (Rep. No. ISBN–1–878380–08–7). Washing- ton, DC: School of Education and Human Development, George Washington University. (ERIC Document Reproduction Service No. ED 336049)

Carkenord, D. M. (1994). Motivating students to read journal arti- cles. Teaching of Psychology, 21, 162–164.

Glynn, S. M. (1990). Token economy approaches for psychiatric pa- tients: Progress and pitfalls over 25 years. Behavior Modification, 14, 383–407.

Gurung, R. (2002, June). Sleeping students don’t talk (or learn): En- hancing active learning via class participation. In P. Price (Chair), Active learning in the classroom: Overview and methods. Symposium conducted at the 14th annual meeting of the American Psycho- logical Society, New Orleans, LA.

Hodge, G. K., & Nelson, N. H. (1991). Demonstrating differential reinforcement by shaping classroom participation. Teaching of Psy- chology, 18, 239–241.

Junn, E. (1994). “Pearls of wisdom”: Enhancing student class partici- pation with an innovative exercise. Journal of Instructional Psychol- ogy, 21, 385–387.

Junn, E. N. (1995). Empowering the marginal student: A skills-based extra-credit assignment. Teaching of Psychology, 22, 189–192.

226 Teaching of Psychology

Kazdin, A. E. (1982). The token economy: A decade later. Journal of Applied Behavior Analysis, 15, 431–445.

Kendall, B. (1994). Moment of silence. In E. Bender, M. Dunn, B. Kendall, C. Larson, & P. Wilkes (Eds.), Quick hits: Successful strat- egies by award winning teachers (p. 18). Bloomington: Indiana Uni- versity Press.

Komaki, J. (1975). Neglected reinforcers in the college classroom. Journal of Higher Education, 46, 63–74.

Lloyd, M. E., & Zylla, T. M. (1981). Self-pacing: Helping students establish and fulfill individualized plans for pacing unit tests. Teaching of Psychology, 8, 100–103.

McKeachie, W. J. (2002). McKeachie’s teaching tips: Strategies, re- search, and theory for college and university teachers (11th ed.). Boston: Houghton Mifflin.

Miltenberger, R. G. (1997). Behavior modification: Principles and pro- cedures. Pacific Grove, CA: Brooks/Cole.

Nation, J. R., & Bourgeois, A. E. (1978). PASS, an alternative method of teaching introductory psychology. Research in Higher Education, 8, 273–282.

Oley, N. (1992). Extra credit and peer tutoring: Impact on the qual- ity of writing in introductory psychology in an open admissions col- lege. Teaching of Psychology, 19, 78–81.

Powers, R. B., Edwards, K. A., & Hoehle, W. F. (1973). Bonus points in a self-paced course facilitates exam-taking. Psychologi- cal Record, 23, 533–538.

Reder, L. M. (1987). Strategy selection in question answering. Cog- nitive Psychology, 19, 90–138.

Notes

1. We thank Bill Lammers and Timothy Johnston for their helpful comments on an earlier draft of this article.

2. Send correspondence to Kurt A. Boniecki, University of Central Arkansas, Department of Psychology and Counseling, 201 Donaghey Avenue, UCA Box 4915, Conway, AR 72035; e-mail: [email protected].

Vol. 30 No. 3, 2003 227

Effects on Content Acquisition of Signaling Key Concepts in Text Material

Jeffrey S. Nevid Jodi L. Lampmann St. John’s University

Eighty college students read textbook passages that either included marginal inserts to signal key concepts or did not include these in- serts. Signaling key concepts enhanced performance on content quiz- zes overall and on subsets of items assessing signaled material. Performance was not affected on subsets of items for nonsignaled content. Students reported preferring the signaled format and found it both clearer and easier to understand than the nonsignaled format. Signaling key concepts by extracting and highlighting them in mar- ginal inserts may facilitate encoding and retention of these concepts.

Even in this day of multimedia enhancements in the class- room, textbooks remain very much at the core of the learning process. In recent years, increasing concerns about declining student competencies in mastering basic subject matter have led to the incorporation of numerous pedagogical aids (Weiten & Wight, 1992), including the SQ3R study method, marginal running glossaries, pronunciation guides, built-in or accompanying study guides, self-scoring quizzes, chap- ter-by-chapter learning objectives, and interactive laboratory demonstrations on CD–ROMs and companion Web sites. Publishers are spending increasing amounts of money pro- ducing textbooks, and this increase is passed along to con- sumers via higher prices (Weiten & Wight, 1992). Despite these changes, it remains unclear whether the benefits of learning enhancements are worth the additional costs. Sur- prisingly, there is little research on the use of pedagogical fea- tures as learning devices.

Most reported studies on textbook pedagogy are limited to student surveys. In one survey, Weiten, Guadagno, and Beck (1996) assessed student familiarity with pedagogical devices, their likelihood of using them, and their perceptions of the de- vices’ value. Students were generally familiar with most peda- gogical aids, but reported they rarely used some of the aids, such as outlines and discussion questions. Among the most highly valued and widely used pedagogical aids were boldfaced technical terms, chapter summaries, and running or chapter glossaries.

Other investigators reported similar findings, with students generally endorsing the value of boldfaced technical terms, running or chapter glossaries, chapter summaries, and self-tests (Marek, Griggs, & Christopher, 1999; Weiten, Deguara, Rehmke, & Sewell, 1999). Students also tend both to value and make greater use of pedagogical devices that take little time to read and those that they perceive as relevant in helping them prepare for course examinations (Marek et al., 1999; Weiten et al., 1996). Students appear to be more con- cerned with meeting course demands and less concerned with developing more elaborate study patterns (Marek et al., 1999).

Weiten and his colleagues (1999) reported small, but sig- nificant positive correlations between grade point averages and students’ ratings of how likely they were to use pedagogi- cal devices. Although correlational links between academic success and use of pedagogical aids may be encouraging, they cannot be used as a basis for drawing cause–effect relations.