You are on page 1of 19

JOURNAL OF APPLIED BEHAVIOR ANALYSIS 1979, 12.

467-485 NUMBER 3 (FALL 1979)

SELF-CONTROL TRAINING IN THE CLASSROOM.


A REVIEW AND CRITIQUE
MICHAEL S. ROSENBAUM AND RONALD S. DRABMAN
UNIVERSITY OF MISSISSIPPI MEDICAL CENTER

Self-control training in classroom settings is becoming widespread. Establishing effective


self- rather than externally controlled behavior modification programs in schools would
enable children to control their own academic and social behavior, while enabling
teachers to devote more time to teaching. The following components of self-control
are reviewed in the present article: self-recording, self-evaluation, self-determination of
contingencies, and self-instruction. Self-control strategies designed for the maintenance
of appropriate classroom behavior, and issues associated with self-control training, such
as the reliability of self-observation, response maintenance, generalization, and the role
of external control, are examined. Finally, suggestions for maximizing the potential
effectiveness of self-control training in the classroom (e.g., teaching self-observational
procedures, teaching students to provide themselves with instructions and praise), as
well as future areas for experimental investigation (e.g., social changes that may be as-
sociated with self-control procedures), are presented.
DESCRIPTORS: self-control, self-management, self-recording, self-regulation, class-
room behavior

Behavior modification programs have demon- tion. Teachers have also been assigned the re-
strated that the academic and social behavior of sponsibility of controlling the social behavior of
school-aged children in classrooms can be effec- children in the classroom. One inevitable result
tively altered through a variety of procedures of this type of educational system is that, be-
(Ayllon and Rosenbaum, 1977; Copeland and cause of involvement with the mechanics associ-
Hall, 1976; Drabman, 1976; O'Leary and Drab- ated with developing academic skills and con-
man, 1971). For the most part, these programs trolling social behavior, the time available for
have focused on using external agents (parents, the teacher to teach becomes restricted.
teachers, peers, therapists) to arrange and admin- Kazdin (1975a) has discussed several potential
ister the contingencies applied in the classroom. disadvantages associated with relying on external
Lovitt (1973) has pointed out that teachers agents to design and administer classroom con-
have traditionally been delegated the func- tingencies: (a) a great deal of behavior may
tions of evaluating the child's performance in occur unnoticed by these external agents, espe-
an academic program and delivering the appro- cially when they work with several children
priate consequences contingent on that evalua- simultaneously; (b) because external agents are
paired with the administration of contingencies,
Preparation of this manuscript was supported in they may acquire discriminative properties and
part by National Institute of Mental Health grant become cues for the performance of target be-
#1 ROI MH 28367-01 to Ronald S. Drabman. The haviors, the occurrence of which may become
authors express their sincere thanks to Stephanie B.
Stolz and Donald M. Baer for their invaluable assis- restricted to the presence of those cues; (c) diffi-
tance throughout the revision process of this manu- culty may be encountered in generalizing behav-
script. Reprints are available from Ronald S. Drab- iors to situations in which externally adminis-
man, Department of Psychiatry and Human Behavior,
University of Mississippi Medical Center, 2500 North tered contingencies have not been applied. To
State Street, Jackson, Mississippi 39216. avoid these disadvantages, researchers (e.g., Kaz-
467
4,-68 MICHAEL S. ROSENBAUM and RONALD S. DRABMAN

din, 1975a; O'Leary and Drabman, 1971; ment (e.g., recording the number of occurrences
O'Leary and O'Leary, 1976) have emphasized of a particular behavior). In self-evaluation, the
the potential importance of self-control proce- self-monitoring process is followed by an evalu-
dures with which children can be taught to ation of the behavior on a subjective basis, usu-
evaluate their performance and administer the ally with an externally provided criterion (e.g.,
appropriate contingencies. These procedures rating behavior according to a scale of 1-10).
might enable children to develop their own
academic and social skills, while teachers devote Self-Recording
more time to teaching and less time to classroom Many authors have found self-recording alone
mechanics and social behavior. to be effective in producing behavior change
Several authors have reviewed and discussed in nonclassroom settings (e.g., Johnson and
self-control processes in children (e.g., Bandura, White, 1971; Kanfer, 1970b, 1971; Kazdin,
1969; Karoly, 1977; Masters and Mokros, 1974; 1974b; Lipinski and Nelson, 1974; Nelson,
O'Leary and O'Leary, 1972, 1976). Whereas Lipinski, and Black, 1976a). This finding has
early studies of self-control with children were also received support from studies conducted in
conducted in laboratory settings, recent research classrooms. The effects of self-recording on aca-
has been conducted in the classroom (McLaugh- demic behavior and classroom disruption were
lin, 1976). The purpose of the present article is investigated by Broden, Hall, and Mitts (1971).
to: (a) present research dealing with self-control In one experiment, self-recording appeared to
training in classrooms, (b) evaluate the effective- be associated with a 48% increase in study be-
ness of this training, (c) discuss factors associated havior, even though self-recording did not occur
with effective self-control training, and (d) sug- in 10 of 23 sessions, and the correlation between
gest some future directions for self-control re- student records and those of an independent
search in the classroom. observer was very low. Self-recorded study be-
havior was also paired with praise from the stu-
Self-Observational Procedures dent's counselor, thereby precluding separate
Behavioral self-observation involves individu- assessment of the effects of self-recording from
als monitoring their own behavior and subse- those of self-recording plus praise. It appears
quently recording that behavior. This is the likely that instructions and the opportunity to
initial and an extremely important step in self- self-record may have been responsible for the
control training. It provides individuals with observed changes in study behavior irrespective
systematic data regarding their behavior, with of the occurrence or accuracy of self-recording.
which they can evaluate the effects of any pro- In a second experiment, a decrease in inap-
posed strategies for change. Self-observation can propriate verbalizations was initially obtained
have a reactive effect on the target behavior with self-recording in the absence of any contin-
being self-observed, whereby the behavior gencies, but these verbalizations gradually in-
changes as a function of the self-observational creased until the rates were equal to those in
process (e.g., Jeffrey, 1974; Kanfer, 1970b, baseline. As in the first experiment, student
Kazdin, 1974a, 1974b; McFall, 1977; Nelson. records were not significantly correlated with
1977). In this capacity, self-observation also those of an observer. The results of this experi-
functions as a behavior-change technique. ment showed that self-recording could be effec-
O'Leary and O'Leary (1976) have made a dis- tive in at least initiating a desirable change in
tinction between self-recording and self-evalua- behavior.
tion, both of which are self-observational pro- Working with potential high school dropouts,
cedures. In self-recording, behavior is monitored Gottman and McFall (1972) found that ap-
and recorded with a minimum amount of judg- propriate classroom participation significantly
SELF-CONTROL IN THE CLASSROOM 469

increased when self-recorded and decreased in a desirable direction (i.e., appropriate be-
when nonparticipation was self-recorded. In haviors have increased and inappropriate behav-
two experiments with elementary school chil- iors have decreased). Although self-recording
dren exhibiting behavior disorders, Lovitt has also been shown not to have an effect on
(1973) found that self-recording was associated target behaviors, there has been no evidence of
with decreases (86% and 55 %, respectively) any deleterious effects.
in inappropriate behaviors. Although these lat-
ter two studies showed that desirable effects Self-Evaluation
could be obtained with self-recording, the period Several authors have emphasized the necessity
of time allotted for self-recording was too brief of an externally provided criterion for self-
to demonstrate any long-term effects of this pro- evaluating performance before behavior change
cedure. can occur (Kanfer, 1970a; Kanfer and Phillips,
Knapczyk and Livingston (1973) found that 1970; Spates and Kanfer, 1977). However,
the addition of self-recording to token rein- when Santogrossi, O'Leary, Romanczyk, and
forcement did not produce any significant Kaufman (1973) had disruptive adolescents in
changes in the reading performance of junior a psychiatric hospital school rate their own be-
high school educable mentally retarded students havior according to rules of classroom conduct,
relative to token reinforcement alone. Their this self-evaluation procedure did not produce
data, however, reflected a decreasing trend when any substantial effects on the level of disruptive
self-recording was removed and an increasing behavior. Teacher-determined reinforcement re-
trend when it was reinstated. It is therefore pos- sulted in an 80% decrease in disruptive behav-
sible that self-recording may have added slightly ior, which was maintained for only a brief pe-
to the effects of token reinforcement. Ballard riod by self-evaluation plus self-determined
and Glynn (1975) found that self-recording had reinforcement. Additionally, Turkewitz, O'Leary,
no effect on the writing skills (e.g., number of and Ironsmith (1975) found that self-evaluation
sentences written) of third graders. When self- had no effect on the level of disruptive behav-
reinforcement was added to self-recording, writ- ior in children manifesting academic and social
ing output more than doubled. problems. Teacher-determined reinforcement
The studies dealing with self-recording in produced a 68% decrease in disruptive behavior
classrooms indicate that it can initiate desirable relative to baseline. This low level of disruption
changes in behavior (Broden et al., 1971; Gott- was maintained by a combined self-evaluation,
man and McFall, 1972; Lovitt, 1973) and en- self-reinforcement, and matching procedure.
hance the effects of other reinforcement proce- Self-evaluation has also been included as a
dures (Knapczyk and Livingston, 1973). The component in a package of techniques designed
results of the Ballard and Glynn (1975) study to teach self-control skills (Bolstad and Johnson,
suggest that the changes obtained with self- 1972; Drabman, Spitalnik, and O'Leary, 1973;
recording may require reinforcing contingencies Robertson, Simon, Pachman, and Drabman, in
in order to maintain these behavioral changes. press). In these studies, the role of self-evalua-
Variables affecting the pattern of behavior tion was unclear, because its effects after being
change associated with self-recording and theo- paired with token reinforcement could not be
retical mechanisms for these changes have been separately assessed.
discussed by several authors (Kazdin, 1974a; Neither self-recording nor the self-evaluation
Nelson, Lipinski, and Black, 1976b). The ap- of behavior according to some criterion can be
plied literature has shown that in those instances performed without self-monitoring of that be-
where self-recording has been effective, the havior. The results of only two studies demon-
changes obtained in target behaviors have been strating no effect associated with self-evaluation
44170 MICHAEL S. ROSENBAUM and RONALD S. DRABMAN

alone do not rule out the possibility that self- to the students, have resulted in high levels of
evaluation may have a desirable effect on target reliability in self-observed behavior (Bolstad and
behaviors. Yet, it seems doubtful that requiring Johnson, 1972; Drabman et al., 1973; Robert-
students to evaluate their behavior relative to son et al., in press; Turkewitz et 41., 1975).
an externally provided criterion will increase the Other authors have reported high levels of re-
effectiveness of self-recording. The failure of liability of self-recorded behavior without using
externally instructed self-evaluation to add to matching procedures (Ballard and Glynn, 1975;
the effectiveness of self-recording is not surpris- Frederiksen and Frederiksen, 1975; Glynn,
ing, since in applied settings instructions to self- Thomas, and Shee, 1973; Knapczyk and Liv-
record probably serve as discriminative stimuli ingston, 1973; Thomas, 1976). These studies
for self-evaluation. either did not assess the role of self-recording or
self-evaluation alone or found the procedures
Reliability of Self-Observation did not have an effect on target behaviors. In
Reliability refers to the consistency of mea- light of the findings described above, it may not
surement between self-recordings of behavior be necessary to program for high levels of ac-
and those of one or more independent observers. curacy in self-recorded behavior in order to ob-
Several self-control studies have included reli- tain desirable changes in target behaviors.
ability measures for observers' records of student
behavior, without assessing the reliability of stu- Conclusions
dents' records (Glynn and Thomas, 1974; Gott- The results of these studies dealing with self-
man and McFall, 1972; Humphrey, Karoly, and observational procedures lead us to the follow-
Kirschenbaum, Note 1). Other studies have ing tentative conclusions: (a) Self-recording ap-
compared self-recordings and observer records pears to have either desirable or no effects on
and have reported low correlations between target behaviors. These desirable effects have,
these two measures (Broden et al., 1971; Kauf- unfortunately, consisted of only modest changes.
man and O'Leary, 1972; Turkewitz et al., 1975). To increase their applied significance, future
In these studies, it is not clear if self-monitoring studies should focus on variables that can pro-
occurred or whether the self-recording of target duce greater magnitudes of behavior change.
behaviors or of any behaviors was responsible for Self-recording has not been shown to have a
the changes obtained in target behaviors. What deleterious effect. (b) Desirable effects associ-
does seem clear is that instructions to self-record ated with self-recording may be short-term,
a target behavior combined with self-recording requiring the addition of reinforcing contingen-
of some behavior(s) can have a desirable effect cies for their maintenance. (c) Self-evaluation
on the target behavior (Broden et al., 1971; does not appear to add to the effectiveness of
Gottman and McFall, 1972; Lovitt, 1973). self-recording. (d) Self-recording does not have
Several self-control studies have included to be accurate in order to produce desirable
elaborate procedures to assure high levels of ac- changes in target behaviors. (e) Further assess-
curacy in self-recorded target behaviors. These ment of accurate versus inaccurate self-record-
procedures have consisted of awarding students ing is needed to determine which, if any,
bonus points for matching or being within a situations will require training for accurate self-
specified range of an external agent's records recording, clearly a more costly procedure than
and penalizing them for being beyond that simply allowing students to self-record regard-
range. Matching procedures, whether abruptly less of their accuracy.
or gradually withdrawn in an effort to transfer Future research could include a multiple-
control of contingencies from an external source baseline-across-behaviors design with initially
SELF-CONTROL IN THE CLASSROOM 471

inaccurate self-recorders to investigate this is- three groups during Phase III and was main-
sue. Changes obtained in target behaviors A tained throughout Phase IV. The differences be-
and B following training in accurate self- tween the two self-regulated groups and the
recording for behavior A would indicate that in- externally regulated group were not significant
accurate self-recording is as effective as accurate in any of these phases. Yet, the self-regulated
self-recording. The failure of behavior B to groups had 42% fewer disruptions than the
change prior to training in accurate self-record- externally regulated group in Phase III and
ing for this behavior would suggest the neces- 39% fewer disruptions in Phase IV. Disruptive
sity for training accurate self-recording in cer- behavior increased for all three groups in Phase
tain situations. (f) Future studies designed to V, with no significant differences between the
investigate the effects of self-observational proce- self-regulated groups and the externally regu-
dures used alone need to provide adequate peri- lated group. Once again, however, the self-
ods of time for assessing these effects. regulated groups averaged 39 % fewer disrup-
tions than the externally regulated group.
Farnum, Brigham, and Johnson (Note 2)
SELF-DETERMINED versus found that although both teacher- and student-
EXTERNALLY DETERMINED determined performance goals were effective in
CONTINGENCIES increasing the math productivity of fifth graders,
A number of studies have been conducted to students consistently completed the most work
compare self- with teacher-determined contin- in the self-determined condition and overwhelm-
gencies. Lovitt and Curtiss (1969) investigated ingly preferred self-determined contingencies.
this strategy in two experiments with the aca- The response-facilitation effect of being able to
demic responses of a student who had behavior choose one's own contingencies has received ad-
problems and found an increase in median re- ditional support from several studies by Brig-
sponse rate (47% and 44%, respectively) when ham and his associates (e.g., Brigham, Note 3;
contingencies were student-determined. Follow- Brigham and Bushell, 1973; Brigham and Sher-
ing baseline (Phase I), Bolstad and Johnson man, 1973; Brigham and Stoerzinger, 1976).
(1972) had each of three groups of disruptive Research by Glynn (1970) involved four dif-
first and second graders receive externally de- ferent ninth-grade classes receiving either ex-
termined reinforcement (Phase II). In Phase III, perimenter-, self-, chance-determined, or no
two of these groups recorded their own behavior reinforcement. Test performance was measured
and received reinforcement for matching an ac- across these groups during.token reinforcement,
curacy criterion (self-regulated). Matching was extinction, and self-reinforcement. The results
withdrawn in Phase IV, in which reinforcement showed that during token reinforcement and
for the two self-regulated groups was deter- extinction, self- and experimenter-determined
mined by their self-ratings of disruptive behav- reinforcement were equally effective in produc-
ior. The third group received externally deter- ing significantly higher test performance than
mined reinforcement throughout Phases III and chance-determined and no reinforcement. When
IV. allowed to determine their own reinforcement,
When reinforcement was withdrawn in Phase students previously in the chance-determined
V (extinction), one self-regulated group con- group performed worse than the other two re-
tinued to self-record disruptive behavior. The inforcement groups and the no-reinforcement
results showed that disruptive behavior de- group. It appeared that an inconsistent rein-
creased at least 67 % for all three groups in forcement history, as in the chance-determined
Phase II. A further decrease was obtained in all group, might impair students' subsequent abili-
472 MICHAEL S. ROSENBAUM and RONALD S. DRABMAN

ties to apply self-determined contingencies ef- lenient standards during reinforcement. During
fectively. extinction, no significant differences were found
Results obtained by Felixbrod and O'Leary between stringent and lenient performance stan-
(1973, 1974) and Frederiksen and Frederiksen dards. Those groups whose standards were previ-
(1975) further supported and extended the find- ously self-determined, however, completed sig-
ings of Glynn (1970), that self- and externally nificantly (p < .05) more math problems than
determined reinforcement can be equally effec- those groups who had externally imposed stan-
tive in producing behavior change. Felixbrod dards, regardless of the stringent versus lenient
and O'Leary (1974) also found that during ex- distinction.
tinction, the differences in academic performance
between a group that had externally imposed
standards and one whose standards were self- Conclusions
determined during a prior reinforcement period The research reviewed in this section leads us
were not significant. Yet, the former group out- to suggest the following tentative conclusions
performed the latter group by at least 40%. regarding self-determined contingencies: (a)
The effect of stringent standards on perfor- Self-determined contingencies can be as effective
mance has been discussed by several authors as or more effective than externally determined
(e.g., Bandura, 1974; Kanfer, 1970a; Liebert, contingencies during periods of reinforcement.
Hanratty, and Hill, 1969; Locke, Cartledge, and (b) The role of prior self-determined contin-
Koeppel, 1968). In laboratory settings, stringent gencies during subsequent extinction periods is
standards have been established through verbal not clear. One study (Felixbrod and O'Leary,
instructions (e.g., Liebert and Allen, 1967; Lie- 1974) found students who previously deter-
bert and Ora, 1968) and modeling (e.g., Ban- mined their own contingencies performed worse
dura, Grusec, and Menlove, 1967; Bandura and than those whose contingencies were externally
Kupers, 1964; Hildebrandt, Feldman, and Dit- determined. Another study (Brownell et al.,
richs, 1973; McMains and Liebert, 1968; Mi- 1977) found the opposite effect. Further re-
schel and Liebert, 1966, 1967). Several studies search is needed to provide more data concern-
conducted in classrooms have shown that stu- ing the effects of prior self-determined con-
dents tend to select more lenient performance tingencies relative to externally determined
standards in self-determined reinforcement rela- contingencies during subsequent periods of ex-
tive to externally imposed standards in exter- tinction. (c) Being able to choose one's own
nally determined reinforcement (Felixbrod and contingencies may function as a response facili-
O'Leary, 1973, 1974; Frederiksen and Frederik- tator. To date, this response facilitation effect
sen, 1975). Brownell, Colletti, Ersner-Hersh- appears to be moderate. Future studies should
field, Hershfield, and Wilson (1977) developed attempt to identify variables resulting in greater
a prompt for selecting stringent standards and magnitude of behavior changes, thereby in-
socially reinforced the selection of these stan- creasing their applied significance. (d) An incon-
dards for the academic performance of third and sistent pattern of externally determined rein-
fourth graders. Their results showed that stu- forcement can have a detrimental effect on
dents in a self-determined performance stan- an individual's subsequent ability to apply self-
dards group presented with the prompt se- determined reinforcement. (e) Students' tenden-
lected more stringent standards than students cies to choose lenient performance standards
in a self-determined group without the prompt. when determining their own reinforcement can
Furthermore, stringent standards, whether self- be altered by prompting the selection of more
or externally determined, produced significantly stringent standards, which can produce greater
(p < .05) greater academic performance than performance than lenient standards.
SELF-CONTROL IN THE CLASSROOM 473
suited in an 80 % decrease in disruptive behavior
SELF-CONTROL STRATEGIES IN relative to baseline. When the students were al-
THE MAINTENANCE OF lowed to determine their own reinforcement,
APPROPRIATE CLASSROOM this low level of disruption was maintained only
BEHAVIOR for a short time, with disruptive behavior even-
Authors of the major reviews of the token tually increasing to baseline levels. During this
economy literature (Kazdin, 1975b, 1977; Kaz- period the students were giving themselves the
din and Bootzin, 1972, 1973; O'Leary and highest possible ratings, even while displaying
Drabman, 1971) have emphasized the use of alarming levels of disruptive behavior. A match-
self-control skills to maintain appropriate be- ing condition, instituted to resolve this problem,
haviors initially obtained with externally de- resulted in higher levels of disruption than dur-
termined reinforcement. Working in a psychi- ing baseline. It appeared that two problems
atric hospital school, Kaufman and O'Leary were associated with this method of introducing
(1972) found that adolescent students in a token matching: (a) a second abrupt shift in stimulus
reinforcement classroom obtained an average control occurred, with reinforcement becoming
82% reduction in disruptive behavior relative to partially controlled by the teacher; and (b) it
nontoken phases. An average 78 % decrease in represented an aversive procedure, in that the
disruptive behavior relative to nontoken phases magnitude of reinforcement was lowered, due to
was found for students in a response cost token the penalty for inaccurate self-evaluations.
classroom. Following several periods of exter- To resolve the problems in the Santogrossi
nally determined contingencies, students in both et al. (1973) study, Drabman et al. (1973) used
classrooms were instructed to rate their own be- the following successive steps to transfer the
havior and decide on the number of tokens they evaluation and reinforcement duties gradually
should receive. Using these self-control proce- from the teacher to the students: (a) students
dures, students in both groups maintained the recorded points awarded by the teacher; (b) stu-
low levels of disruptive behavior previously dents were awarded bonus points for matching
achieved with externally determined contingen- teacher ratings; (c) matching was gradually
cies, while awarding themselves the highest pos- faded in four phases, with successively smaller
sible ratings. numbers of students being required to match the
The authors suggested several possibilities for teacher; and (d) students rated their own behav-
the maintenance of appropriate behavior: (a) ior and determined their own reinforcement
self-administration of contingencies may have independently of the matching criterion. Rela-
become reinforcing; (b) failure to remain well- tive to the introduction of matching in the
behaved might have resulted in a return to Santogrossi et al. (1973) study, this method
teacher ratings, which would have been lower avoided an abrupt shift in stimulus control by
than student ratings, thereby resulting in less enabling the teachers to fade their roles grad-
reinforcement; (c) academic skills may have ually. Additionally, it provided an opportunity
developed during teacher-determined reinforce- for an increased magnitude of reinforcement
ment and may have become intrinsically rein- (bonuses). A 57 % increase in academic per-
forcing; and (d) students may have been adven- formance and a 67 % decrease in disruptive be-
titiously reinforced for high ratings and low havior relative to baseline were obtained under
levels of disruption (Kaufman and O'Leary, teacher-determined reinforcement for students
1972). with academic and behavioral problems. These
In an attempt to replicate these findings with changes in behavior were maintained through-
a similar population, Santogrossi et al. (1973) out matching and fading, and finally were main-
found that teacher-determined reinforcement re- tained by the students themselves. Additionally,
474 MICHAEL S. ROSENBAUM and RONALD S. DRABMAN

these behavioral changes generalized to control man et al. (1973) was developed by Robertson
periods in which token reinforcement was not in et al. (in press) to transfer evaluation and rein-
effect, although student work received teacher forcement functions from the teacher to the
praise. students. Disruptive behavior decreased by 70%;
Drabman et al. (1973) suggested several fac- during teacher-determined reinforcement, rela-
tors that may have contributed to the main- tive to baseline levels. A further decrease of
tenance of appropriate behavior: (a) continuous 209% was obtained when the students evaluated
teacher praise for appropriate student behavior their own behavior and delivered their own re-
may have become more reinforcing over the inforcement. Token reinforcement was faded via
course of the study; (b) peer reinforcement for an "inflation" period, during which the point
appropriate behavior also increased; (c) accurate costs for back-up reinforcers were doubled. This
self-evaluation was praised by the teacher and was followed by a removal of back-up reinforce-
may have acquired conditioned reinforcing ment and a return to baseline. During this
properties, thereby strengthening appropriate return-to-baseline phase, the students maintained
behavior; and (d) improved academic skills in- disruptive behavior at a level 82% lower than
compatible with inappropriate behavior were during the initial baseline. Furthermore, in-
developed. stances of generalization to an afternoon period
In a systematic replication and extension of in the absence of the therapeutic procedure, to
this study, Turkewitz et al. (1975) worked with mornings when the training program was not in
similar children in an after-school token rein- effect, and to the presence of different teachers
forcement program. A 68 % decrease in dis- were noted.
ruptive behavior relative to baseline was Working with second graders, Glynn et al.
achieved with teacher-determined reinforcement. (1973) introduced self-control procedures by
This low level of disruption was maintained having students indicate whether or not their
while the evaluation and reinforcement proce- behavior was on-task when auditory signals
dures were transferred from the teacher to the were randomly presented. Each instance of on-
students, using fading procedures like those of task behavior could be self-reinforced. Prior
Drabman et al. (1973). In the final phase of the teacher-determined reinforcement for the entire
study, the students evaluated their own behavior class being on-task resulted in an average 46%
without exchanging their points for back-up increase in on-task behavior relative to baseline
reinforcement. A 56% decrease in disruptive levels. The self-control procedures were associ-
behavior, relative to an identical phase prior to ated with a further increase of 20% in on-task
teacher-determined reinforcement, was obtained. behavior, in addition to producing more stable
These results also generalized to no-token con- rates of responding. The change to self-control
trol periods, in which goals were assigned by represented a change from group to individual
the teacher. Because the assignment of goals contingencies, which may have contributed to
was also included during token reinforcement, the effectiveness of the self-control procedures.
goals may have acquired conditioned reinforcing Glynn and Thomas (1974) introduced self-
properties, thereby contributing to the main- control procedures for third graders without
tenance. Unfortunately, disruptive behavior for giving them prior exposure to externally de-
these students in their home special education termined reinforcement, and found a 41 % in-
classroom did not differ from that of a control crease in on-task behavior relative to baseline.
group, thereby suggesting limited generalizabil- A cueing device consisting of a chart specifying
ity of the effects of the self-control training the on-task behaviors required for each lesson
procedure. was introduced to eliminate confusion regarding
A variation of the fading procedures of Drab- what constituted on-task behavior. This proce-
SELF-CONTROL IN THE CLASSROOM 475

dure resulted in a 79 % increase in on-task be- reinforcement, at least with normal children in
havior and more stable patterns of responding regular classrooms. (f) Finally, further research
relative to a second baseline instituted after the is needed to compare the effects of an abrupt
initial self-control phase. These results were and a gradual transfer of control over a variety
replicated by Thomas (1976), who demon- of students. To be most effective, this type of
strated the effectiveness of similar self-control research should be conducted across classrooms
procedures for a longer period of time. using a between-groups experimental design,
and should measure long-term maintenance.
Conclusions
Based on the results of the studies in this sec- SELF-DETERMINED
tion, we can suggest the following conclusions: REINFORCEMENT versus
(a) Self-control skills can effectively maintain RESPONSE COST
desirable levels of behavior achieved under ex-
ternally determined contingencies. Relative to Although a wealth of research has focused on
baseline, target-behavior changes obtained in the teaching students to determine their own rein-
studies in this section were of sufficient mag- forcement contingencies, only two studies could
nitude to have applied significance. Mainte- be found that focused on teaching students to
nance, in this sense, refers to enduring changes determine their own punishment contingencies.
in target behaviors for target subjects in the In the first study (Kaufman and O'Leary, 1972),
therapeutic setting with the expressed continua- described in the previous section, students in
tion of some form of program contingencies. a response-cost token-reinforcement classroom
Enduring changes in target behaviors for target maintained disruptive behavior at a low level
subjects in the therapeutic setting after all pro- initially obtained during externally determined
gram contingencies, have been removed (i.e., a response cost, when instructed to determine the
return to baseline conditions) are more appro- number of tokens they should lose. Unfortu-
priately referred to as time generalization (Drab- nately, the effects of self-imposed response cost
man, Hammer, and Rosenbaum, in press). were assessed only for a short period of time.
Whereas all the studies reviewed in this section To investigate response cost further as a
addressed the issue of maintenance, only the method of self-management, Humphrey et al.
Robertson et al. (in press) study addressed the (Note 1) instructed second graders in a self-
issue of time generalization; this study found reward group to reinforce themselves contingent
that self-control skills can effectively achieve on accurate academic performance. Students in a
this class of generalization. (b) With disruptive self-imposed response-cost group were given the
children, an abrupt transfer of control from the maximum number of tokens that could be
teacher to the students may result in high levels earned on a particular day and were instructed
of undesirable behavior. (c) Introduction of a to fine themselves contingent on inaccurate
matching criterion, followed by fading of the academic performance and failure to complete
matching procedures, may produce an effective the daily assignment of reading papers. Follow-
transfer of control from an external agent to ing a return to baseline, the groups switched
the student. (d) Fading back-up reinforcement contingencies. The results showed that both self-
may result in desirable behavior being main- control procedures improved academic perfor-
tained by students themselves after the self- mance (mean increase of 58 %) relative to
control program has been removed (i.e., time baseline, with self-reward producing greater in-
generalization). (e) Self-controlled reinforce- crements than self-imposed response cost. The
ment can initiate and maintain desirable changes level of disruptive behavior decreased when
in the absence of prior externally determined either self-control procedure was in effect (mean
476 MICHAEL S. ROSENBAUM and RONALD S. DRABMAN

decrease of 34%), even though there were no instruction). During training in the first stage,
specified contingencies for disruption. Inspection the adult modeled performance-relevant skills,
of the data, however, revealed considerable over- such as problem definition, planning strategies,
lap of baseline and self-management results for guidance, self-reinforcement, self-evaluation,
both groups. Although this study may indicate and error correction. The child was trained to
that self-imposed response cost can effectively use self-instruction to control nonverbal behav-
control classroom behavior, we feel that further ior in a variety of tasks, ranging from simple
research yielding more conclusive results (i.e., sensorimotor tasks (e.g., copying line figures)
greater magnitude in the differences between to more complex problem solving tasks. Addi-
baseline and self-management phases) is needed tionally, the level of difficulty of the tasks in-
before self-imposed response cost can be advo- creased over the training sessions, with the adult
cated for classroom use. model demonstrating self-verbalizations for each
task, and the child following the fading proce-
dure. The results showed that the cognitive-
SELF-INSTRUCTIONAL TRAINING
training group had significantly improved per-
A potentially important method for develop- formance on psychometric tests designed to
ing self-control is through self-instructional measure behavioral and cognitive impulsivity
training, in which individuals are taught to make (e.g., Matching Familiar Figures, Porteus Maze
suggestions to themselves to guide their own be- Test) relative to two control groups. No signifi-
havior in a manner similar to being guided by cant differences, however, were found on two
another individual. The important role of "talk- measures of appropriate behavior in the class-
ing to oneself" in controlling behavior has been room, thereby bringing into question the applied
emphasized by several authors (e.g., Meichen- significance of this self-instructional training
baum, 1975, 1977; Meichenbaum and Cam- procedure.
eron, 1974; Skinner, 1953). In laboratory set- Working with kindergarten children in their
tings, self-instructions have proven effective with classroom, Robin, Armel, and O'Leary (1975)
discrimination training (Bem, 1967), finger- provided feedback and social praise, in addition
tapping (Meichenbaum and Goodman, 1969), to a training procedure similar to that of
resistance to temptation (Hartig and Kanfer, Meichenbaum and Goodman (1971), to a self-
1973), and instruction-following (Monahan and instruction group. The results showed that this
O'Leary, 1971; O'Leary, 1968). group performed significantly (69%) better on
Meichenbaum and Goodman (1971) worked a writing task (printing different letters of the
with second graders who had behavior problems alphabet) than a no-treatment control group.
such as hyperactivity and poor self-control, and The self-instruction group also significantly
attempted to teach them self-control skills. A (28%) outperformed a direct-training group
cognitive-training group received the following that received only feedback and praise on com-
fading procedure: (a) an adult model performed pletion of each letter copied. The authors listed
a task while talking aloud (cognitive modeling); several limitations associated with self-instruc-
(b) the child performed the same task while tional training: (a) higher rates of self-verbaliza-
being directed by the adult (overt external tions after self-instructional training were not
guidance); (c) the child performed the task correlated with superior writing performance;
while self-instructing aloud (overt self-guid- (b) it was difficult to teach self-instructional re-
ance); (d) the child performed the task while sponding; (c) some children self-instructed cor-
whispering the instructions (faded, overt self- rectly but made incorrect writing responses: (d)
guidance); and (e) the child performed the task the amount of time required to teach self-
with guidance from private speech (covert self- instruction decreased the amount of time avail-
SELF-CONTROL IN THE CLASSROOM 477

able for writing practice; and (e) generalization of self-instructional procedures, after which (b)
from letters that were trained to letters not a well-controlled component analysis separating
trained did not occur. Furthermore, the authors the effects of each step in the self-instructional
seriously questioned the feasibility of teaching package; (c) an identification of those behaviors,
self-instructional responding in applied settings settings, and populations for which self-instruc-
to obtain only a 28% increase in performance tional training may be most appropriate; and
relative to a more cost efficient direct-training (d) an identification of those variables that
procedure. would promote generalization of self-instruc-
Bornstein and Quevillon (1976), working tional training effects.
with disruptive preschool children, introduced a
self-instructional training procedure similar to
that of Meichenbaum and Goodman (1971) and GENERALIZATION
presented reinforcement contingent on complet- Self-control training is a strategy often men-
ing each step in the training sequence. The tioned for promoting generalization of behavior
results showed that, relative to baseline, an changes. In their discussion of the technology
average 65.3 % increase in on-task behavior was of generalization, Stokes and Baer (1977) de-
obtained in the classroom following self-instruc- scribe the possible mediational function of self-
tional training. Follow-up data collected 22.5 control skills in promoting generalization. The
wk following initiation of baseline revealed an ease in transporting these skills should render
average increase of 59 % in on-task behavior them readily accessible in a variety of situations
relative to that baseline. In a systematic repli- to facilitate appropriate responding. According
cation of the Bornstein and Quevillon (1976) to the conceptual analysis of generalization pro-
study, Friedling and O'Leary (1979) worked vided by Drabman et al. (in press), self-control
with hyperactive second and third graders and studies conducted in classroom settings have
failed to obtain any systematic changes in on- investigated only 4 of 15 different classes of
task behavior related to self-instructional train- generalization. Time generalization has already
ing. The subsequent application of token rein- been discussed and refers to changes in the
forcement resulted in increased on-task behavior. target behavior of target subjects in the thera-
In assessing each of four components of self- peutic setting that endure after the program
instructional training (instructions, overt model- contingencies have been withdrawn. Several
ing, self-instructional modeling, training in studies have shown that self-control skills can
task-relevant speech) relative to arithmetic per- produce time generalization (Bolstad and John-
formance assessed by pre- and posttests, Wein son, 1972; Brownell et al., 1977; Felixbrod and
and Nelson (Note 4) presented different com- O'Leary, 1974; Glynn, 1970; Robertson et al.,
binations of these components to different in press). Additionally, laboratory studies on
groups of second graders. The results showed no self-control have demonstrated time generaliza-
significant differences among these groups or tion with visual discrimination tasks (Johnson,
between them and a no-treatment control group 1970; Johnson and Martin, 1973; Kanfer and
on an arithmetic posttest. Duerfeldt, 1967) and with responding to a
game-like apparatus (Weiner and Dubanoski,
Conclusions 1975). The time generalization effects associated
The mixed results from the studies reviewed with self-control have been attributed to the
in this section lead us to conclude that the fol- conditioned reinforcing properties acquired by
lowing future research on self-instructional the self-recording or self-evaluation procedures,
training is needed: (a) further outcome research which may continue on a covert level (self-
to demonstrate the applied significance, if any, monitoring) after the formal program contin-
478 MICHAEL S. ROSENBAUM and RONALD S. DRABMAN

gencies have been removed. Furthermore, the ing setting to the classroom after training was
therapeutic setting itself may acquire discrimina- completed, thereby providing support for setting
tive stimulus properties and function as a cue time generalization. Unfortunately, these results
for the continuation of behavior changes ob- have been difficult to replicate (Friedling and
tained under the formal training program. O'Leary, 1979).
Setting generalization has occurred when Only two studies have investigated response
changes in target subjects' target behaviors ob- generalization, which refers to changes in non-
tained during an intervention transfer to a target behaviors of target subjects in the thera-
different setting, with program contingencies peutic setting during intervention. Both studies
still in effect only in the therapeutic setting. The (Ballard and Glynn, 1975; Robin et al., 1975)
absence of the salient discriminative stimuli failed to find evidence for this class of general-
present in the therapeutic setting is the criterion ization.
for determining whether the new setting is
different. Setting generalization associated with Conclusions
self-control training has been reported in several Based on the studies reviewed in this section,
classroom studies (Drabman et al., 1973; we can conclude that substantial evidence has
Robertson et a/., in press; Turkewitz et al., been provided only for time generalization. Ac-
1975). In each of these studies, the salient dis- cording to Drabman et al. (in press), there are 14
criminative stimuli absent in the new setting additional classes of generalization. Therefore, it
were the formal program contingencies. In the seems clear that further research is needed to
Drabman et a/. (1973) study, it was not clear identify those factors that can effectively pro-
whether maintenance or setting generalization mote generalization in a variety of classes, in-
was obtained, since the teacher praised the stu- cluding setting, setting time, and response gen-
dents for doing their work in token and control eralization. Additionally, a longer period of time
periods. Teacher praise, therefore, was paired for measurement of generalization in self-
with token reinforcement and may have control studies is needed.
acquired conditioned reinforcing properties,
thereby maintaining changes in target behaviors
in the control periods. Some form of the pro- THE ROLE OF
gram contingencies was present during the con- EXTERNAL CONTROL
trol periods, thereby resulting in the mainte- Every study dealing with self-control in the
nance of changes in target behaviors, not setting classroom has included external variables prior
generalization. This same issue is relevant to the to or concurrent with the introduction of self-
Turkewitz et al. (1975) study in which goals control training. The presence of these external
assigned by the teacher were included in both variables has led several authors (e.g., Kazdin,
token and control periods. Broden et al. ( 1971) 1975a, Morgan and Bass, 1973) to raise ques-
failed to find setting generalization associated tions regarding the ability of self-control strate-
with self-monitoring. gies, especially self-reinforcement, to modify
Only one study reviewed in this paper in- and/or maintain behavior in the absence of
vestigated setting time generalization, which external influence.
includes changes in target subjects' target be- Theoretically, according to Jones, Nelson, and
haviors that transfer to a different setting, when Kazdin (1977): "Self-reinforcement occurs
program contingencies are no longer in effect when, in the relative absence of external con-
in the therapeutic setting. Bornstein and Que- trolling influences, an individual has full control
villon (1976) found that the effects of self- over available reinforcers but freely imposes
instructional training transferred from the train- certain contingencies for the self-administration
SELF-CONTROL IN THE CLASSROOM 479
of the reinforcing stimuli" (p. 151). Morgan and example, no one has yet been able to maintain
Bass (1973) add to this theoretical description: appropriate behavior while allowing the pupils
"It is therefore crucial that the subject under- to partake of reinforcement when, where, and
stand and believe that he truly does have com- in amounts they desired. Although further re-
plete control over all contingencies and could, search is needed which focuses on eliminating
if he wished, 'cheat' or reward himself at any the need for as many sources of external control
time without bringing about any other contin- as possible, the issue of external influences on
gent external consequences of either a rein- self-control in applied classroom settings can
forcing or punishing nature" (p. 120). be resolved if self-control in these settings is
Based on the research reviewed in previous "considered a matter of degree where external
sections, a disparity exists between theoretical control is minimal or intermittent" (Kazdin,
self-reinforcement (and self-control) and that 1975a, p. 207). The purpose of self-control
found in applied classroom settings. In fact, the training in the classroom should be to enable
sources of external control present in classroom students to manage as much of their own educa-
studies investigating self-control strategies ren- tion as possible and to enable the teacher to
der theoretical self-reinforcement in the class- devote time to teaching and supervising stu-
room undemonstrated to date. External control dents' work without having to control disruptive
has been manifested by including the introduc- behavior or to provide incentives constantly for
tion of self-determined contingencies following academic performance.
some period of externally determined contingen-
cies (Bolstad and Johnson, 1972; Drabman et
al., 1973; Frederiksen and Frederiksen, 1975; EXPERIMENTAL DESIGN
Glynn et al., 1973; Kaufman and O'Leary, The most frequently used designs in self-
1972; Lovitt and Curtiss, 1969; Robertson et al., control studies in the classroom have been the
in press, Santogrossi et al., 1973; Turkewitz et within-subjects reversal and the between-sub-
al., 1975; Farnum et al., Note 2); requiring jects comparison designs. One problem associ-
students to match self-ratings of behavior with ated with the reversal design is that reversing
those of the teacher or a trained observer (Bol- desirable behavior changes sometimes raises
stad and Johnson, 1972; Drabman et al., 1973; practical and ethical questions (Stolz, 1976). Al-
Robertson et al., in press; Santogrossi et al.. though reversal designs provide important and
1973; Turkewitz et al., 1975); and praising ap- scientifically valid information, they often may
propriate behavior combined with ignoring in- not be appropriate for self-control training.
appropriate behaviors (Broden et al., 1971; In the Robertson et al. (in press) and Turke-
Drabman et al., 1973; Santogrossi et al., 1973; witz et a!. (1]975) research, the lack of a reversal
Turkewitz et al.. 1975). Additional evidence for indicated the durability of the behavior changes
external control has been provided in the limi- obtained, although a causal interpretation of the
tation of the range of points that students could results was difficult. These reports, in addition
award themselves, determination of rules speci- to that of Drabman et a!. (1973), used an ex-
fying how to apply contingencies, administration tended comparison design in which several dif-
of reinforcement, checks on students' behaviors, ferent interventions were successively introduced
specification of performance standards, model- to transfer control gradually from the teacher
ing desirable behaviors, and cueing desirable to the students. Even though this design may
behaviors. make the interpretation of the results difficult,
Based on these findings, we can conclude that we want to consider the reasons for its use.
only a weak form of self-control has been dem- Ayllon and Rosenbaum (1977) mentioned that
onstrated so far in these classroom studies. For an important component contributing to the
480 MICHAEL S. ROSENBAUM and RONALD S. DRABMAN

success of classroom behavior modification is demonstrated that students can be taught to


the cooperation of the teacher. Teachers are observe and record their own behavior, deter-
typically interested in maintaining academic per- mine and administer their own contingencies,
formance and appropriate social behavior in and provide instructions to guide their behavior.
their students. With extremely disruptive stu- The results reviewed suggest that several steps
dents, as in the Drabman et al. (1973), Robert- should be followed to maximize the potential
son et al. (in press), and Turkewitz et al. (1975) effectiveness of self-control training in the class-
studies, the teacher's goal was to decrease the room: (a) students should be taught self-obser-
level of disruption and increase academic per- vational procedures; (b) future research may
formance. It would be difficult to convince a demonstrate that accurate self-observation will
teacher of the importance of conducting a re- increase the efficiency of self-control training. In
versal to demonstrate experimental control, these situations, accurate self-observation can be
since this would include the possibility of stu- prompted by initially requiring students to
dents again becoming highly disruptive. Because match their records with those of the teacher
of its methodological limitations, the extended or a trained observer, followed by gradually fad-
comparison design should be used in combina- ing the matching process; (c) once self-observa-
tion with a control group, if possible. Criticisms tion has been established, externally adminis-
similar to those stated above can also be made tered contingencies for desirable behavior
against using a control group, since nothing change can be introduced; (d) the control of
would be done to improve the academic or these contingencies can then be transferred to
social behavior of students in that group. One the students; (e) at the same time that self-
possible solution to this problem would be to determined contingencies are introduced, stu-
use either a multiple baseline across subjects dents can be taught to provide themselves with
or groups, in which the control subjects would instructions and praise designed to guide their
receive the benefit of an effective intervention behavior; and (f) when the students are reliably
at a later temporal point than the treatment controlling their academic and social behavior,
subjects. explicit contingencies can be gradually with-
Another methodological issue concerns iso- drawn. This multi-component approach would
lating the effects of self-observation from those maximize the probability of maintenance of
of other self-control components in packaged self-control skills after the training program
treatment approaches. Jeffrey (1974) has sug- has been removed. It would also contribute to
gested the inclusion of a self-monitoring or the potential generalization of these skills across
self-evaluation phase following baseline and different situations. Stokes and Baer (1977)
preceding additional self-control components have emphasized the necessity of specifically
within reversal or multiple baseline designs. programming the occurrence of generalization,
This would enable an assessment of self-observa- which implies that teachers must be prepared to
tional effects and their potential contribution train self-control skills in several different set-
to the entire treatment package. tings.
The effectiveness of applied behavioral
change strategies is evaluated relative to their
CONCLUSIONS AND production of socially meaningful changes in
FUTURE DIRECTIONS behavior. In evaluating applied classroom re-
Self-control is a therapeutic goal that is ad- search, target behavior changes are meaningful
vocated by almost all helping professionals, re- relative to the magnitude of change required for
gardless of their discipline or theoretical orien- educational significance. Outcome studies re-
tation. The research reviewed in this article has stricted to increases in appropriate behaviors or
SELF-CONTROL IN THE CLASSROOM 481

decreases in inappropriate behaviors reveal little havior, independent subjective evaluation of


regarding satisfaction of classroom and school this behavior change by the teacher, parents, and
curricula requirements. Therefore, the broader peers is needed. When self-control training is
context of the educational significance of be- conducted to improve the behavior of only a
havior changes in externally and self-controlled few students, then data are needed on the be-
behavior modification programs needs to be ad- havior of well-behaved peers.
dressed. 7. Standardized data collection procedures
In the studies reviewed in this paper, only are needed to render different studies compa-
anecdotal evidence concerning the educational rable.
implications of target behavior changes has been 8. Parametric comparative research focused
reported (e.g., self-control programs enable on the advantages and disadvantages resulting
teachers to devote more time to individual stu- from self-control versus external control tech-
dents). This points to several areas of classroom niques along such dimensions as program ef-
research warranting further investigation in self- fectiveness, social validity, acceptance by teach-
control studies: ers and students, and ease of implementation, is
warranted.
1. Because studies investigating the socio-
metric status of target children have indicated Laboratory research from the developmental
that desirable changes can be obtained from suc- literature has suggested several variables that
cessful teacher-controlled behavior modification may influence self-control. For example, Masters
programs (Drabman and Lahey, 1974; Drab- and Santrock (1976) suggest that cognitive and
man, Spitalnik, and Spitalnik, 1974), those so- affective events associated with a task can affect
cial changes that may be a function of self- subsequent performance of that task. These
control procedures should be studied. authors found that preschoolers instructed to
2. The issue of under what circumstances the think of pleasant events or to emit positive ver-
self-selection of reinforcers is differentially ef- balizations worked significantly longer at a
fective when compared with external selection motor task than children instructed to think of
of reinforcers must be studied. unpleasant events or to emit negative verbaliza-
3. Methods of identifying children for whom tions, or children in a control group. Switzky
self-control is not an appropriate goal need to and Haywood (1974) found that children's
be developed. motivational orientations (i.e., externally or in-
4. How children in unsuccessful self-control trinsically motivated) can affect their on-task
programs are affected should be studied. performance. Intrinsically motivated children
5. The effectiveness of self-control proce- performed significantly more wheel-turning re-
dures needs to be assessed by comparing results sponses than externally motivated children
from the initial baseline with those from a final under self-reinforcement conditions. Opposite
return to baseline. If target-behavior changes results were obtained under external reinforce-
are under the control of the students themselves, ment conditions. Although the results of these
then those changes should be maintained after studies have little, if any, applied significance,
all program contingencies have been removed. the variables studied may have an effect on self-
To date, only one study has included this type of control training in applied settings and do, there-
comparison (i.e., Robertson et al., in press). fore, warrant further investigation.
6. The social validity of behavior changes One important area in which self-control may
obtained with self-control procedures needs to prove useful is the transfer of positive behavior
be demonstrated. In addition to demonstrating changes from one academic year to the next.
a statistically significant improvement in be- Typically, the beginning of each academic year
482 MICHAEL S. ROSENBAUM and RONALD S. DRABMAN

consists of teachers reviewing material sup- Bandura, A. Self-reinforcement processes. In M. J.


posedly learned during the previous year. Chil- Mahoney and C. E. Thoresen (Eds), Self-control:
Power to the person. Monterey, California:
dren often have difficulty recalling this material Brooks/Cole, 1974.
because they have not used it during the summer Bandura, A., Grusec, J. E., and Menlove, F. L. Some
months. With the development of self-control social determinants of self-monitoring reinforce-
ment systems. Journal of Personality and Social
skills, children could be encouraged to review Psychology, 1967, 5, 449-455.
material during the summer vacation. Bandura, A. and Kupers, C. J. Transmission of pat-
Self-control is a relatively new area of class- terns of self-reinforcement through modeling.
room research, yet there is already a substantial Journal of Abnormal and Social Psychology,
1964, 69, 1-9.
literature. The studies reported here represent Bem, S. L. Verbal self-control: The establishment of
only the most rudimentary attempts to achieve effective self-instruction. Journal of Experimental
this important goal. We hope that this review Psychology, 1967, 74, 485-491.
leads to a redoubling of research attempts. If Bolstad, 0. D. and Johnson, S. M. Self-regulation in
the modification of disruptive classroom behavior.
this happens, a technology for teaching most Journal of Applied Behavior Analysis, 1972, 5,
pupils to be responsible for maintaining high 443-454.
levels of appropriate behaviors seems achievable. Bornstein, P. H. and Quevillon, R. P. The effects of
a self-instructional package on overactive pre-
school boys. Journal of Applied Behavior Analy-
REFERENCE NOTES sis, 1976, 9, 179-188.
1. Humphrey, L. L., Karoly, P., and Kirschenbaum, Brigham, T. A. and Bushell, D. Notes on autono-
mous environments: Effects of student-selected
D. S. Self-imposed response cost versus self- and teacher-selected rewards on academic per-
reward in classroom behavior management. Paper formance. Educational Technology, 1973, 13,
presented at the meeting of the Midwestern Psy- 19-22.
chological Association, Chicago, May 1977. Brigham, T. A. and Sherman, J. A. Effects of choice
2. Farnum, M., Brigham, T. A., and Johnson, G.
A comparison of the effects of teacher determined and immediacy of reinforcement on single re-
and student determined contingencies on arithme- sponse and switching behavior of children. Jour-
tic performance. Unpublished manuscript, 1977. nal of the Experimental Analysis of Behavior,
(Available from T. A. Brigham, Department of 1973, 19, 425-435.
Psychology, Washington State University, Pull- Brigham, T. A. and Stoerzinger, A. An experi-
man, Washington 99164.) mental analysis of children's preference for self-
3. Brigham, T. A. The role of choice in a behav- selected rewards. In T. A. Brigham, R. Hawkins,
ioral analysis of freedom. Unpublished manuscript, J. Scott, and T. F. McLaughin (Eds), Behavioral
1977. (Available from T. A. Brigham, Department analysis in education: Self-control and reading.
of Psychology, Washington State University, Pull- Dubuque, Iowa: Kendall/Hunt, 1976.
man, Washington 99164.) Broden, M., Hall, R. V., and Mitts, B. The effect of
4. Wein, K. S. and Nelson, R. 0. The effect of self- self-recording on the classroom behavior of two
instructional training on arithmetic problem solv- eighth-grade students. Journal of Applied Behav-
ing skills. Paper presented at the meeting of the ior Analysis, 1971, 4, 191-199.
Association for Advancement of Behavior Ther- Brownell, K. D., Colletti, G., Ersner-Hershfield, R.,
apy, Atlanta, December 1977. Hershfield, S. M., and Wilson, G. T. Self-
control in school children: Stringency and leni-
REFERENCES ency in self-determined and externally imposed
performance standards. Behavior Therapy, 1977,
Ayllon, T. and Rosenbaum, M. S. The behavioral 8, 442-455.
treatment of disruption and hyperactivity in Copeland, R. and Hall, R. V. Behavior modification
school settings. In B. B. Lahey and A. E. Kazdin in the classroom. In M. Hersen, R. M. Eisler, and
(Eds), Advances in clinical child psychology (Vol. P. M. Miller (Eds), Progress in behavior modifi-
1). New York: Plenum, 1977. cation (Vol. 3). New York: Academic Press,
Ballard, K. D. and Glynn, T. Behavioral self- 1976.
management in story writing with elementary Drabman, R. S. Behavior modification in the class-
school children. Journal of Applied Behavior room. In W. E. Craighead, A. E. Kazdin, and M.
Analysis, 1975, 8, 387-398. J. Mahoney (Eds), Behavior modification: Princi-
Bandura, A. Principles of behavior modification. ples, issues, and applications. Boston: Houghton
New York: Holt, Rinehart & Winston, 1969. Mifflin, 1976.
SELF-CONTROL IN THE CLASSROOM 483

Drabman, R. S., Hammer, D., and Rosenbaum, M. S. Thoresen (Eds), Self-control: Power to the person.
Assessing generalization in behavior modification Monterey, California: Brooks/Cole, 1974.
with children: The generalization map. Behav- Johnson, S. M. Self-reinforcement versus external
ioral Assessment, in press. reinforcement in behavior modification with
Drabman, R. and Lahey, B. B. Feedback in class- children. Developmental Psychology, 1970, 6,
room behavior modification: Effects on the target 147-148.
child and her classmates. Journal of Applied Be- Johnson, S. M. and Martin, S. Developing self-
havior Analysis, 1974, 7, 591-598. evaluation as a conditioned reinforcer. In B. A.
Drabman, R. S., Spitalnik, R., and O'Leary, K. D. Ashem & E. G. Poser (Eds), Adaptive learning:
Teaching self-control to disruptive children. Behavior modification with children. New York:
Journal of Abnormal Psychology, 1973, 82, Pergamon Press, 1973.
10-16. Johnson, S. M. and White, G. Self-observation as
Drabman, R., Spitalnik, R., and Spitalnik, K. Socio- an agent of behavioral change. Behavior Ther-
metric and disruptive behavior as a function of apy, 1971, 2, 488-497.
four types of token economies. Journal of Applied Jones, R. T., Nelson, R. E., and Kazdin, A. E. The
Behavior Analysis, 1974, 7, 93-101. role of external variables in self-reinforcement.
Felixbrod, J. J. and O'Leary, K. D. Effects of rein- Behavior Modification, 1977, 1, 147-178.
forcement on children's academic behavior as a Kanfer, F. H. Self-regulation: Research, issues, and
function of self-determined and externally im- speculations. In C. Neuringer and J. L. Michael
posed contingencies. Journal of Applied Behavior (Eds), Behavior modification in clinical psychol-
Analysis, 1973, 6, 241-250. ogy. New York: Appleton-Century-Crofts, 1970.
Felixbrod, J. J. and O'Leary, K. D. Self-determina- (a)
tion of academic standards by children: Toward Kanfer, F. H. Self-monitoring: Methodological
freedom from external control. Journal of Edu- limitations and clinical applications. Journal of
cational Psychology, 1974, 66, 845-850. Consulting and Clinical Psychology, 1970, 35,
Frederiksen, L. W. and Frederiksen, C. B. Teacher-
148-152. (b)
Kanfer, F. H. The maintenance of behavior by
determined and self-determined token reinforce- self-generated stimuli and reinforcement. In A.
ment in a special education classroom. Behavior Jacobs and L. B. Sachs (Eds), The psychology of
Therapy, 1975, 6, 310-314. private events. New York: Academic Press, 1971.
Friedling, C. and O'Leary, S. G. Effects of self- Kanfer, F. H. and Duerfeldt, P. H. Motivational
instructional training on second- and third-grade properties of self-reinforcement. Perceptual and
hyperactive children: A failure to replicate. Jour- Motor Skills, 1967, 25, 237-246.
nal of Applied Behavior Analysis, 1979, 12, 211- Kanfer, F. H. and Phillips, J. S. Learning founda-
219. tions of behavior therapy. New York: Wiley,
Glynn, E. L. Classroom applications of self-deter- 1970.
mined reinforcement. Journal of Applied Behav- Karoly, P. Behavioral self-management in children:
ior Analysis, 1970, 3, 123-132. Concepts, methods, issues, and directions. In M.
Glynn, E. L. and Thomas, J. D. Effect of cueing on Hersen, R. M. Eisler, and P. M. Miller (Eds),
self-control of classroom behavior. Journal of Progress in behavior modification (Vol. 5). New
Applied Behavior Analysis, 1974, 7, 299-306. York: Academic Press, 1977.
Glynn, E. L., Thomas, J. D., and Shee, S. M. Be- Kaufman, K. F. and O'Leary, K. D. Reward, cost,
havioral self-control of on-task behavior in an and self-evaluation procedures for disruptive ado-
elementary classroom. Journal of Applied Behav- lescents in a psychiatric hospital school. Journal
ior Analysis, 1973, 6, 105-1 13. of Applied Behavior Analysis, 1972, 5, 293-309.
Gottman, J. M. and McFall, R. M. Self-monitoring Kazdin, A. E. Reactive self-monitoring: The effects
effects in a program for potential high school of response desirability, goal setting, and feed-
dropouts: A time-series analysis. Journal of Con- back. Journal of Consulting and Clinical Psy-
sulting and Clinical Psychology, 1972, 39, 273- chology, 1974, 42, 704-716. (a)
281. Kazdin, A. E. Self-monitoring and behavior change.
Hartig, M. and Kanfer, F. H. The role of verbal In M. J. Mahoney and C. E. Thoresen (Eds), Self-
self-instructions in children's resistance to tempta- control: Power to the person. Monterey, Cali-
tion. Journal of Personality and Social Psychology fornia: Brooks/Cole, 1974. (b)
1973, 25, 259-267. Kazdin, A. E. Behavior modification in applied
Hildebrandt, D. E., Feldman, S. E., and Ditrichs, R. settings. Homewood, Illinois: Dorsey Press, 1975.
A. Rules, models, and self-reinforcement in (a)
children. Journal of Personality and Social Psy- Kazdin, A. E. Recent advances in token economy
chology, 1973, 25, 1-5. research. In M. Hersen, R. M. Eisler, and P. M.
Jeffrey, D. B. Self-control: Methodological issues Miller (Eds), Progress in behavior modification
and research trends. In M. J. Mahoney and C. E. (Vol. 1). New York: Academic Press, 1975. (b)
484 MICHAEL S. ROSENBAUM and RONALD S. DRABMAN
Kazdin, A. E. The token economy. New York: Meichenbaum, D. Self-instructional methods. In F.
Plenum Press, 1977. H. Kanfer and A. P. Goldstein (Eds), Helping
Kazdin, A. E. and Bootzin, R. R. The token econ- people change. New York: Pergamon Press, 1975.
omy: An evaluative review. Journal of Applied Meichenbaum, D. Cognitive-behavior modification.
Behavior Analysis, 1972, 5, 343-372. New York: Plenum Press, 1977.
Kazdin, A. E. and Bootzin, R. R. The token econ- Meichenbaum, D. and Cameron, R. The clinical
omy: An examination of issues. In R. D. Rubin, potential of modifying what clients say to them-
J. P. Brady, and J. D. Henderson (Eds), Advances selves. In M. J. Mahoney and C. E. Thoresen
in behavior therapy (Vol. 4). New York: Aca- (Eds), Self-control: Power to the person. Mon-
demic Press, 1973. terey, California: Brooks/Cole, 1974.
Knapczyk, D. R. and Livingston, G. Self-recording Meichenbaum, D. H. and Goodman, J. The de-
and student teacher supervision: Variables within velopmental control of operant motor responding
a token economy structure. Journal of Applied by verbal operants. Journal of Experimental
Behavior Analysis, 1973, 6, 481-486. Child Psychology, 1969, 7, 553-565.
Liebert, R. M. and Allen, M. K. Effects of rule Meichenbaum, D. H. and Goodman, J. Training im-
structure and reward magnitude on the acquisi- pulsive children to talk to themselves: A means of
tion and adoption of self-reward criteria. Psy- developing self-control. Journal of Abnormal
chological Reports, 1967, 21, 445-452. Psychology, 1971, 77, 115-126.
Liebert, R. M., Hanratty, M., and Hill, J. H. Effects Mischel, W. and Liebert, R. M. Effects of discrep-
of rule structure and training method on the ancies between observed and imposed reward
adoption of a self-imposed standard. Child De- criteria on their acquisition and transmission.
velopment, 1969, 40, 93-101. Journal of Personality and Social Psychology,
Liebert, R. M. and Ora, J. P. Children's adoption 1966, 3, 45-53.
of self-reward patterns: Incentive level and Mischel, W. and Liebert, R. M. The role of power
method of transmission. Child Development, in the adoption of self-reward patterns. Child
1968, 39, 537-544. Development, 1967, 38, 673-683.
Lipinski, D. and Nelson, R. The reactivity and un- Monahan, J. and O'Leary, K. D. Effects of self-
reliability of self-recording. Journal of Con- instruction on rule-breaking behavior. Psycho-
Sulting and Clinical Psychology, 1974, 42, 118- logical Reports, 1971, 29, 1059-1066.
123. Morgan, W. G. and Bass, B. A. Self-control through
Locke, E. A., Cartledge, N., and Koeppel, J. Moti- self-mediated rewards. In R. D. Rubin, J. P.
vational effects of knowledge of results: A goal Brady, and J. D. Henderson (Eds), Advances in
setting phenomenon. Psychological Bulletin. behavior therapy (Vol. 4). New York: Academic
1968, 70, 474-485. Press, 1973.
Lovitt, T. C. Self-management projects with chil- Nelson, R. 0. Methodological issues in assessment
dren with behavioral disorders. Journal of Learn- via self-monitoring. In J. D. Cone and R. P.
ing Disabilities, 1973, 6, 138-150. Hawkins (Eds), Behavioral assessment: New di-
Lovitt, T. C. and Curtiss, K. A. Academic response rections in clinical psychology. New York:
rate as a function of teacher- and self-imposed Bruner/Mazel, 1977.
contingencies. Journal of Applied Behavior Anal- Nelson, R. O., Lipinski, D. P., and Black, J. L. The
ysis, 1969, 2, 49-53. relative reactivity of external observations and
Masters, J. C. and Mokros, J. R. Self-reinforcement self-monitoring. Behavior Therapy, 1976, 7, 314-
processes in children. In H. W. Reese (Ed), Ad- 321. (a)
vances in child development and behavior. New Nelson, R. O., Lipinski, D, P., and Black, J. L. The
York: Academic Press, 1974. reactivity of adult retardates' self-monitoring: A
Masters, J. C. and Santrock, J. W. Studies- in the comparison among behaviors of different valences,
self-regulation of behavior: Effects of contingent and a comparison with token reinforcement. Psy-
cognitive and affective events. Developmental chological Record, 1976, 26, 189-201. (b)
Psychology, 1976, 12, 334-348. O'Leary, K. D. The effects of self-instruction on
McFall, R. M. Parameters of self-monitoring. In immoral behavior. Journal of Experimental Child
R. B. Stuart (Ed), Behavioral self-management. Psychology, 1968, 6, 297-301.
New York: Bruner/Mazel, 1977. O'Leary, K. D. and Drabman, R. S. Token rein-
McLaughlin, T. F. Self-control in the classroom. forcement in the classroom: A review. Psycho-
Review of Educational Research, 1976, 46, 631- logical Bulletin, 1971, 75, 379-398.
663. O'Leary, K. D. and O'Leary, S. G. Classroom man-
McMains, M. J. and Liebert, R. M. Influence of agement: The successful use of behavior modifi-
discrepancies between successively modelled self- cation. New York: Pergamon Press, 1972.
reward criteria on the adoption of a self-imposed O'Leary, S. G. and O'Leary, K. D. Behavior modi-
standard. Journal of Personality and Social Psy- fication in the school. In H. Leitenberg (Ed),
chology, 1968, 8, 166-171. Handbook of behavior modification and behavior
SELF-CONTROL IN THE CLASSROOM 485
therapy. Englewood Cliffs, New Jersey: Prentice- haviour Research and Therapy, 1976, 14, 479-
Hall, 1976. 481.
Robertson, S. J., Simon, S. J., Pachman, J. S., and Switzky, H. N. and Haywood, H. C. Motivational
Drabman, R. S. Self-control and generalization orientation and the relative efficacy of self-moni-
procedures in a classroom of disruptive retarded tored and externally imposed reinforcement sys-
children. Child Behavior Therapy, in press. tems in children. Journal of Personality and
Robin, A. L., Armel, S., and O'Leary, K. D. The Social Psychology, 1974, 30, 360-366.
effects of self-instructions on writing deficiencies. Thomas, J. D. Accuracy of self-assessment of on-
Behavior Therapy, 1975, 6, 178-187. task behavior by elementary school children. Jour-
Santogrossi, D. A., O'Leary, K. D., Romanczyk, R. G., nal of Applied Behavior Analysis, 1976, 9, 209-
and Kaufman, K. F. Self-evaluation by ado- 210.
lescents in a psychiatric hospital school token Turkewitz, H., O'Leary, K. D., and Ironsmith, M.
program. Journal of Applied Behavior Analysis, Generalization and maintenance of appropriate
1973, 6, 277-287. behavior through self-control. Journal of Con-
Skinner, B. F. Science and human behavior. New sulting and Clinical Psychology, 1975, 43, 577-
York: Macmillan, 1953. 583.
Spates, C. R. and Kanfer, F. H. Self-monitoring, Weiner, H. R. and Dubanowski, R. A. Resistance
self-evaluation and self-reinforcement in chil- to extinction as a function of self- or externally-
dren's learning: A test of a multistage self-regu- determined schedules of reinforcement. Journal
lation model. Behavior Therapy, 1977, 8, 9-16. of Personality and Social Psychology, 1975, 31,
Stokes, T. F. and Baer, D. M. An implicit tech- 905-910.
nology of generalization. Journal of Applied Be-
havior Analysis, 1977, 10, 349-367.
Stolz, S. B. Evaluation of therapeutic efficacy of be- Received 27 March 1978.
havior modification in a community setting. Be- (Final Acceptance 29 March 1979.)

You might also like