EFFECTS OF ALIEN-REINFORCEMENT AND ITS COMBINED TYPE ON LEARNING BEHAVIOR AND EFFICACY IN RELATION TO PERSONALITY*

Effects of new types of reinforcement, alien type and alien combined with external type (double reinforcement), were examined in comparison with a conventional external one. One hundred and twenty children, 6th grade boys and girls, worked on learning discrimination tasks. Under alien reinforcement, upon correct responses of the child the experimenter received reinforcers from him-(or her-) self thus reinforcing the child responses. Under each condition, alien, external or double, children learned the tasks and their learning efficacy increased. Personality characteristics of children, in terms of extraversion and emotionality, differentiated effects of three reinforcement conditions on learning behavior and efficacy. The effect of alien reinforcement was influenced by the personality factors most, and that of double reinforcement least. In addition, personality factors influenced differently between alien and external conditions. Learning occurred differently under three different conditions of reinforcement, depending on the personality type of learners. Underlying mechanisms of alien reinforcement were different from, and its functions were independent of, those of external reinforcement. None of the effects of alien reinforcement on learning and motivation were contaminated by the intellectual faculties of learners.     

Self-reinforcement: Impact of instructions,criterion setting,and tangible rewards

This investigation was designed to ascertain the effects of instructions, criterion setting, and the presence of tangible rewards on the self-reinforcement process. Fifty-two third- and fourth-grade subjects were assigned to one of four treatment groups: (a) stringent instructions/criterion setting/tangible reward, (b) stringent instructions/criterion setting/no tangible reward, (c) nonstringent instructions/criterion setting/tangible reward, and (d) nonstringent instructions/no tangible reward. In the stringent-instruction conditions, subjects received social reinforcement for selecting stringent performance criteria, whereas in the non-stringent-instruction conditions, social reinforcement was withheld. Subjects in the tangible-reward groups were allowed to select a prize following the successful completion of their self-selected work performance. Subjects in the no-tangible-reward groups received no prizes for their work. All subjects performed an arithmetic task in which the number of correct problems completed, number of problems attempted, and time at task served as dependent variables across five reinforcement and two extinction trials. The results suggest that the condition of stringent instructions, criterion setting, and tangible reward was more effective in producing behavior change than the other three conditions. Perceived task difficulty and previous achievement on arithmetic task performance were shown to affect criteria selected and mathematical performance. The results are discussed in light of the contributory role of instructions, criterion setting, and tangible rewards on the self-reinforcement procedure.This study is based on a doctoral dissertation submitted to the Faculty of Arts and Sciences at the University of Pittsburgh by Helen L. Evans. Dr. Russell T. Jones was the dissertation committee chairperson. It was partially funded by an American Psychological Association Minority Fellowship awarded to the first author. Special thanks are extended to the committee members, Drs. Lloyd Bond, Robert Glaser, Johnny Matson, and Samuel M. Turner, for their assistance. The authors would like to thank Thomas DeVoge, Paul Karoly, and Samuel M. Turner for reading and commenting on an early draft of this study. Portions of this paper were presented at the 1982 APA convention.     

Partial reinforcement and breadth of learning

Two experiments on partial reinforcement were undertaken to test predictions made by a two process model of discrimination learning. In the first experiment rats were trained on a discrimination involving two relevant cues: one group (C) was trained on a 100: o schedule, the other (P) on a 50:0 schedule. Both groups were then given transfer tests with the two cues presented individually; finally all animals were extinguished on the original training stimuli and on the single cue stimuli. During extinction there was a negative correlation between the number of correct responses made by individual subjects of Group C to each single cue; whereas the correlation was positive for subjects of Group P. The second experiment employed basically the same design, but subjects were trained with seven relevant cues. The results of transfer tests showed that subjects of Group P learned to attach the correct response to many more cues than subjects of Group C. This suggests that the breadth of learning is greater under partial than under consistent reinforcement. The results were predicted by the model of discrimination learning under test.     

Errors of Omission and Commission during Alternative Reinforcement of Compliance: The Effects of Varying Levels of Treatment Integrity

We conducted two experiments to evaluate the effects of errors of omission and commission during alternative reinforcement of compliance in young children. In Experiment 1, we evaluated errors of omission by examining two levels of integrity during alternative reinforcement (20 and 60 %) for child compliance following no treatment (baseline) versus treatment at full (i.e., 100 %) integrity. Results indicated that compliance varied according to the level of integrity in place. In addition, compliance in the 60 % integrity condition was high and stable when it followed baseline, but was substantially lower for one participant and slightly lower for a second participant when it followed the full integrity condition. In Experiment 2, we evaluated errors of commission. For three participants, we compared treatment at full integrity to a condition in which errors of commission were made on every trial (i.e., 0 % integrity). For one of these three participants, we also compared treatment at full integrity to baseline and to a condition in which errors of commission were made on 50 % of trials. Results of all four evaluations again indicate that compliance varied according to the level of integrity in place: compliance was low in both the 0 and 50 % integrity conditions, regardless of the preceding condition. These results suggest that during alternative reinforcement of compliance, the effect of occasional errors of omission may depend on the immediately preceding context but that errors of commission are more detrimental.     

Effects of antipsychotic medication on discrimination learning for institutionalized adults who have mental retardation

Reduction in medication levels of drugs used to suppress inappropriate behavior (chlorpromazine, thioridazine and haloperidol) resulted in increases in performance on a discrimination learning task for seven residents of a state institution who are mentally retarded. After training to criterion on a matching-to-sample task, these participants experienced reductions in medication in an N of 1 AB replication paradigm. Two controls were medication free, and one remained on a stable dosage throughout. Improvement of performance on the first dependent variable (number of trials needed to meet criterion) ranged from 13.8% to 53.3% for the seven participants, while the three controls improved less than 1%. Improvement of performance on the second dependent variable (percentage of correct responses) ranged from 2.7% to 19.7% for experimental subjects; five of the seven subjects exhibited a minimum improvement of 8%. The change in percentage of correct responses for the three controls ranged from ?4.9 to 1.3. Treatment implications are discussed.     

The repeated acquisition of behavioral chains

Monkeys were trained with food reinforcement in a chamber containing four groups of three levers. For each session the monkey's task was to learn a new four-response chain by pressing the correct lever in each group. A stable pattern of learning resulted, and the number of errors reached a steady state from session to session. The technique was then used to determine how various durations of timeouts, following errors, affected the acquisition of new chains. With no timeout, the monkeys made a great many errors, due in large part to superstitious responses within the reinforced chain. Timeout durations ranging from 1 sec to 4 min reduced the number of errors substantially. A second experiment investigated the effects upon acquisition errors of presenting a single light (an “instruction” stimulus) over the correct lever. When this light did not influence the monkeys' responses to the three alternatives, the chains were learned as without it. When the light did control responding, the monkey pressed the appropriate sequence of levers but did not learn the sequence. Thus, when the light was removed, the monkey performed as if learning that sequence for the first time.     

Serial rote-learning: Incremental or all-or-none?

If the associative connections in a serial list are acquired in an all-or-none fashion, rather than gradually with every trial adding an increment of associative strength, then changing the serial order of the middle items in the list during the course of practice should have no effect on the rate of learning the list as a whole or even of the particular items that have been interchanged. Thirty subjects learned a serial list by the anticipation method. The middle items of the list were reversed in serial order approximately half-way through the number of trials required for mastery. The subjects took no longer to learn the list and made no more errors than did 30 control subjects for whom there was no change in serial order. The serial-position curves of the two groups were almost identical. It was also shown that the learning “curves” of single items in the series, when plotted for individual subjects do not reveal a gradually increasing probability of the correct response, but show instead a sudden jump on one trial from the chance guessing level to a level close to 100 per cent, correct responses. The results are consistent with a non-incremental theory of serial learning.     

Massed trial Overtraining Effects on Extinction and Reversal Performance in Turtles (Geoclemys reevesii)

Two experiments with turtles (Geoclemys reevesii) examined the overtraining extinction effect (OEE) and the overtraining reversal effect (ORE), under massed training conditions. In Experiment 1, three groups of turtles received 7, 14, or 21 sessions of training in a runway situation for food reinforcement, followed by 15 sessions of extinction. Extinction was faster, the greater the number of acquisition sessions. In Experiment 2, the two groups of turtles learned a spatial discrimination for food reinforcement either to a criterion (19 of 20 correct responses), or to that criterion plus 100 additional trials of overtraining. When the validity of the positions was reversed, learning was faster for the overtrained group than for the group trained to a criterion. This evidence of the OEE and the ORE is the first yet reported for a reptile. The results are discussed in the framework of comparative research on rewardschedule effects.     

THE EFFECTS OF CONTINGENT AND NONCONTINGENT REWARDS ON THE I.Q. SCORES OF CHILDREN OF ABOVE-AVERAGE INTELLIGENCE1

This study investigated the effects of candy reinforcement on I.Q. test scores in first and second graders of above-average intelligence. Thirty-six subjects were randomly assigned to one of three groups and either given candy contingent on correct responses, noncontingently, or not given candy. After measuring all subjects on Form L of the revised Stanford Binet, each subject in the contingent group was given an M & M immediately following each correct answer on Form M, while a yoked-comparison subject received the same number of M & M's before responding to a question and therefore without regard to correctness of answers. It was expected that both types of candy administration would produce higher I.Q. change scores than the no-candy control group, but there were no statistically significant differences among the three treatments. The differences between the first and second test scores were 4.17 (contingent), 4.67 (noncontingent), and 1.00 (no reinforcement). Reasons for this failure to replicate previous findings were discussed.     

Social facilitation in a coaction setting: An examination of the effects over learning trials

Zajonc's proposal that the presence of others facilitates emission of dominant responses was examined in a coaction setting with human maze learning. On a maze where dominant responses were likely to be correct, coacting subjects made fewer errors than those working alone. On a maze where dominant responses were likely to be incorrect, subjects performing alone made fewer errors than those coacting. Investigation of task performance at different stages in learning showed that a change in the rate of learning corresponded to a change in the dominant response from incorrect to correct. It was concluded that the presence of others has a facilitative effect on the dominant response, hindering learning when the dominant response is incorrect and helping learning when the dominant response is correct. Coaction effects were extremely pronounced in females but almost nonexistent in males.     

Setting Generalization during Self-Reinforcement with Educationally Handicapped Children

Eight educationally handicapped boys ranging in age form 10 to 11 years old and described as the worst behavior problems in their class participated in a series of single-subject studies carried out in their self-contained classrooms. Three served as target subjects, using self-reinforcement procedures to increase their sustained on-task behavior in a treatment setting; five served as generalization subjects. Data were collected in three settings within the self-contained classroom; an early morning treatment setting, a late morning generalization setting, and an afternoon generalization setting. The three target subjects averaged a 51% increase over their baseline median levels of sustained on-task behavior in the treatment setting. They averaged 84% and 96% generalization to the late morning and afternoon generalization settings, respectively. These results demonstrate that self-reinforcement can produce setting generalization of treatment effects in the absence of prior externally determined reinforcement. Evidence for subject and subject-setting generalization was also obtained.     

Spatial probability learning by alcoholic Korsakoff patients.

Performance of 10 alcoholic Korsakoff patients was compared to that of 10 normal and 10 alcoholic control subjects on each of three different schedules of spatial probability learning (50:50, 70:30, and 30:70) using monetary reinforcement and a correction procedure. There was some evidence that the Korsakoff patients were less sensitive than normals to the effects of reward: Although on all three schedules, choice ratios by normal subjects approximated the reinforcement ratios, the choice ratios of Korsakoffs on the second and third schedules remained close to the reinforcement ratio acquired with the first schedule. In addition, the Korsakoffs made an abnormal number of perseverative errors early in training. On most measures, performance by alcoholic controls fell between that of the other two groups.     

DIFFERENTIAL REINFORCEMENT OF CORRECT RESPONSES TO PROBES AND PROMPTS IN PICTURE-NAME TRAINING WITH SEVERELY RETARDED CHILDREN

A systematic sequence of prompt and probe trials was used to teach picture names to three severely retarded children. On prompt trials the experimenter presented a picture and said the picture name for the child to imitate; on probe trials the experimenter did not name the picture. A procedure whereby correct responses to prompts and probes were nondifferentially reinforced was compared with procedures whereby correct responses to prompts and probes were differentially reinforced according to separate and independent schedules of primary reinforcement. In Phase 1, correct responses to prompts and probes were reinforced nondifferentially on a fixed ratio (FR) 6 or 8 schedule; in Phase 2, correct responses to prompts were reinforced on the FR schedule and correct responses to probes were reinforced on an FR schedule of the same value; in Phase 3, correct responses to prompts were reinforced on the FR schedule and correct responses to probes were reinforced on a continuous reinforcement (CRF; every correct response reinforced) schedule; in Phase 4, correct responses to prompts were reinforced on a CRF schedule and correct responses to probes were reinforced on the FR schedule; in Phase 5, a reversal to the conditions of Phase 3 was conducted. For all three children, the FR schedule for correct responses to prompts combined with the CRF schedule for correct responses to probes (Phases 3 and 5) generated the highest number of correct responses to probes, the highest accuracy (correct responses relative to correct responses plus errors) on probe trials, and the highest rate of learning to name pictures.     

Substitutability between conditioned and primary reinforcers in discrimination acquisition.

Rats and pigeons were trained on a series of reversals of a conditional simultaneous discrimination. The percentage of reinforcement for correct trials was varied across reversals. When nonreinforced correct trials produced the same feedback as incorrect trials, the number of errors to reach an acquisition criterion was greater for smaller percentages of reinforcement, but the number of reinforcers required was either approximately constant or smaller for the smaller percentages. When a stimulus paired with food (the conditioned reinforcer) was added on nonreinforced correct trials, both measures were substantially decreased. When the same stimulus was presented, but without a history of food pairing, learning rate was similar to when no stimulus was presented on nonreinforced trials. The results provide direct evidence that conditioned reinforcers may substitute, although imperfectly, for a primary reinforcer, and that pairing with the primary reinforcer is a necessary condition for such substitutability to occur.     

Effects of two procedures for varying information transmission on observing responses

Two experiments were conducted with pigeons to examine the effects of procedures that varied information transmission on observing responses. The basic procedure for Experiment I was one in which a trial terminated in either non-contingent reinforcement or timeout. Pecking during a trial produced either green (positive) or red (negative) keylights. If no pecking occurred no differential stimuli appeared. The probability of positive trials was either 0.25, 0.50, or 0.75. Observing response rates and relative frequencies of occurrence were highest when the probability of positive trials was 0.25 and lowest at 0.75. In Experiment II, a modified chain procedure was used in which responding produced either red or green lights. Reinforcement or timeout followed light onset by 15 sec. The correlation between the stimuli and the event at the end of the trial (reinforcement or timeout) was varied. Reinforcement followed green 100%, 90%, 70%, or 50% of the time that green occurred. Since the overall probability of reinforcement remained at 0.50, reinforcement followed red in either 0%, 10%, 30%, or 50% of the time that it occurred. The rate of responses that produced these stimuli varied as a function of the correlation. The greater the probability of reinforcement after green, the higher the response rate.     

The use of programmed materials in the analysis of academic contingencies

Programmed handwriting materials were used to examine the effects of different reinforcement contingencies on the academic performance of six public school kindergarten children. The children's responses to these materials provided an educationally relevant dependent variable for the analysis of factors that affected the accuracy of their responses and the attainment of criterion performances. Variations in the complexity of most academic materials, which confound the analysis of contingencies, were eliminated by the programmed sequence so that the differential effects of three reinforcement conditions were observed. The three conditions were: baseline without tokens, tokens contingent on correct writing responses, and noncontingent tokens. It was consistently observed that the children were more accurate when their correct responses produced tokens, and that noncontingent tokens reduced accuracy below baseline levels.     

Information processing complexity and leadership status

144 Subjects divided into four-person groups participated in three discussions. In each group, one member (Target subject) was preselected as either being high or low in cognitive complexity. The groups were then assigned to one of three conditions. Under Continuous Reinforcement the Target subject received a reinforcing light cue following each verbalization in the second of three discussions (no light cues used in the first and third discussions). Target subjects under Partial Reinforcement received light cues on a 50% variable ratio reinforcement schedule. Control subjects received no light cues in any session. Target subjects in both experimental groups showed conditioning effects on all dependent measures. While no differences were found between reinforcement conditions, results suggest that abstract subjects are able to make more use of feedback cues than concrete subjects.     

Primary and secondary reinforcers as distinctive cues which acquire information value in partial-reinforcement discrimination training

To assess the cue as opposed to incentive effect of reward, rats were trained on an easy visual discrimination with food contingent on 50 or 100% of the correct responses, and white-noise contingent on 0, 50 or 100% of these responses. Additionally, the 50% food and noise schedules were structured for different groups to produce positive, zero, or negative correlations of the two events. Although the addition of noise did not affect learning with 100% food, the slower learning observed with 50% food was increasingly offset by greater percentages of noise, with the 50% groups showing faster learning under the negative than zero correlation and faster under the zero than positive correlation. Together with supporting speed data, these results indicate that a “neutral” stimulus can be substituted for food with little loss in performance. Consequently, the reinforcing effect of food is attributed in part to its function as a cue which, like noise, can increase the discriminability of the alternatives and provide information about the correctness of the response.     

Reinforcement contingencies and signal detection.

Pigeons were trained to discriminate temporal stimuli in a discrete-trial signal-detection procedure. Pecks to one side key were reinforced intermittently after exposure to one duration, and pecks to the other side key were reinforced intermittently after exposure to a different duration. In Experiment I, the allocation of reinforcers was varied systematically for correct responses and for errors, using a procedure that controlled the obtained numbers of reinforcers. When reinforcers were allocated symmetrically, the level of discrimination decreased as the proportion of reinforcers for errors increased. When reinforcers were allocated asymmetrically, the decrease in discrimination was less systematic. Bias toward one or the other side key roughly matched the ratio of reinforcers obtained by pecks at those keys, independent of the level of discrimination. In Experiment II, the overall rate of reinforcement for correct responses was varied both within and between experimental conditions. The level of discrimination was positively related to the overall rate of reinforcement. The discrimination data of both experiments were interpreted in relation to the contingencies of reinforcement and nonreinforcement, characterized by the average difference in reinforcement probability for correct responses and errors.     

SOME EFFECTS OF REINFORCEMENT SCHEDULES IN TEACHING PICTURE NAMES TO RETARDED CHILDREN1

The effects of several different schedules of primary reinforcement were compared in a picture-naming task with retarded children. In Experiment I, number of correct responses and learning rate were higher under fixed-ratio schedules than under continuous reinforcement. In Experiment II, number of correct responses and learning rate tended to be greater under intermediate than under low or high fixed-ratio schedules. In Experiment III, number of correct responses was higher under interlocking schedules, in which the response requirement increased with time following the previous reinforcement, than under comparable fixed-ratio schedules. Learning rates were generally low and, perhaps because of this, not very different under the two types of schedules in this experiment. Accuracy (i.e., proportion of trials on which correct responses occurred) was typically high and insensitive to variations in schedule and schedule parameter throughout each experiment.