Individual differences in human fixed-interval performance.
The fixed-interval (Fl) schedule is one of the most intensively studied schedules (Lowe, 1979). Under the Fl schedule, two types of patterns of responding are usually found in humans. The first of these is a high-rate pattern, which consists of a steady high response rate throughout each interreinforcer interval with little or no postreinforcement pause (e.g., Weiner, 1962). The second is a low-rate pattern, which consists of a very low response rate often with just one or two responses at the end of the interval (e.g., Buskist, Bennett, & Miller, 1981; Matthews, Shimoff, Catania, & Sagvolden, 1977).
However, it may be premature to conclude that there are reliable individual differences in the performance of humans under Fl schedules. With some types of instructions, detecting the individual differences seems to be difficult. For example, subjects instructed with regard to an accurate schedule contingency responded slowly with a pause (Lippman & Meyer, 1967), whereas subjects, whose responding was established not by instructions but by shaping, also revealed the low-rate pattern (Matthews et al., 1977).
Individual differences in Fl performance have been often observed in experiments in which contingencies or how to respond were not instructed but responding was established through instructions (Baron, Kaufman, & Stauber, 1969; Bennett, Miller, & Buskist, 1984; Buskist, Miller, & Bennett, 1980; Leander, Lippman, & Meyer, 1968; Lippman & Meyer, 1967; Weiner, 1964, 1969). Even in such instructional conditions, however, some research has found either the high-rate pattern (Weiner, 1962) or the low-rate pattern (Buskist et al., 1981), suggesting no systematic individual differences in Fl performance within each experiment.
One of the reasons for such inconsistency may be the sample size. Most of the research on Fl schedules has employed small numbers of subjects in each identical experimental condition, usually lower than 10 (e.g., Bennett et al., 1984; Bentall, Lowe, & Beasty, 1985; Buskist et al., 1980; Lippman & Meyer, 1967; Weiner, 1964, 1969; but see Darcheville, Riviere, & Wearden, 1992, 1993; Leander et al., 1968). Using small numbers of subjects may relate to the single-subject tactics employed in behavior analysis. Behavior analysts have attempted to find the effects of experimental manipulations not by increasing the number of subjects, but by increasing the number of observations per subject (of. Skinner, 1938). These tactics can function if behavioral phenomena common across individuals are under investigation. This is not the case, however, when individual differences are under investigation. If half of the population reveals a high rate of responding and the other half a low rate, it is highly probable that all su bjects show the same pattern of responding (i.e., no individual differences) in an experiment using a small sample. Thus, direct observation of individual differences using a large number of subjects within an identical experimental condition is needed to confirm the reliability of the phenomena.
Reliability can also be demonstrated by examining whether a behavioral phenomenon is observed repeatedly within individuals. In terms of individual differences in Fl schedule performance, whether a certain response pattern persists or not must be answered. In those studies arranging a simple Fl schedule in humans, Weiner's (1962) Experiment 2 provides data from the most sustained exposure (15.5 hours). He found a high-rate pattern for one male adult and a moderate steady-rate pattern for another subject under an Fl 60-s schedule. In contrast, extended exposure to an experimental condition arguably reduces the effects of extra-experimental variables that result in inter-individual variability (e.g., Baron & Perone, 1998; Bernstein, 1988). Thus, whether such patterns continue after more extended exposure to such schedules is controversial.
The present study examined individual differences in Fl performance in normal adult humans using a relatively large number of subjects and observations. Thirty-two undergraduates were exposed to an Fl 60-s schedule. Beyond the point where a given stability criterion was met, 4 subjects randomly selected from those 32 subjects were exposed to the schedule for a total of 60 sessions (over 20 hours).
The present study also examined relations between Fl performance and other aspects of individual behavior. Van den Broek, Bradshaw, and Szabadi (1987) found a correlation between performance under a temporal schedule and impulsiveness. In their experiment, adult humans classified as impulsive according to the Matching Familiar Figures Test (MFFT Kagan, 1966) responded at higher rates under a differential-reinforcement-of-low-rate (DRL) schedule than those classified as nonimpulsive. As an extension of van den Broek et al., the present study assessed the possible relation between Fl performance and impulsiveness measured by the MFFT.
Subjects were 9 male and 23 female undergraduates recruited from an introductory psychology class at Osaka Kyoiku University. They were 19-25 years old. None had experience with operant conditioning, experiments. All subjects received course credit for their participation.
The experimental room was 1.70 m wide, 2.20 m deep, and 2.17 m high. A Nihon Electric Company PC-9821AP microcomputer, located in an adjacent room, was used to control the experiment. The subjects sat facing a 25-cm wide by 18-cm high color display monitor equipped with a Micro Touch Systems touch screen on a desk. A 5.5-cm-diameter white circle was presented in the middle left of the display monitor and each touch on the circle (operandum) was defined as a response. Points (reinforcers) accumulated in the session were presented on the top right of the monitor. Responses and reinforcers were accompanied, respectively, by low and high tones through a speaker beneath the desk. All interevent times were recorded with 50-ms resolution, in real time for within-session data analysis.
Subjects 29, 30, 31, and 32, who were selected randomly from all 32 subjects, were asked and agreed to remain in the experiment for a maximum of 15 90-mm experimental periods. The other 28 subjects were informed to remain in the experiment for a maximum of three experimental periods. A 90-mm experimental period was conducted once per day, two times per week. During this 90-mm period, four sessions occurred. Sessions were separated by 2-to-3-min breaks during which the experimenter recorded the data. Upon completion of the experiment, subjects were paid for participation (50 yen per hour, approximately 0.56 U.s. dollars) and performance (0.5 yen per 100 points) and debriefed.
On the first day of the experiment, after being escorted into the room, subjects were asked to read the instructions, The instructions were written in Japanese and their English translation was as follows:
Your task is to earn as many points as you can. Points will be shown in the top right of the display monitor. A hundred points are worth 0.5 yen. In addition, you will be paid 50 yen for every hour you spend in the experiment. Total payment will be made at the end of the experiment.
A white circle will be shown in the middle left of the monitor. If you touch the circle, you may earn points. Every touch, however, may not result in accumulating points.
The words "READY" and "GO" will appear in sequence on the display monitor. When the word "GO" disappears, do the task until the words "GAME OVER" appear on the display monitor.
The typed set of instructions remained on the desk throughout the experiment. Questions regarding the experimental procedure were answered by the experimenter telling the subject to reread the appropriate sections of the instructions. Then the words "READY" and "GO" were presented in sequence in the top left of the display monitor. After the word "GO" disappeared, a white circle, which served as the operandum, was presented in the middle left of the display monitor.
An FI 60-s schedule was in effect throughout the session. That is, the first touch on the circle after 60 s from the previous reinforcement accumulated 100 points on the top right counter. Each session lasted until 20 reinforcers occurred. Thereafter, the circle and points disappeared, and the words "GAME OVER" appeared at the top left of the monitor.
Except for Subjects 29, 30, 31, and 32, the experiment was terminated after 8 sessions or when the following stability criterion was met, whichever came first: the difference between the response rates (number of responses per minute) for the last two and the immediately preceding two sessions had to be less than 15% of the mean rate for the four sessions (Galizio, 1979). In order to assess long-term effects of the schedule, the experiment lasted for 60 sessions beyond the point where the stability criterion was met for Subjects 29, 30, 31, and 32.
Matching Familiar Figures Test (MFFT)
Prior to the experiment, subjects completed the adolescent/adult version of the MFFT. The MFFT was scored in the usual way. Two measures were derived: the mean latency to the first response, averaged across the 12 items of the test, and the total number of errors made on all 12 items. Kagan (1966) argued that impulsive individuals tend to have low latency scores and high error scores.
Individual Differences in FI Performance
Table 1 shows the number of sessions each subject completed. For 30 of 32 subjects, the rates of responding met the stability criterion within eight sessions, whereas the rates did not meet the criterion for Subjects 8 and 11. Figure 1 is a scatter plot of the mean postreinforcement pause (PRP: time to the first response after each reinforcer) and the mean response rate during the last four sessions for each subject (see also Table 1). There were extreme individual differences in PRPs and response rates, and a clear correlation between them: The longer the pauses, the lower the rates.
Figure 2 shows cumulative records illustrating individual differences in FI performance. For Subject 15, response rates were high and steady throughout each interreinforcer interval (IRI) with little or no PRP. For Subject 26, only one or two responses occurred at the end of the IRI with a long PRP. A break-and-run pattern was observed repeatedly for Subject 9, whose averaged response rate and PRP were moderate. For Subject 7, whose averaged response rate was also moderate but whose averaged PRP was short, the cumulative record showed moderate constant response rates throughout each IRI. However, inspection of cumulative records for all subjects revealed that this pattern of responding was not typical. Most subjects with moderate response rates and short PRPs consistently responded once or twice immediately after reinforcement, then paused, and only thereafter emitted a run of responses.
FI Performance and Impulsiveness
Table 1 shows the total number of errors and mean latency to the first response on the MFFT for each subject. The Kendall rank order correlation was computed between measures of the FI performance and those of the MFFT. The number of errors made on the MFFT was negatively correlated with the PRP (r = -.300, p < .05) and was positively correlated with the response rate (r= .361, p < .01). In contrast, the mean latency to the first response on the MFFT was not correlated with the PRP (r = .062, ns) and the response rate (r = -.111, ns). The correlation between the number of errors and the latency on the MFFT was negative and significant (r= -.528, p < .01).
Performance under Sustained FI Exposure
Figures 3 and 4, respectively, show session-by-session response rates and PRPs of each subject exposed to the FI 60-s schedule for 60 sessions. For Subjects 29 and 31, response rates were high and PRPs were not found during all sessions. For Subject 32, response rates were low and PRPs were long. Her PRPs increased for the last 10 sessions. For Subject 30, response rates were moderate and PRPs were short, with some variability in response rates across all sessions and in PRPs during the first 20 sessions. In general, there were consistent individual differences in response rates and PRPs for 60 sessions.
Figure 5 shows cumulative records from the final session when the stability criterion was met and the 60th session for each subject who was exposed to the FI 60-s schedule for 60 sessions. For Subjects 29 and 31, response rates were high and steady throughout each IRI without PRP, whereas Subject 30's response rates were moderate and steady throughout each IRI. The cumulative records of an early session and the final session were indistinguishable for each of these 3 subjects. For Subject 32, low-rate and break-and-run patterns were observed during the 6th session, whereas a low-rate pattern was prominent during the 60th session. In general, however, patterns of responding for each subject did not change drastically through sustained exposure to the FI schedule.
As described in previous research, there were systematic individual differences in FI performance (e.g., Bennett et al., 1984; Bentall et al., 1985; Buskist et al., 1980; Darcheville et al., 1992, 1993; Leander et al., 1968; Lippman & Meyer, 1967; Weiner, 1964, 1969). There was a negative correlation between response rates and PRPs. The present results, obtained from a larger number of subjects, emphasize the reliability of the phenomena.
Weiner (1962) exposed 2 adults to an FI 60-s schedule for 15 1/2 hours and with maximally 930 reinforcers. The present experiment, in contrast, exposed 4 students to an FI 60-s schedule for over 20 hours with 1200 reinforcers. In spite of such an exceptional extended exposure, individual differences in FI performance persisted throughout these extended sessions. Furthermore, the present study found high-, moderate-, and low-rate patterns of responding after sustained exposure to the schedule, whereas Weiner observed only high and moderate patterns. Contrary to the previous view (Baron & Perone, 1998; Bernstein, 1988), in the present experiment simple exposure to FI schedules did not reduce the individual differences, but rather demonstrates the robustness and reliability.
Variability in responding under FI schedules is sometimes said to be caused by inadequate experimental control (Sidman, 1960). One of the variables that can result in this variability is the lack of effectiveness of reinforcers. Especially because subjects generally earned more by attending the sessions than by obtaining points, one may wonder whether the points maintained behavior. In the previous experiment conducted in the same laboratory, students' rates of responding were lower under a DRL schedule than a FI schedule, indicating that the rates were under the control of the schedule contingencies (Okouchi, 1993). Because both absolute and relative earnings by points in Okouchi's experiment were approximately equal to those in the present experiment, one can argue that points functioned as reinforcers in the present study.
The present study confirmed the reliability of the individual differences in human FI performance. However, the data came from limited conditions, and the generality of the results is unknown. As mentioned in the introduction, systematic individual differences may be difficult to detect if the instructions used are different from those employed in the present experiment (cf. Lippman & Meyer, 1967; Matthews et al., 1977).
FI performance for the present subjects was correlated with error scores on the MFFT but not with latency. Impulsive individuals have been discussed as having low latency scores and high error scores on the MFFT (Kagan, 1966), and individuals with low latency scores and high error scores on the MFFT have been often classified as impulsive (e.g., van den Broek et al., 1987). However, there is some evidence suggesting that differential functions exist between the two measures of the MFFT. For example, using 96 undergraduates, Glow, Lange, Glow, and Barnett (1983) found that standardized self-report impulsiveness measures were correlated not with MFFT latency scores but with MFFT error scores. Gjerde, Block, and Block (1985) repeatedly administered the MFFT to 128 children (aged 3 to 11 yr) over a period of 8 years and found that MFFT error scores were more consistent over time than MFFT latency scores. Thus, the present results were correlated with a more reliable and valid measure of impulsiveness.
Both biological factors and different individual histories have been proposed as causes of individual differences in operant responding (Harzem, 1984). One suggested variable causing individual differences in impulsiveness has been the socioeconomic level of the subjects. For example, Bresenham and Shapiro (1972) proposed that the behavior of subjects from lower socioeconomic groups was more impulsive than that of subjects from more favored backgrounds. One of the possible causes of individual differences in FI performance is also historical. The subjects' differing histories of carrying out tasks similar to that required by the schedule have been said to contribute to their schedule performance (Weiner, 1969).
The present study indicates that individual differences are observed reliably in basic, behavior-analytic experimental situations. Although some of the human behavior patterns may be difficult to control through experimental manipulations, the patterns seem to relate in orderly ways to other behavior patterns of the same individual. For example, sensitivity to changing contingencies was correlated with depression (Rosenfarb, Burker, Morris, & Gush, 1993) or with personality rigidity (Wulfert, Greenway, Farkas, Hayes, & Dougher, 1994). Individual differences in behavior under FI schedules have been reported to be correlated with individual differences in self-control in young children (Darcheville et al., 1992) and in infants and toddlers (Darcheville et al., 1993). In their experiments, subjects who produced a high-rate pattern of responding under FI schedules chose smaller immediate reinforcers on the self-control procedure (Darcheville et al., 1992, 1993). Because choice of a smaller immediate reinforcer is regarded as impulsiveness (Logue, 1988), the results of Darcheville et al. may indicate that impulsiveness relates to FI performances. Although impulsiveness measured by the MFFT may not be exactly the same as that measured by Darcheville et al.'s self-control procedure, the present study also found a correlation between FI response patterns and impulsiveness in adults. It should be noted, however, that all these findings show correlation between performance and another, not implying causal relations between them. Nevertheless, explorations along these lines, at least, have led to the discovery of new regularities in human behavior.
[FIGURE 1 OMITTED]
[FIGURE 3 OMITTED]
[FIGURE 4 OMITTED]
Table 1 Number of Sessions, Average Response Rate and Postreinforcement Pause (in sec) over Last Four Sessions, and Total Number of Errors and Average Latency to First Response (in sec) on the Matchng Familiar Figures Test (MFFT) FI performance MFFT Number of Response Number of Subject Sessions Rate PRP Errors Latency 25 4 419.9 .2 9 11.3 15 4 418.6 .2 11 4.9 13 4 364.5 .2 5 10.8 27 4 358.2 .2 5 10.5 16 4 329.6 6.0 6 6.4 28 4 327.3 .3 1 18.7 29 4 (b) 316.0 .2 5 6.9 3 4 315.1 .2 6 4.7 30 8 (b) 293.7 .2 4 17.2 6 4 290.7 .2 2 17.5 23 4 272.7 5.2 5 14.9 2 4 253.0 .5 9 6.8 10 6 247.7 .4 3 15.4 5 4 238.8 .3 2 17.4 20 4 211.9 2.7 7 11.8 18 4 144.8 4.8 8 11.2 31 4 (b) 82.1 1.0 9 8.7 1 5 80.0 2.1 7 3.9 24 6 68.3 24.1 2 10.5 7 7 57.2 2.2 10 4.6 21 8 36.4 .5 0 17.8 9 8 30.9 28.4 4 7.3 11 8 (a) 27.9 1.0 3 11.7 12 6 24.8 25.8 4 6.6 17 5 17.2 37.7 2 5.4 8 8 (a) 9.9 47.4 2 10.1 32 6 (b) 8.2 47.2 1 20.3 19 4 5.4 54.4 2 17.4 4 5 3.1 50.6 6 9.0 22 7 1.9 54.5 2 15.2 14 4 1.2 62.4 3 10.5 26 8 1.2 63.1 1 18.5 Note. (a): Rates of responding did not meet the stability criterion within eight sessions. (b): Although the experiment lasted for 60 sessions, data are shown according to the point where the rates of responses met the stability criterion.
BARON, A., KAUFMAN, A., & STAUBER, K. A. (1969). Effects of instructions and reinforcement feedback on human operant behavior maintained by fixed-interval reinforcement. Journal of the Experimental Analysis of Behavior, 12, 701-712.
BARON, A., & PERONE, M. (1998). Experimental design and analysis in the laboratory study of human operant behavior. In K. A. Lattal & M. Perone (Eds.), Handbook of research methods in human operant behavior (pp. 4591). New York: Plenum Press.
BENNETT, R. H., MILLER, H. L., & BUSKIST, W. F (1984). Human performance on modified fixed-interval schedules. The Psychological Record, 34, 99-114.
BENTALL, R. R, LOWE, C. F, & BEASTY, A. (1985). The role of verbal behavior in human learning: II. Developmental differences. Journal of the Experimental Analysis of Behavior, 43,165-181.
BERNSTEIN, D. J. (1988). Laboratory lore and research practices in the experimental analysis of human behavior: Designing session logistics--How long, how often, how many? The Behavior Analyst, 11, 51-58.
BRESENHAM, J. L., & SHAPIRO, M. M. (1972). Learning strategies in children from different socioeconomic levels. In H. W. Reese (Ed.), Advances in child development and behavior (Vol. 7, pp. 32-79). New York: Academic Press.
BUSKIST, W. F, BENNETT, R. H., & MILLER, H. L. (1981). Effects of instructional constraints on human fixed-interval performance. Journal of the Experimental Analysis of Behavior, 35, 217-225.
BUSKIST, W. F, MILLER, H. L., & BENNETT, R. H. (1980). Fixed-interval performance in humans: Sensitivity to temporal parameters when food is the reinforcer. The Psychological Record, 30, 111-121.
DARCHEVILLE, J. C., RIVIERE, V., & WEARDEN, J. H. (1992). Fixed-interval performance and self-control in children. Journal of the Experimental Analysis of Behavior, 57, 187-199.
DARCHEVILLE, J. C., RIVIERE, V., & WEARDEN, J. H. (1993). Fixed-interval performance and self-control in infants. Journal of the Experimental Analysis of Behavior, 60, 239-254.
GALIZIO, M. (1979). contingency-shaped and rule-governed behavior: Instructional control of human loss avoidance. Journal of the Experimental Analysis of Behavior, 31, 53-70.
GJERDE, P. F., BLOCK, J., & BLOCK, J. H. (1985). Longitudinal consistency of Matching Familiar Figures Test performance from early childhood to preadolescence. Developmental Psychology, 21, 262-271.
GLOW, R. A., LANGE, R. V., GLOW, P. H., & BARNETT, J. A. (1983). Cognitive and self-reported impulsiveness: Comparison of Kagan's MFFT and Eysenck's EPQ impulsiveness measures. Personality and Individual Differences, 4, 179-187.
HARZEM, P. (1984). Experimental analysis of individual differences and personality. Journal of the Experimental Analysis of Behavior, 42, 385-395.
KAGAN, J. (1966). Reflection-impulsivity: The generality and dynamics of conceptual tempo. Journal of Abnormal Psychology, 71, 17-24.
LEANDER, J. D., LIPPMAN, L. G., & MEYER, M. E. (1968). Fixed interval performance as related to subjects' verbalizations of the reinforcement contingency. The Psychological Record, 18, 469-474.
LIPPMAN, L. G., & MEYER, M. E. (1967). Fixed-interval performance as related to instructions and to subjects' verbalizations of the contingency. Psychonomic Science, 8, 135-136.
LOGUE, A. W. (1988). Research on self-control: An integrating framework. Behavioral and Brain Sciences, 11, 665-709.
LOWE, C. F. (1979). Determinants of human operant behaviour. In M. D. Zeiler & P. Harzem (Eds.), Advances in analysis of behavior: Vol. 1. Reinforcement and the organization of behaviour (pp. 159-192). Chichester, England: Wiley.
MATTHEWS, B. A., SHIMOFF, E., CATANIA, A. C., & SAGVOLDEN, T. (1977). Uninstructed human responding: Sensitivity to ratio and interval contingencies. Journal of the Experimental Analysis of Behavior, 27, 453-467.
OKOUCHI, H. (1993). Ningen no schedule performance (2): Fl performance ni oyobosu hannou cost no kouka [Schedule performance in humans (2): Effects of response cost on the fixed-interval performance]. Proceedings of the 11th meeting of the Japanese Association for Behavior Analysis, 57-58.
ROSENFARB, I. S., BURKER, E. J., MORRIS, S. A., & CUSH, D. T. (1993). Effects of changing contingencies on the behavior of depressed and nondepressed individuals. Journal of Abnormal Psychology, 102, 642-646.
SIDMAN, M. (1960). Tactics of scientific research: Evaluating experimental data in psychology. New York: Basic Books.
SKINNER, B. F. (1938). The behavior of organisms. New York: Appleton-Century-Crofts.
SKINNER, B. F. (1956). A case history in scientific method. American Psychologist, 11, 221-233.
VAN DEN BROEK, M. D., BRADSHAW, C. M., & SZABADI, E. (1987). Behaviour of "impulsive" and "non-impulsive" humans in a temporal differentiation schedule of reinforcement. Personality and Individual Differences, 8, 233-239.
WEINER, H. (1962). Some effects of response cost upon human operant behavior. Journal of the Experimental Analysis of Behavior, 5, 201-208.
WEINER, H. (1964). Response cost effects during extinction following fixed-interval reinforcement in humans. Journal of the Experimental Analysis of Behavior, 7, 333-335.
WEINER, H. (1969). Controlling human fixed-interval performance. Journal of the Experimental Analysis of Behavior, 12, 349-373.
WULFERT, E., GREENWAY, D. E., FARKAS, P., HAYES, S. C., & DOUGHER, M. J. (1994). Correlation between self-reported rigidity and rule-governed insensitivity to operant contingencies. Journal of Applied Behavior Analysis, 27, 659-671.
I thank Ai Ishida for assistance in collecting and analyzing the data. Reprints may be obtained from Hiroto Okouchi, Department of Psychology, Osaka Kyoiku University, 4-698-1 Asahigaoka, Kashiwara, Osaka 582-8582 Japan. (E-mail: email@example.com).
|Printer friendly Cite/link Email Feedback|
|Publication:||The Psychological Record|
|Date:||Mar 22, 2002|
|Previous Article:||Unreinforced conditional selection by two-year-olds in a six-comparison matching task.|
|Next Article:||Immediate feedback assessment technique promotes learning and corrects inaccurate first responses.|