THE CHANGING CRITERION DESIGN

This article describes and illustrates with two case studies a relatively novel form of the multiple-baseline design called the changing criterion design. It also presents the design's formal requirements, and suggests target behaviors and circumstances for which the design might be useful.     

The effects of brief-stimulus presentations in fixed-ratio second-order schedules

Pigeons' responses were reinforced according to a three-component multiple schedule. In Component 1, key pecks produced food according to a fixed-ratio second-order schedule with fixed-ratio units. Here, a fixed number of fixed-ratio units produced food, and the brief stimulus terminating each unit also accompanied food. Responses in Component 2 produced food on an identical schedule except that the brief stimulus was not paired with food. Component 3 contained a simple fixed-ratio schedule whose response requirement equaled that of Components 1 and 2. Across conditions the size of the fixed-ratio unit (five, ten, twenty, forty, and eighty responses) and the total number of responses per reinforcement were parametrically manipulated. The highest response rates and shortest preratio pauses were observed in Component 3 (no brief stimulus). The lowest rates and longest pauses were found in the component with paired brief-stimulus presentations, indicating that the food-paired brief stimulus suppressed responding. The suppressive effects were greatest when the fixed-ratio units were small (e.g., fixed-ratio 5) and the total fixed-ratio requirement was large (e.g., fixed-ratio 160). Under no conditions did the paired brief stimulus facilitate responding. The nonpaired brief stimulus also suppressed responding but to a lesser extent. The suppressive effects of nonpaired brief stimuli were greatest when the fixed-ratio units were small and the total response requirement was large. These data suggest that the suppressive effects of the brief stimuli may have masked the conditioned-reinforcing effects reported in other studies, and that conditions that maximize suppression in second-order schedules involve the use of fixed-ratio schedule units and the presentation of many brief stimuli per reinforcer.     

THE GOOD TOOTHBRUSHING GAME: A SCHOOL-BASED DENTAL HYGIENE PROGRAM FOR INCREASING THE TOOTHBRUSHING EFFECTIVENESS OF CHILDREN

There has been a serious lack of experimentally verified, effective dental hygiene programs in the schools. In and of themselves, the instruction-alone programs which comprise children's dental education do not produce proper toothbrushing skills. In the present study, a school-based contingency dental hygiene program designed to increase the effectiveness of children's toothbrushing skills at home was implemented with grade one and two classes. Each class was divided into teams and participated in the “Good Toothbrushing Game.” Each day four children from each team had the cleanliness of their teeth assessed according to the Simplified Oral Hygiene Index (Greene & Vermillion, 1964). The team with the lowest mean oral hygiene score was declared the daily winner. Winning teams received stickers and had their names posted. A multiple baseline across classrooms single-subject group design (Hersen & Barlow, 1976, pp. 228–229) established that the good toothbrushing game greatly increased the effectiveness of children's oral hygiene skills. The treatment terminal level for the grade one scores was 2.0 as compared to a baseline terminal level of 5.0, and for the grade two's was 2.3 compared to 5.7 at the end of baseline. A 9-mo follow-up indicated that these results were maintained. The data strongly suggest that proper implementation of behavioral principles is essential to the success of oral hygiene programs.     

TRAINING AND GENERALIZATION OF LAUNDRY SKILLS: A MULTIPLE PROBE EVALUATION WITH HANDICAPPED PERSONS

An instructional procedure composed of a graded sequence of prompts and token reinforcement was used to train a complex chain of behaviors which included sorting, washing, and drying clothes. A multiple probe design with sequential instruction across seven major components of the laundering routine was used to demonstrate experimental control. Students were taught to launder clothing using machines located in their school and generalization was assessed later on machines located in the public laundromat. A comparison of students' laundry skills with those of normal peers indicated similar levels of proficiency. Follow-up probes demonstrated maintenance of laundry skills over a 10-month period.     

On the effects of component durations and component reinforcement rates in multiple schedules

Four experiments, each using the same six pigeons, investigated the effects of varying component durations and component reinforcement rates in multiple variable-interval schedules. Experiment 1 used unequal component durations in which one component was five times the duration of the other, and the shorter component was varied over conditions from 120 seconds to 5 seconds. The schedules were varied over five values for each pair of component durations. Sensitivity to reinforcement rate changes was the same at all component durations. In Experiment 2, both component durations were 5 seconds, and the schedules were again varied using both one and two response keys. Sensitivity to reinforcement was not different from the values found in Experiment 1. In Experiment 3, various manipulations, including body-weight changes, reinforcer duration changes, blackouts, hopper lights correlated with keylights, and overall reinforcement rate changes were carried out. No reliable increase in reinforcement sensitivity resulted from any manipulation. Finally, in Experiment 4, reinforcement rates in the two components were kept constant and unequal, and the component durations were varied. Shorter components produced significantly increased response rates normally in the higher reinforcement rate component, but schedule reversals at short component durations eliminated the response rate increases. The effects of component duration on multiple schedule performance cannot be interpreted as changing sensitivity to reinforcement nor to changing bias.     

Key-peck durations under behavioral contrast and differential reinforcement

Pigeons were maintained on a multiple schedule in which both components were variable-interval one-minute schedules. When they were switched to a condition in which one component was extinction, behavioral contrast was observed. The median durations of the key pecks in the unchanged component did not decrease in size. The results are incompatible with a theory of behavioral contrast which considers the added pecks to be short-duration responses. In a second experiment, pigeons were required to emit short-duration key pecks in one component of a multiple schedule, and long-duration pecks in the other. Two of three pigeons learned to emit responses appropriate to the requirements of the component in effect, suggesting that the duration of the key-peck response is sensitive to differential reinforcement.     

Effects of ethanol on multiple fixed-interval fixed-ratio schedule performances: dynamic interactions at different fixed-ratio values.

Key pecking by three pigeons was maintained under a multiple fixed-interval fixed-ratio schedule of food presentation. The fixed-interval value remained at 3 minutes, while the fixed-ratio size was increased systematically in 30-response increments from 30 to either 120 (two pigeons) or 150 (one pigeon). At least two lower fixed-ratio values were also redetermined. The effects of ethanol (5 to 2.5 g/kg) were assessed at each of the different schedule parameters. Both overall and running response rates under the fixed-ratio schedule decreased with increases in the size of the fixed-ratio schedule; pause duration under the fixed-ratio schedule was directly related to increases in fixed-ratio size. Overall and running rates of responding under the fixed-interval schedule changed little with increases in the size of the fixed-ratio schedule. Despite the relative invariance of fixed-interval responding across the different fixed-ratio values, the effects of ethanol on responding under the fixed-interval schedule differed depending on the size of the fixed-ratio schedule. Greater increases occurred in both overall and in lower local rates of responding under the fixed-interval schedule when the fixed-ratio value was 120 or 150. The effects of ethanol on responding under the fixed-ratio schedule also depended on the size of the fixed ratio. Increases in responding under the fixed-ratio schedule were typically greater at the higher fixed-ratio values where response rates were lower. When the effects of ethanol were redetermined at the lower fixed-ratio parameter values, rates and patterns of responding were comparable to those obtained initially. However, the dose-effect curves for responding under both fixed-ratio and fixed-interval schedules were shifted up and to the right of those determined during the ascending series. The effects of ethanol can depend on rate or responding, behavioral history, and the context in which behavior occurs.     

Contingency and stimulus change in chained schedules of reinforcement

Higher rates of pecking were maintained by pigeons in the middle component of three-component chained fixed-interval schedules than in that component of corresponding multiple schedules (two extinction components followed by a fixed-interval component). This rate difference did not occur in equivalent tandem and mixed schedules, in which a single stimulus was correlated with the three components. The higher rates in components of chained schedules demonstrate a reinforcing effect of the stimulus correlated with the next component; the acquired functions of this stimulus make the vocabulary of conditioned reinforcement appropriate. Problems in defining conditioned reinforcement arise not from difficulties in demonstrating reinforcing effects but from disagreements about which experimental operations allow such reinforcing effects to be called conditioned.     

Note on the squared multiple correlation as a lower bound to communality

It is demonstrated that the squared multiple correlation of a variable with the remaining variables in a set of variables is a function of the communalities and the squared canonical correlations between the observed variables and common factors. This equation is shown to imply a strict inequality between the squared multiple correlation and communality.     

FUNCTIONAL ANALYSIS OF SEPARATE TOPOGRAPHIES OF ABERRANT BEHAVIOR

We conducted a functional analysis of distinct topographies of aberrant behavior displayed by 4 clients. We first analyzed the behaviors in an aggregate fashion and then separated the behaviors to formulate hypotheses about the maintaining variables for each behavior. The procedures were used in a two-phase experiment. During Phase 1, two extended functional analyses were completed, one in an inpatient unit and one in a special education classroom. During Phase 2, two brief functional analyses were completed in an outpatient clinic. Results indicated that hypotheses of separate functions for distinct behaviors can be generated using both extended and brief functional analyses when the results are graphed in the aggregate and are separated by response topography. The results also suggest that these methods can improve the accuracy of data interpretation and treatment selection.