Method

Participants

Inclusion Criterion:

• At least 18 years of age, or older
• have a diagnosis which qualifies as a disability under the Individuals with Disabilities Education Act (IDEA)
• ability to correctly manipulate all items within the study
• Provide consent before initiating, and throughout the study
  • If participant could not legally provide consent, consent was obtained from the participants guardian and assent was obtained prior to each session

Demographics:

 

Materials

• Table with at least two chairs
• Items required to emit the response (dice or clothespin)
• Plate to place edible reinforcement (when relevant)
• Sound system to play 30s of preferred sound (when relevant)
• Hardware for data collection (i.e., handheld computer)
• Software for data collection (Countee®)
• Participant specific binders
  • Relevant data collection (such as Procedural Fidelity sheets)
  • Summary materials for each phase of the study
  • Sign-In Sheet
  • A 10-whole punch card utilized to track when compensation ($25 gift card) is expected
    • One punch was given for every 20 minutes of work in the study, always rounded up.

Response Measurement

Observers trained on the procedures collected continuous timed-event measures of target responding frequency, reinforcer delivery, latency to reinforcer delivery from session start, and overall session duration.

• During progressive-ratio reinforcer analysis (PRA),  the specific schedule value (such as FR1, FR4, etc.), which was met prior to providing each reinforcer, was recorded.
• During progressive fixed-ratio reinforcer analysis (PFRA), the specific schedule value for which the highest response output occurred, was recorded.

Within this study, reinforcers utilized within the study included either putting an edible item on a plate in front of the participant, or playing a preferred sound (such as a favorite song) for 30 seconds. In order to earn access to the reinforcer, participants had to “pay” by engaging in the target response. This included rolling a dice (66 participants) and manipulating a clothespin (3 participants).

Dependent Variables

The dependent variables within the study were the derivative measures from each analysis.

• In PRA, this included breakpoint, or the schedule of the reinforcer last obtained prior to stopping target response.
• In PFRA, this included the demand, the number of reinforcers obtained prior to discontinuing target responding, as well as the Pmax, or the schedule value in use when the target responding was highest.
• During baseline and extinction sessions of the extinction challenge, we recorded:
  • the baseline response rate by averaging the rate of responding during last 3 sessions of baseline.
  • the peak response rate during extinction by identifying the rate of responding during the session wherein the single highest rate of responding during extinction was obtained
  • the peak response rate during extinction depicted as a proportion of baseline identified by dividing the peak response rate by the average baseline response rate, multiplied by 100.
  • the prevalence of extinction bursts which is identified as occurrences wherein the peak response rate was higher than the average response rate during baseline.
  • the total number of target responses emitted during extinction.
  • the latency to extinction
  • the reinforcers “missed” prior to achieving extinction which was identified by dividing the number of responses in extinction phase by the baseline schedule value.

Experimental Design

Participants recruited in groups of 10 were sorted into two groups of four (tetrad) wherein the smallest range of variation was recorded between participants average breakpoints as determined through intake assessments and PRA. The members of each tetrad were then randomly assigned to one of the four experimental groups:

• cheap EO
• cheap AP
• expensive EO
• expensive AO

The remaining two participants who were not assigned to one of the two tetrads completed the PFRA but were placed on hold to be randomized into an experimental group with members of later tetrads prior to beginning the extinction challenge.

Data Analysis Plan

For the continuous and count dependent variables, which includes all with the exception of extinction bursts, evaluations included:

• the effect of extinction timing which can be determined by examining how close reinforcer consumption prior to extinction was to demand
• the effect of relative price such as cheap or expensive in relation to Pmax
• the price by timing interaction, or whether changes in the effects of timing is dependent on the price

In is known, or predicted that a number of things may effect responding during extinction. Consequently, we aimed to control and balance the variation in these parameters among the test groups through our matching and randomization procedures. These parameters may include:

• the value of the reinforcer
• the value of the baseline schedule
• the response rate during baseline
• the rate at which reinforcement

Therefore, we anticipated showing similarity in the distribution of these parameters by revealing statistically insignificant differences in these values between the groups, with one exception:

• Within test groups wherein unit price was manipulated (i.e., cheap vs. expensive), we expected to see significant differences in baseline schedule value, baseline response rates, and obtained rates of baseline reinforcement.

Furthermore, during extinction, it was predicted that:

• Mean magnitude of peak responding for participants assigned to cheap conditions would be higher than those assigned to expensive conditions, holding time constant.
• Mean magnitude of peak responding would be higher for those assigned to EO as compared to AO, holding price constant.
• A timing by price interaction such that the difference in the mean peak responding in cheap and expensive conditions would be bigger in EO relative to AO.

Procedures

All participants completed the stages of the study in the same progression:

Interview

The interview consisted of obtaining informed consent from the participant (or guardian when relevant), determining appointment schedule, and identification of possible programmed reinforcers and target-response that will be used for the duration of the study, as according to participant’s preference and skill repertoire.

Target Response Training 

All target responses included manipulation and restoration. That is, the last step of the response sets up another opportunity for the participant to complete the response. Eligible responses were:

• simple to do
• unlikely to occur in the absence of reinforcement
• matched to the individual participants skill repertoire.

For training, the correct response was modeled for the participant and then the participant was told to replicate the response. If they could not:

• least-to-most prompting was utilized to guide initial responding and delivered a non-contingent reinforcer, such as verbal praise.
• If they could or would still not independent produce a response, then the response was disqualified and another was evaluated.

Free Access to Reinforcement 

Before beginning, participants were presented with an array of possible edible reinforcers as determined by the intake interview and asked to pick one that they would like to consume for the remainder of their participation in the study. The edible items were then carefully measured and uniform in size. If edible reinforcement was a concern (i.e., dietary restrictions, allergies), then non-edible reinforcement options were presented such as music or simple crafts

Participants were then provided with a contingency review which included informing them that they will begin to work to earn their reinforcer during their next appointment (as to avoid the effects associated with satiation). They are told that at this appointment, they are going to be allowed to have as much of the reinforcer as they want but if they no longer want the item or wish to terminate the session. They are reminded that researchers cannot talk once the session has begun but if they are bored or wish to talk instead, they have the option of terminating the session. During the session, one reinforcer was placed on the plate at a time, and replaced when consumed by the participant.

The session terminated when: participant did not consumer reinforcer for one minute, participants expressed that they wanted to be done (either verbally or by walking away for those with complex communication needs), participant touched the laminated “stop” card, or after one hour had passed.

Progressive-Ratio Reinforcer Analysis (PRA)

During PRA, participants endured three PRA (test) sessions which were interspersed with, and always following control sessions. In effort to avoid satiation, only one test condition could occur in each appointment.

• Control
  • Item used for target responding, such as a dice, and a low-preferred alternative was placed in front of the participant
  • Contingency review provided wherein participant were:
    • told that they will now have the opportunity to to the task but that they should only do so if they want to as nothing will happen as a result
    • reminded that facilitators cannot talk to them but that this is not indicative of their feelings
    • told that if they wish to be done or to talk, they just need to tell the facilitator or touch the “stop” card placed on the table
  • Facilitator then counted down to begin the session and avoided eye contact with the participant.
  • If participant emitted a target response, the facilitator stopped and interrupted
    • Participant was reminded that responding will not get them anything and that is shown when they don’t emit a response for one minute
    • Participant reminded that if they don’t want to wait for the minute, or would rather talk, they just need to let the facilitator know that they are done.
    • Facilitator then reset the timer and restarted the session
  • This continued until a sample was obtain wherein no target responding occurred for one minute or the participant asked to end the session.
• Test
  •  Target responding was reinforced following a progressive-ratio schedule. After the delivery of two reinforcers, the price increased by 3 responses (i.e., FR1, FR4, FR7, etc.)
  • Prior to starting, participants were again provided with a contingency review wherein participants were told:
    • they now have the opportunity to earn their individualized reinforcer
    • earning the reinforcer would be easy at first but would get more difficult
    • reminded that facilitator cannot talk during the session
    • Told that if they stop wanting reinforcers, get bored doing the task, or just want to talk—they can stop doing the task for one minute, say “I’m done” or touch the stop card.
  • Facilitator then counted down to begin the session and avoided eye contact with the participant.
  • If participants asked how many they have until the next one, they were told.
  • Sessions ended either when participants didn’t respond for one minute, stated they wanted to be done, touched the stop card, or after 1- hour had elapsed.

Progressive Fixed-Ratio Analysis (PFRA)

The PFRA was similar to PRA with some exceptions.

• did not include a control condition
• Schedule values were fixed for an entire session (requirements increased by 3 across sessions rather than across reinforcer deliveries)
• only one PFRA session occurred per appointment, to account for satiation.

As in the progressive ratio analysis,

• Contingency review was provided
• Facilitator then counted down to begin the session and avoided eye contact with the participant
• If participants asked how many they have until the next one, they were told.
• Sessions ended either when participants didn’t respond for one minute, stated they wanted to be done, touched the stop card, or after 1- hour had elapsed.

The entire PFRA ended after the number of target responses emitted in a current session fell below that produced by a previous session, which empirically confirms Pmax.

Extinction Challenge

The main objective was to create a history with the baseline reinforcement contingencies in the first place.

The baseline schedule of reinforcement was determined by comparing the Pmax, as determined in PFRA, and breakpoint, as determined by PRA, for each individual.

• For those assigned to the expensive groups, half of the difference between Pmax and breakpoint was added to their Pmax value, and rounded up to the nearest whole number.
• For those assigned to cheap groups, half of the difference between Pmax and breakpoint was subtracted from Pmax, and rounded down to the nearest whole number.

Baseline 

• Participants were provided with the same contingency review as in PFRA (for the first five sessions) and target responding was reinforced according to individualized baseline schedules.
• When participants ended early—the full 5 minute duration was still utilized as the denominator when calculating rates because participants had the opportunity to respond for the entire duration and it allows to standardize our unit of analysis while honoring participants requests to be done or leave.

Extinction 

During extinction, programmed consequences (reinforcers) were no longer produced by engaging in the target response.

Prior to starting the extinction session— participants were prompted to respond until they earn a single reinforcer, and then delivered contingency review.

For participants assigned to AO:

• extinction was introduced after appointment-reinforcer consumption fell one below demand or participants wanted to end
• Extinction was introduced toward the end of the appointment
• If the baseline was completed before appointment-reinforcer consumption levels were reached, then one reinforcer was delivered at a time, noncontingently until consumption fell one below demand, or until the participant wanted to stop.

By contrast, for those assigned to EO:

• extinction was introduced at the start of the appointment.

Extinction was carried out until responding decreased to 10% (or less) of the final three baseline sessions, either over a period equivalent to two consecutive sessions (i.e., 10 minutes) or until participant wanted to stop.