Schedules of Reinforcement: A Practitioner's Guide to CRF, FR, VR, FI, VI, Compound Schedules, and Schedule Thinning

What a "schedule of reinforcement" actually controls

A schedule of reinforcement is the contingency that defines which responses produce the reinforcer — by count, by time, by both, by neither, or by some combination Killeen (2023). Killeen's quantitative integration within the Mathematical Principles of Reinforcement framework shows that response rates under different schedules emerge from three core variables — arousal generated by the reinforcer, coupling between target and reinforcer, and competing responses — and that these predict the characteristic patterns each schedule produces, including post-reinforcement pause length, run rate, contrast effects, and momentum Killeen (2023). Practically, the schedule is what makes thinning predictable rather than guesswork: if a clinician knows the schedule the learner is currently on and the one they are moving toward, the literature predicts the pause patterns and persistence the learner will display along the way Killeen (2023). Baum's molar reanalysis sharpens this by showing that when behavior is measured as time-allocation rather than response count, the apparent differences between FR, VR, FI, and VI flatten into a single orderly function relating obtained reinforcement rate to proportion of session time spent responding Baum (2025). The takeaway is not that the four basic schedules are interchangeable — they aren't — but that the practitioner who tracks both response count and time-on-task gets a cleaner read of what the schedule is doing Baum (2025).

Continuous reinforcement (CRF) and skill acquisition

Continuous reinforcement — every correct response produces the reinforcer (FR-1) — maximizes acquisition speed and minimizes the latency between responding and feedback, which is exactly what a learner needs while a behavior is being shaped or differentiated from a competing response (Cividini-Motta et al., 2024). Cividini-Motta and colleagues' PRISMA-guided systematic review compared dense (DR-1, every independent correct response reinforced) versus leaner (DR-3 or no-DR) arrangements and concluded that DR-1 consistently outperformed leaner schedules, accelerating mastery while suppressing prompted responses (Cividini-Motta et al., 2024). CRF also extinguishes faster than any intermittent schedule: the absence of reinforcement is a maximally salient discriminative change, and the learner's experience instantly differentiates "reinforcement is on" from "off" Killeen (2023). The clinical implication: use CRF (or dense DR-1) during acquisition, then move off it deliberately at mastery, because remaining on CRF after mastery means any future schedule slip immediately registers as extinction-like (Cividini-Motta et al., 2024). Jones and colleagues' translational analogue makes the post-mastery side concrete: when a clinician inadvertently omits some 1:1 reinforcer deliveries after criterion, the schedule has shifted to a low-density VR, and that shift makes the now-imperfect responding more persistent — including persistence of any errors that begin getting reinforced (Jones et al., 2026). Post-mastery fidelity is itself a schedule variable (Jones et al., 2026).

Fixed ratio (FR): post-reinforcement pause, run rate, ratio strain

Under a fixed-ratio schedule, every Nth response produces the reinforcer (FR-5: every fifth correct response). The signature pattern is a brief post-reinforcement pause followed by a high, sustained run rate until the next reinforcer, with pause duration scaling with ratio size and reinforcer magnitude Killeen (2023). Two practical risks define FR programming. Ratio strain: thinning too aggressively (FR-5 to FR-25 in one step) often produces extinction-like response collapse because the learner does not have a reinforcement history dense enough to bridge the new requirement Killeen (2023) (Frank-Crawford et al., 2024). Pre-ratio pause: extended hesitation before initiating the run that emerges with very large fixed ratios and signals the schedule is now functioning as a discriminated effort barrier rather than a productive contingency Killeen (2023). Frank-Crawford and colleagues' single-subject delay-based thinning case in a 5-year-old with tangibly maintained elopement illustrates the safer architecture: thin in small increments (30 s → 60 s → 90 s), require multiple consecutive successful sessions before each step, and back up to the last successful interval whenever problem behavior resurges (Frank-Crawford et al., 2024). The same logic transfers cleanly to FR thinning in skill or output programs (Frank-Crawford et al., 2024).

Variable ratio (VR): high steady rate, resistance to extinction

Variable ratio schedules deliver the reinforcer after an unpredictable number of responses averaged around N (VR-5: average of every fifth response, distribution across roughly 1–9). Behavior under VR is famously high-rate, steady, and resistant to extinction — properties that make VR the default schedule whenever the goal is sustained output rather than discrimination of a specific moment Killeen (2023). The unpredictability eliminates the discriminative cue that signals "extinction is on" under CRF, so when reinforcement is later removed or thinned, behavior persists for substantially longer Killeen (2023). Two corpus findings sharpen the picture. Johnson and colleagues compared FR-5 versus VR-5 training in mice and found that on a subsequent progressive-ratio breakpoint test neither schedule produced higher motivation than the other Johnson et al. (2022). The implication is not that VR is useless; the case for VR over FR has to rest on response-rate and persistence properties rather than on a vague claim that VR raises engagement Johnson et al. (2022). Regnier and colleagues' contingency-management work compared VR-2 versus VR-6 voucher schedules for cocaine abstinence and found equivalent outcomes, meaning leaner VRs can capture most of the behavioral effect of richer ones at lower programmatic cost Regnier et al. (2022). VR is also the schedule that most reliably produces resurgence when withdrawn; Fisher and colleagues showed across four children that dense baseline reinforcement (FR-1 or VI 2-s) for problem behavior produced larger resurgence after FCT than leaner baselines, so treatment over an FR-1 or short-VI baseline needs upfront relapse-mitigation Fisher et al. (2019).

Fixed interval (FI): the scallop and why it's clinically rare

Under a fixed-interval schedule, the first response after a fixed amount of time (e.g., FI-30 s) produces the reinforcer; responses during the interval do nothing. The classic pattern is the scallop: a long post-reinforcement pause, then accelerating responding as the interval ends, peaking right before the reinforcer becomes available Killeen (2023). FI dominates laboratory work but is rare in pure form in clinical practice because it produces neither the high run rate of ratio schedules nor the steady moderate output of VI Killeen (2023). Where FI matters most for practitioners is diagnostic: when a learner's responding shows a scalloped pattern under what was supposed to be a different schedule, the schedule has slipped toward FI-like control, typically through inconsistent timing of consequence delivery Killeen (2023). Pause patterns are predictable from schedule parameters; observing a scallop where you didn't program one signals that consequence-delivery timing has become the controlling variable Killeen (2023).

Variable interval (VI): moderate steady rate, common in OBM and maintenance

Variable interval schedules deliver the first response after an unpredictable elapsed time averaged around N (VI-30 s: average 30 seconds, distribution across roughly 5–60 s). The signature pattern is a moderate, steady response rate without the pauses of FR/FI and without the very high run rate of VR — exactly the profile most useful for sustained engagement, maintenance, and OBM applications where output volume matters less than consistent presence Killeen (2023). Baum's molar reanalysis showed that VI and VR data points fall on the same orderly time-allocation function, suggesting that in time-on-task terms the practical difference between VI and VR is smaller than the response-rate literature implies Baum (2025). Several applied studies rest on VI-style parameters Diaz de Villegas et al. (2020). Diaz de Villegas and colleagues used synchronous (continuous, duration-yoked) reinforcement to boost on-task behavior in preschoolers and found that synchronous delivery — effectively a VI-zero/duration-matched arrangement — outperformed accumulated-token alternatives in both on-task percentage and child preference Diaz de Villegas et al. (2020) Diaz de Villegas et al. (2024). Gomes and colleagues extended the synchronous-schedule logic to a fourth-grade group contingency: synchronizing the reinforcer to simultaneous on-task behavior across the class increased rule-following while merely delivering the same incentive non-contingently produced no change Gomes et al. (2025). Baruni and colleagues' five-component mixed schedule using high-preference music reliably controlled treadmill-walking speed in 72% of 25 college participants Baruni et al. (2025). McLean and colleagues' rapidly-changing concurrent-VI work in rats adds an important caveat: within-session matching to the prevailing VI ratio is fast, but across-session sensitivity weakens with continued unsignaled change — frequent unannounced VI shifts degrade rather than sharpen schedule control McLean et al. (2018).

Compound schedules: chained, tandem, mixed, multiple, concurrent

Past single schedules, the four useful compound forms are chained, tandem, mixed, and multiple, plus the concurrent arrangement that runs two schedules simultaneously Saini et al. (2016). A chained schedule strings two or more components in a fixed order, each signaled by a distinct SD, with the terminal reinforcer after the final component. A tandem schedule is identical except components are not signaled. A mixed schedule alternates components without signals (unsignaled multiple). A multiple schedule alternates components each signaled by its own SD, so the learner can discriminate which component is in effect Saini et al. (2016).

The applied literature on multiple schedules is dense and high-utility. Saini, Miller, and Fisher's narrative review of 147 practical multiple-schedule applications — with 52 analyzed in depth — found that multiple schedules consistently produced stimulus control allowing dense FCT reinforcement in one component while thinning or withholding reinforcement in the other, maintaining low problem behavior without returning to baseline Saini et al. (2016). Two tactics emerged as especially effective: (1) start with a long reinforcement (SD) component and brief extinction (S∆) period, then gradually lengthen S∆; and (2) pair S∆ with competing stimuli that provide alternate reinforcement, which speeds discrimination and suppresses resurgence Saini et al. (2016).

Two single-subject studies operationalize chained-schedule logic in clinical practice Torres‐Viso et al. (2018). Torres-Viso and colleagues used a chained 60-second SD/S∆ schedule with three adolescents whose destructive behavior was reinforced by adult compliance with mands for environmental rearrangement; alternating reinforcement and extinction components within sessions reduced destructive behavior while maintaining mand use Torres‐Viso et al. (2018). Owen and colleagues extended the same architecture to thin caregiver compliance from continuous to VR-24 through VR-120 (or 240-second S∆ periods) in five children with severe destructive behavior, keeping destructive behavior near zero by signaling each schedule transition and lengthening S∆ duration only after stable performance Owen et al. (2020).

Concurrent schedules present two response options simultaneously, each on its own schedule; the matching law describes how organisms allocate behavior in proportion to the obtained reinforcement rates McLean et al. (2018). Concurrent VI–EXT and concurrent VI–VI arrangements are foundational to FCT: the alternative communicative response is on a rich VI (or FR-1) while the problem behavior moves to extinction or a much leaner schedule, and the learner allocates behavior toward the richer alternative Saini et al. (2016) (Ruiz Méndez, 2024). Ruiz Méndez's preliminary undergraduate work using concurrent VI–EXT multiple schedules confirmed that participants quickly allocated responding toward the VI alternative associated with the higher reinforcement rate, and that this preference persisted when competing rules were later in effect (Ruiz Méndez, 2024).

Differential reinforcement schedules (DRO/DRA/DRI/DRL/DRH)

Differential reinforcement procedures — DRO, DRA, DRI, DRL, DRH — are formally schedules in their own right: each defines a rule for which responses (or which absences) contact the reinforcer. The full taxonomic treatment lives on the Differential Reinforcement page (link pending). For a schedules-of-reinforcement page, the operational point is that DR procedures are reinforcement schedules layered onto specific response classes (zero occurrences for DRO, alternative responses for DRA/DRI, low rates for DRL, high rates for DRH) and obey the same acquisition, persistence, and resurgence dynamics as any other schedule (Cividini-Motta et al., 2024) Fisher et al. (2019) Nevin et al. (2016). Brown and colleagues' DRA-with-versus-without-extinction comparison is instructive: keeping the same reinforcement schedule for problem behavior during DRA produced substantially less resurgence when reinforcement was later withdrawn than DRA-with-extinction did, isolating the schedule rather than the contingency as the active variable Brown et al. (2020).

Schedule thinning: from CRF or rich to leaner schedules

Schedule thinning is where most clinical schedules-of-reinforcement decisions actually happen. The corpus pattern is consistent: thin gradually, document each step, build in stimulus control via discriminative signals, and back up the moment problem behavior resurges Saini et al. (2016) (Frank-Crawford et al., 2024) Owen et al. (2020). Phillips and colleagues' retrospective review of 27 NCR applications reported systematic thinning was achieved in 7 of 27 cases, with the remaining 20 retaining dense NCR — only about a quarter of cases reach substantial leanness even in specialized settings Phillips et al. (2017). Frank-Crawford and colleagues' delay-based thinning case (30 s → 150 s in graduated steps with competing stimuli) added two rules: (1) multiple consecutive successful sessions before each delay increase, and (2) return to the last successful delay and advance in smaller increments when problem behavior emerges (Frank-Crawford et al., 2024). Owen and colleagues' multiple-schedule caregiver-compliance thinning (continuous → VR 24–120) added a third: discriminative stimuli that explicitly signal SD versus S∆ let the learner map the schedule transition, materially reducing resurgence Owen et al. (2020). Shahan and colleagues' rat-model work identified the converse failure mode: abrupt downshifts in alternative-reinforcement rate produce the strongest resurgence, so step size matters as much as the endpoint Shahan et al. (2020).

Behavioral momentum and persistence

Behavioral momentum theory predicts that the resistance of a behavior to disruption — extinction, distraction, schedule thinning — is a function of the reinforcement rate historically associated with the discriminative context in which the behavior occurs Killeen (2023) Lattal & Miles (2024). Killeen's MPR synthesis treats momentum as a quantitatively predictable property of schedule history: higher reinforcement rates create higher response strength, manifesting as greater persistence under disruption Killeen (2023). Clinically, this creates a tension: dense schedules build momentum that helps the target behavior persist when conditions change, but they also build momentum for problem behavior when problem behavior is what's being densely reinforced — exactly Fisher and colleagues' finding that dense (FR-1 or VI 2-s) baseline reinforcement of destructive behavior produced the largest post-treatment resurgence Fisher et al. (2019). Lattal and Miles argue that resurgence and behavioral contrast are two faces of the same schedule-alternation phenomenon, and that the size of the reinforcement-rate change between phases — not just the contingency change — drives the magnitude of both effects Lattal & Miles (2024). The practical translation: thinning should preserve enough reinforcement density in the alternative response to maintain its momentum, while making the schedule shift small enough that the learner doesn't experience it as a contrast event Lattal & Miles (2024) Shahan et al. (2020).

Resistance to extinction across schedules

Resistance to extinction follows a well-established ordering: CRF extinguishes fastest, FR slower, VR slowest, with interval schedules typically falling in between based on rate parameters Killeen (2023). The corpus complicates this in practice. King and colleagues' systematic review of resurgence across seven experiments and five schedule types (RR, VR, RT, VT, continuous) found that response-independent time-based schedules (RT/VT) delivered during the relapse-test phase reliably produced less resurgence than continued ratio schedules — meaning the schedule in effect during the extinction or thinning phase is itself a predictor of later relapse King et al. (2025). Brown and colleagues' DRA-with-versus-without-extinction comparison reinforced this: maintained-rate (no extinction) produced substantially less resurgence than full extinction Brown et al. (2020). King and colleagues' human-operant follow-up demonstrated that the discriminative properties of the reinforcer itself modulate resurgence size after extinction, implicating stimulus-control variables as legitimate schedule-thinning levers King et al. (2025). Nevin and colleagues' signaled-versus-unsignaled DRA work across children and pigeons showed that higher-rate alternative-reinforcement schedules suppressed problem behavior faster but produced larger resurgence when reinforcement ceased Nevin et al. (2016). Craig and Shahan's translational ethanol work added a useful warning: pairing alternative non-drug reinforcement with the same discriminative stimuli that maintained the drug-seeking response increased rather than decreased extinction-resistance, so DRA arrangements need to keep the alternative schedule discriminably separated from the target schedule's stimulus context Craig & Shahan (2022).

Progressive-ratio schedules as assessment tools

Progressive-ratio (PR) schedules — where the response requirement increases systematically across deliveries (e.g., FR-1, FR-3, FR-5, FR-9, FR-15…) until responding ceases at the breakpoint — function as practical assessments of reinforcer value rather than as treatment schedules in their own right Wilson & Gratz (2016) (Lambert et al., 2026). Wilson and Gratz's case study used a PR assessment to set the initial treatment reinforcement schedule for a functional communicative response and again post-treatment to demonstrate that the breakpoint had escalated with treatment Wilson & Gratz (2016). Lambert and colleagues' three-participant analysis showed that PR breakpoint data tracked behavioral-economic demand intensity (Q0) and elasticity, supporting PR as a proxy for full economic-demand analyses when those are operationally impractical (Lambert et al., 2026). Russell and colleagues' alternating-treatments comparison of tokens, edibles, and leisure items on PR across three children with ASD/DD found that tokens maintained higher breakpoints than edibles or leisure items in two of three children — even with presession edible satiation — supporting tokens as potent generalized conditioned reinforcers Russell et al. (2018). The operational role of PR is the same in each study: use it briefly to set or verify a treatment schedule, not as the treatment schedule itself (Lambert et al., 2026).

Time-based (NCR) and synchronous schedules

Two non-traditional families deserve specific attention because they're increasingly common in applied work. Fixed-time (FT) and variable-time (VT) schedules deliver the reinforcer on a temporal rule independent of any response — the foundation of NCR procedures Phillips et al. (2017). López-Tolsa and Pellón's rat work showed that under FT food schedules, schedule-induced drinking emerges at the end of the inter-food interval and can become an operant if drinking is later required for food delivery — illustrating that FT/VT schedules can produce adventitious operant responses if the practitioner isn't watching López-Tolsa & Pellón (2025). The applied implication for NCR is concrete: monitor for problem behavior that begins occurring near reinforcer delivery, because that proximity can convert FT-delivered reinforcement into adventitious reinforcement of the maladaptive response López-Tolsa & Pellón (2025). Phillips and colleagues' 27-case NCR series confirms this risk is worth managing — NCR thinning is feasible in roughly a quarter of cases but requires explicit FT-interval mastery criteria and supplementary procedures Phillips et al. (2017). Martínez-Herrada and colleagues added that the temporal distribution of schedule-induced behavior under FT depends on the reinforcer's hedonic value, with less-preferred reinforcers producing more sustained adjunctive behavior Martínez‐Herrada et al. (2025).

Synchronous reinforcement schedules deliver the reinforcer continuously while the target response is occurring, so reinforcement duration covaries with behavior duration Diaz de Villegas et al. (2020). Diaz de Villegas and colleagues' two preschool studies demonstrated that synchronous schedules outperformed yoked accumulated-token schedules for on-task behavior and were generally preferred by participants Diaz de Villegas et al. (2020) Diaz de Villegas et al. (2024). Gomes and colleagues' fourth-grade replication showed that the synchrony itself is the active variable: the same incentive delivered non-contingently produced no change Gomes et al. (2025). Baruni and colleagues' treadmill study extended the schedule to adult exercise contexts Baruni et al. (2025). Synchronous schedules sit between CRF and continuous-VI — particularly useful when the goal is sustained engagement measured in time rather than discrete responses Diaz de Villegas et al. (2020).

Schedules of social reinforcement and in functional analysis

Most BCBA work involves social, not automatic, reinforcers, and the schedule literature has begun to model this explicitly. Benvenuti and colleagues' theoretical paper recasts social interaction as occurring on interdependent schedules, where consequences each person receives depend partly on others' behavior, and shows that ratio size, DRL interval, and response-rate requirements quantitatively predict when cooperation versus competition emerges Benvenuti et al. (2024). The translation for group contingencies and parent-training programs: the schedule is the social structure — small ratios and short DRL windows raise cooperation, while large ratios and long delays create choice conflicts that lower social responding Benvenuti et al. (2024). Frampton and colleagues underscore a complementary point: the reinforcer's momentary value (which the schedule samples from) is manipulable through brief deprivation, so the schedule and the EO interact rather than functioning independently (Frampton et al., 2024). Gallistel's reanalysis of Herrnstein-era concurrent-VI data argues that organisms' allocation of behavior across schedules is best explained by rate-based computations — practitioners should provide clear rate signals if they want behavior to track the intended schedule Gallistel (2025). Schedule logic also shows up in modern functional analysis: Jessel and colleagues' performance-based IISCA validation used systematically alternating reinforcer-present and reinforcer-absent intervals to demonstrate schedule control of problem behavior, with brief well-designed extinction segments producing diagnostic schedule effects without prolonged deprivation (Jessel et al., 2024). Mattson and colleagues' preschool cooperative-vocal-exchange study illustrates the classroom analogue: a postsession contingent group reinforcer paired with an activity schedule maintained schedule-following after the visual schedule was faded, demonstrating that schedule-plus-contingent-reinforcement packages outperform either component alone (Mattson et al., 2024).

A foundation page should be honest about where the evidence comes from Killeen (2023). The schedule-of-reinforcement literature spans theoretical synthesis, SCED, narrative and systematic reviews of applied procedures, and a smaller band of case-series and translational work Saini et al. (2016).

Theoretical and quantitative integration. Killeen's Mathematical Principles of Reinforcement synthesis predicts response rates, post-reinforcement pauses, contrast, momentum, and progressive-ratio behavior from three core variables and provides the quantitative scaffolding most practitioners use implicitly when reasoning about thinning Killeen (2023). Baum's molar-time-allocation reanalysis offers a complementary lens that flattens FR/VR/FI/VI distinctions when behavior is measured in time rather than counts; Gallistel's rate-computation reframing argues that organisms compute and respond to scalar reinforcement rates rather than associative strengths Baum (2025) Gallistel (2025). Lattal and Miles read multiple-schedule alternation as the active variable behind both resurgence and contrast Lattal & Miles (2024). Benvenuti and colleagues extend schedule theory to social interaction via interdependent schedules Benvenuti et al. (2024).

Systematic and narrative reviews. Cividini-Motta and colleagues' PRISMA-guided review anchors the CRF/DR-1 evidence for acquisition (Cividini-Motta et al., 2024). Saini, Miller, and Fisher's review of 147 multiple-schedule applications — with 52 analyzed in depth — anchors the FCT-thinning literature Saini et al. (2016). King and colleagues' systematic review of seven resurgence experiments across five schedule types anchors the resurgence-by-schedule findings King et al. (2025).

Single-subject experimental designs (the bulk of applied evidence). Fisher and colleagues on baseline-rate effects on resurgence (n=4); Brown and colleagues on DRA with/without extinction (n=3); Owen and colleagues on multiple-schedule caregiver-compliance thinning (n=5); Torres-Viso and colleagues on chained-schedule mand treatment (n=3); Frank-Crawford and colleagues on delay-based thinning (n=1); Wilson and Gratz on PR-as-assessment (n=1); Lambert and colleagues on PR-and-demand (n=3); Russell and colleagues on tokens-as-generalized-reinforcers under PR (n=3); Diaz de Villegas and colleagues on synchronous reinforcement (n=3, then n=7); Gomes and colleagues on synchronous group contingencies (n=45 across three classes); Baruni and colleagues on synchronous treadmill schedules (n=25); Jessel and colleagues on IISCA schedule-controlled responding (n=5); Mattson and colleagues on activity schedules plus contingent reinforcement (n=6); Nevin and colleagues on signaled-vs-unsignaled DRA (n=4 children); Phillips and colleagues' NCR case-series (n=27); Regnier and colleagues' VR-2 vs VR-6 voucher comparison (n=6) Fisher et al. (2019) Brown et al. (2020) Owen et al. (2020) Torres‐Viso et al. (2018) (Frank-Crawford et al., 2024) Wilson & Gratz (2016) (Lambert et al., 2026) Russell et al. (2018) Diaz de Villegas et al. (2020) Diaz de Villegas et al. (2024) Gomes et al. (2025) Baruni et al. (2025) (Jessel et al., 2024) (Mattson et al., 2024) Nevin et al. (2016) Phillips et al. (2017) Regnier et al. (2022).

Translational and basic-science work. Several findings rest on well-controlled non-human or human-operant analogues: Jones and colleagues on post-mastery fidelity drift (n=4 adults); McLean and colleagues on within- vs across-session VI sensitivity (n=8 rats); Johnson and colleagues on FR-vs-VR experience effects (n=20 mice); Shahan and colleagues on resurgence after alternative-rate downshifts (n=24 rats); López-Tolsa and Pellón on schedule-induced drinking under FT (n=12 rats); Martínez-Herrada and colleagues on hedonic value × FT (n=8 rats); Craig and Shahan on alternative-reinforcement and ethanol behavior (n=7 rats); King and colleagues' human-operant resurgence (n=3); Ruiz Méndez's concurrent VI–EXT undergraduate work (n=8) (Jones et al., 2026) McLean et al. (2018) Johnson et al. (2022) Shahan et al. (2020) López-Tolsa & Pellón (2025) Martínez‐Herrada et al. (2025) Craig & Shahan (2022) King et al. (2025) (Ruiz Méndez, 2024). Translational analogues are mechanistic evidence rather than direct treatment recommendations, but Jones on post-mastery fidelity and Shahan on rate-downshift resurgence translate cleanly to clinical thinning (Jones et al., 2026) Shahan et al. (2020).

Bottom line. The convergent picture is strong for the operational claims this page makes: CRF/DR-1 maximizes acquisition, VR produces high steady output and slow extinction, multiple schedules with discriminative stimuli are the workhorse for FCT thinning, gradual rate reductions outperform abrupt downshifts, and baseline schedule density predicts post-treatment resurgence (Cividini-Motta et al., 2024) Saini et al. (2016) Fisher et al. (2019) Shahan et al. (2020) Owen et al. (2020). It is weaker for specific claims about exact thinning step sizes, optimal VR/VI parameters for a given population, or head-to-head superiority of one compound schedule over another in matched patient samples Regnier et al. (2022) Phillips et al. (2017).

Schedule decisions a senior practitioner makes are not "ratio or interval" so much as "which schedule, at what density, with which stimulus control, for the goal of this phase." A defensible logic, drawn from the corpus:

1. New skill being acquired. Use CRF (FR-1) or dense DR-1 for independent correct responses; withhold reinforcement for prompted responses (Cividini-Motta et al., 2024). Move off CRF at mastery — staying past mastery means any future fidelity slip registers as extinction-like Killeen (2023).
2. Mastered skill needing high output rate. Move to VR — high steady rate, slow extinction. Be explicit about average ratio and upper bound to avoid ratio strain Killeen (2023). Reasonable start: VR-3 to VR-5, then graduated steps based on response-rate stability Regnier et al. (2022).
3. Mastered skill needing sustained engagement rather than rate. Move to VI for moderate steady output, or a synchronous schedule when the target behavior is measured in duration (on-task, treadmill speed, group rule-following) Baum (2025) Diaz de Villegas et al. (2020) Gomes et al. (2025).
4. Sequential behavior chain. Use a chained schedule where each component is signaled and the terminal reinforcer follows the final component; tandem (unsignaled) chains work but provide weaker discriminative control during fading Saini et al. (2016) Torres‐Viso et al. (2018).
5. FCT or DRA thinning. Use a multiple schedule with explicit SD/S∆ stimuli; start with long SD and brief S∆, then gradually lengthen S∆; pair S∆ with competing stimuli; require multiple consecutive successful sessions per step; back up after any resurgence Saini et al. (2016) Owen et al. (2020) (Frank-Crawford et al., 2024) Shahan et al. (2020).
6. DRA over a baseline of dense problem-behavior reinforcement. Plan for resurgence upfront — dense baseline (FR-1, very short VI) for problem behavior produces materially larger resurgence than leaner baselines Fisher et al. (2019). Consider DRA-without-extinction (keeping the baseline schedule for problem behavior) during early phases when extinction risk is high; this trades faster suppression for later durability Brown et al. (2020).
7. NCR (FT/VT) thinning. Document baseline FT, set explicit mastery criteria per interval increase, monitor for adventitious operant responses near reinforcer delivery, and combine with supplementary procedures — only about a quarter of NCR cases reach substantial leanness without supplementation Phillips et al. (2017) López-Tolsa & Pellón (2025).
8. OBM and contingency-management. Default to VR for output-rate goals; lean schedules (VR-6 vs VR-2) can preserve treatment effects at lower cost, but parameters need empirical verification Regnier et al. (2022). For sustained-engagement metrics, VI and synchronous schedules outperform ratio arrangements Baum (2025) Baruni et al. (2025).
9. Need to assess reinforcer value before setting a schedule. Run a brief PR assessment to identify the breakpoint, set the initial schedule below it, and re-run PR post-treatment to demonstrate escalation Wilson & Gratz (2016) (Lambert et al., 2026).
10. Multiply controlled or context-discriminated behavior. Use multiple schedules with explicit discriminative stimuli to teach component discrimination; concurrent schedules let the matching law work in your favor when shaping allocation Saini et al. (2016) McLean et al. (2018) (Ruiz Méndez, 2024).
11. Group or parent-training program. Treat the schedule as the social structure — small ratios and short DRL windows raise cooperation; interdependent group contingencies require synchronous delivery to function Benvenuti et al. (2024) Gomes et al. (2025). Token systems operate as compound schedules (token earned on one schedule, exchange on another) and require explicit accounting of both Russell et al. (2018).

Clinic skill acquisition

Outpatient and university clinics are where the CRF→thin acquisition pattern lives most cleanly. The typical sequence: dense DR-1 differential reinforcement during acquisition, criterion-based transition to VR-3 or VR-5 after mastery, then graduated thinning with explicit step criteria (Cividini-Motta et al., 2024). Clinics are where progressive-ratio assessment is most operationally feasible — use PR briefly at the start to set an empirically-grounded initial schedule and again at the end to demonstrate treatment effect Wilson & Gratz (2016) (Lambert et al., 2026). The risk specific to this setting is post-mastery fidelity drift: Jones and colleagues showed that intermittent omissions of 1:1 reinforcer delivery after mastery shift the schedule to a low-density VR and make the now-imperfect responding more — not less — persistent, including persistence of any errors that begin getting reinforced (Jones et al., 2026). Consequence-fidelity probes belong in the maintenance phase, not just acquisition (Jones et al., 2026).

Schools (K–12)

In schools, the dominant schedules are VI-style for sustained engagement (on-task, attending, rule-following) and multiple-schedule arrangements for behavior-reduction work tied to FCT Saini et al. (2016) Diaz de Villegas et al. (2020). Synchronous schedules — reinforcer access yoked to the duration of target behavior — fit classroom on-task and group-contingency work, with preschool and elementary demonstrations showing synchronous delivery outperforms accumulated-token arrangements Diaz de Villegas et al. (2020) Diaz de Villegas et al. (2024). Gomes and colleagues' fourth-grade group-contingency study established that the synchrony itself is the active variable: the same incentive delivered non-contingently produced no behavior change Gomes et al. (2025). For school-based FCT and DRA, the multiple-schedule architecture with explicit SD/S∆ cues — colored cards, schedule boards, contingency-specifying rules — has the strongest empirical support Saini et al. (2016). Mattson and colleagues showed that activity schedules paired with postsession contingent group reinforcement maintained schedule-following after the visual schedule was faded (Mattson et al., 2024).

Home and parent training

Home-based schedules typically end up as compound arrangements even when no one writes them down: a token earned on one schedule, exchanged on another; an FCR reinforced on caregiver compliance, itself on a thinning schedule Russell et al. (2018). Frampton and colleagues' tutorial argues for brief reinforcer-deprivation periods (5–15 minutes) that raise the value of the reinforcer the schedule is sampling from — a complement to schedule density rather than a substitute (Frampton et al., 2024). Owen and colleagues' multiple-schedule caregiver-compliance thinning provides the operational template for fading from continuous to VR-24 through VR-120 (or 240-second S∆ periods) without escalating destructive behavior, using discriminative stimuli to signal each transition Owen et al. (2020). Frank-Crawford and colleagues' delay-based thinning gives the small-step rule for transitions: 30 s → 60 s → 90 s with three elopement-free sessions before each step, returning to the last successful delay if problem behavior resurges (Frank-Crawford et al., 2024). Token systems implicitly use multiple schedules and need both schedules made explicit in the parent-training plan Russell et al. (2018).

OBM and adult performance

OBM leans heavily on VR for output-rate goals, with the corpus supporting leaner-than-typical VRs for many applied contexts Regnier et al. (2022). Regnier and colleagues' contingency-management work with adults in cocaine treatment found VR-2 vs VR-6 voucher schedules produced equivalent abstinence outcomes, meaning programs can often be made leaner at lower cost without sacrificing effect Regnier et al. (2022). Baruni and colleagues' five-component synchronous-reinforcement treadmill study with 25 college participants demonstrated that high-preference music functioned as an effective reinforcer to control walking speed, with 72% of participants showing schedule-appropriate responding Baruni et al. (2025). For OBM applications where the goal is sustained engagement rather than discrete output, VI and synchronous schedules outperform ratio arrangements, and the molar-time-allocation framework gives a defensible measurement structure Baum (2025). Johnson and colleagues' FR-vs-VR comparison is a useful caution: switching to VR purely to "boost motivation" may not produce additive effects on breakpoint or output Johnson et al. (2022).

Residential and severe-behavior units

Residential and inpatient settings concentrate the highest-stakes schedule decisions: severe destructive behavior, dense baseline reinforcement, NCR thinning that doesn't always succeed, and multiple-schedule FCT under variable staff coverage Phillips et al. (2017). Phillips and colleagues' 27-case NCR retrospective found that systematic schedule thinning was successfully achieved in 7 of 27 cases by lengthening the FT interval or increasing demands between non-contingent escape periods, while the remaining 20 retained dense NCR — a sobering benchmark even for specialized settings Phillips et al. (2017). Owen and colleagues' multiple-schedule caregiver-compliance thinning shows VR-24 to VR-120 (or 240-second S∆) schedules can be reached when transitions are signaled and steps are gradual Owen et al. (2020). Fisher and colleagues' baseline-rate work establishes the upfront planning rule: dense baseline reinforcement (FR-1 or VI 2-s) for problem behavior predicts larger post-treatment resurgence, so cases coming from dense baselines need explicit relapse-mitigation — leaner baselines if possible, thicker FCR schedules if not, and gradual rather than abrupt thinning at every step Fisher et al. (2019).

• Confusing CRF with reinforcement-on-everything. CRF is FR-1 — every correct response of the target class produces the reinforcer. It is not "reinforcement is generally available." Mislabeling a dense but inconsistent schedule as CRF means thinning starts from the wrong baseline and post-mastery fidelity drift goes undiagnosed (Cividini-Motta et al., 2024) (Jones et al., 2026).
• Ratio strain from too-fast thinning. Jumping FR-5 to FR-25 or VI-30s to VI-120s in a single step typically produces extinction-like collapse rather than continued steady output. Use graduated steps with explicit mastery criteria, and back up after any failure (Frank-Crawford et al., 2024) Shahan et al. (2020).
• Chains of forgotten schedules. Token systems and compound contingencies often embed two or three schedules — earning the token, exchanging it, caregiver delivery — and BIPs often document only one. When the program drifts, no one can identify which schedule slipped Saini et al. (2016) Russell et al. (2018).
• Undocumented schedule shifts during fading. Schedules slip silently when staff change or sessions get rushed. Without written parameters and per-step mastery criteria, the team can't detect that the operating schedule is no longer what the BIP says (Jones et al., 2026) Phillips et al. (2017).
• Treating CRF + extinction as a "schedule shift." CRF interrupted by extinction is not a schedule change; it is a contingency reversal. Calling it "moving to a leaner schedule" hides the actual reversal from the supervision record Killeen (2023).
• Assuming VR raises engagement on its own. Johnson and colleagues' FR-5 vs VR-5 comparison produced equal subsequent PR breakpoints — VR did not boost motivation over FR in that preparation Johnson et al. (2022). The case for VR rests on response-rate and persistence properties, not a global "VR = more motivation" claim Johnson et al. (2022).
• Pairing alternative-reinforcement schedules with the target's stimulus context. Craig and Shahan's translational ethanol work shows that delivering rich alternative reinforcement under the same discriminative stimuli that maintained the target response increases rather than decreases extinction-resistance and reinstatement Craig & Shahan (2022). Alternative-reinforcement schedules need to be discriminably separated from the target-behavior context Craig & Shahan (2022).
• Ignoring adventitious operant responses under FT/VT. López-Tolsa and Pellón showed that schedule-induced behaviors emerging near reinforcer delivery under FT can become reinforced operants if not monitored López-Tolsa & Pellón (2025). Plan NCR with explicit observation of behavior occurring in the seconds before each FT delivery López-Tolsa & Pellón (2025).
• Assuming a schedule that built behavior in clinic will hold under different ambient schedules elsewhere. McLean and colleagues showed that across-session sensitivity to changing concurrent schedules weakens with continued unsignaled change — when generalization fails, the underlying schedules in the new setting are often the variable McLean et al. (2018).

• Severe destructive behavior on dense baseline reinforcement. Cases coming in with FR-1 or very short VI baselines for the problem behavior carry materially elevated post-treatment resurgence risk. If your setting cannot support multiple-schedule thinning with explicit SD/S∆ cues and graduated steps, refer to a specialized behavior unit Fisher et al. (2019) Owen et al. (2020).
• NCR cases that don't thin. Roughly three-quarters of NCR applications in a specialized inpatient series did not reach substantial schedule leanness without supplementation. If your setting has tried two thinning cycles without progress, refer for consultation rather than continuing dense NCR indefinitely Phillips et al. (2017).
• Cases requiring formal economic-demand assessment beyond progressive-ratio breakpoints. When PR breakpoints don't map cleanly to demand intensity or elasticity for the case, refer for a setting with capacity for full closed-economy preparations (Lambert et al., 2026).
• Substance-treatment programs where the alternative-reinforcement schedule may share stimulus context with the target behavior. Craig and Shahan's findings suggest specific risk in substance-use treatment when alternative non-drug schedules occupy the same stimulus context as drug-seeking; this is specialist territory Craig & Shahan (2022).
• Cases where staff training cannot reach reliable consequence-fidelity probes during the maintenance phase. Jones and colleagues' translational analogue is direct: post-mastery omissions effectively shift CRF to low-density VR and make errors more persistent. If fidelity probes can't be staffed, the case needs a setting that can (Jones et al., 2026).

The convergent picture across schedule research is strong for the operational claims this page makes: CRF/DR-1 maximizes acquisition speed, VR produces high steady-state output and slow extinction, multiple schedules with explicit discriminative stimuli are the workhorse for FCT thinning, gradual rate reductions outperform abrupt downshifts, and baseline schedule density predicts post-treatment resurgence (Cividini-Motta et al., 2024) Killeen (2023) Saini et al. (2016) Fisher et al. (2019). The literature is weaker for specific questions practitioners answer every day. There is no large head-to-head trial of optimal step-size for VR or VI thinning across matched populations; most thinning evidence is SCED with case-specific step rules (Frank-Crawford et al., 2024) Owen et al. (2020). Regnier and colleagues' VR-2-vs-VR-6 contingency-management comparison suggests leaner schedules can preserve treatment effect in some contexts, but cross-population replication is needed Regnier et al. (2022). The synchronous-reinforcement literature is promising but built on small-N preschool and college samples; pediatric clinical and adult-disability replications are the obvious next step Diaz de Villegas et al. (2020) Diaz de Villegas et al. (2024) Gomes et al. (2025). Interdependent-schedule modeling of social behavior offers quantitative predictions for cooperation and competition, but empirical tests of those predictions in dyads and groups are still few Benvenuti et al. (2024). The post-mastery fidelity-drift finding needs longitudinal field replication that tracks consequence delivery across months in actual programs (Jones et al., 2026).

1. Track the schedule explicitly in every BIP and skill program. Document the exact schedule (FR-1, VR-3, VI-30s, multiple 60s SD/60s S∆) the learner is on now and the schedule the next phase moves to. "Reinforcement provided" is not a schedule (Cividini-Motta et al., 2024) Saini et al. (2016).
2. Use CRF (or dense DR-1) during acquisition, then move off deliberately at mastery. Staying on CRF post-mastery means any future fidelity slip registers as extinction-like and degrades responding (Cividini-Motta et al., 2024) (Jones et al., 2026).
3. Default to VR for output-rate goals, VI for sustained engagement. VR produces high steady rate and slow extinction; VI produces moderate steady output without ratio-schedule pauses Killeen (2023) Baum (2025).
4. Use multiple schedules with explicit SD/S∆ stimuli for FCT and DRA thinning. Long SD, brief S∆ at the start, then gradual S∆ lengthening; pair S∆ with competing stimuli to suppress resurgence Saini et al. (2016) Owen et al. (2020).
5. Thin in small steps with mastery criteria; back up after any failure. Frank-Crawford and Owen support 30%–50% increments at most, three consecutive successful sessions per step, and immediate return to the last successful schedule if problem behavior resurges (Frank-Crawford et al., 2024) Owen et al. (2020) Shahan et al. (2020).
6. Plan for resurgence based on baseline schedule density. Dense (FR-1 or short-VI) baselines predict larger post-treatment resurgence; build in relapse-mitigation upfront Fisher et al. (2019) Nevin et al. (2016).
7. Run a brief PR assessment to set initial schedules empirically. PR breakpoints map onto reinforcer value and demand intensity well enough to ground starting parameters Wilson & Gratz (2016) (Lambert et al., 2026).
8. Build consequence-fidelity probes into the maintenance phase. Post-mastery omissions shift CRF to low-density VR and make errors more, not less, persistent — antecedent-only fidelity checks miss this (Jones et al., 2026).
9. Use synchronous schedules when the target behavior is measured in duration. On-task, group rule-following, treadmill speed — synchronous delivery outperforms accumulated-token alternatives and is generally preferred Diaz de Villegas et al. (2020) Diaz de Villegas et al. (2024) Gomes et al. (2025).
10. For chained behaviors, signal each component. Signaled chains outperform tandem chains during fading; the SD provides the transition cue the learner uses to organize the sequence Saini et al. (2016) Torres‐Viso et al. (2018).
11. Document both schedules in token systems. Earning operates on one schedule; exchange operates on another. BIPs that document only one leave the program vulnerable to silent drift Russell et al. (2018).
12. Keep alternative-reinforcement schedules discriminably separated from the target's stimulus context. Pairing rich alternative reinforcement with the same stimuli that maintained the target can paradoxically increase target persistence Craig & Shahan (2022).
13. Monitor for adventitious operant responses under FT/VT. Behavior occurring in the seconds before each FT delivery can become reinforced operant responding; explicit monitoring catches this López-Tolsa & Pellón (2025).
14. Treat lean schedules as a feature, not a goal. Regnier's VR-2 vs VR-6 comparison shows leaner schedules can preserve effect at lower cost — but the burden is to verify the parameter empirically, not assume it Regnier et al. (2022).
15. For social and group contingencies, treat the schedule as the social structure. Interdependent ratios, DRL windows, and synchronous group contingencies directly predict cooperation versus competition Benvenuti et al. (2024) Gomes et al. (2025).

Primary research synthesized in this guide. DOIs link to the original source.

• Cividini-Motta, C., Livingston, C., & Efaw, H. (2024). Systematic review of differential reinforcement in skill acquisition. Behavior Analysis in Practice, 17, 401–416. https://doi.org/10.1007/s40617-023-00903-z https://doi.org/10.1007/s40617-023-00903-z
• Jones, L., Brand, D., Bensemann, J., Heinicke, M. R., Penrod, B., & Burlison, S. (2026). A translational approach to investigating the effects of consequence-based procedural fidelity errors postmastery. Journal of Applied Behavior Analysis, 59(1), e70042. https://doi.org/10.1002/jaba.70042 https://doi.org/10.1002/jaba.70042
• Killeen, P. R. (2023). Theory of reinforcement schedules. Journal of the Experimental Analysis of Behavior, 120(3), 289-319. https://doi.org/10.1002/jeab.880 https://doi.org/10.1002/jeab.880
• Lambert, J. M., Osina, M. A., Staubitz, J. L., Reed, D. D., & Madden, G. J. (2026). Evaluating contributions of progressive ratio analysis to economic metrics of demand. Journal of the Experimental Analysis of Behavior, 125(1), e70077. https://doi.org/10.1002/jeab.70077 https://doi.org/10.1002/jeab.70077
• Benvenuti, M. F., Bernardy, J. L., de Toledo, T. N., Santiago, J., Siqueira, J. d. O., & Silveira, P. S. P. (2024). The role of schedules of reinforcement in the experimental analysis of social behavior. Perspectives on Behavior Science, 47(4), 717-738. https://doi.org/10.1007/s40614-024-00421-5 https://doi.org/10.1007/s40614-024-00421-5
• Wilson, A. N. & Gratz, O. H. (2016). Using a Progressive Ratio Schedule of Reinforcement as an Assessment Tool to Inform Treatment. Behavior Analysis in Practice, 9(3), 257-260. https://doi.org/10.1007/s40617-016-0107-2 https://doi.org/10.1007/s40617-016-0107-2
• López-Tolsa, G. E. & Pellón, R. (2025). Dynamic Interactions between Induction and Reinforcement in the Organization of Behavior. Perspectives on Behavior Science, 48(2), 315-339. https://doi.org/10.1007/s40614-025-00453-5 https://doi.org/10.1007/s40614-025-00453-5
• Diaz de Villegas, S. C., Dozier, C. L., Jess, R. L., & Foley, E. A. (2020). An evaluation of synchronous reinforcement for increasing on‐task behavior in preschool children. Journal of Applied Behavior Analysis, 53(3), 1660-1673. https://doi.org/10.1002/jaba.696 https://doi.org/10.1002/jaba.696
• King, H. C., Kranak, M. P., Falligant, J. M., & Spink, C. (2025). Discriminative Properties of Reinforcers Modulate Resurgence: Evidence from a Systematic Review. Perspectives on Behavior Science, 48(4), 711-729. https://doi.org/10.1007/s40614-025-00482-0 https://doi.org/10.1007/s40614-025-00482-0
• Gomes, G., Linton, E., & Donaldson, J. M. (2025). Synchronous reinforcement arranged as a group contingency increases rule following in elementary classes. Journal of Applied Behavior Analysis, 58(3), 490-503. https://doi.org/10.1002/jaba.70018 https://doi.org/10.1002/jaba.70018
• Saini, V., Miller, S. A., & Fisher, W. W. (2016). Multiple schedules in practical application: Research trends and implications for future investigation. Journal of Applied Behavior Analysis, 49(2), 421-444. https://doi.org/10.1002/jaba.300 https://doi.org/10.1002/jaba.300
• Baum, W. M. (2025). Explaining Performance on Interval and Ratio Schedules with a Molar View of Behavior. Perspectives on Behavior Science, 48(2), 173-202. https://doi.org/10.1007/s40614-025-00455-3 https://doi.org/10.1007/s40614-025-00455-3
• Fisher, W. W., Saini, V., Greer, B. D., Sullivan, W. E., Roane, H. S., Fuhrman, A. M., Craig, A. R., & Kimball, R. T. (2019). Baseline reinforcement rate and resurgence of destructive behavior. Journal of the Experimental Analysis of Behavior, 111(1), 75-93. https://doi.org/10.1002/jeab.488 https://doi.org/10.1002/jeab.488
• Frank-Crawford, M. A., Tate, S. A., Goetzel, A., & Finn, M. (2024). Using functional communication and competing stimuli to gradually increase the distance of transitions in the treatment of tangibly maintained elopement. Behavior Analysis in Practice, 17, 921–925. https://doi.org/10.1007/s40617-024-00957-7 https://doi.org/10.1007/s40617-024-00957-7
• Ruiz Méndez, D. (2024). Toward a procedure to study rule-governed choice: preliminary data. The Analysis of Verbal Behavior, 40, 280–305. https://doi.org/10.1007/s40616-024-00206-6 https://doi.org/10.1007/s40616-024-00206-6
• Johnson, A. R., Christensen, B. A., Kelly, S. J., & Calipari, E. S. (2022). The influence of reinforcement schedule on experience‐dependent changes in motivation. Journal of the Experimental Analysis of Behavior, 117(3), 320-330. https://doi.org/10.1002/jeab.755 https://doi.org/10.1002/jeab.755
• Gallistel, C. R. (2025). Reconceptualized Associative Learning. Perspectives on Behavior Science, 48(2), 203-239. https://doi.org/10.1007/s40614-025-00442-8 https://doi.org/10.1007/s40614-025-00442-8
• Diaz de Villegas, S. C., Dozier, C. L., Kanaman, K. C., Leslie, S. C., & Kamlowsky, M. E. (2024). Comparison of synchronous reinforcement and accumulated reinforcement for increasing on‐task behavior in preschoolers. Journal of Applied Behavior Analysis, 57(3), 725-741. https://doi.org/10.1002/jaba.1080 https://doi.org/10.1002/jaba.1080
• Brown, K. R., Greer, B. D., Craig, A. R., Sullivan, W. E., Fisher, W. W., & Roane, H. S. (2020). Resurgence following differential reinforcement of alternative behavior implemented with and without extinction. Journal of the Experimental Analysis of Behavior, 113(2), 449-467. https://doi.org/10.1002/jeab.588 https://doi.org/10.1002/jeab.588
• McLean, A. P., Grace, R. C., Shevchouk, O. T., & Cording, J. R. (2018). Rat choice in rapidly changing concurrent schedules. Journal of the Experimental Analysis of Behavior, 109(2), 313-335. https://doi.org/10.1002/jeab.314 https://doi.org/10.1002/jeab.314
• Phillips, C. L., Iannaccone, J. A., Rooker, G. W., & Hagopian, L. P. (2017). Noncontingent reinforcement for the treatment of severe problem behavior: An analysis of 27 consecutive applications. Journal of Applied Behavior Analysis, 50(2), 357-376. https://doi.org/10.1002/jaba.376 https://doi.org/10.1002/jaba.376
• Jessel, J., Fruchtman, T., Raghunauth‑Zaman, N., Leyman, A., Lemos, F. M., Costa Val, H., Howard, M., & Hanley, G. P. (2024). A two step validation of the performance‑based IISCA: A trauma‑ informed functional analysis model. Behavior Analysis in Practice, 17, 727–745. https://doi.org/10.1007/s40617-023-00792-2 https://doi.org/10.1007/s40617-023-00792-2
• Regnier, S. D., Strickland, J. C., & Stoops, W. W. (2022). A preliminary investigation of schedule parameters on cocaine abstinence in contingency management. Journal of the Experimental Analysis of Behavior, 118(1), 83-95. https://doi.org/10.1002/jeab.770 https://doi.org/10.1002/jeab.770
• Russell, D., Ingvarsson, E. T., Haggar, J. L., & Jessel, J. (2018). Using progressive ratio schedules to evaluate tokens as generalized conditioned reinforcers. Journal of Applied Behavior Analysis, 51(1), 40-52. https://doi.org/10.1002/jaba.424 https://doi.org/10.1002/jaba.424
• Baruni, R. R., Cook, J. L., Pinkston, J. W., Rapp, J. T., Miltenberger, R. G., Walker, E., Deshmukh, S., & Sheridan, D. J. (2025). Schedule control with a synchronous reinforcement treadmill preparation: A replication and extension. Journal of Applied Behavior Analysis, 58(2), 422-432. https://doi.org/10.1002/jaba.70008 https://doi.org/10.1002/jaba.70008
• Martínez‐Herrada, A., Pellón, R., & López‐Tolsa, G. E. (2025). Temporal distribution of schedule‐induced behavior depends on the essential value of the reinforcer. Journal of the Experimental Analysis of Behavior, 123(1), 10-29. https://doi.org/10.1002/jeab.4235 https://doi.org/10.1002/jeab.4235
• Shahan, T. A., Browning, K. O., Nist, A. N., & Sutton, G. M. (2020). Resurgence and downshifts in alternative reinforcement rate. Journal of the Experimental Analysis of Behavior, 114(2), 163-178. https://doi.org/10.1002/jeab.625 https://doi.org/10.1002/jeab.625
• Torres‐Viso, M., Strohmeier, C. W., & Zarcone, J. R. (2018). Functional analysis and treatment of problem behavior related to mands for rearrangement. Journal of Applied Behavior Analysis, 51(1), 158-165. https://doi.org/10.1002/jaba.437 https://doi.org/10.1002/jaba.437
• Owen, T. M., Fisher, W. W., Akers, J. S., Sullivan, W. E., Falcomata, T. S., Greer, B. D., Roane, H. S., & Zangrillo, A. N. (2020). Treating destructive behavior reinforced by increased caregiver compliance with the participant's mands. Journal of Applied Behavior Analysis, 53(3), 1494-1513. https://doi.org/10.1002/jaba.674 https://doi.org/10.1002/jaba.674
• King, H. C., Spink, C., & Falligant, J. M. (2025). Discriminative properties of reinforcers modulate resurgence: A human‐operant demonstration. Journal of the Experimental Analysis of Behavior, 123(3), 526-538. https://doi.org/10.1002/jeab.70010 https://doi.org/10.1002/jeab.70010
• Craig, A. R. & Shahan, T. A. (2022). Nondrug reinforcers contingent on alternative behavior or abstinence increase resistance to extinction and reinstatement of ethanol‐maintained behavior. Journal of the Experimental Analysis of Behavior, 118(3), 353-375. https://doi.org/10.1002/jeab.801 https://doi.org/10.1002/jeab.801
• Frampton, S. E., Davis, C. R., Meleshkevich, O., & Axe, J. B. (2024). A clinical tutorial on methods to capture and contrive establishing operations to teach mands. Behavior Analysis in Practice, 17, 1270–1282. https://doi.org/10.1007/s40617-024-00985-3 https://doi.org/10.1007/s40617-024-00985-3
• Mattson, S. L., Higbee, T. S., Nichols, B., Aguilar, J., & Campbell, V. E. (2024). Effects of an activity schedule intervention package on cooperative vocal exchanges during learning centers. Behavior Analysis in Practice, 17, 1175–1190. https://doi.org/10.1007/s40617-024-00939-9 https://doi.org/10.1007/s40617-024-00939-9
• Nevin, J. A., Mace, F. C., DeLeon, I. G., Shahan, T. A., Shamlian, K. D., Lit, K., Sheehan, T., Frank‐Crawford, M. A., Trauschke, S. L., Sweeney, M. M., Tarver, D. R., & Craig, A. R. (2016). Effects of signaled and unsignaled alternative reinforcement on persistence and relapse in children and pigeons. Journal of the Experimental Analysis of Behavior, 106(1), 34-57. https://doi.org/10.1002/jeab.213 https://doi.org/10.1002/jeab.213
• Lattal, K. A. & Miles, A. K. (2024). Resurgence and Behavioral Contrast, Compared and Contrasted. Perspectives on Behavior Science, 47(2), 315-334. https://doi.org/10.1007/s40614-024-00408-2 https://doi.org/10.1007/s40614-024-00408-2