University of South FloridaApplied Behavior Analysis
TALCON LabTalhelm and Concepcion, USF Applied Behavior Analysis

Draft for the lab's review. Items in brackets are placeholders until the lab confirms them.

Four areas, twelve projects

Research

Four lines of work guide the lab's studies. The graphs on this page show the designs we use to answer them, drawn with illustrative data.

Graphs on this page use illustrative data,
not results from the lab.

A  Resurgence (three phases)

Relapse after treatment

Why behavior that treatment reduced can come back, and how to make that less likely.

Treatment often works by reinforcing a new, safer behavior. When that reinforcement stops, or the setting changes, the old behavior can return: resurgence and renewal. We study these effects in translational studies with adult volunteers, then ask what they mean for clinical treatment.

Design
Resurgence (three phases)
Measure
Responses per minute
Projects
History effects on resurgence; Multiple-context training and renewal; Conditioned reinforcing properties of tokens
Fig. 1. Relapse after treatmentIllustrative data
0510152025Responses per minute15101520SessionsBaselineTreatmentExtinctionTarget behaviorAlternative behaviorResurgence
Fig. 1. A three-phase resurgence arrangement. Filled circles: target behavior. Open squares: the alternative behavior. Dashed lines mark each phase change. Illustrative data, responses per minute. Target behavior: baseline 18, 20, 17, 21, 19, 20; treatment 9, 5, 3, 1, 1, 0, 1, 0; extinction 10, 13, 8, 6, 3, 2. Alternative behavior: baseline 0, 1, 0, 0, 1, 0; treatment 11, 16, 19, 21, 20, 22, 21, 22; extinction 9, 5, 3, 2, 1, 1.

B  Multielement functional analysis

Severe problem behavior

Finding why self-injury and aggression happen, then matching treatment to that reason.

Both directors assessed and treated severe self-injury and aggression at the Kennedy Krieger Institute's Neurobehavioral Unit. A functional analysis tests each possible reason in turn; the condition where behavior is highest points to its function and to the treatment most likely to help.

Design
Multielement functional analysis
Measure
Problem behavior per minute
Projects
Protective equipment in behavior analysis; Self-injury in an academic context
Fig. 2. Severe problem behaviorIllustrative data
012345Problem behavior per minute15101520SessionsFunctional analysisEscapeAttention, tangible, play
Fig. 2. A multielement functional analysis. Squares: escape sessions. Filled circles: attention. Triangles: tangible. Open circles: play. Illustrative data, problem behavior per minute in each condition across five rounds. Escape: 3.4, 4.1, 3.7, 4.4, 4.0. Attention: 0.5, 0.3, 0.6, 0.2, 0.3. Tangible: 0.2, 0.6, 0.3, 0.4, 0.1. Play: 0, 0.1, 0, 0, 0.

C  Multiple baseline across people

Performance and skills training

Video feedback and coaching for lifting form, esports, public speaking and safety skills.

Behavioral skills training and video feedback are the lab's main tools here. Feedback starts at a different time for each person, so a rise that follows each start shows the feedback is doing the work.

Design
Multiple baseline across people
Measure
Steps done with correct form (%)
Projects
Video feedback for weight training form; Video feedback for esports; Automated feedback for speech disfluencies; Computerized firearm safety training for practitioners
Fig. 3. Performance and skills trainingIllustrative data
050100Athlete 1050100Athlete 2050100Athlete 3BaselineVideo feedback151015SessionsSteps done with correct form (%)
Fig. 3. A multiple baseline across three athletes. The stepped dashed line shows video feedback starting after sessions 5, 8 and 11. Illustrative data, percent of steps done with correct form. Athlete 1: baseline 30, 38, 25, 33, 35; video feedback 64, 78, 86, 90, 94, 92, 96, 98, 95, 97, 100, 98, 97. Athlete 2: baseline 28, 22, 34, 30, 26, 31, 27, 33; video feedback 60, 74, 85, 91, 93, 96, 94, 98, 100, 97. Athlete 3: baseline 36, 30, 40, 28, 34, 38, 31, 29, 35, 33, 30; video feedback 58, 76, 88, 92, 95, 97, 98.

D  Cumulative record

Learning in higher education

Procrastination, discussion boards and group contingencies in online college courses.

Work that piles up before each deadline draws a scallop on a cumulative record, much like a fixed-interval schedule. We study what predicts that pattern, including delay discounting, and course arrangements that spread work across the week.

Design
Cumulative record
Measure
Assignment steps finished, cumulative
Projects
Academic procrastination and delay discounting; Group contingencies in online courses; Response modality in online discussion boards
Fig. 4. Learning in higher educationIllustrative data
015304560Assignment steps finished, cumulativeDueDueDueDueDueDueDaysBaselineDaily goals
Fig. 4. A cumulative record of coursework. Open circles mark each weekly deadline in baseline; filled circles mark deadlines with daily goals. Illustrative data, assignment steps finished each day. Baseline, three weeks: 0, 0, 0, 1, 1, 3, 5, 0, 0, 0, 1, 1, 3, 5, 0, 0, 0, 1, 1, 3, 5. Daily goals, three weeks: 1, 2, 1, 2, 1, 2, 1, 1, 2, 1, 2, 1, 2, 1, 1, 2, 1, 2, 1, 2, 1. Cumulative total: 30 after baseline, 60 at the end.

Projects

Recent and ongoing projects, each marked with its area's symbol. Collaborators from other labs are listed with the lab's directors.

  • A. Relapse after treatment
  • B. Severe problem behavior
  • C. Performance and skills training
  • D. Learning in higher education

Relapse after treatment

  • A. Relapse

    History effects on resurgence

    A translational evaluation of how a behavior's history affects its return.

    People
    Talhelm, Bloom, Concepcion and Samaha
    Shared at
    ABAI 2019 and 2020; FABA 2019
  • A. Relapse

    Multiple-context training and renewal

    Whether training in several settings reduces renewal after differential reinforcement of alternative behavior.

    People
    Talhelm, DeLeon, Cormier, Ritchey and Podlesnik
    Shared at
    FABA 2023; ABAI 2024
  • A. Relapse

    Conditioned reinforcing properties of tokens

    When tokens work as reinforcers on their own.

    People
    Concepcion, Sy and Talhelm
    Shared at
    ABAI 2017; FABA 2018

Severe problem behavior

  • B. Severe

    Protective equipment in behavior analysis

    A review of how protective equipment is used during treatment and its idiosyncratic effects.

    People
    Raaymakers, Talhelm and Concepcion
    Shared at
    ABAI 2022; FABA 2022
  • B. Severe

    Self-injury in an academic context

    Analysis and treatment of self-injurious behavior that occurred during academic work.

    People
    Concepcion, Talhelm, Rey and Zarcone
    Shared at
    ABAI 2017

Performance and skills training

  • C. Performance

    Video feedback for weight training form

    Peers trained with behavioral skills training gave video feedback on deadlift form.

    People
    Cochrane, Miltenberger and Concepcion
    Shared at
    Journal of Applied Behavior Analysis, 2022
  • C. Performance

    Video feedback for esports

    Video feedback to improve play in Street Fighter V.

    People
    White, Miltenberger and Concepcion
    Shared at
    Behavioral Interventions, 2024
  • C. Performance

    Automated feedback for speech disfluencies

    An automated feedback tool to reduce filler words in public speaking.

    People
    Medlin, Drew, Concepcion and Talhelm
    Shared at
    CoFABA 2026
  • C. Performance

    Computerized firearm safety training for practitioners

    Interactive computerized training to teach BCBAs to run firearm safety skills training.

    People
    Baruni, Miltenberger, Cook, Concepcion and Maxfield
    Shared at
    Journal of Applied Behavior Analysis, 2025

Learning in higher education

  • D. Learning

    Academic procrastination and delay discounting

    Whether discounting predicts how close to the deadline students submit work.

    People
    Concepcion, Scalzo and Crosland
    Shared at
    ABAI 2020 and 2021
  • D. Learning

    Group contingencies in online courses

    Group contingencies to improve academic performance in online graduate courses.

    People
    Evans, Scalzo and Concepcion
    Shared at
    ABAI 2021
  • D. Learning

    Response modality in online discussion boards

    Written versus recorded responses in online discussion boards.

    People
    Dunn, Concepcion and Crosland
    Shared at
    ABAI 2022

[Confirm each project's current status: recruiting, collecting data, writing up or complete]