ASSESSING PREFERENCES AND IDENTIFYING REINFORCERS KEVIN P KLATT

ASSESSING PREFERENCES AND IDENTIFYING REINFORCERS KEVIN P. KLATT, PH. D. , BCBA-D PROFESSOR, UWEC

OBJECTIVES TODAY • PREFERENCE ASSESSMENT VS REINFORCER ASSESSMENT • CURRENT PRACTICES IN THE COMMUNITY • METHODS TO IDENTIFY PREFERENCES • FACTORS THAT AFFECT REINFORCER EFFECTIVENESS • RECOMMENDATIONS

MOST COMMON CALLS I GET FROM MEMBERS OF THE COMMUNITY (E. G. , TEACHERS AND PARENTS)

I’VE TRIED EVERYTHING, AND NOTHING WORKS.

I TRIED REINFORCEMENT AND IT DOESN’T WORK FOR THIS KID.

I SHOULD NOT HAVE TO GIVE SOMETHING SPECIAL TO THIS KID BUT NOT OTHER KIDS…NOT FAIR.

WE CAN’T AFFORD TO BE BUYING STUFF FOR KIDS ALL THE TIME.

IRONICALLY… • THE SOLUTION TO ALMOST ALL OF THE CALLSI GET INVOLVES IDENTIFYING REINFORCERS. • NOT THE ONLY SOLUTION…BUT PART OF IT.

Antecedent Behavior Consequences ?

• THE CHALLENGE FOR THE BEHAVIOR ANALYST/TEACHER/PARENT IS TO IDENTIFY STIMULI THAT WILL FUNCTION AS A REINFORCER. • “THE USE OF A PLANNED REINFORCEMENT SYSTEM MAY BE THE MOST IMPORTANT THING THAT YOU DO TO HELP YOUR CHILD LEARN. ” (ANDERSON, TARAS & CANNON, 1996).

UNFORTUNATELY, NOT UNDERSTANDING REINFORCEMENT LEADS TO MISTAKES.

TWO MISTAKES ARE COMMON AND EASY TO MAKE

TWO COMMON MISCONCEPTIONS • 1. CERTAIN STIMULI (E. G. , ITEMS LIKEM&M’S AND PRAISE) ARE “INHERENT” POSITIVE REINFORCERS. • 2. CERTAIN STIMULI THAT SEEM AVERSIVE TO MOST PEOPLE (E. G. , YELLING) ARE NOT REINFORCERS.

• THIS ASSUMES A STIMULUS FUNCTIONS AS A REINFORCER BASED ON TOPOGRAPHY OF THE STIMULUS. • REINFORCEMENT, HOWEVER, IS DEFINED BYITS EFFECT ON BEHAVIOR.

PREFERENCE ASSESSMENT OR REINFORCER ASSESSMENT

LET’S START WITH PREFERENCE ASSESSMENT

• PREFERENCE ASSESSMENTS PROVIDE DATA-BASED METHODS FOR IDENTIFYING PREFERRED STIMULI ON AN INDIVIDUAL BASIS. • THE VALUE OF PREFERENCE ASSESSMENTS IS THE EXTENT TO WHICH THE PREFERRED ITEMSFUNCTION AS REINFORCERS.

Antecedent Behavior Consequences ?

• PRIOR TO 1985, THERAPISTS SELECTED POTENTIAL POSITIVE REINFORCERS SOMEWHAT ARBITRARILY. • THAT IS, WITHOUT KNOWING WHETHER THE STIMULUS WOULD FUNCTION AS REINFORCEMENT.

HOW MUCH HAS CHANGED SINCE THE 1980’S?

GRAFF & KARSTEN (2012) • SURVEYED THE CURRENT PRACTICES OF HOW PROFESSIONALS ASSESS PREFERENCES FOR PEOPLE WITH DEVELOPMENTAL DISABILITIES. • 406 PEOPLE COMPLETED THESURVEY

PARTICIPANTS • 57% WORKED INSCHOOLS (OTHER 43% FROM A VARIETY OF SETTINGS) • 32% WERE CERTIFIED BEHAVIOR ANALYSTS • 64% HAD MASTERS DEGREE • SPECIAL ED (43%), GENERAL ED (30%), BEHAVIOR ANALYSIS (27%), PSYCHOLOGY (22%)

• STIMULUS PREFERENCE ASSESSMENTS CAN BE CONDUCTED THREE DIFFERENT WAYS: • THE FIRST TWO WAYS AREINDIRECT ASSESSMENTS OF PREFERENCES.

• 1. ASK • THE CAREGIVERS • THE CLIENT • OFFER A PRETASK CHOICE (THIS IS EASY TO DO IN A CLINICAL PROGRAM. )

THERE ARE TOOLS AVAILABLE…

RAISD (1996) The Reinforcement Assessment for Individuals with Severe Disabilities • FISHER, W. , PIAZZA, C. C. , BOWMAN, L. G. , & AMARI, A. (1996). INTEGRATING CAREGIVER REPORT WITH SYSTEMATIC CHOICE ASSESSMENT TO ENHANCE REINFORCER IDENTIFICATION. AMERICAN JOURNAL OF MENTAL RETARDATION, 101, 15– 25.

BUT BE CAREFUL…

• RESEARCH HAS CONSISTENTLY SHOWN THAT TEACHERS AND CAREGIVERS ARE NOT ALWAYS ACCURATE IN WHAT THEY THINK INDIVIDUALS PREFER. • (E. G. , GREEN, REID, WHITE, HALFORD, BRITTAIN, &GARDNER, 1988)

• 2. OBSERVE THE CLIENT INTERACT WITH THE ITEMS IN NATURAL SITUATIONS. • THIS IS FAIRLY EASY TO DO IN A CLINICAL PROGRAM. • THIS, HOWEVER, IS NOT TYPICALLY DONE SYSTEMATICALLY. • THEREFORE MAY NOT BE RELIABLE.

THE THIRD WAY TO ASSESS PREFERENCES ISDIRECT MEASUREMENT

• 3. MEASURE THE CLIENT’S RESPONSE DURING SPECIFIC TESTS FOR PREFERENCE.

TRIAL BASED PREFERENCE ASSESSMENTS SINGLE STIMULUS PAIRED STIMULUS MULTIPLE STIMULUS TIME BASED PREFERENCE ASSESSMENT FREE OPERANT

TRIAL BASED ASSESSMENTS…

SINGLE STIMULUS PREFERENCE ASSESSMENT PACE, IVANCIC, EDWARDS, IWATA & PAGE (1985)

SINGLE STIMULUS • THIS ASSESSMENT INVOLVES PRESENTING STIMULI INDIVIDUALLY. • FOR EXAMPLE, A TOY IS PLACED ON THE TABLE IN FRONT OF THE CHILD.

SINGLE STIMULUS PACE, IVANCIC, EDWARDS, IWATA & PAGE (1985) • IF THE CHILD APPROACHES (SELECTS) THE TOY SHE IS ALLOWED 5 -30 SEC. OF ACCESS TO IT. • IF THE CHILD DOES NOT APPROACH WITHIN 5 SEC. SHE IS PROMPTED TO SAMPLE THE ITEM (PLAY WITH IT). • ITEMS ARE PRESENTED NUMEROUS TIMES. • ITEMS SELECTED MOST ARE CONSIDERED HIGH PREFERRED AND “POTENTIAL” REINFORCERS.

PAIRED STIMULUS (MITHAUG AND HANAWALT, 1978; FISHER, PIAZZA, BOWMAN, HAGOPIAN, OWENS & SLEVIN (1992). • IN THE PAIRED STIMULUS ASSESSMENT, 2 ITEMS ARE PLACED EQUAL DISTANCE IN FRONT OF THE PERSON.

PAIRED STIMULUS PREFERENCE ASSESSMENT • THE PERSON IS TOLD TO “CHOOSE” AN ITEM. • WHEN THE PERSON CHOOSES AN ITEM HE IS ALLOWED TO PLAY WITH IT FOR A FEW SECONDS. • IF HE DOESN’T CHOOSE HE IS PROMPTED TO SAMPLE THE ITEM AND THE TRIAL IS RUN AGAIN.

DATA SHEET

PAIRED STIMULUS PREFERENCE ASSESSMENT • ALL ITEMS ARE PAIRED WITH EACH OTHER THROUGHOUT THE ASSESSMENT. • A HIERARCHY OF MOST-TO-LEAST PREFERRED ITEMS CAN BE GRAPHED.

PAIRED STIMULUS • A DRAWBACK OF THIS ASSESSMENT IS THAT IT CAN BE VERY TIME CONSUMING.

PAIRED STIMULUS • A STUDY COMPARING THESS AND PS FOUND THAT THE SS TENDED TO OVERESTIMATE PREFERENCES COMPARED TO THE PS. • BOTH THE SS AND PS CAN RESULT IN PROBLEM BEHAVIORS WHEN CONDUCTED WITH PEOPLE WITH DISABILITIES. WHY?

MULTIPLE STIMULUS (WINDSOR, PICHE, & LOCKE, 1994) • THIS PROCEDURE INVOLVES PRESENTING (MULTIPLE) ITEMS IN FRONT OF THE PERSON.

MULTIPLE STIMULUS (WINDSOR, PICHE, & LOCKE, 1994)

• THE PERSON IS TOLD TO CHOOSE AN ITEM AND DATA ARE RECORDED ON WHICH ITEM(S) ARE SELECTED ACROSS TRIALS. • THE PERSON CAN HAVE ACCESS TO THE ITEM FOR ABOUT 20 SECONDS (OR A SMALL AMOUNT OF THE DRINK). • THESE TRIALS ARE REPEATED SEVERAL TIMES.

• THIS ASSESSMENT TAKES LESS TIME TO CONDUCT THAN THE SS AND PS. • THIS ASSESSMENT, HOWEVER, CAN RESULT IN ONLY ONE ITEM BEING CHOSEN ON ALL TRIALS (THUS UNDERESTIMATING OTHER ITEMS).

We don’t know much about these items relative to each other.

MULTIPLE STIMULUS W/OUT REPLACEMENT (MSWO) DELEON & IWATA (1996) • THIS PROCEDURE IS THE SAME AS THE MULTIPLE STIMULUS EXCEPT THAT AS ITEMS ARE CHOSEN THEY ARE REMOVED AND NOT REPLACED.

16 14 12 10 Number times 8 selected 6 4 2 0 juice box pop kool-aid Toys milk

DELEON AND IWATA, 1996 • A STUDY COMPARING THESE METHODS FOUND: • THE PAIRED STIMULUS PROCEDURE TOOK AN AVERAGE OF 53 MINUTES. • WHILE THE MSWO TOOK 16. 5 MINUTES.

TIME BASED ASSESSMENT…

BRIEF FREE OPERANT ROANE, VOLLMER, RINGDAHL, & MARCUS (1998) • MULTIPLE STIMULI ARE PLACED ON A TABLETOP, AND PARTICIPANTS ARE FREE TO ENGAGE WITH ANY OF THE ITEMS FOR 5 MIN. • DURATION OF ENGAGEMENT WITH EACH ITEM IS MEASURED.

IN GENERAL, USE • SINGLE STIMULUS FOR CLIENTS WHO HAVE DIFFICULTY MAKING A CHOICE. • PAIRED STIMULUS TO OBTAIN A HIERARCHY OF PREFERENCES. • MSWO AND FREE OPERANT FOR TIME EFFICIENCY.

COMBINE TRIAL BASED AND FREE OPERANT?

RESPONSE RESTRICTION ANALYSIS • IN TRIAL FORMATS (SS AND PS) THERE IS LIMITED ACCESS TO THE ITEMS. ON THE OTHER HAND… • IN A FREE OPERANT ASSESSMENT RESPONDING MIGHT OCCUR WITH ONLY ONE ITEM. • BOTH OF THESE LIMITATIONS MIGHT BE ALLEVIATED BY COMBINING FORMATS.

RESPONSE RESTRICTION ANALYSIS: I. ASSESSMENT OF ACTIVITY PREFERENCES • HANLEY, IWATA, LINDBERG AND CONNERS, 2003 • JOURNAL OF APPLIED BEHAVIOR ANALYSIS

7 ACTIVITIES CHOSEN FOR EACH CHILD

HOW? • 1. START WITH A FREE OPERANT ASSESSMENT • THIS OVERCOMES PROBLEMS WITH LIMITED ACCESS TO ITEMS. • PROBABLY REDUCES TIME INVOLVED BECAUSE YOU DO NOT HAVE TO PRESENT ITEMS NUMEROUS TIMES. • MIGHT REDUCE PROBLEMS BECAUSE YOU ARE NOT TAKING AWAY ITEMS.

• 2. AT SOME POINT YOU BEGIN TAKING AWAY THE MOST PREFERRED ITEM(S) TO PREVENT EXCLUSIVE SELECTION OF THOSE ITEM(S). • RESPONSE RESTRICTION ANALYSIS HAS BEEN USED IN BASIC RESEARCH WITH HUMANS AND NONHUMANS.

EXAMPLES OF RULES FOR WHEN TO TAKE AWAY TOP ITEMS • 1. IF ITEM IS INTERACTED WITH ON 60% OR MORE OF INTERVALS FOR TWO SESSIONS, REMOVE IT FROM THE ARRAY. • 2. IF RESPONDING IS CONSISTENTLY DISTRIBUTED AMONG A SMALL GROUP OF ITEMS, THAN THOSE ITEMS ARE REMOVED.

Item removed First Assessment Second Assessment

COMPARE FREE OPERANT TO RESPONSE RESTRICTION

WHAT IS BEING ASSESSED IN PREFERENCE ASSESSMENTS?

WHAT ABOUT ASSESSING PREFERRED ACTIVITIES?

ASSESSING PREFERENCES FOR SOCIAL INTERACTION IN CHILDREN DIAGNOSED WITH AUTISM NUERNBERGER, SMITH, CZAPAR, &KLATT BEHAVIORAL INTERVENTIONS (2012)

GENERAL METHOD • PHASE 1: § POTENTIAL REINFORCING ACTIVITIES (FORMS OF PHYSICAL ATTENTION) WERE IDENTIFIED. • PHASE 2: § PICTURES OF THE ACTIVITIES WERE TESTED (TAUGHT) FOR RECEPTIVE AND EXPRESSIVE IDENTIFICATION • PHASE 3: § MULTIPLE-STIMULUS WITHOUT REPLACEMENT (MSWO) PREFERENCE ASSESSMENT WAS CONDUCTED USING THE PICTURES. • PHASE 4: § REINFORCER ASSESSMENT WAS CONDUCTED.

PHASE 1: IDENTIFY PREFERRED ACTIVITIES • PARENTS AND THERAPISTS WERE ASKED TO IDENTIFY PHYSICAL ACTIVITIES THAT WERE POTENTIALLY REINFORCING. • SIX ACTIVITIES WERE SELECTED FOR EACH PARTICIPANT. • PICTURES OF THE CHILD ENGAGING IN EACH ACTIVITY WERE PRINTED ON 3 X 5 INCH CARD STOCK. • PARTICIPANTS HAD TO DEMONSTRATE THE ABILITY TO IDENTIFY ACTIVITIES BOTH EXPRESSIVELY AND RECEPTIVELY.

Spin Blast Off Swing Horse Ride Chase Carousel

PHASE 3: PREFERENCE ASSESSMENT • A MSWO PREFERENCEASSESSMENT WAS CONDUCTED EVERY SESSION. • THE EXPERIMENTER PLACED ALL PICTURE CARDS ON THE TABLE IN FRONT OF THE CHILD AND GAVE THE INSTRUCTION “PICK ONE. ” • SELECTING A PICTURE WITHIN 15 SECONDS RESULTED IN 15 SECONDS ACCESS TO THE ACTIVITY DEPICTED IN THE PICTURE. • THE REMAINING CARDS WERE SHUFFLED AND REPRESENTED FOR THE NEXT TRIAL, AND CONTINUED UNTIL ALL ACTIVITIES WERE SELECTED. • AT LEAST 6 PREFERENCE ASSESSMENTS WERE CONDUCTED PER CHILD.

WHAT ABOUT REINFORCER ASSESSMENTS? • HOW DO WE KNOW WHETHER PREFERENCE ASSESSMENTS ACCURATELY PREDICT WHAT WILL FUNCTION AS A REINFORCER?

“THE ONLY WAY TO TELL WHETHER A GIVEN EVENT IS REINFORCING TO A GIVEN ORGANISM IS TO MAKE A DIRECT TEST. ” B. F. SKINNER

PHASE 4: REINFORCER ASSESSMENT • A SORTING TASK THAT WAS PREVIOUSLY MASTERED WAS SELECTED FOR EACH PARTICIPANT. • SORTING WAS RECORDED USING A 30 S PARTIAL INTERVAL RECORDING SYSTEM. • THE CHILD WAS INSTRUCTED TO SORT THROUGHOUT A 5 MINUTE SESSION UNTIL ONE OF THE FOLLOWING CRITERIA WAS MET. • THE CHILD ENGAGED IN PROBLEM BEHAVIOR • 20 SECONDS OF NO SORTING • 5 MINUTES ELAPSED

HOW OFTEN SHOULD A PREFERENCE ASSESSMENT BE CONDUCTED?

FACTORS THAT MAY INFLUENCE THE EFFECTIVENESS OF REINFORCEMENT • CHANGES IN PREFERENCE OVER TIME • SATIATION VS. DEPRIVATION

CHANGES IN PREFERENCE OVER TIME • THE RESULTS OF PREVIOUS RESEARCH IS MIXED ON HOW STABLE PREFERENCES ARE OVER TIME. • TWO STUDIES SUGGEST PREFERENCES REMAIN FAIRLY STABLE. • HANLEY, IWATA &ROSCOE, 2006 • KELLEY, SHILLINGSBURG, &BOWEN, 2016

• BOTH STUDIES IMPLEMENTED NUMEROUS “PAIRED-CHOICE” PREFERENCE ASSESSMENTS ACROSS TIME. • THEN THE PREFERENCE ASSESSMENTS WERE COMPARED TO EACH OTHER TO LOOK FOR STABILITY. • CORRELATION COEFFICIENTS ABOVE. 58 WERE CONSIDERED STABLE.

16 of 21 participants had stable preferences

WHAT ABOUT CHANGES WITHIN AND ACROSS DAYS?

PREFERENCES, DEPRIVATION AND PRESESSION EXPOSURE (HARTMAN & KLATT, 2005)

PHASE 1 • PREFERENCE ASSESSMENT • PAIRED STIMULUS (SEAN) • MSWO (BILLY)

PHASE 2 • TEACH MANDING (REQUESTING) • TWO VARIABLES INVESTIGATED: 1. 5 MIN OF PRESESSION EXPOSURE (SATIATION) VS 23 HOUR DEPRIVATION, AND 2. HIGH VS LOW PREFERENCE TOYS

SATIATION VS DEPRIVATION • LIMITING ACCESS TO POTENTIAL REINFORCERS OUTSIDE OF TRAINING OR TREATMENT SITUATIONS IS A COMMONLY RECOMMENDED STRATEGY.

WHAT ABOUT ASSESSING PREFERENCE FOR TREATMENT PROCEDURES?

INVESTIGATING THE PREFERENCE FOR SIMULTANEOUS PROMPTING AND CONSTANT PROMPT DELAY PROCEDURES FOR TEACHING SKILLS TO CHILDREN WITH AUTISM (ASHLEY NIEBAUER)

WE USED A “CONCURRENT-CHAINS” PROCEDURE TO DETERMINE PREFERENCE BETWEEN TWO PROMPTING PROCEDURES.

METHOD PROCEDURE • PREFERENCE ASSESSMENT: CONCURRENT-CHAINS PROCEDURE • Each child assigned 3 color cards • Determined by color preference assessment

METHOD PROCEDURE • PREFERENCE ASSESSMENT: CONCURRENT-CHAINS PROCEDURE • Each color randomly assigned to Constant Prompt Delay, Simultaneous Prompting, or Control conditions

METHOD PROCEDURE • PREFERENCE ASSESSMENT: CONCURRENT-CHAINS PROCEDURE Control • Each color randomly assigned to CPD, SP or control conditions

METHOD PROCEDURE • PREFERENCE ASSESSMENT: CONCURRENT-CHAINS PROCEDURE Control Simultaneous • Each color randomly assigned to CPD, SP or control conditions

METHOD PROCEDURE • PREFERENCE ASSESSMENT: CONCURRENT-CHAINS PROCEDURE Control Simultaneous Constant Prompt Delay • Each color randomly assigned to CPD, SP or control conditions

• 3 COLOR CARDS PLACED ON THE TABLE IN A RANDOM ARRAY EACH SESSION • CHILD CHOOSES A CARD & CORRESPONDING PROCEDURE IS PUT IN PLACE • EACH DAY OF RESEARCH = 3 FORCED CONDITIONS AND 4 FREE-CHOICE CONDITIONS Control Simultaneous Constant Prompt Delay

SAMPLE RESULT

RESULTS

OVERALL CONCLUSIONS AND RECOMMENDATIONS • 1. PREFERENCE ASSESSMENTS ARE CONDUCTED TO IDENTIFY POTENTIAL REINFORCERS D. ON’T ASSUME.

OVERALL CONCLUSIONS AND RECOMMENDATIONS • 2. TEACHING SKILLS, INCREASING APPROPRIATE BEHAVIORS, AND REPLACING INAPPROPRIATE BEHAVIORS ALL REQUIRE IDENTIFYING REINFORCERS. • FUNCTIONAL BEHAVIORAL ASSESSMENT IS A PROCESS TO IDENTIFY REINFORCERS FOR PROBLEM BEHAVIOR.

OVERALL CONCLUSIONS AND RECOMMENDATIONS • 3. RESULTS FROM THE VAST MAJORITY OF RESEARCH INDICATES THAT PREFERENCE IS A RELATIVELY GOOD PREDICTOR OF REINFORCEMENT EFFICACY. • PREFERENCE ASSESSMENT, HOWEVER, MAY UNDERESTIMATE THE EFFECTIVENESS OF LOWER PREFERRED STIMULI AS REINFORCERS.

OVERALL CONCLUSIONS AND RECOMMENDATIONS • 4. REINFORCERS INCLUDE MANY MORE THINGS THAN FOOD OR TOYS. • BREAKS • ACTIVITIES

OVERALL CONCLUSIONS AND RECOMMENDATIONS • 5. IN GENERAL, PREFERENCES APPEAR TO REMAINSTABLE. • THEREFORE, CLINICIANS/TEACHERS CAN LIKELY SPEND TIME IN BEGINNING PHASES OF TREATMENT OR SCHOOL YEAR ESTABLISHING THE STABILITY OF EACH PERSON’SPREFERENCES.

OVERALL CONCLUSIONS AND RECOMMENDATIONS • 6. IF POSSIBLE AND ETHICAL, LIMIT POWERFUL REINFORCERS OUTSIDE OF TRAINING. • 7. TRY TO FIND OR TEACH NEW REINFORCERS ON A FREQUENT BASIS. • 8. USE A COMBINATION OF ASSESSMENTS, DIRECT AND INDIRECT… • 9. BUT CONSTANTLY BE THINKING ABOUT WHETHER YOU HAVE STIMULI THAT FUNCTION AS REINFORCERS. • 10. REMEMBER REINFORCEMENT CANNOT FAIL!

THANK YOU • SOUTHWEST CENTRAL COOPERATIVE • AMBER BRUNS

THANK YOU KLATTKP@UWEC. EDU