Tony Nevin – Behavioral Momentum

•    Invited much feedback

•    “I work with pigeons” – very different engagement with the field

•    share the central concern – an interest in establishing durable behavior change

•    persistence is key

•    metaphor – a class of behavior in a given S situation can develop the functional equivalent of a mass of an object

o    trucks rolling down the road – identical

•    1 –Styrofoam

•    1 – concrete blocks

o    remove feet from the accelerator

•    heavier truck loses speed less rapidly than the lighter truck

•    can figure out the ratio of their loads is

o    can quantify the mass-like qualities of behavior

o    weight at the moment of disruption – determines how the truck will carry on

•    multiple schedule paradigm – basic momentum equation

o    (delta)B = – x/m – the change in the motion directly proportional  to force over mass

o    Behavioral mass – reinforcer rate in target situation/overall average Sr rate (life/ multiple schedules) * .5

o    If that multiple schedule is large,  that little Sr rate isn’t going to be as strong as if things are “boring”

o    Multiple schedules = successive discriminations

o    More frequent Sr = more frequent R

•    Add the baselines in  - adds the momentum

•    Braking effect of providing free access – reduces responding in components

•    Reduces responding more in components where Sr is leaner

o    Behavioral mass is the square root of the reinforcer ratio

o    Important for development of a model with integrity

o    Has the properties of a physical scale

•    Additive

•    Ratio scale

•    Persistence/reinforcement – is a scientific construct that is grounded in the most basic of what is in our field

•    Generality – replicated with goldfish, humans, monkeys

o    Disrupters – food, extinction, prefeeding, Sp equally given

•    Has to be the same for both components – yoked

•    Like Dave Wacker’s list (see his presentation)

o    Bud Mace – replicated with humans in a group home – sorting red dishes and green dishes

•    Popcorn on a VI schedule

•    Then turns on the TV as a disruptor

•    Goes down on the more lean schedule

o    Cohen (1996) – typing words by college students – replicated

•    Disruptor – where’s waldo

o    Dube – also contributed

•    Momentum – important with response rate and changes in response rate

o    Not limited to rate as a measure

o    Also applicable to quality (how accurate)

•    Doesn’t equate as well as velocity

•    Mass part remains solid

o    (Nevin, Milo, Odum, & Shahan, 2003) – MTS rich v. lean

•    Conditional probability .8 v .2

•    Quantity and quality go hand in hand

•    Log responding – better than % correct – no ceiling effects?  Check this (proportion of baseline log d)

o    Delayed MTS – studying short term remembering

•    Results replicated for reinforcement

•    Standard forgetting – pretty much the same in rich and lean

•    Sheer level of accuracy related to level of reinforcement

•    More frequent Sr – more accuracy, greater response rate, greater resistance to change

•    Application

o    Must teach to fluency – rapid, accurate, persistent, retentive

o    Lindsley – do it faster

•    That’s not just it

•    Reinforcer is absolutely crucial – could it be intrinsic?

•    Reinforcement and fluency are absolutely hand in hand

•    Intense, frequent, powerful

•    Applications from the audience

o    Might  FT schedules actually build momentum in the natural environment? (at least this is how I understand it)

o    Reversing the contingencies – slower and continue to make errors more frequently

•    Can make it harder to learn the reverse performance

•    Goes against malleability

o    When do you switch the contingencies to avoid overselective responding?

•    Extinction – we use it as a test, but as a model of desirable behavior

o    Why would you use extinction as a test of desirable behavior? Shouldn’t that be resistant to extinction?

•    Especially if the behavior is “never reinforced”?

•    Might not be an ethically appropriate test for persistence

•    Will it have comparable maintaining contingencies outside the setting

•    Applications – precision teaching

o    Done with math and reading problems

o    Seems to be most relevant in special education settings – academic and classroom based

•    Eliminating and reducing undesirable behaviors

o    Self-injury – hard to place on extinction w/o restraint and blocking

o    Arrange for alternative Sr for desirable behavior

o    Adding NCR to the environment – will reduce the frequency of “key pecking” (self-injury as it relates to basic research)

o    Ends the frequency of contingent responding – noncontingent responding increases

o    Prefeeding also reduces the strength of the Sr

o    More Sr for alternative responding – more persistence

o    Yoked reinforcement schedules – response allocation follows predictable pattern  (alternative response increases)

o    Resistance – identical

•    Sr explicitly defined alternative responses – lower response rate and increases resistance to change

•    After DRA – (Mace, 1991) – problem behavior still persists

•    How can we tinker with the DRA paradigm so that change of responding is quicker?

o    Model:  alternative reinforcement disrupts target behavior via the numerator and at the same time increases is persistence via the denominator

o    Enhance resistance – put all of the Sr in the denominator

o    Reduce problem – put all of the Sr in the numerator

•    A potentially general paradigm

o    Baseline – behavior gets a reinforcer

o    Baseline – other behaviors that get other reinforcers

o    Contingency – target response gets the reinforcer

o    Alternative – gets access to a second situation – alternative, qualitatively different reinforcers for different responses (different setting)

•    Might not have the effect of strengthening the responding in setting 1

•    NEED TO REVIEW THIS – NOT AS GOOD AT UNDERSTANDING IT AS I WOULD LIKE TO BE

•    Partial Sr and Extinction

o    Is intermittent Sr more resistant to extinction?  How does that relate to momentum?

•    Not a problem for the model

•    Assume the estimation of the behavioral mass that the value of X is the same for both conditions

•    How long does it take you to figure out that the world has changed?

•    Disruptors are not equal

•    Breaks X into parts

•    Suspending the contingency

•    Generalization decrement as a function of removal of Sr – standard result

•    Which one is going to have the larger decrement?

•    Actually allows us to predict the resistance to extinction

•    Says nothing about the mass – resurgence problem

•    Whatever Sr arrangements to strengthen multiple schedules predicts what will happen with resurgence

•    Application – must think about post-extinction resurgence – how solidly did you establish the behavioral mass?