Learning: The Behavioral Perspective

This is Unit Four, part two, on the behavioral perspective in practice: schedules of reinforcement, the behavioral teacher, and the discipline of measuring consequences. Part one closed on intermittent reinforcement, which makes behavior slower to fade. Part two turns that finding into classroom practice. Carry one idea through this part: a consequence counts as a reinforcer only when it measurably increases the behavior it follows. At the entry on rewards and reinforcers, the part shows why many classroom reward systems never find out whether they work. By the end you will be able to identify a reinforcement schedule. To tell primary from secondary reinforcers. To break a skill into a learning hierarchy. To explain why a reward is not always a reinforcer. And to describe how behavioral programs fade to natural reinforcers.

Six entries on schedules and kinds of reinforcement, from continuous and partial schedules to tokens and time-out.

Every time, or only some of the time.

EDP 0324 Continuous and Partial Reinforcement

A continuous schedule reinforces a behavior every time it occurs, bringing fast learning but rapid extinction once reinforcement stops. A partial schedule reinforces it only some of the time, by count of responses or by elapsed time: ratio schedules count responses, interval schedules time them, and each is fixed or variable.

Identifying the four partial schedules.

EDP 0325 Identifying a Reinforcement Schedule

There are four partial schedules. Fixed-ratio reinforces after a set number of responses, as with pay per item produced. Variable-ratio reinforces after an average but unpredictable number, as with a slot machine. Fixed-interval reinforces the first response after a set time, as with a monthly salary. Variable-interval reinforces the first response after an average but unpredictable time, as with checking email. A professor who gives unscheduled quizzes, three one week and none for two, puts studying on a variable-interval schedule. Decide ratio or interval first, responses or time, then fixed or variable. Ratio schedules produce faster responding, and variable schedules stronger responding.

Innate and learned reinforcers.

EDP 0326 Primary and Secondary Reinforcers

Primary reinforcers satisfy innate needs, such as food, water and touch; secondary reinforcers are learned, gaining value through association, such as praise, money and stickers, so a sticker means nothing to a toddler until it is linked with approval. Most classroom reinforcers are secondary, so their value depends on what they stand for.

Token economies.

EDP 0327 Token Economies

A token economy gives students tokens, such as stickers or points, for target behaviors, which they later exchange for rewards they value, as a sticker chart earns points toward extra free time at the week's end. Immediate reinforcement works better than delayed, which is why tokens are given at once and exchanged later.

Where reinforcement comes from.

EDP 0328 Many Sources of Reinforcement

In classrooms, reinforcement comes from teachers, from classmates and from the activity itself, often several at once, and it can support intrinsic as well as extrinsic motivation, as a runner who times herself is reinforced by beating her own record. When the activity reinforces itself, the motivation is intrinsic, developed in unit seven.

Time-out.

EDP 0329 Time-Out

Time-out is a form of negative punishment in which a child is removed from a desirable activity for a short period after misbehavior, as a child who grabs toys sits apart briefly while the game continues. Guidance on calm delivery accompanies it, but durations are not rules; the course describes the practice without prescribing it.

Seven entries on teaching behaviorally, from prerequisites and task analysis to prompts and informational feedback.

Teaching the behavioral way.

EDP 0330 The Behavioral Teacher

A teacher working behaviorally focuses on observable performance, makes sure prerequisite skills are in place, elicits rapid correct responses and follows them with fitting consequences, as a teacher checks that students can add before introducing multiplication. Behaviorists measure performance, not unseen learning, which critics say favors objectives that are readily measured.

Missing prerequisites, not missing ability.

EDP 0331 Prerequisites, Not Ability

The behavioral view traces learning failures to missing prerequisites or poor instruction rather than to a lack of ability: a student failing algebra may turn out to lack fluency with fractions. Prerequisites must become automatic, not merely familiar, before they support new learning and transfer well.

Breaking a skill into its parts.

EDP 0332 Task Analysis

Task analysis breaks a skill into the smaller skills it depends on and arranges them in a learning hierarchy, a sequence in which each skill is a prerequisite for the next, as Gagné described. Start with the target skill and ask what a learner must already be able to do; ask the same of each of those skills, working down until you reach skills the learners have mastered. Teaching then runs back up. Long division requires estimating, multiplying and subtracting, which require multiplication facts, so a student missing the facts starts there. Teaching a skill before its prerequisite produces errors that look like low ability.

Aiming for no errors.

EDP 0333 Errorless Learning

Behavioral instruction aims for errorless learning through stimulus control: specific directions stripped of extraneous information, so students respond correctly from the start, as in circle every verb rather than look over the sentences and think about the words. Rapid pacing suits basic skills, keeping students responding often and correctly.

Every student responding.

EDP 0334 Active Responding

Active responding, where students write, say or do something rather than only listen, predicts achievement, as Greenwood, Delquadri and Hall found; every student writing an answer on a whiteboard beats one student answering aloud. The authors also recommend that most practice responses be both active and correct.

Three kinds of prompt.

EDP 0335 Prompts

Prompts help a learner give the right response and come in three kinds, verbal, gestural and physical, used from least to most intrusive: a teacher first says check the sign, then points to it, then guides the student's pencil. Verbal prompts come first because they are the easiest to fade, the goal of every prompt.

Feedback that informs.

EDP 0336 Informational Feedback

Informational feedback tells a student that a response is correct and why, as in right, both sides changed equally; for errors it gives the correct response or rule and has the learner try again. Attention can reinforce errors: circling only correct letters reduced children's letter and digit reversals, and scolding is avoided.

Five entries on measuring consequences, ending with the payoff: why a reward is not always a reinforcer.

Consequence, or reinforcer?

EDP 0337 Consequence or Reinforcer?

A positive consequence counts as a reinforcer only if it is shown to increase the behavior it follows; otherwise it is only a pleasant event. A class party offered for quiet work, when the room stays as noisy as before, reinforced nothing. The same test applies to punishment: a punisher must reduce the behavior.

Doing reinforcement properly.

EDP 0338 Doing Reinforcement Properly

Using reinforcement fully means measuring the behavior first, finding out what the student actually prefers, reinforcing immediately and continuously, then fading to natural reinforcers, the behavior's own consequences, as reading for its own sake eventually replaces points earned for pages read. Most school reward systems never measure whether they reinforce at all.

Consequences meant to reduce behavior.

EDP 0339 Negative Consequences and Punishers

Consequences meant to reduce behavior include verbal reprimands, overcorrection, response cost and exclusion, each a punisher only if the behavior actually declines; losing a privilege after each outburst is response cost. Punishment can eliminate a behavior, but so can positive reinforcement of an alternative, and the course prescribes neither.

The payoff: a reward is not a reinforcer.

EDP 0340 A Reward Is Not a Reinforcer

Here the part's idea is at its sharpest. A reward is whatever the teacher intends as pleasant; a reinforcer is whatever actually increases the behavior it follows, and the two coincide less often than classrooms assume. Stickers, parties and praise are secondary reinforcers, so the same sticker can strengthen one child's behavior and do nothing for another's. Behaviorists therefore measure a baseline first, find out what the student prefers, deliver the consequence immediately and at first continuously, and check whether the behavior rises. Take a pizza party promised for a week of homework. The teacher first counts assignments from the previous two weeks. For a student who dislikes parties, nothing changes; for one who wants the teacher's attention more, a quiet note home may work instead. If the count rises and stays up, the party reinforced homework. When reward systems skip the measurement, the conclusion drawn is that reinforcement does not work, when the consequence was never a reinforcer. Measuring first keeps behavioral methods honest.

Fading to natural reinforcers.

EDP 0341 Fading to Natural Reinforcers

Behavioral programs aim to fade artificial reinforcers so that behavior comes under the control of natural reinforcers, the rewards the activity itself provides, as a student who once practiced piano for points now practices because playing well is satisfying. Skinner treated these natural consequences as the goal, the point where extra rewards are no longer needed.

Look at what you can now do. You can identify a reinforcement schedule by asking whether it depends on responses or on time, and whether it is predictable. You can tell primary reinforcers, which meet innate needs, from secondary ones, which gain value through association. You can break a skill into a learning hierarchy by working down to what students have mastered, then teaching back up. You can explain why a reward is not always a reinforcer: only a measured increase in the behavior shows reinforcement. And you can describe the goal of fading: behavior maintained by the rewards the activity itself provides. Unit Five turns to the cognitive perspective on learning.