Chapter 6: Aversive Regulation of Behavior

1
Chapter 6Aversive Regulation of Behavior
2
Classes of Reinforcing and Punishing Stimuli
3
Contingencies of Punishment

• Positive punishment occurs when a stimulus is
  presented following an operant and the rate of
  response decreases. ex. Spanking
• Negative punishment occurs when a stimulus is
  removed reliant on a response.

4
Contingencies of Punishment

• The Premack principle states that the opportunity
  to engage in a higher frequency behavior will
  reinforce a lower frequency response. Essentially
  reinforcement is relative and not absolute. This
  also applies to punishment.

5
Reinforcement and Punishment

• A reinforcement can be either the presentation of
  a desirable item such as money or food, or the
  removal of an unpleasant stimulus, such as verbal
  nagging or physical pain.
• A punishment can be the removal of a desirable
  condition such as driving privileges or the
  presentation of an unpleasant condition such as
  physical pain.

6
Reinforcement and Punishment

• All things being equal, most people will respond
  better to both immediate reinforcement and
  immediate punishment.
• Most punishments in American society are given
  for behaviors that are immediately reinforcing,
  (drug use) while the threat of the punishments
  for these deeds is delayed and uncertain.

7
Reinforcement and Punishment

• Punishment, by itself, tends to be ineffective
  except for temporarily suppressing undesirable
  behavior. Only very severe punishment can produce
  long-term suppression of behavior
• Mild, logical and consistent punishment can be
  informative and helpful.

8
Punishment A process that makes a behavior less
likely to occur again (suppression)
Decreases chances of
Behavior
followed by
Punishment
9
Does punishment work? Both Thorndike and
Skinner and others concluded that it does
not. Skinner said it only temporally suppressed
behavior but did not suppress or eliminate it
long-term. Ex) Skinner (1938)
10
(No Transcript)
11
Effectiveness of Punishment

• Azrin and Holz 1966
• Manner of introduction
• Immediacy of punishment
• Schedule of punishment
• Schedule of reinforcement
• Motivational variables
• Availability of other reinforcers
• Punisher as discriminative stimulus

12
Making Punishment Most Effective

• Abrupt Introduction of punishment
• Pigeons suddenly received a moderate shock
  (80V)which irreversibly suppressed responding
  whereas a gradually increasing the shock
  intensity (60V-130V) responding would continue
  past the moderate shock intensity.
• Intensity of punishment
• High intensity positive punishment may
  permanently suppress behavior due to the
  organisms discontinue of responding
• If organisms do respond after punishment,
  behavior eventually recovers to prepunishment
  levels.

13
But maybe the intensity of the punisher (slap)
was not aversive enough. Boe and Church (1967)
replicated Skinner and used shock instead of
slaps. Six groups of rats received a different
intensity of shock.
14
(No Transcript)
15
Williams et al (1993) study with a severely
mentally retarded child emitting self-destructive
behavior. Mild shock was not effective but more
intense shock was. A secondary punisher (verbal
reprimand) was all that was needed to maintain
the suppression. Response ---gt Reprimand
Shock Response ---gt Reprimand
16
Making Punishment Most Effective

• Immediacy of punishment
• Punishment is most effective at reducing
  responses when it is presented shortly after the
  behavior
• Schedule of punishment
• Punishment delivered continuously is more
  effective versus intermittently. As the rate of
  punishment increases the response decreases.

17
Delay of Punishment The longer the delay between
response and punisher, the less effective the
punishment will be. Response -------------------
------gt Punisher Ex) Solomon et al (1968) with
dogs 15 sec delay --gt Suppression for 3
minutes 5 sec delay --gt 8 days 0 sec delay
--gt 2 weeks
18
Procrastination Humans (and animals) tend to
behave in ways that produce delayed punishment
over immediate punishment, even if the delayed
punisher is of greater intensity. Ex) Cramming
for an exam Study some now -------gt Small
punisher now Cram a lot later -------gt Large
punisher later
19
Consistency (Schedule) of Punishment Low ratio
schedules of punishment are more effective than
high ratio schedules. Ex) Punishment after
every response (FR 1 schedule) compared with
punishment after every 10th response (FR 9
schedule). Human examples?
20
Stimulus Control of Punishment Suppression of
operant responding in presence of specific
stimuli that have been present during
punishment. Honvig and Silva (1964) Pigeons
were punished for pecking at a 550 nm light but
not for pecking other colors.
21
(No Transcript)
22
At first, suppression generalized to other
wavelengths, but later the generalization
reversed (discrimination occurred). Suppression
had come under stimulus control of the 550 nm
light. Human examples? Ex) Bucher and Lovaas
(1968) with an autistic child and
self-destructive behavior
23
(No Transcript)
24
If using punishment

• If you are punishing a specific behavior, the
  reinforcement for that behavior should be
  discontinued,or at least reduced or made
  available contingent upon some other appropriate
  behavior.
• Punishment only teaches one thing What NOT to
  do. What kind of teacher would only try to teach
  by just saying No!?

25
Motivation and Punishment

• When motivation is reduced the punishment is most
  effective.
• Azrin trained pigeons to peck a key on a VI
  schedule and later they introduced 160V shock for
  every 100th response. The response rate continued
  at 60, 65, 70, 75 and 85 of the pigeons free
  feeding weight. At 85 of free feeding weight the
  responding stopped.
• Behavior may be completely suppressed when the
  motivation to respond is low.
• Research suggests when motivation is increased
  after punishment, responding will not recover to
  prepunishment levels.

26
Negative Reinforcement
Escape Avoidance Discriminated avoidance Non-discr
iminated avoidance
27
Escape vs. Avoidance

• Negative reinforcement occurs whenever an operant
  results in the removal or prevention of a
  stimulus and the operant subsequently increases
  in rate
• Negative reinforcement contingencies control
  escape and avoidance
• The actual distinction between what is escape and
  what is avoidance is difficult to maintain as the
  aversive stimulus from which a subject may escape
  will re-occur with high probability in an operant
  chamber (the shock-shock interval) and with
  somewhat less probability in the natural
  environment
• In general, a subject will acquire escape before
  avoidance
• Avoidance responses are VERY resistant to
  extinction

28
Escape vs. Avoidance

• Escape responding is reactive, it is acquired
  more readily than avoidance responding.
• Avoidance responding is proactive and will only
  be acquired after a history of escape.

29
Types of Avoidance

• If a warning signal precedes the aversive
  stimulus and a response during this warning
  stimulus prevents the aversive stimulus delivery,
  this constitutes discriminated avoidance.
• Discriminated avoidance can be very difficult to
  acquire, as the warning stimulus comes to be a CS
  that will elicit respondents that can interfere
  with the avoidance.
• If the avoidance response is part of an
  organisms Species Specific Defensive Responses
  (SSDRs), the avoidance response will be much more
  readily acquired than otherwise. Example Pigeons
  typically fly away from a threat or an aversive
  stimulus, hence flapping wings or flying will be
  easily acquired as an avoidance response.
  Pecking a key or pressing a foot-treadle would be
  much more difficult to acquire as an avoidance
  response than wing flapping or flying. Likewise
  rats innately freeze in response to a threat
  before running away. During the warning
  stimulus, a rat will freeze which is incompatible
  with lever pressing lever pressing to avoid
  electric shock is acquired slowly at best with
  the rat as subject.

30
A species specific defensive response of an
opossum, freezing and gaping.
31
Nondiscriminated Avoidance

• If no warning stimulus precedes the aversive
  stimulus which itself occurs periodically, this
  constitutes nondiscriminated avoidance. In the
  operant chamber, the subject must perform the
  avoidance response periodically, once every 60
  seconds or so, to prevent the aversive stimulus
  from occurring.
• Many behaviors of clinical interest are avoidance
  responses to unseen or unsignalled aversive
  events, most obvious in obsessive compulsive
  disorder. The behaviors of Freuds patients were
  avoidance responses controlled by interpersonal
  (family, relationship, marriage) aversive events.
  The behaviors were labeled as dissociative or
  conversion but they cannot be explained as being
  due to the condition. A label for a behavior
  cannot be used to explain the behavior. A person
  who displays/performs different personalities is
  said to have dissociative identity disorder.
  When asked why the person displays the different
  personalities, one cannot logically say it is due
  to the dissociative identity disorder.

32
What maintains avoidance responding

• After an avoidance response, nothing happens, so
  why is it maintained?
• Two factor theory
• Classical Conditioning The warning stimulus
  acquires aversive effects and elicits fear
• Operant Conditioning responding during the
  warning stimulus reduction of fear
• Evidence supporting two-factor theory
• Second order conditioning, the warning stimulus
  can be used to condition another aversive
  stimulus
• Problems with two-factor theory
• Avoidance without fear or fear reduction, the
  animal subject may shown no overt signs of fear,
  the fear is inferred
• Extinction of Avoidance
• Where is the extinction? If the warning stimulus
  is a conditioned aversive stimulus, then since
  responding to it prevents the shock US, these are
  in effect, extinction trials but avoidance
  responses are very resistant to extinction

33
One-Factor Theory

• Operant Conditioning Only
• Sidman aka nondiscriminated avoidance, avoidance
  is acquired with no discernible warning stimulus
  to become a conditioned aversive stimulus
• Shock-Shock or S-S interval 10 secs for example
• Response-Shock or R-S interval 15 seconds for
  example, thus a response delays a shock but some
  responses still followed by shock
• If the R-S interval is greater than the S-S
  interval, there will be an overall reduction in
  frequency of shocks. Avoidance is maintained by
  an overall reduction in frequency of shock
• The R-S interval has to be greater than the S-S
  interval for avoidance responding to be learned.

34
Herrnstein and Hineline

• In the previous example, once could still argue
  that the passage of time was an aversive
  stimulus, hence, two factor theory was still in
  effect.
• Hineline and Herrnstein programmed contingencies
  such that in a two-second interval, the p. of a
  shock was .3 in the absence of a response but was
  reduced to .1 in the next two seconds if a
  response occurred. Hence responses were still
  followed by shock but at lower probability. With
  shock occurring at random intervals, these
  contingencies produced avoidance responding in 17
  of 18 subjects. Overall reduction in shock
  frequency WAS enough to maintain avoidance
  responding. Hence two-factor theory is not
  necessary to account for avoidance.

35
Shock Frequency and Avoidance Behavior

• From a molecular perspective, the
  moment-to-moment time between shocks (S-S) and
  the time from response to shock (R-S) represent
  the essential variables regulating avoidance
  responding.
• Nonetheless, the bulk of the evidence supports a
  molar perspective, suggesting that the molar
  variable, overall reduction in shock frequency
  (or sensitivity to rates of shock), can produce
  and maintain operant avoidance.

36
Reasons to not use punishment

• Behavioral persistence Since punishment can be
  so effective, it is reinforcing for the user, the
  user can come to rely on it too much.
• Seligman
• Repeated exposure to aversive events that are
  unpredictable and uncontrollable can have
  debilitating effects, learned helplessness and
  depression.
• Other emotional behavior occurs in response to
  punishment, pain-elicited aggression and
  counterattack, the person who delivers punishment
  is feared
• Behaviors involving fear or anxiety can be
  exacerbated
• Social disruption, the subject who is punished
  drops out of school, society, family.

37
Learned helplessness

• An animal experiences learned helplessness when
  it is exposed to an aversive stimulus and are
  unable to escape. After several pairings of this
  condition the animal gives up and stops
  attempting to escape. When given the opportunity
  to escape learned helplessness is demonstrated by
  not attempting to escape.

38
Helplessness and Depression

• Helplessness is involved with and is a model for
  depression. Depression may arise when a person
  feels inescapable of abuse.
• Helpless dogs (have previously learned
  helplessness) which are forced to make a response
  that escapes shock begin to make that response on
  their own. Depressed individuals may go through
  treatment in which they are not allowed to fail.
• In this situation, the person may learn to emit
  appropriate responses in the presence of aversive
  events

39
Helplessness and Depression

• Although animal experiments shed light on human
  behavior, there are differences
• Human verbal behavior allows a person to talk
  about his/her problems and attribute them to a
  cause (internal or external)
• Attributing a problem to internal causes (e.g.,
  I am a failure) could function as a
  discriminative stimuli for giving up or losing
  hope

40
Aggression

• Respondent aggression occurs when painful stimuli
  are presented to two organisms and the organisms
  attack each other. This may also be known as
  pain-elicited aggression. The probability of
  aggression increased as more and more shocks were
  presented. This result of aggression to aversive
  stimuli applies to humans

41
Operant Aggression

• When one person punishes anothers behavior, the
  punished individual may retaliate, a strategy
  known as operant aggression
• One way to escape aversive stimulation is to
  eliminate/neutralize the person who is delivering
  it
• Operant aggression is shaped and maintained
  through negative reinforcement

42
Aggression breeds more aggression

• Operant and respondent principles suggest that
  the presentation of an aversive stimulus may
  elicit or set the occasion for aggressive
  behavior
• Consider the interactions of the Israelis and the
  Palestinians For every Palestinian attack, there
  is an Israeli counterattack to punish and
  prevent any further such attacks.
• Every Israeli attempt at punishment and
  prevention simply leads to yet another
  Palestinian attack.
• Are the Israelis technically using punishment
  upon a specific target behavior and specific
  individual?
• Skinners principles are confirmed by controlled
  experiments showing that both physical and verbal
  provocation produces aggression

43
Social Disruption

• When punishment is used to decrease behavior, the
  attempt is usually made to stop a particular
  response
• Hopefully unpunished behavior is not affected,
  but two factors work against this
• The person delivering the punishment
• The Setting
• Both become conditioned aversive stimuli (Save)
• This negative side effect is known as Social
  Disruption
• Also, a social agent who frequently uses
  punishment becomes a conditioned punishing
  stimulus, who presence can disrupt all ongoing
  operant behavior

44
Coercion and its Fallout

• Murray Sidman, a prominent behavior analyst, has
  researched the social disruptive effects of
  aversive control
• Coercion Use of punishment and the threat of
  punishment to get others to act as we would like,
  and to our practice of rewarding people just by
  letting them escape from our punishments and
  threats
• Involves the basic contingencies of punishment
  and negative reinforcement

45
Coercion and its Fallout

• Dropping out is an escape contingency, and a
  major social problem
• People drop out of education, family, personal
  and community responsibility, citizenship,
  society, and even life
• Sidman notes the common element b/w these forms
  of conduct is negative reinforcement
• Once involved in an aversive situation, a person
  can get out by removing themselves from the
  situation, thus strengthening the behavior of
  dropping out