Professional Documents
Culture Documents
Student: Alan Cummins 1165236 Course: PSY286 Lecturer: Dr. Sinead Eccles
Contents
Sniffy had previously been shaped to associate the sound of the foot-dispenser with the
dropping of a food pellet. That is Sniffy had already been magazine trained. There were two
reinforcers in use: Food as the primary reinforcer and sound as the secondary reinforcer. The
operant-level or base-line level behaviour for Sniffy was that of mixed exploration, sniffing and
rearing up against the cage walls. This operant level had its rate and pattern of emitted
responses noted. Based on Skinner, 1938 operant conditioning was carried out, where the
animal was presented with a reinforcing stimulus, that of food, immediately following the
The cumulative records (See Figure 1 and Figure 2) consist of sections split into 5 minute
segments with 75 responses till he market returns to its base position. Reinforced responses
Conditioning occurs if and only if the response rate increases in rate of occurrence, magnitude
of Sniffy was carried out using reinforcement of successive approximations of the desired target
behaviour. As indicated in Table 1 actions such as rearing up, paw lifting and so on were
reinforced to train Sniffy to move closer to the food dispenser and finally to associate the action
of bar pressing with the reward of food. Looking at Figure 1 it can be seen that Sniffy began
with occasional spontaneous bar presses causing food dispensation but spent most time
was reinforced with food pellets. The action strength increased thereby strengthening the
linking of behaviour to obtaining food. Four major behavioural changes can be seen on the
cumulative chart. There was an increase in response frequency, noted by the number of
reinforced responses as the training progressed and also by the steepness of the cumulative
line. Other behaviour such as locomotion and grooming decreased as paw lifting and bar
pressing increased. There was sequential change in the responding. That is there was a steady
rate of advancement to the desired behaviour. Finally there was a change in response
variability. Sniffy began carrying out many different behaviours but via shaping decreased these
down till eventually he was consumed with bar pressing alone by the end of the cumulative
report.
Extinction is behaviour changes that occur when a previously reinforced behaviour no longer
produces reinforcement. Looking at Figure 2 the association of bar pressing and sound was
strongly associated with food dispensation. Once the bar press triggering release of food pellets
and the sound of the bar press itself were turned of, extinction began. There was an extinction
burst at the start where Sniffy repeatedly hit the bar. He doesn’t step down to examine the bar
or the cage because there is still a strong association with food and bar pressing. Extinction
occurs quickly as the steepness of the cumulative graph decreases rapidly. This is due to
continuous reinforcement having previously been used to shape Sniffy. A fixed ratio of one bar
press to one food pellet release means Sniffy immediately begins to notice when no food is no
variability of the form and magnitude of the response. As indicated in Table 2 there is an
increase in exploration and grooming. This is due t the disruption in the loop of behaviour that
characterised the reinforced operant. By the end of the cumulative graph extinction has
completely occurred and Sniffy makes no bar press and spends his time exploring and
grooming. In this software based experiment it means that Sniffy has returned to his base-line
operant level.
Record of Behaviour from Operant Conditioning Training Session (See Table 3 for Key)
Record of Behaviour from Operant Conditioning Training Session (See Table 3 for Key)
Key Description
PL Paw Lift
FT Face Touch
S Sniff
L Locomotion
HSE Hopper Sniff, Eat
D Drinking
BP Bar Press
BPE Bar Press Eat
HL Head Lower
Table 3 - Sniffy Behavioral Repertoire Key
The Sniffy experiment of Shaping and extinction does agree with previous research and in
applied situations. The following are criticisms and strengths of the experimentation carried out
and its application in the real world via previous research studies carried out.
The Sniffy experiment for shaping uses operant conditioning. Operant behaviour
operant behaviour. There is a high degree of similarity between humans and other species. The
Sniffy experiment can be related to previous research in animals and then in applied
behavioural analysis with human participants. Operant conditioning has important adaptive
mechanism that has been favoured by natural selection. While the Sniffy experiment makes use
human behaviour. Hall, Lund and Jackson, 1968 have carried out research on disruptive
children in school and now operant conditioning can be used to increase wanted behaviour.
Although many species in the studies such as pigeons, rats and dogs generate similar and
difficult to observe and evaluate. Humans have different verbal rules for describing
one or many humans becomes one of determining what the exact reinforcement stimulus is.
This can be overcome by shaping successive guesses which change both the verbal rule being
1990 for example of such. Bentall, Lowe and Beasty, 1985 have noted that young children
produce patterns of behaviour typical of other species. In applied settings such as classroom,
home, hospitals, institutions and society behavioural programs rely heavily on principles of
operant conditioning. It should be noted that Behaviourists, in the main, are interested in
‘learning’ rather than in instinct, the behavioural predispositions that an individual organism
possesses by virtue of its membership of a particular species. Research has begun to examine
ontogenetic influences on behaviour and this work should be considered in relation to operant
conditioning as a whole. Reinforcement as applied in Sniffy has been extended out into society
in many research studies. Garcia, Guess and Byrnes, 1973 looked at improving simple language
skills of retarded children by giving sweets as reinforcement when correct responses where
elicited. Cartwright and d’Orso, 1993 took an inner city school and transformed it into a
dynamic learning centre via behavioural shifts with operant conditioning. Tokenized economies
were put in place by Ayllon and Azrin, 1965 to promote positive and active engagement in
behaviour with institutionalised groups which had been suffering from lethargy, inaction and
depression. Hall et al, 1968 focused on disruptive children of 6-8 year old and made attention
and praise contingent on studying thereby sharply increasing the amount of study in class and
There are several different schedules of reinforcement that may be used within
behavioural analysis. The Sniffy experiment makes use of a continuous fixed ratio of one
reinforcement per desired behaviour. Other schedules include those such as a ratio scheduling
where reinforcement if given at fixed or intermittent points once a certain number of desired
behaviours have been noted, or interval scheduling where reinforcement is given after a certain
period of time has occurred. Intermittent reinforcement is generally of more benefit for
Intermittent reinforcement resembles many situations in every day life where behaviour is
maintained by a reinforcer that occurs only occasionally and is highly resistant to extinction.
While Sniffy had a steady fixed rate of reinforcement Cohen, Chelland, Ball and LeMura, 2002
have found that responses become faster as you increase the ratio requirement. This means
Sniffy could have been modified to use intermittent reinforcement where the amount of
desired behaviour before reinforcement would be increased till an optimum level at which
reinforcement rates at maximised. Stephens, Pear, Wray and Jackson, 1975 found that while
working with a child, Sidney that his learning rate increased as they increased the interval ratio
reinforcement, however does have a place within applied behavioural analysis. Continuous
reinforcement if administered after every desired or correct response is the most effective
pattern of reinforcement in which some portion of correct responses are reinforced can then
intermittent reinforcement have great utility for generating stable long term baselines of
learned behaviour against which effect of drugs, physiological manipulations, emotional stimuli
and motivational factors can be studied. Intermittent reinforcement produces reliable and
Extinction Alone
Extinction in the Sniffy experiment simply worked in isolation. The pairing of behaviour
and reward was removed immediately. The base-line operant level returned to its original level
as if no shaping had occurred. This is not the case in research using live animals or human
participants. There is never a full return to the base-level. Extinction is much more effective
when used in conjunction with reinforcement. Extinction alongside positive reinforcement was
used by Pinkston, Reese, LeBlanc and Baer, 1973 while observing preschool children and the
aggressive behaviour of one student called Cain. They ignored the aggressive behaviour,
essentially extinguishing the reinforcement of attention that he had previously been getting
and reinforced only whenever good behaviour was noted. Sniffy is an isolated and simple case
and it should be noted that resistance to extinction is high when a large number of responses
have been reinforced and is low when responses require great effort. Extinction can be
confused and obfuscated by behavioural processes such as fatigue, habituation, satiation and
punishment. Extinction brings with it increased behaviour in other areas once the reinforced
behaviour has been extinguished. Eckerman and Lanson, 1969 noted that response rates
decreased but that there was a change in behaviour as well as their experimental subjects,
applied in applied settings such Allyon, Michael, 1959 where there was a need to reduce visits
of a psychiatric patient to nurses, in France and Hudson, 1990 where children had difficulties
sleeping and parents where asked to extinguish attention to reduce awakening in middle of
night. That is extinction useful in situations in which reinforcer maintains behaviour can be
readily identified and controlled. However, extinction doesn’t develop new pro-social
behaviours to replace responses that have been extinguished. There can be increased
aggression if extinction occurs in an immediate and absolute way such as in Sniffy. Aggression
upon extinction of a shaped response within pigeons was noted by Azrin, Hutchinson and Hake,
1966 with previously unseen aggressive pecking taking the place of the extinguished behaviour.
Extinction is not recommended for behaviours that are dangerous or highly disruptive, such as
in self-harm. Sniffy gave example of an extinction burst. With self-harm it is too dangerous to
the client to simply ignore behaviours to try to extinguish them. Furthermore the amount of
responding in extinction is affected by number of reinforcers and the effortful ness of the
response during the previous period of reinforcement. Even more powerful influence on the
resistance to extinction is the schedule on which reinforcers were previously delivered. Partial
reinforcement extinction effect has been used to overcome persistent problem of treatment
gains that have not been maintained after behaviour intervention has been withdrawn. Kazdin,
1994, Nation and Woods, 1978 and Tierney and Smith, 1988 have all researched training at a
continuous schedule until it occurs at a high rate and then intermittent reinforcement is
invoked, which is where the schedule is incrementally gradually till the client is responding on a
learning difficulties.
Sniffy is a simplified version of a rat with a much reduced behavioural repertoire. This
makes it simpler to identify and classify actions and successive steps towards or from a desired
behaviour. The behaviour to be reinforced is close to an action that would be carried out
anyway. It becomes more difficult in applied situations. In such situations there is greater
difficulty in identifying rewards and punishments. Particularly in the case of humans whom
have different ideational concepts of what constitutes a reward for one as compared to
another. Primary reinforcers such as food in the Sniffy experiment are generally not used and
secondary reinforcers have to be identified instead. Reinforcers are identified by their function.
Primary or unconditioned reinforcers and conditioned reinforcers come into play with a
reinforcement, Fantino 1977. Sniffy suffers none of the emotional and physical limits that
hinder human behavioural analysis. Sniffy is never satiated, tired, bored or frustrated. The
reinforcer of food never grows old for Sniffy whereas in applied situations reinforcers do lose
value over time and need to be modified and extended over a period of analysis. Difficulties
arise in trying to make internal changes with human clients. The cycle of When, do, get must be
broken so as the behavioural analyst and in turn the primary care-giver do not become
Much research has been carried out into response shaping. Sniffy is shaped by
successively reinforcing closer and closer approximations to a desired behaviour, namely hat of
bar pressing. Pear and Legris, 1987 carried out response shaping by reinforcing successive
closer approximations of pigeons head to a spatial position. Similarly Midgley, Lea and Kirby,
1989 looked at shaping rats to deposit ball bearings by successive approximations of the
desired behaviour. This mimics what occurred within the Sniffy experiment carried out.
human problems. Many such problems can be characterised as behavioural deficits. The
individual concerned does not succeed in making normal responses and thus his or her
behaviour is not maintained by the social consequences that influence the behaviour of others.
Galbicka, 1994 used response shaping in applied settings with the use percentile reinforcement
schedules while Howie and Woods, 1982 used shaping to increase fluency in stuttering. It
should be noted that a non-zero operant level is essential for simple operant conditioning. That
is there must be a base-line operant behaviour or set of behaviours that can be shaped. An
example of such is the many studies regarding speech development in children and adults as
compared to studies such as Hayes, 1951 which tried to shape apes to increase speech skills.
Ayllon, T., and Michael, J. (1959). The psychiatric nurse as a behavioral engineer. Journal of the
Experimental Analysis of Behavior, 2, 323-334.
Azrin, N.H., Hutchinson, R.R., and Hake, D.F. (1966). Extinction-induced aggression. Journal of
the Experimental Analysis of Behavior, 9, 191-204.
Bentall, R.P., Lowe, C.F., and Beasty, A. (1985). The role of verbal behaviour in human learning.
Journal of the Experimental Analysis of Behavior, 43, 165-183.
Cartwright, M., and D’Orso, M. (1993). For the Children: Lessons from a Visionary Principal.
Doubleday.
Catania, A.C., Matthews, B.A., and Shimoff, E. (1990). Properties of rule-governed behaviour
and their implications. In D.E. Blackman and H. Lejeune (Eds.), Behaviour analysis in
theory and practice. Hillsdale, New Jersey:Erlbaum.
Cohen, S.L., Chelland, S.A., Ball, K.T., and LeMura, L.M. (2002). The effects of fixed-ratio
schedules of reinforcement on exercise by college students. Perceptual and Motor
Skills, 94, 1177-1186.
Eckerman, D.A., and Lanson, R.N. (1969). Variability of response location for pigeons responding
under continuous reinforcement, intermittent reinforcement and extinction. Journal of
the Experimental Analysis of Behavior, 12, 73-80.
France, K.G., and Hudson, S.M. (1990). Behavior management of infant sleep disturbance.
Journal of Applied Behavioral Analysis, 23, 91-98.
Galbicka, G. (1994). Shaping in the 21st century: Moving percentile schedules in applied settings.
Journal of Applied Behavioral Analysis, 27, 739-760.
Goldstein, A.P., and Krasner, L. (1987). Modern applied psychology. New York: Pergamon.
Hall, R.V., Lund, D., and Jackson, D. (1968). Effects of Teacher attention on study behavior.
Journal of Applied Behavioral Analysis, 1,1-12.
Hayes, C. (1951). The ape in our house. New York: Harper and Row.
Howie, P.M., and Woods, C.L. (1982). Token reinforcement during the instatement and shaping
of fluency in the treatment of stuttering. Journal of Applied Behavioral Analysis, 15, 55-
64.
Kazdin, A.E., and Polster, R. (1973). Intermittent token reinforcement and response
maintenance in extinction. Behavior Therapy, 4, 386-391.
Kazdin, A.E. (1994). Behavior modification in applied settings. Fifth Edition. Pacific Grove,
California: Brooks/Cole.
Midgely, M., Lea, S.E.G., and Kirby, R.M. (1989). Algorithmic shaping and misbehaviour in the
acquisition of token deposits by rats. Journal of the Experimental Analysis of Behavior,
52, 27-40.
Nation, J.R., and Woods, P. (1978). Persistence: The role of partial reinforcement in
psychotherapy. Journal of Experimental Psychology:General, 109, 175-207.
Pear, J.J., and Legris, J.A. (1987). Shaping by automated tracking of an automated response.
Journal of the Experimental Analysis of Behavior, 47, 241-247.
Pinkston, E.M., Reese, N.M., LeBlanc, J.M., and Baer, D.M. Independent control of a pre-school
child’s aggression and peer interaction by contingent teacher attention. Journal of
Applied Behavior Analysis, 6, 115-124.
Stephens, C.E., Pear, J.J., Wray, L.D., and Jackson, G.C. (1975). Some effects of reinforcement
schedules in teaching picture names to retarded children. Journal of Applied Behavioral
Analysis, 8(4), 435-447.
Tierney, K.J., and Smith, H.V. (1988). The effect of different combinations of continuous and
partial reinforcement on response persistence in mentally handicapped children.
Behavior Psychotherapy, 16, 23-37.