Asignatura: Psicologa del Aprendizaje.

Curso: 2014/2015.

ACTIVIDADES FORMATIVAS COMPLEMENTARIAS.

A sugerencia del equipo docente se realizarn una serie de actividades formativas que

sern complementarias al estudio del texto base. Estas actividades se irn desarrollando a

lo largo del curso y sern supervisadas por los Profesores Tutores en los Centros

Asociados. Estas actividades estn organizadas en dos bloques y a continuacin

presentamos una breve descripcin de las mismas:

1. ANLISIS DE VDEOS DIDCTICOS.

Los alumnos tendrn a su disposicin en la pgina web de la asignatura una serie de vdeos

didcticos que podrn utilizar como ayuda al estudio de la asignatura. Los Profesores

Tutores en los Centros Asociados o en su espacio de tutora del curso virtual podrn

orientar a los alumnos sobre los conocimientos que han adquirido a partir de la

visualizacin de estos vdeos. El equipo docente facilitar, a travs del curso virtual, los

cuestionarios mediante los cuales se evaluar el progreso de los alumnos. Durante este

curso se evaluar el vdeo: Programas de reforzamiento.

1. Describa los cuatro programas simples de reforzamiento.

2. Compare los programas de razn y de intervalo.

3. En qu se diferencian, si es que se diferencian, los programas de intervalo y de

tiempo?

4. Defina y compare los programas de reforzamiento diferencial de tasas bajas y tasas

altas.

5. Clasificacin y definicin de los programas complejos de reforzamiento.

2. ANLISIS DE ARTCULO CIENTFICO.

El artculo propuesto para su anlisis durante el curso 2014-15 es: Herrnstein, R.J. (1961).

Relative and absolute strength of response as a function of frequency of reinforcement.

Journal of the Experimental Analysis of Behavior, 4, 267-274. Se trata del experimento

original donde informa sobre la ley de igualacin, y que se describe en el libro de texto en

las pginas 239-240.

1. Cul es la pregunta que se pretende responder con este experimento?

2. Describa brevemente el Mtodo del experimento.

3. Cul es el principal resultado que se obtuvo y cmo fue interpretado?

RELATIVE AND ABSOLUTE. STRENGTH OF RESPONSE AS A FUNCTIONOF FREQUENCY OF REINFORCEMENT1, 2

R. J. HERRNSTEINHARVARD UNIVERSITY

A previous paper (Herrnstein, 1958) reported howpigeons behave on a concurrent schedule under whichthey peck at either of two response-keys. The signifi-cant finding of this investigation was that the relativefrequency of responding to each of the keys may becontrolled within narrow limits by adjustments in an in-dependent variable. In brief, the requirement for rein-forcement in this procedure is the emission of a mini-mum number of pecks to each of the keys. The pigeonreceives food when it completes the requirement onboth keys. The frequency of responding to each keywas a close approximation to the minimum re-quirement.The present experiment explores the relative fre-

quency of responding further. In the earlier study itwas shown that the output of behavior to each of twokeys may be controlled by specific requirements of out-puts. Now we are investigating output as a functionof frequency of reinforcement. The earlier experimentmay be considered a study of differential reinforcement;the present one, a study of strength of response. Bothexperiments are attempts to elucidate the properties ofrdlative frequency of responding as a dependent vari-able.

MI,THODSubjectsThree adult, male, White Carneaux pigeons, main-

tained at 80% of free-feeding weights, and experi-mentally naive at the start of the study, were used.ApparatusA conventional experimental chamber for pigeons

(Ferster & Skinner, 1957) was modified to contain tworesponse-keys. Each key was a hinged, translucentPlexiglas plate mounted behind a hole in the centerpartition of the chamber. The pigeons pecked at acircular.area (diameter = 0.75 inch) of the plate, and aforce of at least 15 grams was necessary to activate thecontrolling circuitry. Any effective response operatedan audible relay behind the center partition; it has beenfound that the resulting auditory feedback stabilizes thetopography of pecking. Behind each key was a groupof Christmas-tree lamps of various colors, each groupmounted in such a way that it cast significant amountsof light through only one key. The two keys were

11 wish to express my indebtedness to Dr. Douglas Anger ofthe Upjohn Company for his valuable comments concerningthe interpretation of the data in this experiment.

2The work reported in this article was supported by GrantG-6435 from the National Science Foundation, Washington,D. C. to Harvard University.

4.5 inches apart (center-to-center) around the verticalmidline of the center partition and on a horizontal lineabout 9 inches from the floor of the chamber. Througha 2-inch-square hole in the center partition, 2 inchesfrom the floor, the pigeon occasionally received the re-inforcer-4 seconds' access to grain.A masking noise and a low level of general illumina-

tion were provided.Procedure

Preliminary training lasted for two sessions of 60 re-inforcements each. During these sessions, a peck toeither key was reinforced only when the just-previousreinforcement was for a peck to the other key. Thisalternating pattern of reinforcement led rapidly to apattern of responding that consisted of almost perfectalternation between the two keys. The left key wasalways red; the right, always white.During the experiment proper, responding to either

key was reinforced on a variable-interval schedule. Theschedule for one key was independent of the schedulefor the other. Thus, at any given moment, reinforce-ment could be made available on neither key, on onekey or the other, or on both keys. A reinforcedresponse to one key had no effect on the programmerthat scheduled reinforcements on the other.The primary independent variable was the mean time

interval between reinforcements on each key. Theseintervals were chosen so that the mean interval of re-inforcement for the two keys taken together was heldconstant at 1.5 minutes.3 The over-all average value of1.5 minutes was produced by a number of pairs ofvalues for the two keys. The combined frequency ofreinforcement from independent variable-intervalschedules will be a constant if the values for each of thetwo keys are chosen according to the hyperbolicrelationship:

1 1 1_ + _ = -

x y c

in which x is the mean interval on one key, y is the meaninterval on the other, and c is the combined mean

3It should be noted that, by convention, the mean of avariable-interval schedule refers to the minimum average inter-reinforcement time and not to the actual inter-reinforcementtime obtained under conditions of responding. Thus, if aparticular animal responds very slowly, the actual meaninterval of reinforcement may be larger than the value desig-nated by the experimenter. The value designated is a mini-mum that is closely approached in practice because theanimal's rate of responding is ordinarily high in comparisonto the intervals in the reinforcement schedule.

267

R. J. HERRNSTEIN

interval for the two keys taken together. The pairs ofvalues used were VI(3) VI(3); VI(2.25) VI(4.5); VI(1.8)VI(9); and VI(I.5) VI(-)-i.e., extinction on one ofthe keys.During most of the experiment, the pigeons were

penalized for switching from one key to the other.Each time a peck to one key followed a peck to theother key, no reinforcement was possible for 1.5 sec-onds. Thus, the pigeon never got fed immediately afterchanging keys. When the pigeon switched keys beforethe 1.5-second period was completed, the period simplystarted anew. At least two consecutive pecks on a givenkey were necessary before reinforcement was possible:the first peck to start the period, and the second after itwas completed. This penalty for alternation will bereferred to as the "change-over delay of 1.5 seconds,"or COD (1.5").The sequence of pairs of values of the variable-

interval schedules and the number of sessions at eachpair of values are shown in Table 1. Key A is the left,red key; Key B is the right, white key. Sessions lastedfor 60 reinforcements, which required approximately90 minutes since the over-all mean interval of reinforce-ment was always 1.5 minutes. Whether the COD waspresent or absent is also shown.

RESULTS

Figure I shows the relative frequency with which thepigeon pecked on Key A as a function of the relativefrequency with which it was reinforced on that key.Each point on the graph is a mean of the last five ses-sions under a given pair of values of the variable-interval schedule. The COD operated on all these ses-

100

-05590 . 0

'--- 231

X-64180/

70 AW

60-

50zo 40-0-w7ce 30

20 30 40 50 60 70 8% REINFORCEMENTS ON KEY A

Fig. 1. Relative frequency of responding to Key A as afunction of relative frequency of reinforcement on Key A, forthree pigeons; COD (1.5") is present throughout.

sions; the results without the COD will be given later.The ordinate and abscissa values were calculated bycomparable methods. The number of responses(ordinate) or reinforcements (abscissa) on Key A wasdivided by the total number of responses or reinforce-ments, respectively. The five last sessions were pooledto make this computation.The diagonal line with a slope equal to 1.0 in Fig. I

shows the function that would be obtained if the rela-tive frequency of responding were exactly equal to therelative frequency of reinforcement. The empiricalvalues approximate the theoretical function with amaximum discrepancy of only about 8%. There seemsto be no regular pattern to the deviations from thetheoretical function.The absolute rate of responding on each of the keys is

shown in Fig. 2. Responses per hour are plotted againstreinforcements per hour, for each key separately andfor the two pigeons (231 and 055) that had an apprecia-ble range of the independent variable. Data from thesame sessions are plotted in Fig. I and 2. With oneexception (Pigeon 055, Key A, at 40 reinforcements perhour), the points in Fig. 2 approximate a linear func-tion that passes through the origin. It will be shownlater that this relation between absolute rate of respond-ing and absolute rate of reinforcement is the simplestone that is compatible with the relative-frequency func-tion presented in Fig. 1.

0I

wa-

(U)wU)z00.U)w

10 20 30REINFORCEMENTS PER HOUR

Fig. 2. Rate of responding on each key as a function ofrate of reinforcement on that key, for two pigeons; COD(1.5") is piesent throughout.

268

RELA TIVE AND ABSOLUTE STRENGTH OF RESPONSE

Table ISequence of Procedures

VI on VI onKeyA KeyB No.of

Subject (min) (min) Sessions COD055 3 3 20 no

2.25 4.5 18 no2.25 4.5 43 yes3 3 44 yes3 3 25 no9 1.8 35 yes1.5 ext* 37 yes9 1.8 20 yes1.8. 9 39 yes

231 3 3 35 yes3 3 17 no9 1.8 35 yes1.5 ext* 37 yes9 1.8 17 yes1.8 9 40 yes4.5 2.25 38 yes

641 3 3 17 no2.25 4.5 16 no2.25 4.5 45 yes3 3 34 yes3 3 16 no

*extinction

The number of times a pigeon changed keys de-pended on the difference in frequency of reinforcementon the two keys. Figure 3 shows this relation for thethree pigeons. The abscissa gives the difference, with-out regard to sign, between per cent of total reinforce-ment on one key and that on the other. Thus, whenthe two keys are characterized by equal relative fre-quencies of reinforcement, the value on the abscissa is 0;when the responding to one key is extinguished, thevalue is 100, and so on. The ordinate gives simply theaverage number of times the pigeon switched fromKey A to Key B, or vice versa. Once again, the dataare from the same sessions that supplied those in Fig. 1.It should be noted, however, that in Fig. 3 there areonly four values for Pigeons 055 and 231 whereas therewere six in Fig. 1 and -2. This is the result of com-bining the three pairs of variable-interval schedules in-volving mean intervals of 9 minutes and 1.8 minutes.(See Table 1.) The data at abscissa values of about70 per cent are, therefore, based on means of 15, in-stead of 5, sessions. The functions in Fig. 3 are lessconsistent than those in Fig. 1 and 2, but the fre-quency of alternations between keys clearly decreasesas the two keys are associated with increasingly differ-ent relative frequencies of reinforcement.The relation shown in Fig. 3 is found only when the

COD is in operation. Figure 4 shows the frequency of

z "\%W800

w

.\a.~~~~~~~~~~~~~~~~.

w

z

~400-w

200N

O 20 40 60 80 IOC% KEY-A REINF. - % KEY-B REINF.

Fig. 3. Number of alternations between the two keys as afunction of the absolute difference between the per cent ofreinforcements on each key, for three pigeons; COD (1.5") ispresent throughout.

key changes with and without the COD when reinforce-ment frequency is either equally or unequally dis-tributed between the two keys. The data fromPigeon 231 are omitted from this figure, because this

055 641

6000 _ RfA 9 Rf8Z RfA=Rfe0U) 5000U)

R OR

iX 4000 -Bw0.

~~~~~~~RfARf8U1) ACD 3000z

w 2000 nf

COD NO C ONO COD CNO COD CNODCOD 0COD NOD NODFig. 4. Number of alternations between the two keys when

the COD was present or absent and when reinforcements wereequally or unequally distributed between the two keys, fortwo pigeons.

269

R. J. HERRNSTEIN

bird was-not exposed to any procedure that combinedunequal frequencies of reinforcement with no COD.Two facts are evident in Fig. 4. One is that the CODmarkedly reduces the frequency of alternations betweenthe keys. The other is that unequal reinforcementfrequencies on the two keys reduce alternation onlywhen theCOD (1.5") is present.The COD also seems to play a role in the production

of the relation shown in Fig. 1, namely, the tendencyof the relative frequency of responding to match therelative frequency of reinforcement. Pigeons 055 and641 were both exposed to procedures in which the CODwas absent and'the relative frequency of reinforcementon Key A was about 66%. The relative frequencies ofresponding on Key A were 50% and 56% for the twopigeons, respectively. In both these cases the de-partures from matching are outside the range ofdepartures obtained when the COD is present. (SeeFig. 1.)

DISCUSSION

The major problem posed by the present experimentis to explain the simple correspondence in Fig. I be-tween the relative frequency of reinforcement and therelative frequency of responding. In a sense, this cor-respondence is readily explained by the curves in Fig. 2,which suggest that the relation between the absoluterate of responding and the absolute rate of reinforce-ment is a linear function that passes through the origin.If this relation is represented asp = ke, in which p and edenote the absolute frequencies of pecking and eating,then the simple matching function, of Fig. I may beexpected to follow the form

pi ke,pi + p2 k(e, + e2)

The constant, k, drops out and the remaining expres-sions on each side of the equation denote relative fre-quencies of responding and reinforcement. Theequality of these two relative frequencies ma-y thus beregarded as a consequence of a linear relation, of anyslope and zero intercept, between' the absolute fre-quencies. Moreover, this relation between the absolutefaites of responding and reinforcement is one that isconsonant with a plausible view of response strength:Rate, of responding is a linear measure of responsestrength, which is itself a linear function of frequencyof reinforcement. The correspondence in Fig. 1 wouldthereby result from the fact that the behavior on eachof the two keys obeys a simple linear rule governingstrength of response. According to this point of view,thea.jm Js match relative.frquency of responding torelative fquency of reinforementnot. because theyt'ke into account what is happening on the two keys,buit-b@vause they". respond to the two, keys in-dependently.-The' critical relation, p = ke, has been assertedbefore. Skinner (1938, p. 130) has discussed a

quantity called the extinction ratio, which is the totalnumber of responses divided by the number of rein-forced responses in a fixed-interval schedule of re-inforcement.4 He presented a small amount of data thatindicated that this quantity remained constant as thesize of the fixed interval was varied. The constancy ofthe extinction ratio is merely another form, p/e = k,of the function we find.

Perhaps the greatest vulnerability of the foregoingaccount lies in its simplicity. If it were true that therate of responding is so simply related to the frequencyof reinforcement, the fact ought to have been wellestablished by now. We should expect that behavior ina single-key situation would reveal the same linear re-lation shown in Fig. 2, and that with all the work donewith the single-key problem, the nature of the relationbetween rate of responding and frequency of reinforce-ment would be known. Unfortunately, this informationis not'available. In few studies has the frequency ofreinforcement been varied over an adequately widerange. Those which have done so have usually also in-volved manipulations in other, and possibly con-taminating, variables.A small amount of relevant material is shown in

Fig. 5. These curves are adapted from earlier studiesby Clark (1958), Wilson (1954), and Herrnstein (1955).These three experimenters observed the convention ofplotting the independent variable as inter-reinforcementtime, rather than frequency of reinforcement. Clarkand Wilson used rats (Wilson used fixed-interval, in-stead of variable-interval, schedules); Herrnstein usedpigeons. Rate of responding clearly increases withfrequency of reinforcement. In these one-responsesituations, however, we do not obtain the linear, func-tion with zero intercept that was shown in Fig. 2. Therelation suggested by Fig. 5 has downward concavity.Even if this concavity is taken to represent nothingmore than a natural ceiling on the rate of responding,the function is still inappropriate, because the interceptis greater than zero.

Perhaps a more relevant comparison can be madewith some data of Findley (1958), who devised a modifi-cation of concurrent scheduling not unlike the presentprocedure. A pigeon responds to a key and is rein-forced on a variable-interval schedule. By pecking asecond key, the pigeon alters the color of the firs't key.Each color on the first key signifies a particular valueof the variable-interval schedule.' The two variable-interval schedules are independent, just as in the presentstudy. The difference between Findley's procedure and

4Skinner defined the extinction ratio as the number of un-reinforced responses divided by the number of reinforcedresponses, but in actual computation he used the total numberof responses divided by the number of reinforced responses.The difference is of no significance for the present discussionsince both definitions imply a linear relation with zero inter-cept between absolute rate of responding and absolute rate ofreinforcement.

270

RELATIVE AND ABSOLUTE STRENGTH OF RESPONSE

600

cr sooD0I

r 400(w

w 300z0C,,wcr 2004

1004

0 20 40 60 80 oo 120 160 200 240 280 320 360REINFORCEMENTS PER HOUR

Fig. 5. Data from previous experiments replotted to showrate of responding as a function of frequency of reinforcement.

ours is in the character of the switching response.Switching required a peck on a second key in Findley'sexperiment, whereas in ours the pigeon had only tomove over. In the present experiment, the discrimina-tive stimuli for both schedules were concurrently visi-ble; in Findley's, only one was present at a time, butthe other was always available via a switching response.Figure 6 shows the relation between absolute rate ofresponding and absolute frequency of reinforcementobtained by Findley. Findley did not keep the totalfrequency of reinforcement constant as he varied theaverage inter-reinforcement interval associated with thetwo colors. Pigeons 5 and 6 had a constant value of6 minutes in one color, and values ranging from 2 to20 minutes in the other. The graphs are for the vary-ing component. Responding in the other componentwas not, however, constant. Reynolds (1961) has dem-onstrated a similar kind of interaction in an ordinarymultiple schedule. For Pigeons 2 and 4, the scheduleswere varied in both components. Only for Pigeon 5does the function appear linear with an intercept ofzero. For the three other pigeons, the pattern was thesame as in Fig. 5: The relation is either concave down-wards or linear with an intercept greater than zero.Our results suggest that a relative-frequency function

with a slope of less than 1.0 over part of the rangewould have been obtained if it were not for the COD.The precise correspondence between relative frequencyof responding and relative frequency of reinforcementbroke down when the COD was omitted. When therelative-frequency relationship has a slope of less than1.0, then the absolute-frequency relationship must

either be concave downwards or linear with a positiveintercept. Thus, the present experiment shows excellentmatching, on the one hand, and atypical absolute ratefunctions, on the other, probablybecause of the COD.

It remains to be explained why the COD has the effectof bringing the empirical points closer to the perfect-matching function. The data in Fig. 4 show that theCOD greatly reduces the frequency of alternation be-tween the two keys. Without the COD, switching isreinforced by the occasions when, the first peck to akey produces food; with the COD, these occasionsnever occur. In a sense, then, switching is a thirdoperant in the situation and is extinguished by theCOD. The abundance of switching with no CODwould tend to make the frequency of responding to thetwo keys more nearly equal than they would be ifswitching were not being reinforced. The reduction ofswitching by the COD probably does not, however, ex-plain why the absolute rate of responding in the presentexperiment follows the simple linear function. At best,one would expect the absolute rate to behave the way itdoes in single-key experiments. Single-key experiments,including Findley's version of the concurrent schedule,yield functions between absolute rate of responding andabsolute frequency of reinforcement that predict non-matching functions between relative frequency of re-sponding and relative frequency of reinforcement. Away of characterizing this finding is to say that in thesingle-key situation, the animal responds too much atthe low frequencies of reinforcement or too little at thehigh. Thus, the curves in Fig. 5 and 6 have interceptsgreater than zero or are concave downwards. The samewould apparently have been true in the present experi-

6000 -

I

LU0

Cr)LUJ(I)z0al.C'-)LUJa:

5000

4000

3000

2000 _

1000

0 Lo l0 20

REINFORCEMENTS PER30

HOUR

Fig. 6. Data from Findley's experiment (1958) replotted toshow rate of responding as a function of frequency of rein-forcement. Numbers on curves refer to individual subjects(pigeons).

0O /H ERRNSTEIN (1955)

0

0

0

0 CLARK (1958) WILSON (1954)

n , I I '

RI-

/

/

271

272 R. J. HERRNSTEIN

ment had it not been for the COD. Why the COD hasthis effect is intuitively, if not scientifically, obvious.With a COD, two things are likely to happen; bothfollow from the fact that once the animal has switchedto a key, it is likely to stay there for at least the dura-tion of the COD. First of all, if the tendency to re-spond to a key is low, then the COD will probablypush the tendency even lower because switching to thekey calls for not one but a number of responses. Sec-ond, if the tendency is high, then the total number ofpecks to the key will probably be increased becauseeach switch to the key guarantees a number of re-sponses. These presumed effects would change a func-tion that is concave downwards or linear with a posi-tive intercept toward a function that is linear with anintercept of zero. As an analogy, separating the twokeys spatially may have effects similar to those of theCOD. If the two keys were far apart, the animal wouldprobably be less likely to go to the less lucrative keythan if only a small distance were involved, and re-sponding to the more lucrative key would be increased.A COD whose duration is longer than the 1.5 secondsused here might give a matching function with a slopegreater than 1.0 and absolute-rate functions that areconcave upwards or linear with intercepts less thanzero.The suggestion of the present discussion is that the

surprisingly precise correspondence between relativefrequency of responding and relative frequency of rein-forcement arises from the function relating absolute fre-quency of responding and absolute frequency of rein-forcement. When this function is Jinear -with1anintercept of zero, matching is found. In singk-k4ysituations, this linear relation is not obtained; and iLisalso not obtained under concurrent schedules unlesssome additional procedural factor reduces the pigeon'stendency to over-respond at low frequencies of rein-forcement and under-respond at high. The COD issuch a procedural factor; but others, such as distancebetween keys or effort involved in the response, mayalso be satisfactory. The duration of the COD may ormay not be critical in the effect it has on the slope ofthe relative-frequency function. If a broad range ofdurations of the COD all give approximately perfectriatching, then it seems correct to say that the con-current procedure is a good one for studying absolute,

as well as relative, strength of responding. In single-key situations, the rate of responding is not very sensi-tive to frequency of reinforcement. This insensitivityprobably weakens our interest in the concept of strengthof response. It may be that the concept can be givensignificant empirical support in multiple-key situations.

SUMMARY

A two-key, concurrent procedure involving avariable-interval schedule on each key was used. Thevalue of the mean interval on each key was varied overa range from 1.5 to 9 minutes, but the total frequencyof reinforcement for the two keys taken together washeld constant. The pigeon was penalized for alternatingin response between the two keys by making reinforce-ment impossible for 1.5 seconds after every alternation.It was found that the relative frequency of respondingon a given key closely approximated the relative fre-quency of reinforcement on that key.

REFERENCES

Clark, F. C. The effect of deprivation and frequency of rein-forcement on variable-interval responding. J. exp. Anal.Behav., 1958, 1, 221-228.

Ferster, C. B., and Skinner, B. F. Schedules of reinforcement.New York: Appleton-Century-Crofts, 1957.

Findley, J. D. Preference and switching under concurrentscheduling. J. exp. Anal. Behav., 1958,1, 123-144.

Herrnstein, R. J. Behavioral consequences of the removal ofa discriminative stimulus associated with variable-intervalreinforcement. Unpublished doctoral dissertation,Harvard Univer., 1955.

Herrnstein, R. J. Some factors influencing behavior in a two-response situation. Trans. N. Y. Acad. Sci., 1958, 21,35-45.

Reynolds, G. S. Relativity of response rate and reinforcementfrequency in a multiple schedule. J. exp. Anal. Behav.,1961, 2, 179-184.

Skinner, B. F. The behavior of organisms. New York: D.Appleton Century Co., 1938.

Wilson, M. P. Periodic reinforcement interval and number ofperiodic reinforcements as parameters of responsestrength. J. comp. physiol. Psychol., 1954, 47, 51-56.

Received November 14, 1960.

Herrnstein, R. J. (1961). Relative and absolute strength of response as a function of

frequency of reinforcement. Journal of the Experimental Analysis of Behavior, 4,

267-274.

Mtodo

Sujetos

Los sujetos experimentales fueron 3 palomas macho adultas experimentalmente

ingenuas y mantenidas al 80% de su peso ad libitum.

Aparatos

Se us una cmara experimental para el condicionamiento de palomas con dos

teclas de respuesta.

Procedimiento

El entrenamiento previo dur dos sesiones con 60 reforzadores cada una. Durante

estas sesiones se reforzaba un picotazo a la tecla slo cuando el reforzador previo se

haba conseguido por responder a la otra tecla. Con este procedimiento se consegua

rpidamente un patrn de alternancia casi perfecto. La tecla izquierda siempre se

ilumin de rojo y la derecha siempre de blanco.

Durante el experimento, responder a las teclas fue reforzado con un programa de

intervalo variable (IV). El programa para una tecla fue independiente del programa para

la otra. As, en un momento dado, el reforzador poda estar disponible en ambas teclas,

en una, en la otra o en ninguna. Una respuesta reforzada en una tecla no tena efecto

sobre el programa que funcionaba para la otra tecla.

La principal variable independiente fue el valor del programa de IV utilizado en

cada tecla. Los pares de valores usados fueron: IV 3-min IV 3-min; IV 2.25-min IV 4.5-

min; IV 1.8-min IV 9-min; IV 1.5-min Extincin.

Durante la mayor parte del experimento las palomas fueron penalizadas por

cambiar el picoteo constantemente de una tecla a la otra. Cada vez que se picaba a una

tecla y luego a la otra, el reforzador dejaba de estar disponible durante 1.5 segundos.

Esto es lo que se llama demora por el cambio de 1.5 seg (abreviadamente DPC1).

La secuencia de pares de valores de los programas de IV y el nmero de sesiones

para cada par de valores se muestra en la Tabla 1. La Tecla A es la roja izquierda y la

Tecla B es la blanca derecha. Tambin se indica si la DPC estuvo presente o no. Las

sesiones terminaban despus de 60 reforzadores.

Tabla 1

Resultados

La Figura 1 muestra la frecuencia relativa con la que la paloma picoteaba la Tecla

A en funcin de la frecuencia relativa con la que era reforzada en esa tecla. Cada punto

de la grfica es una media de las ltimas cinco sesiones bajo un par dado de valores de

los programas de IV. La DPC operaba en todas estas sesiones; los resultados sin DPC se

vern ms tarde. Los valores de los ejes de ordenadas y abscisas se calcularon con

mtodos similares. El nmero de respuestas (ordenadas) o de reforzadores (abscisas) se

divida por el nmero total de respuestas o de reforzadores, respectivamente. Para hacer

este cmputo se usaron las cinco ltimas sesiones.

Figura 1

La lnea diagonal con valor de 1.0 en la Figura 1 muestra la funcin que se

obtendra si la frecuencia relativa de respuesta fuera exactamente igual a la frecuencia

relativa de reforzamiento. Los valores empricos se aproximan a esta funcin terica

con una discrepancia mxima del 8%.

1 La Demora Por el Cambio es un procedimiento que se superpone al programa de

reforzamiento en curso, y por el que no se refuerza la respuesta de cambio durante un

tiempo (1,5 segundos en el caso del presente experimento). Dado que se utilizan

programas de IV, el cambio puede coincidir con la disponibilidad de reforzamiento. De

no utilizar la contingencia DPC se podra reforzar indirectamente la conducta de

alternancia.

En la Figura 2 se muestra la tasa absoluta de respuesta en cada tecla. En la grfica

aparecen respuestas por hora y reforzadores por hora para cada tecla por separado en las

dos palomas (231 y 055) que fueron expuestas a un mayor rango de combinaciones de

programas de reforzamiento. En las Figuras 1 y 2 aparecen datos de las mismas

sesiones. Con una excepcin (Paloma 055, Tecla A, con 40 reforzadores por hora) los

puntos de la Figura 2 son una funcin lineal que pasa a travs del origen. Esta relacin

entre la tasa absoluta de respuesta y la tasa absoluta de reforzamiento es la ms simple

que sea compatible con la funcin de frecuencia relativa presentada en la Figura 1.

Figura 2

El nmero de veces que una paloma cambiaba de tecla dependa de la diferencia de

frecuencia de reforzamiento entre las teclas. La Figura 3 muestra esta relacin para las

tres palomas. El eje de abscisas es la diferencia, en valores absolutos, entre el porcentaje

de reforzadores totales en una tecla y la otra. As, cuando las dos teclas tenan la misma

frecuencia relativa de reforzamiento, el valor de la abscisa es cero; mientras que cuando

las respuestas a la Tecla B no se reforzaron (extincin), el valor es 100. El eje de

ordenadas es el nmero medio de veces que la paloma cambia de la Tecla A a la B, o

viceversa. Los datos de los valores del eje de abscisas en la Figura 3 son menos

consistentes que los de las Figuras 1 y 2, pero la frecuencia de cambios entre las teclas

claramente disminuye cuando los programas de reforzamiento asociados a cada tecla

presentan importantes diferencias en cuanto a frecuencia de reforzamiento.

Figura 3

La relacin mostrada en la Figura 3 se encuentra slo cuando opera la DPC. La

Figura 4 muestra la frecuencia de cambios con y sin DPC cuando la frecuencia de

reforzamiento estuvo igual o desigualmente distribuida entre ambas teclas. Los datos de

esta figura son claros. La DPC reduce marcadamente la frecuencia de cambios entre las

teclas. La distribucin desigual de frecuencia de reforzamiento en las dos teclas reduce

los cambios slo cuando la DPC estuvo presente.

Figura 4

La DPC tambin parece jugar un importante papel en la relacin mostrada en la

Figura 1, es decir, en la tendencia a igualar la frecuencia relativa de respuesta con la

frecuencia relativa de reforzamiento. Las Palomas 055 y 641 fueron expuestas a

procedimientos en los que la DPC estuvo ausente y la frecuencia relativa de

reforzamiento para la Tecla A fue del 66%. Las frecuencias relativas de respuesta para

la Tecla A fueron del 50% y del 56%, respectivamente en las dos palomas. En ambos

casos se separaban de los resultados de igualacin obtenidos cuando la DPC estuvo

presente (vase la Figura 1).