8,ha,ioural Processes. 18 (1989) 17-34 Elsevier BEllAVIOURl\L CORRELl\TES
or
17
EXTlUNSlCALLY REINFORCED VIDEO GlIIIE PlAY
Department of PsychOlogy, University of Melaide, 1Ulelaide, s.A., 5000 (Austra.lia) SChool of SOCial Studies, SOuth Australian Institute of Technology, Melaid!!, S .A'
r
5000 C1U1Stralial ( Acecpted J October t 988 )
l\BS'l'RJ\C'l'
Dalziel, r.n., Metzer, J.e. aoo waters, L.K., 1989. Behavioura.l Correlates of Extrinsically ReinforcecfVideo Game Play.
Behav.
Process. IS: 17-34
Three experirnents which investigate human observers r conceptions of the nature of extrinsic reinforcement correlated hehaviour (KReS) are reported. in the first two experiments. observers viewed players at a video game under conditions of extt,insic and no extrinsic reinforcement. Observers' tasks 'oJ.ere to identifY'the extrinsic rcinforcemant periods and to provide evidence for these by referring to any player behaviour. Ttle observer statements were submitted to content analyses from which emerged a nWilber of EReB categories, tile most frequently · report.etl ones being Icontrol grip', ltime taken between games', 'movements' and 'smiling/grinning I. In the third experi!lent the 'smi ling/grinning' cue was Rlallipulated lly instruction ,to an actor in order to simulate extrinsic and no extrinsic reinforcement COnditions. Observers were able reliably to identify lIiIRUIated -extrinsic reinforcen".ent com1itions particularly when they reported (and presUll'.ably had attended. tot the changes in the 'smiling/griMing' cue. various implications arisinq from the suggested existence of rRCJJ are diSCUSSed including its place in motivational effects on (tevelopnental changes, socialisation processes and cross-cultural differences. KEY OORDS
Correlates of reinfOrcet:lent, play. video gar.es, intrinsic, extrinsic
INTRODUCTION MiDIS (1984) has dra'ffl attention to the desirability of directing research towards movement sequences and their constituent movement segments as well as to acts if important behavioural information is not to be lost. In the context of theories of reinforcement this means paying morc attention to features of the subject'5 behaviour other than the rising rate with WIlich the reinforcement contingency is met. Such features have inteI'l'littently been m.tters of interest in the past. Going back to the l1ery origins of instrUllental conditioning, Thorndike (1898) \las certainly:as much. concerned ..,ith t he growing stereotypy of his animals' escape responses iJ'J !lith the reduction in their latency. I
0376·6357/S9/SO).SO ® 1989 Ebevier 5 . 10), suggesting'that 9aOO
perfo11l'lance is not affected by extrinsic reinforceNllt", ro,:ty observers \fete gene~allY Wlable' to identify correctly ",hJ.c~ phase was paid i22 correct ind.ications; l 'S incorrect ' indications; Sign test: p) 0.3) .• . They aiel, ~owever, 'offer "a large DUJnber of statem«lts' lJitJl re~t to their obs'orvations ',about .the ,~layerS' performanc'o to support their d"e cision .
louestlon. 3 and 7; Instructio.sl. ill of the euthors (jUdges) independently conductod content· aJlalyses on tlie stat·OJII)nts by fono-,Ing ·the guideline. of III ""cl.)'\ln9
~ ll
abstractions and interpretations, and .(2)
grouping the remaining observations into categories of meaningfUl ' behavioural ~imilarity. 1\Ielve behaviour~ categories accounted for aU s tatements. Tlies'e 12 behavioural categories were '(with e~les of statement s in brackets): Ca) COntrol grip (-Tight grip on controlS") (b) MOvements (-IncreaSed leg Movements"; "EXaggcrated hand movements' ) (c)
Game length
(nplay~
longer. each Uraoll )
("Wiping s~eat off palm of narnl") Smiling/Grinning ("8ig grin lihen successful-: "Often smiles when lOses")
(d, . S~reating
(el (g}
Vocali sation ("Looked Hke i1e 'Was coWlting"' Time taken betue~n games (1I0Ui~ game restarts ")
(h)
Grimaces
(il
Attention to vidoo screen ("Kept eyes continually on screen") Posture (nuead. bent .t.:ell forward")
(II
(j)
(kl (1)
(aGr~raaCed
when hI)
fir~")
Eyes ("Eyes vide open")
Sig'hing/Yawning {"Decreased amount of sighing ....hen bis lIlen were shot dO'm
24 Table I.
Number of observer statements in response to Que,s tions 3 and 7 (Instructions) independently divided behavioural categories by two judgBsin Experiment 1.
_~_ _ _ _into:
Behavioural Categories
Table 1 presents a sllIlIllary of the t.-o judges' frequency tallies (independently conducted) of statements falling into these 12 behavi oural categories. The above content analyses were imposed on ohservers I statements irrespective of the correctness of their decisions. lnterjudge reliability UJR) \las 80.4\ for all 51 statements analysed by the judges. In sununary. although observers apparently could not identify the phases in which payment occurred, their observations of players J behaviour in tlie two phases were sufficiently consistent to enable 'the two jUdges to classify t.h~ statement!> into the same 12 behavioural categories. EXPERIMENT 2
The lack of clear evidence that observers in Experiment I could reliably distinguistl between paid and unpaid phases may have been an artifact of the An and BA. designs in w11ich observers were exposed to only one change of poase. If each observer were exposed to a larger number of phase changes and also to both directions of phase change Ii.e', frOll A. to D and fr01:'l B to A) reliability of correct identification of A and B phases might improve. In addition, to explore the generality of the verbal responses obtaine~ from observers in EXperiment 1 r a second level of extrinsic reinforcement was introduced by doubling t11e rate of payment to players. METHOD
Design Ten experimental' sessions consisted of 4 5-minute phases, precea8d The sessions were ei ther of the design ABAB or BASA: in A phases money was earned for performance. and in B phases no money was earned. In all, 43 neW' observers viewed players in Experiment 2. by a 5-minute warm-up period.
25
~T;i~~~
empl~yed
recording 'equipment used was the same as that in Experiment, 1. However. instead of a I,.{~rning bell. IoIhich might ))9,ve been difficUlt to hoar, a single blue 40 watt globe placed directly', in front of the player and clearly visible to observers served to indicate changes in phases. 'Light on' indicated money phases and 'light off' indicated no ..money phases. 'l'he light was Controlled by timers auudeSe the experimental room. Instructions Observers received preliminary instructions before tho players arrived. They then completed identical anlwer sheets during eaCh phase. They answered the fo11olling qucGtions: 1. "Is the plnyer playing for money no'o1?" 2,. "What is it about his/her behav)t')ur that m3de you 91 ve the above ansloIer?" ' Mter the exporiment the observers anslolercd a final question: "Havo you any fUrther comments on, what it was about the player's behaviour which cuggested he/she was or loIas not playing for money when tho light was either on or Off?1I Players r.eceived instructions abOUt tM game and about the experimental design as in Experiment 1. [lrocedure }\ftar the observers were seated the player was intror;1uced to the oxperimental room as in Experiment 1. The warm-I.lp period thon began. If the 1\ Phase was first , the signal light wns turned off during the warm-up: if the e phase was first, the light was on. Mter the warmup 1\ and D phases alternated. The nwrber of points earned in each phase \laG computed and money earned in 1\ phases at' the rate of 2 cents for each 10 points was set aside for colle satisfaction upon lIaking a hit, she conveys self pleasure"; and for tl Sl}e looked al.Jl.ost happy when the cannon ....as shot do'm " ). (3) Off-cue behavioural state:nents (e.g. "She kept ber eyes on the Phasa B:
scroenlt ~ .
{41
Off-cue abstractlonal staterr.ents le.g. "She has beCOr"e more inVOlVed
in tile game" l . If an obso[ver made at least one statement in a category a score of f l' was assigned to that category" and if no statenents vere D!ade in a category a score of '0' ;.ras ass igned to that category. The total number of poss i rsie patterns of responding for observers in the abo...·(! fOllr categories thereby became 16 . Response patterns f[om observers C0111d consequently be i dentified as I cue-Used , patterns (if onc or more statements •...C[C classi fied in categories (1) and/or (Z» or as fClleooNot-Used' patterns fif ,no"statement
-as classified in either category
(1)
or 12)). Independent ,."'lyses were
conducted for A and Po phases for the t ...'O sets of statement~ '" (as judged by
tile t\lO judges) . Hence :four separate aruuyses lIere con(!ucted: to determine the associations betvcen the cue-Used, cwHlot-Uscd patterns and the correctness of each observer's deCision. itJese arc presented in Table IV. Chi squared analyses supported. the hypothesis that observers reporting the relevant cue were more likely to be correct in their decision that "a phase \las 'paid' or lunpaid'. The final experil1l9nt in this series deraonstrates that ....hen nonverbal bel1aviour deriving UltimatelY fron the verbal behavi our of obServers in the fi rst t\JO experiments is presented to fresh observers, they 'appropriately detect' t he reinforcen:.ent circumstances uru:tcr Uhich the behaviour is generated. In particular r observers WilD report.atl attending to the manipulated cue of contingent SlIiling on vins ~re rore likely to identify extrinsic reinforcement conditions than observers -":00 did not report attending to this cue.
31
Table IV.
Relationsh.ips between patterns of observers' statements (Cue-Used vs . cue-Nat-Used) and the correctness of observers' decisions. as
dBtermined for 1\ and B phases by tve. judges in EXperiment 3.
--------------------Judge 1
Judge 2
A Phases
CUG-Used
Observers' Decision
Correct
44
Cue~Not-Usec1
Cue-UIl(:d cue-Nat-Used
42
17
Incorrect _ - ' ' -_ _-'-'17
19
15
i= 16 .0** B Phases Cue-Used
Observers' Decision
Correct Incorrect
33
CUe-Nat-Used
12
CDEl'-Used Cue-Nat-Used 31
14
28
_-"-_~_-,,,,29
ir;;.19.7·~ · • p ( 0.01 !2-taUed) u p ( 0.001 (2-tailedl
GENER1\L 'DISCUSSION
OVera:1,l '," it WO~ld appear from this serios of experilTlents that thece 'i s indeed · reasonable evldenc(f for the existence of a concept of the human £Rca. Further, the ' behaviour of 'Smiling/Grinning' at apPfI>priate points in"the schedule of reinforcement can be regarded ,as an important part of 'this concept wi thin this particular experimental scttin;.' ,While in these experiments no direct response measure was taken, a reasonable judgement of tile constancy of the players' response ' rat:es during paid and unpaid phases of the game may be derived from an' examination of ' tho point scores, given that players had 6ufficient ' pra'ctlc~ dUring wannup. This evidence wholly in accorc;1.. with the: assumptions which ,the study ,Is based t;hat a strongly intrinsically reinforced response would: not show a further rise in the rate of response if extrinsic reinforcement were.to be appl1e'~ . . If the concept of !!:RCB indeed corresponds to an actual J:lICJ3. and certainly this is tlle next 1011ca1 step of investigat~on, that Is requt'red, then the data obtained in these experln\~ts , m~y be exp;ained rather c.onvenient'ly as follows. 1'he finding that observors 'wcre able ' to,
is
on
ns are nol, of CO'D:rse r an itmnediate (!onc:ern,
i\ll tba.t has so fa.~ been da:nonstratcd iSI firstly, human observers 'Will generate p:.rticular verbal ~~haviour \lOd!!l' conditions Where, if there ls a human
ERCB, it should fonn part of the stimulus controlling that verbalbationi, secondly, wilen part of that verbalisation is translated back into nanverbal behaviour and W;':d as a stimulus, it evokes an appropriate verbal response 'in hunan obsorvers. H. is still possible that ilther parts of the or19'ln8.'. Verbalisation [i.e. parts other thM 'Sr.tiling/GrilUlinq') might not evoke a similar response. and this IlUSt be te~ted. MSO, other reinforcers or levelS, of reinforcement rri9ht not give simlar results, this being a ~ssible explanation for the major findings Of EKp&iment 1 and 2. Most UrlJOrt,a ntly. using trained observers, it has to be soo'olI\ that the alleged !!ReB coes, in fact, rise in frequency wi ttl reinforce~t, i. e. that the ERCB is truly an Mca and not just a CO/lllTlOn huraan concept of •...hat F.RCB Jli9bt be. The present study suggests a profitable behavioural starti[}g' p:>int for sUC'h an ill'Ye5tiqation.
34 REFERENCES
Adams', J.1\., 1984. Learnin9 movement sequences. PsyChological Bullethl, 96 : 1-28.
1\tari Inc.
Instructions accompanying ' space Invaders' games.
talk, J.L.. 1961 . production of polydipsia in normal rats by an intermittent food schedule. science, 133 : 195-196.
HaracktewiC:l, J.ftf., Manderlink, G. and Sansone, C • • 1984.
Rewarding pinl;)aU
wizardry: Effects of evaluation and cue value on intrinsic interest. Journal of Personality and Social PsychOlOgy, 47 : 287-300. Hearst, E., 198,4. ,i\bsence ' &s . information: Some implications , of l earning. performance, and represantatlonal processes.
In H.L.
Roit~lat.
'r.G. eever
and· U.S. Terrace (Eds.), 1\nimal Cognition. Hillsdale, N.J. : Erlbaum. Lepper, . M.a.. , Greene , ,D. Md Nisbett, R.E., ,1973. Undermining children's intrinsic interests with extrinsic rewards : 1\ test of the "overjustificationfl hypotheSis. Journal of personality and Social Psychology, 28:
129-137.
Millcnson, J.R. and Leslie, J .C. , 1979. principles of Behavioral Analysis. Neu York: Macmillan Publishing co., Inc. Newman, J., Wolff, W.T. ami Hearst, E., 1980. The feature positive effect in adult human subjects. 30.urnal of EXperimental PsyChOlogy: Human learning and Hemery , 6, 6)0-650. Notterman, J . M. and Mintz, D.E., 1965 . DynamiCS of Response. New York: John Wiley and Sons, Inc . . Sch...... rtz, 8., 1982 . Reinforcem.ent-induced bel1avioral stereotypy: HoW' not to teaCh peOplQ to discover rules. Journal of. EXperimental PsychOlOgy: General" 111: 23-59. Skinne~, n.F. and Morse, w.n . , 1958. Fixed-interval re inforcement of running in a wheel. Journal of the Experimental Analysis, of Behavior. 1: 371"379.
Thorndike, E. t . ,
189B. Animal Intelligence.
.MOnograph Supplement, No . 8.
PSYChOlogical Review:
waters, L.K., 1980 . social reinforcement of the work beha\.'!our of .' retardates and normals. Unpubli!:ihed doctoral dissertation', uni versity of AdelaidO. . Wilson, M.P. and Keller, F.S. , 1953. On the select.ive roinforcement of spaced . responses. Journal of Comparative and Physiological Psychology, 461 19D-193.
or
17
EXTlUNSlCALLY REINFORCED VIDEO GlIIIE PlAY
Department of PsychOlogy, University of Melaide, 1Ulelaide, s.A., 5000 (Austra.lia) SChool of SOCial Studies, SOuth Australian Institute of Technology, Melaid!!, S .A'
r
5000 C1U1Stralial ( Acecpted J October t 988 )
l\BS'l'RJ\C'l'
Dalziel, r.n., Metzer, J.e. aoo waters, L.K., 1989. Behavioura.l Correlates of Extrinsically ReinforcecfVideo Game Play.
Behav.
Process. IS: 17-34
Three experirnents which investigate human observers r conceptions of the nature of extrinsic reinforcement correlated hehaviour (KReS) are reported. in the first two experiments. observers viewed players at a video game under conditions of extt,insic and no extrinsic reinforcement. Observers' tasks 'oJ.ere to identifY'the extrinsic rcinforcemant periods and to provide evidence for these by referring to any player behaviour. Ttle observer statements were submitted to content analyses from which emerged a nWilber of EReB categories, tile most frequently · report.etl ones being Icontrol grip', ltime taken between games', 'movements' and 'smiling/grinning I. In the third experi!lent the 'smi ling/grinning' cue was Rlallipulated lly instruction ,to an actor in order to simulate extrinsic and no extrinsic reinforcement COnditions. Observers were able reliably to identify lIiIRUIated -extrinsic reinforcen".ent com1itions particularly when they reported (and presUll'.ably had attended. tot the changes in the 'smiling/griMing' cue. various implications arisinq from the suggested existence of rRCJJ are diSCUSSed including its place in motivational effects on (tevelopnental changes, socialisation processes and cross-cultural differences. KEY OORDS
Correlates of reinfOrcet:lent, play. video gar.es, intrinsic, extrinsic
INTRODUCTION MiDIS (1984) has dra'ffl attention to the desirability of directing research towards movement sequences and their constituent movement segments as well as to acts if important behavioural information is not to be lost. In the context of theories of reinforcement this means paying morc attention to features of the subject'5 behaviour other than the rising rate with WIlich the reinforcement contingency is met. Such features have inteI'l'littently been m.tters of interest in the past. Going back to the l1ery origins of instrUllental conditioning, Thorndike (1898) \las certainly:as much. concerned ..,ith t he growing stereotypy of his animals' escape responses iJ'J !lith the reduction in their latency. I
0376·6357/S9/SO).SO ® 1989 Ebevier 5 . 10), suggesting'that 9aOO
perfo11l'lance is not affected by extrinsic reinforceNllt", ro,:ty observers \fete gene~allY Wlable' to identify correctly ",hJ.c~ phase was paid i22 correct ind.ications; l 'S incorrect ' indications; Sign test: p) 0.3) .• . They aiel, ~owever, 'offer "a large DUJnber of statem«lts' lJitJl re~t to their obs'orvations ',about .the ,~layerS' performanc'o to support their d"e cision .
louestlon. 3 and 7; Instructio.sl. ill of the euthors (jUdges) independently conductod content· aJlalyses on tlie stat·OJII)nts by fono-,Ing ·the guideline. of III ""cl.)'\ln9
~ ll
abstractions and interpretations, and .(2)
grouping the remaining observations into categories of meaningfUl ' behavioural ~imilarity. 1\Ielve behaviour~ categories accounted for aU s tatements. Tlies'e 12 behavioural categories were '(with e~les of statement s in brackets): Ca) COntrol grip (-Tight grip on controlS") (b) MOvements (-IncreaSed leg Movements"; "EXaggcrated hand movements' ) (c)
Game length
(nplay~
longer. each Uraoll )
("Wiping s~eat off palm of narnl") Smiling/Grinning ("8ig grin lihen successful-: "Often smiles when lOses")
(d, . S~reating
(el (g}
Vocali sation ("Looked Hke i1e 'Was coWlting"' Time taken betue~n games (1I0Ui~ game restarts ")
(h)
Grimaces
(il
Attention to vidoo screen ("Kept eyes continually on screen") Posture (nuead. bent .t.:ell forward")
(II
(j)
(kl (1)
(aGr~raaCed
when hI)
fir~")
Eyes ("Eyes vide open")
Sig'hing/Yawning {"Decreased amount of sighing ....hen bis lIlen were shot dO'm
24 Table I.
Number of observer statements in response to Que,s tions 3 and 7 (Instructions) independently divided behavioural categories by two judgBsin Experiment 1.
_~_ _ _ _into:
Behavioural Categories
Table 1 presents a sllIlIllary of the t.-o judges' frequency tallies (independently conducted) of statements falling into these 12 behavi oural categories. The above content analyses were imposed on ohservers I statements irrespective of the correctness of their decisions. lnterjudge reliability UJR) \las 80.4\ for all 51 statements analysed by the judges. In sununary. although observers apparently could not identify the phases in which payment occurred, their observations of players J behaviour in tlie two phases were sufficiently consistent to enable 'the two jUdges to classify t.h~ statement!> into the same 12 behavioural categories. EXPERIMENT 2
The lack of clear evidence that observers in Experiment I could reliably distinguistl between paid and unpaid phases may have been an artifact of the An and BA. designs in w11ich observers were exposed to only one change of poase. If each observer were exposed to a larger number of phase changes and also to both directions of phase change Ii.e', frOll A. to D and fr01:'l B to A) reliability of correct identification of A and B phases might improve. In addition, to explore the generality of the verbal responses obtaine~ from observers in EXperiment 1 r a second level of extrinsic reinforcement was introduced by doubling t11e rate of payment to players. METHOD
Design Ten experimental' sessions consisted of 4 5-minute phases, precea8d The sessions were ei ther of the design ABAB or BASA: in A phases money was earned for performance. and in B phases no money was earned. In all, 43 neW' observers viewed players in Experiment 2. by a 5-minute warm-up period.
25
~T;i~~~
empl~yed
recording 'equipment used was the same as that in Experiment, 1. However. instead of a I,.{~rning bell. IoIhich might ))9,ve been difficUlt to hoar, a single blue 40 watt globe placed directly', in front of the player and clearly visible to observers served to indicate changes in phases. 'Light on' indicated money phases and 'light off' indicated no ..money phases. 'l'he light was Controlled by timers auudeSe the experimental room. Instructions Observers received preliminary instructions before tho players arrived. They then completed identical anlwer sheets during eaCh phase. They answered the fo11olling qucGtions: 1. "Is the plnyer playing for money no'o1?" 2,. "What is it about his/her behav)t')ur that m3de you 91 ve the above ansloIer?" ' Mter the exporiment the observers anslolercd a final question: "Havo you any fUrther comments on, what it was about the player's behaviour which cuggested he/she was or loIas not playing for money when tho light was either on or Off?1I Players r.eceived instructions abOUt tM game and about the experimental design as in Experiment 1. [lrocedure }\ftar the observers were seated the player was intror;1uced to the oxperimental room as in Experiment 1. The warm-I.lp period thon began. If the 1\ Phase was first , the signal light wns turned off during the warm-up: if the e phase was first, the light was on. Mter the warmup 1\ and D phases alternated. The nwrber of points earned in each phase \laG computed and money earned in 1\ phases at' the rate of 2 cents for each 10 points was set aside for colle satisfaction upon lIaking a hit, she conveys self pleasure"; and for tl Sl}e looked al.Jl.ost happy when the cannon ....as shot do'm " ). (3) Off-cue behavioural state:nents (e.g. "She kept ber eyes on the Phasa B:
scroenlt ~ .
{41
Off-cue abstractlonal staterr.ents le.g. "She has beCOr"e more inVOlVed
in tile game" l . If an obso[ver made at least one statement in a category a score of f l' was assigned to that category" and if no statenents vere D!ade in a category a score of '0' ;.ras ass igned to that category. The total number of poss i rsie patterns of responding for observers in the abo...·(! fOllr categories thereby became 16 . Response patterns f[om observers C0111d consequently be i dentified as I cue-Used , patterns (if onc or more statements •...C[C classi fied in categories (1) and/or (Z» or as fClleooNot-Used' patterns fif ,no"statement
-as classified in either category
(1)
or 12)). Independent ,."'lyses were
conducted for A and Po phases for the t ...'O sets of statement~ '" (as judged by
tile t\lO judges) . Hence :four separate aruuyses lIere con(!ucted: to determine the associations betvcen the cue-Used, cwHlot-Uscd patterns and the correctness of each observer's deCision. itJese arc presented in Table IV. Chi squared analyses supported. the hypothesis that observers reporting the relevant cue were more likely to be correct in their decision that "a phase \las 'paid' or lunpaid'. The final experil1l9nt in this series deraonstrates that ....hen nonverbal bel1aviour deriving UltimatelY fron the verbal behavi our of obServers in the fi rst t\JO experiments is presented to fresh observers, they 'appropriately detect' t he reinforcen:.ent circumstances uru:tcr Uhich the behaviour is generated. In particular r observers WilD report.atl attending to the manipulated cue of contingent SlIiling on vins ~re rore likely to identify extrinsic reinforcement conditions than observers -":00 did not report attending to this cue.
31
Table IV.
Relationsh.ips between patterns of observers' statements (Cue-Used vs . cue-Nat-Used) and the correctness of observers' decisions. as
dBtermined for 1\ and B phases by tve. judges in EXperiment 3.
--------------------Judge 1
Judge 2
A Phases
CUG-Used
Observers' Decision
Correct
44
Cue~Not-Usec1
Cue-UIl(:d cue-Nat-Used
42
17
Incorrect _ - ' ' -_ _-'-'17
19
15
i= 16 .0** B Phases Cue-Used
Observers' Decision
Correct Incorrect
33
CUe-Nat-Used
12
CDEl'-Used Cue-Nat-Used 31
14
28
_-"-_~_-,,,,29
ir;;.19.7·~ · • p ( 0.01 !2-taUed) u p ( 0.001 (2-tailedl
GENER1\L 'DISCUSSION
OVera:1,l '," it WO~ld appear from this serios of experilTlents that thece 'i s indeed · reasonable evldenc(f for the existence of a concept of the human £Rca. Further, the ' behaviour of 'Smiling/Grinning' at apPfI>priate points in"the schedule of reinforcement can be regarded ,as an important part of 'this concept wi thin this particular experimental scttin;.' ,While in these experiments no direct response measure was taken, a reasonable judgement of tile constancy of the players' response ' rat:es during paid and unpaid phases of the game may be derived from an' examination of ' tho point scores, given that players had 6ufficient ' pra'ctlc~ dUring wannup. This evidence wholly in accorc;1.. with the: assumptions which ,the study ,Is based t;hat a strongly intrinsically reinforced response would: not show a further rise in the rate of response if extrinsic reinforcement were.to be appl1e'~ . . If the concept of !!:RCB indeed corresponds to an actual J:lICJ3. and certainly this is tlle next 1011ca1 step of investigat~on, that Is requt'red, then the data obtained in these experln\~ts , m~y be exp;ained rather c.onvenient'ly as follows. 1'he finding that observors 'wcre able ' to,
is
on
ns are nol, of CO'D:rse r an itmnediate (!onc:ern,
i\ll tba.t has so fa.~ been da:nonstratcd iSI firstly, human observers 'Will generate p:.rticular verbal ~~haviour \lOd!!l' conditions Where, if there ls a human
ERCB, it should fonn part of the stimulus controlling that verbalbationi, secondly, wilen part of that verbalisation is translated back into nanverbal behaviour and W;':d as a stimulus, it evokes an appropriate verbal response 'in hunan obsorvers. H. is still possible that ilther parts of the or19'ln8.'. Verbalisation [i.e. parts other thM 'Sr.tiling/GrilUlinq') might not evoke a similar response. and this IlUSt be te~ted. MSO, other reinforcers or levelS, of reinforcement rri9ht not give simlar results, this being a ~ssible explanation for the major findings Of EKp&iment 1 and 2. Most UrlJOrt,a ntly. using trained observers, it has to be soo'olI\ that the alleged !!ReB coes, in fact, rise in frequency wi ttl reinforce~t, i. e. that the ERCB is truly an Mca and not just a CO/lllTlOn huraan concept of •...hat F.RCB Jli9bt be. The present study suggests a profitable behavioural starti[}g' p:>int for sUC'h an ill'Ye5tiqation.
34 REFERENCES
Adams', J.1\., 1984. Learnin9 movement sequences. PsyChological Bullethl, 96 : 1-28.
1\tari Inc.
Instructions accompanying ' space Invaders' games.
talk, J.L.. 1961 . production of polydipsia in normal rats by an intermittent food schedule. science, 133 : 195-196.
HaracktewiC:l, J.ftf., Manderlink, G. and Sansone, C • • 1984.
Rewarding pinl;)aU
wizardry: Effects of evaluation and cue value on intrinsic interest. Journal of Personality and Social PsychOlOgy, 47 : 287-300. Hearst, E., 198,4. ,i\bsence ' &s . information: Some implications , of l earning. performance, and represantatlonal processes.
In H.L.
Roit~lat.
'r.G. eever
and· U.S. Terrace (Eds.), 1\nimal Cognition. Hillsdale, N.J. : Erlbaum. Lepper, . M.a.. , Greene , ,D. Md Nisbett, R.E., ,1973. Undermining children's intrinsic interests with extrinsic rewards : 1\ test of the "overjustificationfl hypotheSis. Journal of personality and Social Psychology, 28:
129-137.
Millcnson, J.R. and Leslie, J .C. , 1979. principles of Behavioral Analysis. Neu York: Macmillan Publishing co., Inc. Newman, J., Wolff, W.T. ami Hearst, E., 1980. The feature positive effect in adult human subjects. 30.urnal of EXperimental PsyChOlogy: Human learning and Hemery , 6, 6)0-650. Notterman, J . M. and Mintz, D.E., 1965 . DynamiCS of Response. New York: John Wiley and Sons, Inc . . Sch...... rtz, 8., 1982 . Reinforcem.ent-induced bel1avioral stereotypy: HoW' not to teaCh peOplQ to discover rules. Journal of. EXperimental PsychOlOgy: General" 111: 23-59. Skinne~, n.F. and Morse, w.n . , 1958. Fixed-interval re inforcement of running in a wheel. Journal of the Experimental Analysis, of Behavior. 1: 371"379.
Thorndike, E. t . ,
189B. Animal Intelligence.
.MOnograph Supplement, No . 8.
PSYChOlogical Review:
waters, L.K., 1980 . social reinforcement of the work beha\.'!our of .' retardates and normals. Unpubli!:ihed doctoral dissertation', uni versity of AdelaidO. . Wilson, M.P. and Keller, F.S. , 1953. On the select.ive roinforcement of spaced . responses. Journal of Comparative and Physiological Psychology, 461 19D-193.
