Cognitive Therapy and Research, VoL 3, No. 4, 1979, pp. 407-413
Development of Children's Use of Plans for Self-Control' D. Bruce Carler, CharloIle J. Pallerson,: and Susan J. Quasebarth University o f Vir~inta
Although the. ability to ,,lan and organize lengthy sequences of beha,cior in pursuit o f personal objectives is a hallmark of the mature adult (Miller, Galanter, & Pribram, 1960), this kind of behavior is much less characteristic or" the young child. A number of investigators have begun to study the role of I~ians in children's sell-control, and it has become clear that even young children are often able to employ plans presented by an experimenter in order to facilitate self-control (Mischel & Patterson, 1978; Pressley, 19"79). However, relatively little is known about how children learn to formulate their own plans. As a preliminary step toward exploration of this topic, the present study investigated children's ability to elaborate and use partially formulated plans for self-control. In an analysis of the nature or' plans, it has been argued (Niischei & Patterson, 1976, 19781 that two distinctions should be made. First. a distraction should be noted between the organizational level or structure of a plan and its content or substance. Substantive elements of a plan may vary independently of the plan's structural characteristics, and the reverse is also true. Second, even at the simplest level, a plan invariably consists of at least t\~o structural components: a cue [which indicates when the plan is to be executed) and a c o m m a n d (\vhich dictates the nature of the action to be perlormed wilen the plan is executed). ¢ o m p l t : t l o u ol thi~, w o r k x'-a', t a c l l i l a l c d by N I M H g r a n t XIH 2 " 6 ~ q a n d by a \Vile, o n G e e h>,ti[tl[C gl':llll IO ( h a r l o l l e [~[HICF~,OlI. \ \ c \~.i~,[l It) IJl~lllk l);.P,'e DelllarCsl ;.llld ~[¢v~.' ROIll311ciI~.i f o r Ihuir ilelp m CoIIdI.IC'ilI~ this ~'[ldy. ~,V¢ (IJS¢I ~.~,iSJl ~,1 t h a n k l)r. J o i m Bilh.'r, p r i n c i p a l of the [ ] o l l w n e a d .'-;chool, a n d M r s . ( h r i s t m e M c C a u l e y , d i r c c w r o f the E a r l y C h i l d h o o d D e v e i o p IIICIll ('CllleI" ( b o l h o1" ( ' h a r l o l l c s v i l l c , V i r g i n i a h for t h e i r oulsl;.mdillg c o o p e r a t i o n . SLI.¢,a,II (.)llilSCJ'~/ilIJl I l o o d is f~ow at Ihe l{hlg N u r s e r y %~:hoolo .'~ltlllfOrd U n i v e r s i t y . \ddrc~,s all c o r r c ~ p o n c i c l l c c i~ ( h ; . t r l o l t c I)atlcrsoi1. l')~p;.lrllllcnI o f P N y c h o l o g y , U n i v e r s i t y ot V i r g i n i a . ( h m h m c ~ , ~ i l l c . V i r g i n i a 22901. 407
O[-ll~M) l(i/lq/1200-0407'503.00/o
' t. 9 1 9 Plelltlln PUDliSIIIIlg Corporation
408
Carter, Patterson, and Quasebarth
Research reported by Mischel and Patterson (1978) suggests that the degree o f elaboration of both cue and c o m m a n d components is potentially important in determining the efficacy of a plan for self-control. For example, while preschool children were able to employ highly structured, fully elaborated plans to facilitate resistance to temptation, th@ seemed quite unable to use more generalized or unelaborated variants (which did not fully specify the cue or did not fully specify the command) to any advantage (Mischel & Patterson, 1976). One might expect that as children grow older, they become more capable of employing such generalized plans to facilitate self-control. The present study assessed the ability of children at three different ages to use generalized plans to facilitate self-control. The design systematically varied the presence and absence of highly structured, elaborated cue and c o m m a n d components of a verbal plan for self-control. In the experimental situation, children were required to persist at a lengthy task in order to win attractive rewards; however, they were free to stop working at any time and receive a less attracn~c prize. Dependent measures assessed how well children in each condition were able to sustain task-oriented behavior in pursuit of desired goals.
METHOD
The study employed a 3 X 2 X 2 fully factorial design that varied the subjects' age (preschool, kindergarten, second grad e) and the nature ot the instructions (generalized vs. elaborated) about the cue and c o m m a n d components of a plan for self-control. Thus, at each age. the experimental groups consisted of a full plan (cue + command) condition, a cue-only condition, a command-only condition, and a control (no plan instructionsl condition. An additional irrelevant plan control group, in \'~hich subjects received a fully elaborated plan irrelevant to the self control situation, was also included at each age level.
Subjects Fhe ~ubjects were 122 children: 40 preschoolers (mean age 4.07 ~;ears). 42 kindergartners (mean age 5.58 .,,'ears), and 41) ~,ccond-graders (mean age S.() years); they were evenly divided by sex at each age level. {Subjects were randomly assigned Iwithin age and sex), 5 to the no plan control condition, 5 to the irrelevant plan control condition, and 10 to each of the other conditions. Fhe two additional kindergarmers were assigned to the no plan
Children's Plans for Self-Control
409
control condition. The experimenters were four adults, one female and three males, each of whom tested approximately equal proportions of males and females in each condition at each age level.
Procedure In the experimental room, a table and chair were placed directly in front of a one-way observation mirror. A desk bell, a small scoop, and an apparatus called " B a b y Bird" were also placed on the table° Bal0y Bird is a jar that has been painted and otherwise altered to resemble a hungry young bird: it is described in more detail elsewhere (Patterson & Carter, 1979). Six yellow lines have been marked at even intervals on the front of the jar. The apparatus is mounted on a painted wooden base; a small (75-ml) cup is also mounted on the base, next to Baby Bird. ,~ large container of marbles was r",,.~d c'a a second table across the room. t , c ~.ward objects were certificates called " G o o d Player A w a r d s . " Space was provided on each certificate for up to six gold foil stars. Subjects were escorted individually to the experimental room and seated at the table in front of Baby Bird. The experimenter explained that they were going to play a game, and that the game involved teeding Baby Bird. Each child was shown the marbles, and told that they were Baby Bird's food. The experimenter demonstrated how the scoop should be used to transport marbles from the large container to the small cup next to Baby Bird, and how the marbles should be "'ted" individually to Baby Bird. Tile child was told how to get a scoopful of marbles, put them in the cup, feed each one to Baby Bird, return for another scoopful, and so on. At this point, the experimenter introduced the reward contingency. He displayed a Good Plaver Award, and told the child that s, he would receive one for participating in the game. The experimenter explained that. if he wanted to, the child could win gold stars to put on his, her G o o d Player Award by feeding marbles to Baby Bird. If the child put in enough marbles to reach the first line, s/he would get one star: if s/he put in marble.s up to the second line, s/he would get two stars: and so on, up to a maximum o | six stars. All children demonstrated their comprehension of the reward contingency and expressed their desire to win six stars. Ihe experimenter then explained that he had some work to do. and that hc would leave the room while the child was playing the game. rile child was instructed that s/he should play the game only as long as s, he warned, and that s, he could stop playing at anv time and receive the number of stars s/he had already won. When s, he was done playing the game, the child was It) summon the experimenter bv ringing the desk bell.
410
Carter. Patterson. and Quasebarth
The experimenter, who had remained blind to the subject's experimental assignment until this point, consulted a concealed slip of paper that informed him of the subject's condition. After a standard introduction, he presented the instructions for the appropriate experimental condition. Children in the cue-only condition ~ were told: " W h e n you've Fed Baby Bird all the marbles in the cup, then you can think of something to s a y . " In the c o m m a n d - o n l y condition, the experimenter suggested: " W h e n e v e r you want to, you can tell Baby Bird, 'Baby bird, I ' m going to feed you more m a r b l e s . ' " In the cue + c o m m a n d (full plan) condition, subjects were told: " W h e n you've led Baby Bird atl the marbles in the cup, you can say, ' B a b y Bird, I'm going to feed you more m a r b l e s . ' " Finally, in the irrelevant plan condition, children were told: " W h e n you've fed Baby Bird all the marbles in the cup, you can say ' B a b y Bird, hickory, dickorv, d o c k . ' " Children in the no plan control condition were given no instructions about a plan. After etaborp.ging and repeating the preceding instructions, the experimenter delivered siandardized concluding instructions (which reiterated the experimental contingencies) and then departed. The two main dependent measures, amount of time spent working (in minutes) and the a m o u n t of work completed (number of marbles " f e d " to Baby Bird), were recorded by an observer seated behind the one-way mirror.
RESULTS Preliminary analysis indicated that there were no effects due to sex o~ ~,ubject or sex of experimenter (Fs < 1), so these factors were eliminated From further analyses. In addition, there were no significant differences in performance among subjects in the irrelevant plan condition and ti~e no plan condition (rs < 1}; hence ttle data for these ~wo groups were collapsed to form a singie control condition for further analyses. Since the data for alllounl o f time spent working and amotmt of work completed were highly correlated {r = .82), only the time data are presented here. Fhe data For mean amounts of time spent working by subjects in each condition are shown in Figure t. A 3 X 2 X 2 analysiS;, or" variance on these data revealed significant effects of age ( F = 7.58. d.t = 2. 1 it.), ,t~< .0Ol ), presence versus absence o f the COllllllalld c o m p o n e n t (F" = l-'.S2, J..t = [.
;If is xxorlh nOIill~ lh,II lhis parlicular cuc l\\'hcn yml'~e I¢d Baby Bird all ~hc marbles in ~hc Ctlp. ) '.MIs ",cI¢,..'Icd ,,m Ihu" I'~asis c,I cxlcnsi'.c pilol observation:., tha~ indic:.ucd tt~.m ,. i r t u a l I y thv Lmiy Iimc ~.luh.lrcn r0.ne Ihc hell was ',',hcrl d'~c cup wa.,, empty.. Fhc cmp~y cup, then. :,cemcd [(} nCF\L" [IN ~i ~[ I(~I]~ kJUC [() ~'I('IT3x'.orkill~i hcllu'c, il provi~.l~'d ~lll oplinla] duc for CXCCtI[iOH 0[" ,
.
lq:mn lot ,cII-cLmIIol.
Children's Plans for Self-Control
411
INSTRUCTIONAL CONDITION
3O Z
~]
or"
o
F.Z L.~ ca C~
I
25
FULLPLAN COMMAND
OMPONENT
ONLY
D CUECOMPONENT ONLY D
CONTROL
C,r)
N 2o 7
ol 5 c~ rn
t
Z
t0 PRESCHOOL
KINDERGARTEN
SECOND GRADE
I,'i~. 1. ,,\mount of time spent working as a function of grade and instructional condition.
110, p< .001), and presence versus absence of the cue component ( F = 5.85, df = 1, 110, p< .02); none of the interactions were significant. Paralleling the findings from earlier studies on preschool children IMischel & Patterson, 1976), individual comparisons indicated that at the preschool level, only subjects in the full plan condition spent more time ~orking than those in the control condition (t = 2.46, d f = 18. p< .053. Neither the cue-component-only nor the c o m m a n d - c o m p o n e n t - o n l y condition differed from the controls at the preschool level (both [s < 1). .ks expected, individual comparisons indicated that among kindergarten subjects, children in both the full plan (t = 3.78, cii"= 20. p< .01) and the c o m m a n d - c o m p o n e n t - o n l y conditions (t = 2.28, df = 20, p< .05) ,pent more time working than those in the control condition. The time spent ~orking bv those in the cue-component-only condition did not differ from that ,~pent by control subjects (t = 1.08, n.s.). LNt the ~,econd-grade level, there were no significant differences anaong the conditions: all subjects persisted for approximately the ,ame length of t into regardless of condition. In an effort to assess the relationship between subjects" verbalization t)[ lhc plan for self-control and their actual self-control behavior, subjects ~erc divided into those who completed the task (i.e., those who put in 300 or more marbles, thereby reacimlg the top line and winning six stars) and
412
Carter, Patterson, and Quasebarth
those who did not. The number of subjects who verbalized the plan in each of these groups was calculated. Data on verbalization o f the plan and task performance seemed to parallel those for task persistence. For the sample as a whole, there was a strong relationship between verbalization of the plan and task completion (x' = 8.11, df = t, p < .005). When the components of the plan were analyzed separately, task completion was found to be significantly affected by use of both the cue ~ : = 20.05, df --- I, p < .001) and c o m m a n d (~: = 8.04, df = 1, p < .005) c o m p o n e n t s ? Subjects who verbalized the plan were more likely to complete the task than were those who did not verbalize the plan. This relationship is evident in the data for preschool and kindergarten subjects iX" = 7.96, df = 1, p < .005). A m o n g the second-graders, however, there was no relationship between verbalization and task completion ~ : = .11, df = I, n.s.). There were no significant relationships between task persistence and verbalization of the irrelevant plan at any age.
DISCUSSION The results of this study indicate that whether or not a particular plan will facilitate self-control depends both on the structural level at which the plan is presented and on the age of the person who is to employ it. The tendency to use generalized plans for self-control in an effective manner increased from preschool to kindergarten: kindergartners were able ~o elaborate cues for the execution of verbal plans but not to specify the contents of actions to be undertaken. Self-control performance or' the second-graders, who apparently found the task quite easy, was unaffected bv instructions about plans. These data suggest that it may be useful to think of plans as t'alling on a continuum with respect to their structural characteristics, ranging from generalized intention statements at one extreme (e.g., "'1'll v, in all the gold stars for my Good Player A w a r d " ) to detailed plans and strategies for achieving these goals at the other (e.g., "1"11 put this marble in Baby Bird's mouth, and t h e n . . . " ) . In this context, the present findings suggest that older children are better able to utilize plans that fall toward the generalized
' \ \ h e n c h i l d r e n e m p l o y e d t h e c u e \~e h a d ~ u g g e s l c d . o ~ c r ')S~'~ o l t h e m cI,,~pioyed ,l c o m m a ~ l d cOitlpOllCIit q m i l a r o r i d e n t i c a l [o [Jl:.lI ~.l,tliC|l ~.le h~.ld , p e c i f i e d in [i1¢ full p l a n c o n d i t i o n . t . h i l d r c n also l c n d c d to use the.. COlllIlltlnd COlllpOIll.'Ill OIlty ill rcspO~lS¢ [o tile ~pcciI'ied cu~'. t,~,ith (ully 9t)% o f the c t l i l d r e n ~ c r b a l i z i n g tile O.)llllll~tild tit this p o i n t at least OtlCe..-ktnoilg l llONC t~llo d i d 11o[ tlSi2 t h e s p e c i f i e d c u e , tile illOSt COlllillt)ll reslpolts¢ Wtl:,; [llOrO ['rc.'tll.i~.,ll[ ~ e r b a l iI:ltlOll o r t]lC COlllIililnd. w i t h ~,0111C c h i l d r e n v e r b a l i z i n g the." cOlnlll;.lnd COtllpOii~.'ii[ o [ the pl[tll ( ) ~ c r 201) [I[IlCS.
Children's Plans for Self-Control
413
end of this continuum than younger children are. In addition, however, the data suggest that knowledge about when some change in one's behavior is called for seems to precede knowledge about what, exactly, should be done, at least in the present situation. It is possible that knowledge about when new action should be taken may provide the initial step toward use of generalized plans, and indeed toward the invention of plans to guide one's own behavior. Future research will be needed to assess both the generality of the present findings and the adequacy of these interpretations. In s u m m a r y , this study investigated the development of children'sability to use plans that are presented at different levels of generality. Kindergarten children were more successful than preschool children in using generalized plans, but only if it was the cue c o m p o n e n t that had been left une[aborated. Second-graders' self-control performances were unaffected by the presentation of plans. These results indicate that both the structural level at which a 7 ~"-, i~ presented and the age of the individuals who are to use it can be i,-:;.~,.k_-,r c, eterminants of a plan's ability to facilitate selfcontrol,
REFERENCES Xlillcr, G. A., Galanter~ E,, & Pribram. K. H. Plans and the structure o f behavior. Ne,a "fork : Holt. Rinehart & Winston. 1960. Mischet. \V., & Patterson, C. J. Substantive and structural d e m e n t s of cttecti,,c plans for self-control. Journal ~I Personality and Social P,~ychok~ey. 1976, 34. 9,-t2-95(t. Mischet, W.. & Patterson. C. J. Effective plans for self-control in children. 111 W. A. Coilin, (Ed,), 31innesota .vymposta ott chiht p.wcholouy (Vot. ! I). Hillsdale, New Jersey: Erlbaum, [978. Patterson, C. J., & Carter, D. B. Attentional determinants of children's ',eft-control in working and ~aiting situations. Child Development. 1979. 50. 272-275. Prc~,~,lcy. G. 3,1. Increasing children's self-control through cognitive inter;'entitms. Review ,~t Educational Research, 197% ill press.