'01;

d

0 s 0I g V 4, (.0 0P U 0 g sir

rid cn

11) o, 0)til0

0 0 - 0) .di$) 0) .r1 r'l 0cl, PI ni rd 0 iril

!- .0, rd t-I g 0 P ItN ra .ril IW r,1 g P 0 .I.) 0 0 .ri 4) x u) .4-) Iro 0 .4.)

0.P

0 0 RI .0 r4 1CD ° CV 91 r-i 4,-Ji 3 tn tr, CI)P () W ri .ri 0r0+341-10.04I00 arINtrhO 000,0

I 01 0 M C.) O 4.) 4-1 '1ZU it 0 ihi 01 I 'P 0 .r1 I thl 4 0 .1-I ,rI or40 0.

0 14 /41 0 t) P 0 0) r0 g I

> 0. 0 C) 0 En 0 Lk it 4.$ WI W 0 .id 4.1r01.1-1 OUItttrri mil 41 0 d) 114 4-1 4.') 011 4-1 14

44 0 4.1 tn g g 4 0 5 (0 4 40 ' M II W A, >1 0 0 (I) WH ..,4 0,0 DT V) 0

g ra E .H WI Pi g to 't7 Cji g wi tr.d uril P go 4t) X 0 r4 in En 0 A 0er-Y gl Pi 4 Po C6 0 0 4) 0 .rel a) .ri 04 14 , rIt 01 I 0:11 CP r-il En CI 14 4) 0

0 Ica tr.) .1,1 LI W g W 73 ril 04 (1) (V g P .F1 PI (1) (V PI ft >11 (.)I g LC 4-1 cal 0 0 > ro0r0 co'14 r0 0 $-4 t0 PI 030 5 g 0 A I-4

U 0 E g E > ni 4u A > 3 0 0 0 g oil W 1+4 V g 14),p4 tr 0 (121 'ri ra in ori P 4-I r6 Pt 44 4.; 0 0 I tn > W

rd 40 .F1 0 % (51 4..t al a) HI 4.1 W 0 a, g rd tn fa4, 0 ..1. opt 0 0P () rd >1 ,r4 03 P Ks 004)14 ..-1 10)0 7:500>i 04 c0 Ott E,rt tt5 it W tr) .10 .. to 4.1110.r41 COtri .,r11:71104) Z 4-10VW40)ri P ..1 H 4) .ril 4 0 H 0, g' 14-4 > dr) rd 0 00 Pri e-l' 0 H k = . 0 CO ,0,11 4) CO g 0H a) It 0 r,I 0 0 Pi4 RS 1 W 44 ird g W eel g V W .01 I) rd tr) 'PI 114 ,r4 4.1

Pi -I U E 4) 4, ,r1, 0 g P LVrd .I.) CH-I '0 0 tiQ W tP 1.0 73 Ir000 X

0 vIIC1) r41CD CS 2 0-1 4 4-' rri $.4 tri 0 .r1.1.jr000) 001r1k

(1) It g 'Pi 0 a Pe, '0 0 Cr, IP Ind 14-11.1 0HNO001(60 4) E-i ,H fr-4 0 >10 .0-1 ri, ?-I 0 0 0 00 Lkil P 0TIOMW P CJ) C W .r-II gril Irli 0 0 en 0 4-, .0 0 ,paraa)a) / 4 0 JPj1r0 0 1.1 '0 0 0 0 0 g P (1) 0 4,1 0 0 0 IP010 E4L).0> W 0 0 g g PIE GO Fil 0 0., ,r, Cr) 0, e* k 014 .1-1 03 0 P 101 M 01 I.H 0, r-4 0 5 rd 0) c) .1-I >1 trI cn cl) A

X / 5 N a) co to mr,0 .0 rd .0 0) 4) 0 00 .H 0> 0 ,y,p,a) 4:),-10,4-Joa),L4g>g144(,nofa rd 4.1 020' 0 0 0. ,..4 0 Ini C) WI ,-I, W E-1 co, cui .0 rid 0 0 4) W LI ',PI >1 E P nj to ,,..1(1) 1J 74, gr GO g E-1 F-4 P I '04 0 4

0 0 U I ri 10 IE-1 IP V0 $-1 0 Id a) 00 tn

0 0 >4 al ,14: 17) 0, 17, X 4) 144 enti-1 0 4.) tq g gg 0 0 0) g 111 t; HI RI 0 > 1r0 44(110.ggi:24:g4)4IHkr.) 0 ID grl 0 0

*HI I rl in E 0 '1.0 Z it g rI iv. ini 41-1 0 iri M CO 0 Irl, It Irl W 1 :: 'CI LH 0) 4 P (..) ri1-.1 P `r4 0) Prl .1-1 LI; P 4 0 (.0 k IJ V) .ri It rd CV I W X Q) 10 4) al CI) ow 4.) a dp

ra 1.41 01 4J1 Z 0 in 0 0r4) »75 oa) 1.41.0Hilt 4 ,rc 0 rd $4 JP g $4 4'.. 4j IV

1 ) PI 5 .H I D ,r1 .11J it .P 44 W W O O tr) to it 0'0 ,r4 11 0 1..1 0 W op P 1> f.0 0 .1-1 01 Ci; r17.1 0 rd () P g rd A I rd E (0 3 0) 0 0 0 .0 4.)O 44 Eli4V4-1014 ocr,0to .1.J 0 a; 0, .r4 Z Z a, ail 1.4 $.1 cn 1 0 411 r-I (n rd 0 .1.) ,r-ilIfJ) .r1 0 g a 111-1 .r11 1 134 .0 01 0 06 1'6 0 irl 4) ail E gi ItI.0 1-4 4-1 rd /19 .' 4J ti) g E0 114 1Cf)

I 4) 0 r-i: 0 n3 tO.VW41-) >a)g a)T-lociaola).,-Iviroa) o a) 0 1,rc,p4A a) r- 44 14 C:14 CV F-I U 0 .0(0 (0>*01 rt10.).140)(1) ,H4A111M0Z X gt CZ 0 0 0 0 rs.1 x * ,p, * z * * El w E 1..1 .1,..1 H 0 li-1! (0 0 5 ',In >4 67 4.1 1 P 4 0)

4 tr 0 ILO il) PI it e 4-) 0 4,' A 0-1 (.0 r4 0 4i g.ri .1-1 cn 0 4) 41 05 P000e0)1050.113Hrd>RJW 41,14.)0)00 M0L0.0g0g

Milan k 4) 41) r4 r-II P 4.J ,r4 .11 rd Ft 0 r'l>4 g PI 4 1-1 g 10 4-) C) 0) g riHrd C) :7-1

Z U M M 0 (0 (V * 0 W .44 Ft W 171 FIS 4oncn W /C Z Irx1 CY % .r1 11.0 0 40 E W M >1 g 14, a) El 'P 4) 14 to 144 0 0() 0 41 4) 4) 0 0) cn 4) 0

f-i 0 41 t-i IP H IF 00 'n-i, 0 0 g 0 thi 113 I'd Pri .11J (1 ajg Fi' 8 ° ° '4riIP 0:0 E-4 (Z 44 IN 0 P g irli E CD P .ri WI 5 0 r4 0 (I) E rt .ri 4-.1 0) 0

124 (.4 < IN H H 1 WI 4-1' P 0 $4 iii PI 0 0 0 10 P 0r1 W 4) Ct4 4 ttO 41 H t» a ix, E-1, ix, Iri 0 F-11 (I) rd 4-1 0.1H 0 4-1 0 4 0) rd P E Itt WI 4.1 P

P Z ILI M U Z Ei it .,11 0 P ri 0 a5 la t,w m v.1 4) z o a)EH EA [00131F Zr.C1 41 rn W ft1 '111 W ,4 tri 4-) .ol ta 0 0 .k ,111 04 CO 4) 0 Ca 4)1.) H Z fl. t.-", 0 0 fil 0 m 0 wl r.ti.g .g a) a) $4 ..000.00P.CX05.H0c00

gt El H u) ri, z w 0 H

rang

U S. IMPARTMENT OF HEALTH.EDUCATION Ed WELFAREOFFICE OF EDUCATION

THIS DOCUMENT HAS BEEN REPRODUCED EXACTLY AS RECEIVED FROMTHE PERSON OR ORGANIZATION ORIG-INATING I I POINTS or VIEW OR OPIN-IONS STATED DO NOT NECESSARILYREPRESEN I OFFICIAL OFFICE OF EDU-CATION POSITION OR POLICY

MODIFICATION OF SOCIAL INTERACTION

IN FIVE-YEAR-OLDS: A COMPARISON OF TRAINING PARADIGMS

BASED ON REINFORCEMENT, MODELING, RULE CONFORMITY, OR COOPERATION1

Linden Nelson and Millard C. Madsen

University of California, Los Angeles

December 1968

1The work reported herein was carried out with the support

Office of Economic Opportunity, Contract .110. 4117.

MODIFICATION OF SOCIAL INTERACTION

IN FIVE-YEAR-OLDS: A COMPARISON OF TRAINING PARADIGMS

BASED ON REINFORCEMENT, MODELING, RULE CONFORMITY, OR COOPERATION1

Linden Nelson and Millard C. Madsen

University of California, Los Angeles

At least four basic processes have been hypothesized by psychologists

to account for the development of moral-rational social behavior in the

child. The reinforcement of socially appropriate behavior by adults and

the punishment of inappropriate behavior is, of course, one of these pro-

cesses. Another basic process is modeling. In modeling the child acquires

social responses on the basis of his imitation of adults.

The third basic process, rule conformity, involves acceptance of a

rule on the basis of convention or authority and the capacity to govern be-

havior in accordance with the rule. Rule conforming behavior is a-rational

and inflexible and involves learning a somewhat arbitrary connection between

a stimulus condition and a rule. The basis for this connection between a

stimulus and acting according to a rule is heteronomous. The four h possible

basis for moral-rational social behavior is cooperation. Cooperation requires

the acquisition of a principle for rational adjustment to social situations.

It involves a sensitivity to the desirability of mutual assistance in certain

social situations. Cooperation, as defined here, may involve conformity to

rules; but in this case the conformity is autonomously based. That'is, in

cooperation there is a reason for conforming that is inherent to the inter,

personal situation. When cooperation involves a rule, the connection between

stimulus and rule is far from arbitrary.

1

2

2

Previous investigations of the developmental bases of moral-rational

behavior have utilized case study methods (Peck, Havighurst, Cooper, Libien-

thal, and More, 1960), or interviews about hypothetical events (Piaget, 1932;

Kohlberg, 1963b), or inquiries into the childrearing techniques used by par-

ents (see Kohlberg, 1963a, for a review). Peck, Piaget, and Kohlberg seem to

agree in general that the bases for moral behavior change during development

in their relative importance in the following sequence: reinforcement rule

conformity, and rationality-cooperation.

Another approach to the study of the developmental bases of moral-

rational behavior is to investigate the_ responsiveness of children,to behavior

modification. Chittenden (1942) found that a group of preschool children

were more cooperative than controls in their play with other children follow-

ing a series of training sessions in which the examiner and each child ana-

lyzed the "play" of dolls in social interactions and discussed appropriate

social responses. It cannot be determined from this study whether the basis

for learning cooperative behavior was reinforcement, rule conformity, an

understanding of cooperation, or some combination of these. Bandura and

McDonald (1963) found that modeling was more effective in changing the moral

judgments of five- to eleven-year-olds than was social approval. Crowley

(1968) found that labeling and rewarding appropriate moral judgments in train-

ing sessions for first graders resulted in their making more appropriate judg-

ments than untrained controls when tested 18 days later. A group whose train-

ing involved discussion of the basis for labeling judgments as appropriate

(also given labeling and reward) showed no more change than the labelingL

reward group.

The present study was a further attempt to understand the developmental

bases of moral-rational behavior by investigating the effectiveness of vari-

ous training paradigms. The behavior studied was that of taking-turns in

situations where mutual assistance was necessary in order for either of two

children to receive prizes. The measuring instruments were two-person games

in which only one child could get a prtze on each trial. Moral-rational be-

havior in these games would involve taking-turns helping each other get prizes.

If each child attempted to get a prize for himself on every trial, neither

child would get any prizes.

Four training paradigms were compared. In the reinforcement treatment

the children received prizes for cooperating. The children were instructed to

behave such that they took turns in helping each other get prizes. This be-

havior was rewarded in that one child received a prize on each trial. The I

also expressed verbal approval after every trial. The modeling treatment

involved having the children observe adult models taking-turns in getting

prinzes. The rule conformity treatment involved labeling taking-turns and

bringing about practice of the behavior by explaining how to take-turns and

telling the children to take-turns. This could be considered a reinforcement

plus labeling condition. The cooperation treatment involved labeling taking-

turns and identifying taking-turns as a solution for the limited reward pro-

blem. Cooperation training emphasized "No one gets prizes unless you help

each other;" and "If you take-turns, you will both get prizes." Only in co-

operation training was a reason for taking-turns pointed out to the children.

The most interesting measure of the effectiveness of training concerns

the degree to which training transfers to new situations. It might be

4

expected that training based on cooperation would transfer to new situations

better than the other methods of training. This would be expected because

children trained to take-turns on the basis of cooperation should be more re-

sponsive to the relevant cues for taking-turns, and they should connect this

cue with taking-turns as a rational solution. This prediction assumes that

five-year-olds have the cognitive mediational capacity to recognize a rather

complex cue and to assimilate the taking-turns schema. The observations of

Piaget (1932) seem to suggest that five-year-olds do not have this capacity.

The present experiment provides evidence relevant to thia question and in

general yields information concerning the advantages of training social be-

havior on a conceptual basis to five-year-olds.

Various theorists (for example, Bandura & Walters, 1963) have suggested

that modeling is a particularly important process in the learning of social

behavior by young children. The present experiment investigated the useful-

ness of modeling as a training technique compared to simple reinforcement,

training of a rule, and training of a concept.

Method

Subjects

Children from eight Extended Day Children's Centers in Los Angeles County

were matched on the basis of sex, race, and age into 69 pairs At each center

the pairs were rlandomly selected for training treatments. The number of pairs

and the mean age of chi dren in each condition is reported in Table 1.

INSERT TABLE 1 ABOUT HERE

5

All of the children were four or five years old. Except for the model-

ing condition, pairs from four or five of the centers were represented in

every condition. The modeling group had pairs from only two centers. There

were 10 Negro pairs and 59 Caucasian pairs; and there were 36 all boy pairs,

29 all girl pairs, and four boy-girl pairs. Thr representation of pairs by

race and sex was reasonably the same in all conditions. Careful inspection

of the results suggested that there existed no significant differences at-

tributable to race or sex among pairs.

The Games

Each of the five games listed below is a two-person game which requires

mutual assistance for solution. In each game the players have conflicting

goals. For every trial there is only one prize which is given to the person

whose goal is reached. Since it is impossible to reach either goal without

mutual assistance no prizec are obtained on trials in which Ss fail to co-

operate in the allotted time. For all of the games, the rational solution

is for the Ss to take-turns in helping one another get prizes from one trial

to the next.

1. Choice game. There are two spots.on a board. Each spot is a goal

for one S. The prize goes to the S whose goal is first touched simultaneously

by the index finger of both Ss. There is a ten-second limit per trial.

2. Marble-pull Game. (Madsen paper in preparation;)

3. Cooperation Board game. (Nelson & Madsen, 1968.)

4. Circle-matrix Game. (Kagan & Madsen, 1968.)

5. Pull-block game. This game is similar to the Cooperation Game de-

veloped by the present authors (Madsen, 1967). The Pull-block Game involves

6

pulling ropes through a small square opening. There are two ropes each

with two plastic cubes attacned to them. A prize is given to the S pUlling

his rope completely through the opening first. Because of the closeness of

fit in pulling a block through the opening, it is impossible to pull either

rope completely through the opening if both Ss pull at once. Cooperative

interaction is required in order for an S to pull his rope through the open-

ing within the 20-second limit.

Procedure

All pairs of children in the four training conditions participated in

four experimental sessions on four days. The first three sessions were on

consecutive days and the fourth session was seven days after the third ses-

sion. The E for the third and fourth sessions was a different person than

the E for sessions one and two. A description of the four experimental

sessions follows.

I. First Session.

A. Practice in taking-turns.

Reinforcement group. Each child was given six marbles,

and the pair was told that all of the marbles were to be put into

a box. Child One was told to put a marbde in a hole in the top of

the box, then Child Two was told to put a marble in the hole, and

then Child One was told to put another marble in the box, etc. The

children alternated in this fashion until each had placed six marbles

in the box. After each child responded, the E said "Good." The same

procedure was followed as the pairs alternated in placing X's in lZ

squares on a sheet of paper when given only one marking pen. The

same procedure was again followed as the children threw one avail-

able ball at a target box six times each.

2. Modeling group. Pairs of children watched a male and fe-

male E taking-turns in the above situations.2

3. Rule conformity group. This treatment was the same as

for the reinforcement group, except the children were repeatedly

told that they were taking-turns.

4. Cooperation group. This treatment was the same as for the

rule conformity group, except the E explained several times the

reason for taking-turns, e.g., "There is only one pen, so how can

both of you mark X's?" "You are taking-turns so that you both will

have a chance."

B Choice game

Immediately after the taking-turns practice just described,

every pair in the training conditions played the choice game. The

game was explained in the same way to all pairs and there were

eight trials with eight plastic bugs as prizes for all conditions.

After the explanation of the game, the procedure varied depending

on the condition as follows:

1. Reinforcement group. E told the children where to point

their fingers on the first and second trials, forcing them to take-

turns in getting prizes. Thereafter E asked "Where will you point

next?" E continued to direct the action such that the Ss took-turns

for eight trials. After every trial E said "Good" and gave a prize

to one of the pair.

2. Modeling group. E said "Watch how we play the game."

A male and female E took-turns getting prizes for eight trials

wh'ile the Ss watched. After each trial, E asked "Who gets the

prize?" This was to assure that the Ss were paying attention

and understood the game. After finishing the game, E gave four

prizes to each child "..:for helping us with these games."

3. Rule conformity group. This treatment was the same as

for the reinforcement group, except E repeatedly labeled whose

turn it had been and whose turn it was to be. Before the first

trial E said "I want you to take-turns getting prizes." After

each trial E asked "Whose turn is it to get a prize now?" and

said "You are taking-turns getting prizes."

4. Cooperation group. After explaining the game, E said

"G " for the first trial with no further comments. After each

of the first three trials E said "S One (name) touched this spot

and S Two (name) touched this spot, so S (or so no one ) gets a

prize." Then E said either "S helped S get a prize" or "See what

happens if you don't help each other?..no one gets prizes." After

trials four through eight, E made one or all of the following

statements: "No one gets prizes unless you help each other." "If

you take-turns you will both get prizes." "Whose turn should It be

now?"

II. Second Session.

A. Marble pull game There were six testing trials followed by six

trials with training. The instructions and procedure for the te§t

9

trials were the same for all training groups and for control group I.

The procedure for the six training trials varied depending on the

condition and was essentially the same training procedure for each

condition as described above for the choice game. The control group

was given no further instructions or training for trials 1-12. The

12 prizes were marbles.

B. Cooperation board game. The procedure was the same as for the marble

pull game. There were four testing trials followed by four trials

with training. Control group I played for eight trials with no train-

ing. There were eight different prizes (rings, plastic toys, whistles,

etc.

III. Third session.

All of the training groups and control group II played the circle

matrix game for eight trials. The procedure was the same for all groups,

Although this was the third day of participation for the training groups,

none of the Ss had seen the E before, since he was a different person

than the E for the previous sessions. This new E had only the names of

pairs he was to test, and he did not know to which conditions the pairs

had been assigned. The eight prizes were different from one another

and different from those previously used.

Fourth Session.3Iv.

The training groups and control group II played the pull-block game

for eight trials This session occurred seven days following session III

and the E was the same as for the third session. All groups were tested

in the same way, and the E was still ignorant as to the assignment of

pairs into conditions. Eight new and different prizes were used.

10

Results

The data was collected during sessions II, III, and IV. The firSt half

of trials of the marble pull game and of the cooperation board game were

testing trials, and all eight trials of the circle matrix game and of the

pull-block game were for the purpose of measuring cooperative behavior and

comparing groups,. When a member of a pair tested during the second session

was absent from school for the third or forth session, the data from the

second session was used in the analysis of results even though no data from

that pair was available from sessions III or IV. Thus no data actually ob-

tained was excluded from the analysis. The only exception to this rule was

the exclusion of data from three pairs, each from a different condition, in

which one or both children refused to obey instructions or to play a game.

Marble pull. The amount of taking-turns on the marble pull game, as for

the other games, may be represented by a score which is the number of prizes

obtained by the S in a pair who obtained the fewest prizes over trials. Since

there were six test trials with the marble pull game, a score of 3 means that

both Ss received three prizes. A score of 2 means that one S got two prizes

and the other S got either two, three, or four prizes. Although the taking-

turns score does not provide any information about the sequence of getting

prizes, it is a good measure of the degree to which Ss interacted by helping

each other get prizes. For one S to get the first three prizes is "taking-

turns" to the same degree as to alternate in getting prizes from one trial to

the next. Because cooperation was required in order for anyone to get any

prizes the "taking-turns" measure

11

as well as sharing over trials. A high score, then, means high in cooper-

ation and sharing.

The mean taking-turns scores for each training group and for the control

group is reported in Table 2. An analysis of variance showed that the dif-

- - -

INSERT TABLE 2 ABOUT HERE

ference between groups WaS significant at the .05 level F(4,51)=3.63.

A Newman-Keuls test suggested that the cooperation group differed significantly

at the .05 level from each of the other four groups and that no other differ-

ences were significant at the .05 level. Children whose training on the first

day was based on principles of cooperation were more cooperative in their in-

teraction in a new situation on the.second day than were children trained by

other methods. Furthermore, unlike cooperative training, the other training

methods did not increase cooperative interaction above the level for the un-

trained control group.

Cooperation Board Game. For this game there are two measures of cooper-

ation. In addition to the taking-turns measure, time Co solution bn each

trial is of interest. The time limit for a trial was 30 seconds, but Ss who

were fully cooperating with one another could easily complete the trial in

three or four seconds. The more competitive and conflicting the interaction,

and the mean times to solution

Table-3 reports the.-mean:takingturns scores.

fOr.-each .group..

12

The analysis of variance on taking-turns scores suggests nonsignificant

differences at the .05 level, F(4,51)=.91. However, the hypotheses made a

priori as well as the results on the marble pull game give some justification

for a separate comparison of the cooperationrgroup with the other groUps. The

difference between the cooperation group and the control group is signifticaht

only at.the .10 level, F(1,51)-3.3.

The results for the time to solution measure are in close agreement with

the taking-turns results; however, the analysis of variance on time to solu-

tion scores suggests more significance in the differences between groups.

Although the main effect is not significant at the .05 level, F(4,51)=1.77,;

a comparison between the cooperation group and the control group shows this

difference to be significant at the .05 level,.F(1,51)e5.0. A comparison be-

tween the cooperation group and the other three training groups considered

together also suggests a significant difference at the .05 level, F(1,51)=4.3.

Taken together, the results from the two measures suggest that the pairs

who had received training based on a concept of cooperation were more coopera-

tive on the cooperation board than the other groups.

Circle Matrix Game. The measure of cooperation for this game is the

taking-turns score. These scores are listed by condition in Table 4,

INSERT- TAPLE 4 -ARQUT KRE

An analysis of variance did not find a significant difference between, the five

groubs'at.the,.Q5 level, .F 4.-49 Th4 -resultS from the.tWo OreVfOes. games*,

13

justify comparisons between the cooperation groups and the other groups.

The difference between the cooperation group and the control group is signifi-

cant at the .05 level, F_(1,49)=5.2. The difference between the cooperation

group and the other three training groups considered together is also signifi-

cant at the .05 level, F(1,49)=4.3.

Thus, in a new situation folTowing training on two previous days, only

the cooperation group stands out as being more cooperative in a comparison of

the groups.

Pull-Block Game. This was another game with a time limit for each trial;

so time to solution scores as well as taking-turns scores are reported in

Table 5. Considering first the taking-turns scores, the main effect of treatment

INSERT TABLE 5 ABOUT HERE

is not significant at the .05 level, F(4,40)-1.6. However, the cooperation

group differs significantly from the control group at the .05 level, F(1,40)=5.1.

The difference between the cooperation group and the other three training groups

considered together is significant only at the .10 level, F(1,40)=3.0.

The time to solution scores differ between groups significantly at the 05

level, F(4,40)=3.2 The cooperation group differs significantly from the other

three training groups at the .05 level, T(1,40)=7.0.

As'a whole, the results from the 01.144410ame---iMelY:that:bleht-AayS'

after the 'final training session the.ehildren whose training emphasized concep

ually a reason for cooperating were more cooperative than children who were

.

eithee Untraieed-o other'.-MethedS'.. . . . . _ . . .

14

Discussion

Psychologists who have studied moral and social development would seem

to suggest that behavioral change (socialization) for five-year-olds is al-

most exclusively the product of reinforcement contingencies, obedience to

adult rules, and imitation of adult behavior (Bandura & Walters, 1963-

Kohlberg, 1963; Piaget, 1932). This orientation is at least partially a

consequence of viewing the five-year-old as a child whose conceptual and

mediational capacities are severely limited. For example, the limited poten-

tial of the five-year-old for reversible thought is said to severely limit

the child's capacity for cooperation (Piaget, 1932). To the degree that this

view implies that conceptual training is futile with five-year-olds, the re-

sults of the experiment described here suggest that the five-year-old is not

so handicapped conceptually as was thought. In fact, teaching a concept of

cooperation (namely, prizes can be obtained only by taking-turns) was the

only training method which consistently led to more cooperative interaction

in new situations.

The interpretation of the results favored by the authors is that simply

being reinforced for taking-turns, or simply observing models taking-turns,

or simply learning a label for taking-turns while being reinforced for it was

not sufficient training in taking-turns to result in transfer of the inter-

action pattern to new and somewhat different games. The five-year-olds in

the cooperation group were, however, capable of learning the concept that

prizes could be obtained only by taking-turns. This concept provided a basis'

for transfer of the taking-turns interaction to new games whenever the chil-

dr'en reCognized.-the.hewHgameat titUatibns in-WhiChprj4e$'Cdirld-06_:Obt0-ned..--

15

only by takino-turns. The children in the other training conditions did not

have a concept which connected their desire to get prizes with the strategy

of taking-turns on games of the sort presented to them.

Unfortunately, at least oneuother interpretation of the results cannot

be ruled out until further experiments are completed. It may be that coop-

eration training was most effective simply because the E exercised more

authority and unintentionally showed himself to be more concerned for the

cooperation group than for the other groups. It is true that the E made a

greater number of comments to pairs in the cooperation group than to pai s

in the other groups. Ss in the cooperation group, compared to other Ss, may

have better understood and more strongly felt that the adult wanted them to

take-turns. This interpretation will be tested by an experiment in which the

rule conformity and cooperation groups are equated for the number and inten-

sity of adult commands and comments which they experience. Tape recorded in-

structions will be used.

The present experiment was not designed to provide information about in-

cremental learning effects which might have occurred in the course of being

trained on various games. The children were trained on several games in order

to increase the likelihood of transfer of training. The results do suggest

that the first day of training was sufficient to result in significant differ-

ences between the cooperation group and the other groups when tested on the

second day. Whether further training served to increase or to actively main-

tain, or to have any effect at all on these differences cannot be determined.

16

Two other methodologjcal limitations are important. It would have been

desirable to have a control group in which Ss played all of the games, but

without training. Such a control, however, would be difficult to distinguish

from the reinforcement condition. Because the possibility of reinforcement

is inherent to the games, playing the games at all entails some training.

Another methodological shortcoming was the failure to incorporate measures

of learning other than the tests for transfer. It is of interest to know,

for example, whether the training procedures for a given game resulted in any

learning on that game. This is different from asking whether the training pro-

cedure affected behavior on a new and different game. Although all of these

questions could not be answered in an exploratory study of the sort completed,

future experiments might advantageously measure cooperative behavior before and

after training on a given game in addition to testing for transfer with different

games.

FOOTNOTES

1.. The authors are indebted to Dr. Carolyn STern, Director ofthe UCLA Head Start EValuation and REsearch Center for her con-tinuing interest and helpful comments. Srencer Kagan made manyconstructive sUggestions.

2. Harriet Braiker served as the female model in Treatment 2.

3. Edward Chaney helped in the testing in Sessions III and IV.

17

References

Bandura, A., & McDonald, F. G. The influence of social reinforcement and

the behavior of models in shaping children's moral judgments. Journal

of Abnormal and_Socialsychplogy, 1963, 67, 274-282.

Bandura, A., & Walters, R. Social learning_ and personality deyel_opment,

New York: Holt, Rinehart, & Winston 1963.

Chittenden, G. An experimental study in measuring and modifying assertive

behavior in young children. Monograph_ of:the _Society_for.iResearch'in

Child Development 1943 7 No. 31.

Crowley, P. M. Effect of training upon objectivity of moral judgment in

grade-school children. Journal of Personality and Social Psychology,

1968, 8, 228-232.

Kagan, S. & Madsen, M. E. Research proposal for University of California,

Los Angeles, Head Start Evaluation and Research Center August 1968.

Kohlberg, L. Moral development and identification. In H. W. Stevenson (Ed.)

National Society for the Study of Education Yearbook, Child Psychology,

Chicago: University of Chicago Press, 1963.

Kohlberg, L. The development of children's orientations toward a moral orde

I. Sequence in the development of moral thought. Vita Humana, 1963, 6,

11-33.

Madsen, M. C. Cultural differences with a marble-pull game. (In preparation.)

Nelson, L. & Madsen M. C. Cooperative and Competitive behavior of preschool

children as a function of reward condition, sex, and ethnic background.

In C. Stern (Ed.) Annual Report University of California, Los Angeles,

Head Start Evaluation and Research Office,-NPvember, 1967.

18

Peck, R. F., Havighurst, R. J , Cooper, R., Libienthal, J., & More, D.

-11-ny-_it2,1:g).-jx_L,f_sLi___p-ac-terclevelornent, New York: John Wiley, 1960.

Piaget, J. The moral judgment of the child. New York: Harcourt, Brace,

& _o., 1932.

ControlsI II

19

Table 1

Mean Ages of Children by Treatment

Reinforcement Modeling Rule Conformity Cooperation

Numberof Pairs 10 13 10

Mean Agein Months 62 63 60

Table 2

11 12

.62 62

Taking-Turns Scores on Marble Pull:Mean No. Prizes Obtained by Ss Getting Fewest Prizes During Six Trials

..60

Coctrol I

NI= 10Reinfo cement

N = 10ModelingN = 11

Rule ConformityN = 12

CooperationN = 13

1.6 1.5 .8

Table

1.1 2.5

Taking-Turns Scores And Tfmes to Solution Scores on CoOPeration.

Control I ReinforcementN = 10 N = 10

ModelingN = 11

Rule onformity Cooperation= 12 N = 13

Mean number ofprizes obtainedby Ss gettingfewFst prizesduring 4 trials

Mean times tosolution inseconds

20

Table 4

aking-Turns Scores on Circle Matrix Game: Mean Number of Prizes Obtained by Ss GettingFewest Prizes During Eight Trials

Control II

N1

= 13

Reinforcement Modeling Rule Conformity Cooperation

N = 9 N = 10 N = 10 N = 12

1.3 1.4 2.0

Table 5

1.5 2.8

Taking-Turns Scores and Times to Solution Scores on Pull-Block Game

Control IIN = 11

2an nUmber ofwizes obtained)y Ss- gettingFewest PrizesJuring .8 trials.-

2an times to;olution in 15 14 15

;econds

1.2

Reinforcement ModelingN . 8 N = 8

2.1 1.4

Rule Conformity Cooperation= 7 N = 13

1.6

14

2.6