UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Balance and agreement in children's social perception Gutman, Gloria Margaret 1970

You don't seem to have a PDF reader installed, try download the pdf

Item Metadata

Download

Media
[if-you-see-this-DO-NOT-CLICK]
UBC_1970_A1 G88.pdf [ 5.69MB ]
Metadata
JSON: 1.0093329.json
JSON-LD: 1.0093329+ld.json
RDF/XML (Pretty): 1.0093329.xml
RDF/JSON: 1.0093329+rdf.json
Turtle: 1.0093329+rdf-turtle.txt
N-Triples: 1.0093329+rdf-ntriples.txt
Original Record: 1.0093329 +original-record.json
Full Text
1.0093329.txt
Citation
1.0093329.ris

Full Text

BALANCE AND AGREEMENT IN CHILDREN'S SOCIAL PERCEPTION  by  GLORIA MARGARET GUTMAN B.A., University of B r i t i s h Columbia, 1961 M.A., University of Alberta, 1964  A THESIS SUBMITTED IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY  i n the Department of Psychology  We accept this thesis as conforming to the required  standard  THE UNIVERSITY OF BRITISH COLUMBIA September, 1970  In p r e s e n t i n g t h i s  thesis  an advanced degree at  further  fulfilment  o f the  requirements  the U n i v e r s i t y of B r i t i s h Columbia, I agree  the L i b r a r y s h a l l make i t I  in p a r t i a l  freely  available  for  this  thesis  f o r s c h o l a r l y purposes may be granted by the Head o f my Department  of  this  thesis for  It  financial  that  r e f e r e n c e and s t u d y .  agree t h a t p e r m i s s i o n f o r e x t e n s i v e copying o f  by h i s r e p r e s e n t a t i v e s .  for  or  i s understood that c o p y i n g o r p u b l i c a t i o n g a i n s h a l l not be allowed without my  written permission.  The U n i v e r s i t y o f B r i t i s h Columbia Vancouver 8, Canada  ABSTRACT The primary purpose of this study was bases of pleasantness and consistency ship between the two years.  to investigate the s t r u c t u r a l  ratings and to determine the r e l a t i o n -  types of judgement i n children ranging i n age from  A secondary purpose of t i e study was  5-12  to determine whether the r e s u l t s  of studies by Atwood (1969) and by Storm and Knox (1969) using a prediction procedure to investigate the developmental course of cognitive balance would generalize to a d i f f e r e n t dependent measure. Subjects i n the study were 80 children, 20 (10 males and 10 females) from each of the following age groups:  5-6,  7-8,  9-10,  and 11-12  years.  They rated hypothetical s o c i a l situations both f o r pleasantness and for consistency.  The situations were of the P-O-X  type, consisting of the  subject,  another person, and an unspecified, but important "thing." On the assumption that a f f e c t influences the s o c i a l perceptions of younger children more than considerations  of consistency  i t was  predicted  that i n their ratings of s o c i a l situations younger children would d i f f e r entiate l i t t l e between pleasantness and consistency as pleasant would also be rated as consistent).  ( i . e . , situations rated  Relative to the youngest  children, older children were expected to d i f f e r e n t i a t e more between pleasantness and  consistency.  Thus, i t was  predicted that as a function of increas-  ing age,  correlations between pleasantness and consistency  monotonically decrease across the successive Further, i t was  ratings would  age groups i n the study.  predicted that children at a l l age l e v e l s would attach  greater weight to agreement than to balance when making pleasantness ratings and  that younger children would also base consistency  ment than on balance.  However, balance was  ratings more on agree-  expected to exert greater i n -  ii  fluence than -agreement on the consistency ratings of older children.  This  follows from Zajone's (1968) review and i t s extension which suggest that agreement i s more important than balance when the dependent measure relates to affect whereas balance exerts greatest influence when the task relates to psychological consistency. The r e s u l t s f a i l e d to y i e l d evidence of age differences i n d i f f e r e n t i a tion between pleasantness and consistency. of ratings were high i n a l l groups.  Correlations between the two types  These were also no age differences i n the  r e l a t i v e weighting of balance and agreement.  Children i n a l l groups u t i l i z e d  balance to a s l i g h t l y greater extent than agreement when pleasantness was  the  c r i t e r i o n ; agreement was used to a s l i g h t l y greater extent than balance when the children rated for consistency. very small, however, i n comparison  The effects of balance and agreement were  to those of a t t r a c t i o n .  Children i n a l l  age groups appeared to base both pleasantness and consistency ratings primarily on a t t r a c t i o n ( i . e . , on the sign of the P/0  bond).  A cross-validation study conducted concurrently with the p r i n c i p a l study by an independent and  ,;  naive" IS yielded the same pattern of r e s u l t s .  Differences i n results obtained with children i n the rating s i t u a t i o n vs the prediction s i t u a t i o n were tentatively attributed to d i f f e r e n t i a l task complexity.  It was suggested that differences between adults ( c f . , Zajonc,  1968) and children i n the rating s i t u a t i o n may be due to differences i n i n formation processing a b i l i t i e s and/or to differences i n the strength of the balance "schema. ' 1  That i s , the "schema" or i m p l i c i t code for balance may  more firmly established i n adults than i n children.  This could perhaps  account for the fact that although adults u t i l i z e balance to a greater extent than agreement or a t t r a c t i o n i n the prediction s i t u a t i o n and when  be  iii  rating for consistency, strong balance effects among children are obtained only i n the easier prediction s i t u a t i o n .  The balance "schema" i n children,  i n other words, may not be of s u f f i c i e n t strength to withstand the competition of alternative biases such as a t t r a c t i o n , agreement, and p o s i t i v i t y when the more complex rating task i s used.  iv  TABLE OF CONTENTS CHAPTER I.  Page Introduction . . . . . . . . . . . . . .  1  (a) Basic p r i n c i p l e s of Balance Theory (b) Empirical studies of Cognitive Balance ( i ) i n adults ( i i ) i n children (c) Statement of purpose (d) Hypotheses II.  Method  28  (a) Subjects (b) Design (c) Procedure (i) training ( i i ) pleasantness  ratings  ( i i i ) tests of conservation (iv) consistency ratings III.  Assessment of the Procedures  44  (a) Q u a l i t a t i v e observations (b) R e l i a b i l i t y of the ratings (c) Performance on the interpolated tests of conservation IV. V. VI. VII. VIII.  Results and Discussion of Hypotheses  49  A n c i l l a r y Findings and their Implications  68  Results of the Cross-validation Study  73  Further Discussion  83  Summary  91  V  Page BI BLIOGRAPHY  94  APPENDICES Appendix 1.  P i l o t studies  Appendix 2.  A diagrammatic representation of the structures included i n l e v e l 1 and l e v e l 2 of factor 0 . . . . ,  104  Mean pleasantness and consistency ratings f o r each structure  106  Appendix 3. Appendix 4.  98  Multiple regression analysis of the ratings  108  (a) the p r i n c i p a l study (b) the cross-validation study Appendix 5.  Summary of analyses of variance of pleasantness, consistency, and tension ratings obtained from adult Ss (Gutman, 1969)  116  vi  LIST OF TABLES Page Table 1  A comparison of balance, agreement, and a t t r a c t i o n (Zajonc, 1968). .  13  Reorganization and extension of Zajonc's table comparing balance, agreement, and attraction  15  Percent of balanced and t r a n s i t i v e responses and explanations i n children ranging i n age from 5-12 years (Storm and Knox, 1969) . .  21  Balance and agreement i n prediction of p o l i t i c a l preferences of a l i k e d and a d i s l i k e d other (Knox and Gutman, 1968)  23  Table 5  Training cards 1-9  35  Table 6  Mean r e l i a b i l i t y c o e f f i c i e n t s f o r pleasantness and consistency ratings  46  Percent of Ss i n each age group showing quantity and volume conservation  47  Table 2  Table 3  Table 4  Table 7  Table 8  Table 9 Table 10  Mean pleasantness-consistency correlations f o r each age group  50  Design of the analyses of variance  53  Summary of the analysis of variance of pleasant-  Table 11  Zajonc indices f o r pleasantness  53 58  Table 12  Mean pleasantness ratings f o r structures containing agreement and disagreement  60  XI6 S S ITS t X I l ^ S  Table 13  «  «  «  e  e  e  a  *  a  o  e  *  *  «  4  «  «  o  «  o  Summary of analysis of variance of consistency ratings  . . . . .  62  Table 14  Zajonc indices f o r consistency  65  Table 15  Mean pleasantness and consistency ratings relevant to the agreement by structures i n t e r action  70  Mean r e l i a b i l i t y c o e f f i c i e n t s f o r pleasantness and consistency (Cross-validation study) . . . . . .  74  Table 16  vii Page Table 17 Table 18 Table 19 Table 20  Mean pleasantness-consistency correlations for each age group (Cross-validation study)  . . .  Summary of analysis of variance of pleasantness ratings (Cross-validation study)  77  Summary of analysis of variance of consistency ratings (Cross-validation study)  80  Rank order of mean pleasantness and consistency ratings  85  Table 21  Zajonc indices f o r P i l o t Study II  Table 22  Mean pleasantness and consistency ratings f o r each structure . Average regression equations f o r pleasantness and consistency  Table 23  75  . . . . . . .  .  103  107 110  Table 24  Kendall c o e f f i c i e n t s v>f concordance  Table 25  Number of Ss for whom predictors are s i g n i f i c a n t components of pleasantness and/or components of consistency ratings . . . . . . . .  113  Average regression equations for pleasantness and consistency (Cross-validation study) . .  114  Summary of analysis of variance of pleasantness ratings (Gutman, 1969) . . .  117  Summary of analysis of variance of consistency ratings (Gutman, 1969)  118  Summary of analysis of variance of tension ratings (Gutman, 1969) . . .  119  Table 26 Table 27a  Table J7b  Table 27c  Ill  viii  LIST OF FIGURES Page Figure 1  Diagrammatic representation of balanced and imbalanced s;ructures  3  Figure 2  Quadrants u t i l i z e d i n computing Zajonc indices . . .  Figure 3  Diagi ammatic representation of the experimental design . . . . .  . . . . . . . . . . .  10  30  Figure 4  Rating scale I  Figure 5  Rating scale II  Figure 6  Content of t r a i n i n g cards 10-17  36  Figure 7  Rating scale I I I  37  Figure 8  Rating scale IV  41  . . . . . . . . . . . .  32 33  ±x  ACKNOWLEDGEMENTS I would l i k e to extend sincere thanks to my advisor, Dr. R. Knox, for his generous help and encouragement at a l l stages of the research, I would also l i k e to thank the other members of the doctoral committee Dr. R. Potashin, Dr. E. Signori, Dr. D. Smith, Dr. T. Storm and Dr. F. V a l l e for their valuable comments and suggestions. Thanks go also to S i s t e r Mary Neva of Holy T r i n i t y School, Mr. C. S i d d a l l of McKechnie School, and Mr. P. Bak of the Vancouver Talmud Torah f o r providing subjects; to Mr. Dave Malcom for a s s i s t i n g with the s t a t i s t i c a l analysis of the data; and to Miss Carole Hilstrom for typing the manuscript. Special thanks go to Miss E l s i e Eccles for conducting the crossv a l i d a t i o n study.  1  CHAPTER I INTRODUCTION Considerable attention has been focused, i n recent year, on the development and extension of theories of "cognitive consistency."  Abelson  and Rosenberg (1958), Cartwright and Harary (1956), Festinger (1957), Heider (1946, 1958), McGuire (1960), Newcomb (1953) and Osgood and Tannenbaum (1955) are among those who have proposed theories of this type.  Common to  a l l i s the postulation of tendencies toward meaningful and harmonious organization of the individual's thought, b e l i e f s , attitudes, and behaviour. In common also i s the postulation of tendencies toward reduction of i n consistencies between elements of the individual's cognitive system.  These  theories d i f f e r , however, i n terminology, i n r i g o r of d e f i n i t i o n , and i n the type of s i t u a t i o n to which they most t y p i c a l l y apply. The present study relates most d i r e c t l y to the formulations of Heider (1946, 1958).  In Heider's Balance model, analysis i s focused on the  P-O-X  t r i a d , consisting of P (the perceiver), 0 (another person) and X (either a third person or an impersonal e n t i t y ) . are of two types:  Relations between t r i a d i c  elements  sentiment relations ( l i k e or d i s l i k e ) and unit r e l a t i o n s  (associated with or not associated with). . . . separate e n t i t i e s comprise a unit when they are perceived as belonging together. For example, members of a family are seen as a unit; a person and h i s deeds belong together. (Heider, 1958, p.  176)  Units are formed on the basis of perceived s i m i l a r i t y , proximity, f a m i l i a r i t y , ownership, causality, or kinship.  Balance i s defined i n terms of the number  of positive and negative r e l a t i o n s i n the P-O-X  triad.  Like (L) and  associative relations (U) are c l a s s i f i e d as p o s i t i v e ; d i s l i k e (DL) and  non-  2  associative relations (not-U) are c l a s s i f i e d as negative.  A t r i a d i c system  i s considered to be i n a state of balance i f three relations are p o s i t i v e or i f two are negative and one i s p o s i t i v e .  Structures 1, 2, 7, and 8 i n  Figure 1 s a t i s f y the c r i t e r i a for balance.  Triads consisting of two  Figure 1  positive and one negative r e l a t i o n  are considered imbalanced.  Heider states,  though with some equivocation, that imbalance also exists i f a l l three relations are negative. . . . Four [structures] are balanced, containing three positive r e l a t i o n s , or one positive and two negative. Four are unbalanced, with three negative or two positive relations. . . . I f two negative relations are given, balance can be obtained either when the t h i r d r e l a t i o n i s p o s i t i v e or when i t i s negative; although there appears to be a preference for the positive a l t e r n a t i v e . (Heider, 1958, Pp. 204-206) The fundamental assumptions of Balance Theory are (a) that sentiment and unit relations tend toward balance, and (b) that imbalanced states produce tension and generate forces to restore balance.  Heider views the  balanced state as "a s i t u a t i o n i n which the perceived units and experienced sentiments co-exist without stress; there i s thus no pressure toward change either i n the cognitive organization or i n the sentiment" (Heider, 1958, p. 176). Cartwright and Harary  (1956) have extended the range of situations  to which Balance Theory i s applicable by defining balance i n graph theoretical terms.  According to their formulation, a system i s balanced  3  Balanced  P  *0  \/  P  >0  \/  (1)  structures  P  jO  \/  (2)  (7)  P---*,0  \/ (8)  Imbalanced structures p \  ,0 /  v  (3)  F i g u r e 1:  P—f  P---»,(> v (4)  (5)  ijias;rar.: 7atic representation i  structures.  v  of balanced and  (6)  imbalanced  (Solid lines indicate p o s i t i v e r e l a t i o n s ;  dashed l i n e s indicate negative relations.)  4  i f a l l the semicycles within the system are p o s i t i v e .  (A setnicycle i s  defined as a c o l l e c t i o n of two or more l i n e s forming a closed path i n a graph.  For example, the structure Pc'-' "*0 contains three  PO^PO; PO,OX~,XP; and PO,OX,XP.  semicycles:  The sign of a semicycle i s p o s i t i v e i f the  product of the signs of the l i n e s forming the semicycle i s positive.) This d e f i n i t i o n of balance i s applicable to structures containing any f i n i t e number of r e l a t i o n s . There i s no l i m i t a t i o n on the type and number of relations defined on a p a r t i c u l a r set of elements.  In addition, the  Cartwright and Harary formulation provides a means of handling  non-reciprocal  relations between s t r u c t u r a l elements (e.g., a s i t u a t i o n where P l i k e s 0, but 0 does not l i k e P).  Heider states that non-reciprocal l i k i n g i s  imbalanced, but f a i l s to include such situations i n h i s s t r u c t u r a l d e f i n i t i o n of balance.  The d e f i n i t i o n of balance i n terms of semicycles has the added  advantage of permitting degrees of balance to be s p e c i f i e d . This i s accomplished by computing the proportion of p o s i t i v e semicycles number of semicycles  to t o t a l  i n the system.  Empirical Studies of Cognitive Balance A number of techniques  have been devised to test Heider's  Balance theory, and i t s extensions.  (1946, 1958)  One method follows from the assumption  that states of imbalance generate tension and are unpleasant ( c f . , Heider, 1958,  p. 207), and requires that subjects rate hypothetical s o c i a l situations  for subjective feelings of "unpleasantness."  Jordan (1953), f o r example,  had subjects rate sixty-four t r i a d i c structures f o r unpleasantness.  The  structures represented a l l possible combinations of L, DL, U, and not-U relations and were of the following form:  "I d i s l i k e 0; I l i k e X; 0 has  5  no sort of bond or relationship with X."  The subject was instructed to  imagine himself i n the s i t u a t i o n , playing the r o l e of " I " and then to rate the s i t u a t i o n f o r pleasantness or unpleasantness on a ninety-point scale. The results of Jordan's study supported Heider's hypothesis—there was a s t a t i s t i c a l l y s i g n i f i c a n t tendency f o r balanced structures to be rated more pleasant than imbalanced structures.  Contrary to prediction however,  subjects tended to distinguish between balanced structures containing positive relationships between P and 0, and those i n which the P/0 r e l a t i o n ship was negative, the l a t t e r being rated considerably more unpleasant than the former.  In fact, the mean rating f o r balanced structures with negative  P/0 bonds was hardly distinguishable from that given to imbalanced  structures.  Subsequent studies which have used pleasantness-unpleasantness ratings to test derivatives of the balance hypothesis show a s i m i l a r discrepancy between prediction and results (e.g., Hershkowitz, 1954; Price, Harburg, and Newcomb, 1966; Rodrigues, 1966; Steiner and Spaulding, 1966). Jordan (1953) attempted to account f o r t h i s , and other discrepancies between prediction and r e s u l t s , by suggesting that "the o r i g i n a l coordinating d e f i n i t i o n of balance and pleasure i s f a u l t y . "  He recommends instead that  balance "be coordinated with the concept of a strong or good 'gestalt.'  A  strong gestalt i s characterized i n p r a c t i c a l l y a l l 'gestalt' t h e o r e t i c a l l i t e r a t u r e as a most proper, inner-necessary state.  Balanced situations can  therefore be considered to be experienced as more proper than imbalanced situations independent of their degree of experienced pleasantness. i s not synonymous with pleasantness.  Propriety  For many, r e t r i b u t i o n f o r sins i s  proper, but few i f any consider i t to be pleasant" (p. 282). Following this lead, Knox (1963) had subjects (n=10) rate a number  6  of hypothetical s o c i a l situations of the P-O-X for 'consistency."  Consistency was  type f o r pleasantness  defined as "a state of l o g i c a l  congruity among the persons and thinj» i n the [P-O-X] s i t u a t i o n . subject was  and  !i  The  instructed to rate situations that seemed "to hang together  i n a l o g i c a l , sensible, and r a t i o n a l manner" toward the consistent end of the scale.  He was  instructed to rate s i t u a t i o n s that seemed '"out of k i l t e r , '  i l l o g i c a l , and irrational,!' toward the inconsistent end of the scale.  The  results showed a weak, but s i g n i f i c a n t p o s i t i v e r e l a t i o n s h i p between pleasantness  ratings and consistency ratings (r = +.20).  Knox concludes  that pleasantness and consistency are not equivalent terms (p. 81). subsequent study by Gutman (1969) was  designed  A  to r e p l i c a t e Knox's (1963)  findings and to determine the r e l a t i o n s h i p or possible equivalence of the concepts of pleasantness, consistency, and tension.  The subjects i n this  study (n=84) rated the eight structures shoxm i n Figure 1 on 9-point scales anchored by the terms unpleasant-pleasant, tension-very strong tension. c o r r e l a t i o n of +.47  inconsistent-consistent, and no  The results yielded a s i g n i f i c a n t Pearson  between the pleasantness  The correlations between pleasantness  and consistency ratings.^"  and tension ratings and between  consistency and tension ratings were -.81  and. -.42  respectively.  findings indicate that subjects define as unpleasant that they define as tension-provoking.  These  the same s i t u a t i o n s  The results suggest also that  ^"Differences i n the time i n t e r v a l between pleasantness and consistency ratings may perhaps account f o r the difference i n magnitude of the c o r r e l a t i o n between pleasantness and consistency i n the Knox (1963) and the Gutman (1969) studies. In Knox's study, pleasantness and consistency ratings were collected i n four sessions held on alternate days i n an ABBA design. In Gutman's study, pleasantness, consistency, and tension ratings were collected i n a single one hour session.  7  subjects d i f f e r e n t i a t e between situations that are unpleasant and tensionprovoking and those that are \sychologically inconsistent. Rather than a t t r i b u t i n g the d i s t i n c t i o n i n the rated pleasantness of balanced structures with positive P/0 bonds and those with negative P/0 bonds to a f a u l t i n the coordinating d e f i n i t i o n of balance and pleasantness, Rodrigues  (1965) has suggested that "agreement" may act as an  independent source of "cognitive bias," c o n f l i c t i n g with tendencies toward balance i n the P/0 negative case.  As shown i n Figure 1, where P l i k e s 0,  balance i s achieved when P and 0 have the same attitude toward X. d i s l i k e s 0, balance results from disagreement  regarding X.  Where P  Thus, i n the  P/0 p o s i t i v e case preferences for balance and agreement work i n the same direction. directions.  In the P/0 negative case, agreement and balance work i n opposite Jordan's results suggest that subjects' ratings are affected by  this c o n f l i c t of forces.  Mean unpleasantness ratings were lowest i n  balanced structures with agreement and highest (most unpleasant) i n balanced structures with disagreement.  Imbalanced structures containing  agreement were rated somewhat more pleasant than imbalanced containing disagreement.  structures  Rodrigues (1965) obtained s i m i l a r r e s u l t s when  he had subjects rate t r i a d i c structures for tension. His data show an increase i n mean tension ratings as one goes from balanced structures with agreement, through imbalanced structures with agreement and structures with disagreement, to balanced structures with  imbalanced  disagreement.  Newcomb (1968), on the other hand, feels that agreement or disagreement between P and 0 may have l i t t l e e f f e c t on the perceived pleasantness of a s i t u a t i o n i n which P and 0 d i s l i k e one another.  He states that the  negative P/0 bond "engenders i t s own tension, which i s independent of the  8  kind of tension that i s i n t r i n s i c to the notion of balance, as defined by Heider and by others who  have followed him"  (p. 33).  Consistent with this  l i n e of reasoning, Jordan (1966) reports that a considerable portion of the variance i n the 1953  study i s accounted for by the sign of the P/0 bond.  The results of studies by Steiner and Spaulding (1966) and Hershkowitz (1954) also suggest that the sign of the P/0 bond could be an important determinent of pleasantness ratings.  The question that remains i s which  factor contributes most to s o c i a l perception i n the t r i a d i c s i t u a t i o n : balance, agreement, or a t t r a c t i o n ( i . e . , sign of the P/0 bond)? Zajonc (1968) analyzed  data from a number of empirical studies of  s o c i a l perception and reports that approximately h a l f favor agreement over balance as the more c r i t i c a l determinant of responses.  None favor a t t r a c t i o n .  Included i n his analysis were several studies involving  pleasantness-  unpleasantness ratings (Hershkowitz, 1954;  P r i c e , Harburg,  and Newcomb, 1966;  Rodrigues, 1966;  Jordan, 1953;  Steiner and Spaulding,  1966), a study  which required prediction of missing relations (Morrissette, 1958), and two so-called "ease-of-learning" studies (Zajonc and Burnstein, 1965a, 1965b). In the prediction study, subjects were required to role-play a move into an apartment. task was  Sentiments among some of the roommates were given.  The  subject's  to predict the remaining sentiments and to rate how much tension he  would f e e l i n such a s i t u a t i o n . The two ease-of-learning studies u t i l i z e d a paired-associates  technique.  The dependent measure was  errors made i n learning the signs of relations i n P-O-X  the number of  and P-O-X-Y structures.  (Structures of type P-O-X-Y involve two persons and their attitudes toward two  issues.) The method of analysis u t i l i z e d by Zajonc involved c a l c u l a t i o n of  9  separate indices for balance, agreement, and a t t r a c t i o n .  These indices are  derived by ordering the eight t r i a d s shown i n Figure 1 into four quadrants. As shown i n Figure 2, the quadrants are derived by considering the sign of  the P/0 bond (+ or -) and the presence or absence of agreement between  P and 0 with regard to X.  Quadrant A contains structures i n which there  Figure 2  i s a p o s i t i v e bond between P and 0 and agreement with regard to X ( i . e . , structures  P  *0  and P~——40) ; quadrant B contains structures i n  *x*  x-  which there i s a negative bond between P and 0, and agreement concerning X (structures P  $0 and P  V  >0). v  Structures P*>  >0 and P  V  x*  V  represented i n quadrant C; structures P- - -» 0 and P — X i n quadrant D.  ~»0 are  -j 0 are represented X  Only quadrants A and D contain balanced structures. The  effects of balance are estimated by the r a t i o of average scores (on whatever dependent measure was used i n a p a r t i c u l a r study) f o r structures i n balanced quadrants  to the average scores for structures i n imbalanced  ( i . e . , quadrants  quadrants  A + D „ ) . The effects of agreement are estimated by the  r a t i o of quadrants  containing agreement to those containing disagreement  A + B ( )•  The effects of a t t r a c t i o n are estimated by the r a t i o of P/0 A + C positive to P/0 negative quadrants ( _ , _ . ) . The r e l a t i v e strength of each c  +  p  D  T  D  s t r u c t u r a l factor i s determined by comparing the magnitude of indices computed from these r a t i o s .  I f the effects of agreement are stronger than  10  Attraction  PL 0  PL~0  +  Agreement  A  B  C  D  Disagreement  Figure 2:  Quadrants u t i l i z e d i n computing Zajonc indices.  11  those of balance, the agreement index i s greater than that for balance. The s i t u a t i o n i s reversed i f the effects of balance are stronger than those of agreement.  The index for a t t r a c t i o n would be greatest i f this were the  most important determinant of the subjects' responses. A concrete example may perhaps c l a r i f y the procedure following i n computing Zajonc indices.  Data i n the example come from Jordan's (1953)  study. Quadrant  Structures i n Quadrant  Mean rating for each structure  Mean rating for each quadrant  A  +++  22.1 33.8  27.9  B  -++  59.1 69.9  64.5  C  +++-+  64.5 67.1  65.8  -+-  64.2 71.5  67.8  D  +—  —+  Attraction PL 0  PL*"o  27.9  64,.5  65.8  67,.8  +  •P  Agreement Disagreement  \  Index of Agreement: Index of A t t r a c t i o n : Index of Balance:  27.9 + 64.5 65.8 + 67.8 " . 27.9 + 65.8 _ 64.5 + 67.8 27.9 + 67.8 64.5 + 65.8 "  12  The results of Zajonc's (1968) analysis are shown i n Table 1.  As  he reports, approximately half of the studies i n the'table favor agreement over balance as the more c r i t i c a l determinant of s o c i a l perception.  As  Knox (1969) has pointed out, however, the bulk of evidence i n favor of agreement i n the table comes from studies which u t i l i z e pleasantness ratings.  Reorganization and extension of Zajonc's table (see Table 2)  Table 1  shows that indices of balance are greater than those of agreement and attraction i n tasks involving prediction  of missing r e l a t i o n s , ease-of-  learning, consistency ratings, or s t a b i l i t y assessment.  (Studies of the  l a t t e r type require the subject to indicate whether he would expect a p a r t i c u l a r structure to remain the same or to change over time.)  Thus,  balance seems to be the more important factor when the dependent measure relates to psychological consistency.  Agreement would seem to be more  Table 2  important when the measure i s based on a f f e c t .  13  Table 1: A comparison of balance, agreement, and a t t r a c t i o n (Zajonc, 1968, p. 348) Effect r a t i o  Study, measure, and condition  Hershkowitz  p and o agree  p and ,o disagree  +  C  0)  0  8 < u M  c o  •H  J-i  a cd  u u *->  o c CO  tH  CO  pL o  pL o  pL o  pL o  20.1 24.3  54.2 54.7  66.6 60.6  62.9 62.9  1.74 1.35 1.46 1.56 1.38 1.32  27.9  64.5  65.8  67.8  1.44 1.41 1.36  73%  37%  27%  63%  1.22  —  2.12  6%  41%  87%  36%  2.61 —  3.04  27.5 22.5 34.3  65.6 55.0 45.0  63.0 73.1 61.7  57.0 64.5 58.9  1 .29 1 .35 1.52 1 .77 1 .25 1 .47 1 .52 1.08 1 .14  21.9  58.8  64.9  73.3  1 .71 1 .52 1 .40  33.2  50.9  60.1  67.7  1 .52 1 .27 1 .10  <  <  «  (1954)  Unpleasantness scores on a 100point rating scale a) Values b) Objects Jordan (1953) Unpleasantness scores on a 90point rating scale Morrissette  (1958)  Percent of Ss predicting a positive or negative r e l a t i o n between o and x. Price, Harburg, and Newcomb,(1966) Percent of j>s reporting unpleasant affect Rodrigues  (1966)  Unpleasantness scores on a 90point rating scale a) b) c) d)  Control ( r e p l i c a t i o n of Jordan) Strong r e l a t i o n among peers Weak r e l a t i o n among peers Strong r e l a t i o n between p and an expert e) Weak r e l a t i o n between p and an expert  Zajonc and Burnstein (1965a) Errors i n learning of structures a) Important issue b) T r i v i a l issue  1.87 3.87  2.58 2.47  3.07 2.80  1.85 3.00  1 .10 0 .90 1 .53 0.91 0 .82 0 .77  Zajonc and Burnstein (1965b) Errors i n learning of structures  2.45  3.35  3.06  1.91  0 .85 0 .95 1 .47  14  Table 1 Continued  Study, measure, and condition  p and o agree  +  -  p and o disagree  +  -  pL o  pL o  pL o  pL o  31.20 30.47 55,93 58.32  17.68 14.79 33.75 34.32  17.88 19.47 33.23 34.00  22.37 22.47 43.25 42.68  Effect ratio c u o •H C 4J cu CJ o e cd c I-I CO t-1  an  <  JJ 4-1  rH CO  1.22 1.34 1.16 1.20  1.51 1.54 1.48 1.47  <5  Steiner and Spaulding (1966) Pleasantness ratings from 1 to 18 (sum of four items) a) b) c) d)  Wisconsin sample males Wisconsin sample females I l l i n o i s sample (8 items) males I l l i n o i s sample (8 items) females  1.21 1.08 1.17 1.21  15  Table 2:  Reorganization and extension of Zajonc's table comparing balance, agreement, and attraction E f f e c t Ratio p and o agree  Study, measure, and condition  p and o disagree  4-1  c 0) e <D  01 U ID  C  o •H 4-1 O  « u  4-> 4J  O  c  rt i-H rt CO  pL o  pL o  pL o  pL o  6.91  3.40  4.17  1.74  1.74 2.15 1.14  6.77  3.55  4.16  4.46  1.20 1.36 1.46  .76  3.98  3.66  6.14  2.07 2.29 1.11  <  <  I. Pleasantness, tension, or consistency ratings Gutman Q.969) a) Pleasantness scores oh a 9-point scale b) Consistency scores on a 9-point scale c) Tension scores on a 9-point scale Hershkowitz (1954) Unpleasantness scores on a 100-point scale a) values b) objects  20.1 24.3  54.2 54.7  66.6 60.6  62.9 62.9  1.74 1.35 1.46 1.56 1.38 1.32  27.9  64.5  65.8  67.8  1.44 1.41 1.36  Jordan (1953) Unpleasantness scores on a 90-point scale Knox (1963) a) Pleasantness scores on a 9-point scale (reciprocated sentiments only) 1) positive unit relations 2) negative unit relations  8.69 5.97  2.90 4.45  6.45 5.10  1.33 3.94  1.49 3.58 1.07 1.15 1.32 1.04  8.52 6.74  3.60 4.39  4.41 4.28  7.35 6.92  1.03 1.18 1.98 .99 .97 1.58  6%  41%  87%  36%  2.61  b) Consistency scores on a 9-point scale (reciprocated sentiments only) 1) positive unit relations 2) negative unit relations Price, Harburg, and Newcomb (1966) Percent of S_9 reporting unpleasant affect  —  3.04  16 Table 2 Continued  Effect Ratio  Study, measure, and condition  p and o agree  C  c o  •rl  s  4J CJ CO  Qi  U  QJ  r-l  4-1 4-1  a> u a  CO rH CO PQ  pL o  pL o  pL o  1.5  2.3  3.1  4.2  1.92 1.41 .95  27.5 22.5 34.3  65.6 55.0 45.0  63.0 73.1 61.7  57.0 64.5 58.9  1.29 1.35 1.52 1.77 1.25 1.47 1.52 1.08 1.14  21.9  58.8  64.9  73.3  1.71 1.52 1.40  33.2  50.9  60.1  67.7  1.52 1.27 1.10  17.88 19.47 33.23 34.00  22.37 22.47 43.25 42.68  1.21 1.08 1.17 1.21  27%  63%  1.22  +  OC  <  <!  (1965)  Tension ratings on a 7-point scale Rodrigues  4J  pL o +  Rodrigues  p and o disagree  (1966)  Unpleasantness scores on a 90-point scale a) b) c) d)  Replication of Jordan Strong relations among peers Waak relations among peers Strong relations between p and an expert e) Weak relations between p and an expert Steiner and Spaulding (1966) Pleasantness ratings from 1 to 18 (sum of 4 items)  a) b) c) d)  Wisconsin males Wisconsin females I l l i n o i s males I l l i n o i s females  31.20 17.68 30.47 14.79 55.93 33.75 58.32 34.32  1.22 1.51 1.34 1.54 1.16 1.48 1.20 1.47  I I . Prediction of missing relations Morrissette (1958) Percent of Ss predicting a'' positive or negative r e l a t i o n between o and . x  73%  37%  —  2.12  III. Ease-of-learning studies Zajonc and Burnstein (1965a) Errors i n learning of structures a) important issue b) t r i v i a l issue  1.87 3.87  2.58 2.47  3.07 2.80  1.85 3.00  1.10 0.90 1.53 0.91 0.82 0.77  11 Table 2 Continued  Effect Ratio  Study, measure, and condition  p and o agree  p and o disagree  u G 0)  e •  U O ra  u  4-1  <  <  <u <u  pL o  pL o  pL o  pL o  2.45  3.35  3.06  1.91  91%  34%  43%  85%  +  +  c  o -M  CO  u  4J  cu  u c CO t-i  CO PQ  Zajonc and Burnstein (1965b) Errors i n learning of structures  0.85 0.95  1.47  IV. S t a b i l i t y of structures Burnstein (1967) Percent of subjects who predict no change  .98 1.21 2.29  18  Balance i n Children The majority of empirical studies of cognitive balance have been conducted with adolescent or adult subjects.  Few investigators have  studied balance i n children, although such studies could have important implications for the o r i g i n of tendencies toward cognitive consistency. Atwood (1969) attributes this neglect to the conceptual r e l a t i o n between Balance Theory and Gestalt psychology.  He states:  " I t i s t y p i c a l l y assumed  (albeit i m p l i c i t l y ) that the regulative patterns of equilibrium i n a t t i t u d i n a l thought are rooted i n cognitive pregnanz and do not depend upon s p e c i a l lines of development" (p. 74). has come into question.  Recently, however, this assumption  Atwood (1969), for example, feels that an under2  standing of t r a n s i t i v i t y i s a necessary precondition for balance.  Accord-  ing to the theory and research of Piaget (1950), the a b i l i t y to make t r a n s i t i v e inferences presupposes a reversible grouping of mental operations and does not, i n general, appear i n c h i l d r e n before 7-8 years of age. Atwood concludes  that thinking according to the balance model must therefore  a r i s e during the period of concrete operations (7-11 years) or thereafter. Storm and Knox (1967) reached a s i m i l a r conclusion.  They, too, base their  argument on the operational s i m i l a r i t y between balance and  transitivity,  noted e a r l i e r by Heider (1958, p. 206). Studies by Atwood (1969) and by Storm and Knox (1969) were designed to test these contentions.  Both examined cognitive balance by means of  tasks which involved p r e d i c t i o n of missing relations i n three person 2 A r e l a t i o n i s designated t r a n s i t i v e i f (aRb) and (bRc) imply (aRc). An example frequently cited i s the r e l a t i o n "greater than:" i f a i s greater than b, and b i s greater than c, then a i s greater than c.  19  ( i . e . , P-O-Q) systems.  Atwood (1969) preselected subjects on the basis of  their Piagetian "stage" as determined by (a) their chronological age, and (b) t h e i r performance on tests of conservation of quantity and volume and on tests of concrete and verbal s e r i a t i o n .  Those nursery school  children (5-6 years of age) unable to conserve quantity or to seriate on a concrete plane were designated " i n t u i t i v e . "  Children aged 8-9 who were  able to conserve quantity and s e r i a t e on a concrete plane but who  failed in  volume conservation and verbal s e r i a t i o n were designated "concrete operational."  Children aged 11-12 who were able to conserve volume and  seriate verbally were designated "formal operational." Atwood found that i n t u i t i v e subjects f a i l e d to balance any of the three-person systems that, were presented to them.  Concrete operational subjects "conformed to the  balance model with surprising r i g o r , often reacting as i f the implication of each r e l a t i o n within a t r i a d i c system for the other two were purely a matter of l o g i c .  Formal operational subjects consistently balanced the t r i a d s ,  but recognized that a s i t u a t i o n of cognitive balance represents only one of many possible r e l a t i o n a l arrangements i n a given three-person system" (Atwood, 1969, p. 73). Storm and Knox (1969) presented balance and t r a n s i t i v i t y items to f i v e groups of children (n=20 i n each group) selected solely on the basis of age (5-12 years).  In each item, relations between two persons were  given and the subject's task was j u s t i f y his answer.  to predict the third r e l a t i o n and to  As shown i n Table 3 there was  a marked and s t a t i s t i c a l l y  s i g n i f i c a n t (p<.01) increase i n the percent of balanced responses  and  balanced and t r a n s i t i v e explanations between the ages of 6-7 and 8-9 years, the approximate age of t r a n s i t i o n from pre-operational to concrete  20  Table 3  operational functioning (Piaget, 1950,  chapter 5).  Knox (1969). has attempted to explain the developmental course of balance by postulating a s h i f t i n the influence of a f f e c t between the preoperational and concrete operational stages.  He suggests that the s o c i a l  perception of younger, pre-operational children may  be influenced more by  affect than by a desire for or tendency toward consistency.  This hypothesis  derives from the observation that young children seem to predict missing s o c i a l r e l a t i o n s i n a manner designed to produce a happy or pleasant, though not necessarily balanced ending, whereas older children seem to rely more on p r i n c i p l e s of balance i n predicting missing s o c i a l r e l a t i o n s . This f i t s w e l l with Piaget's stage.  (1930) characterization of the  pre-operational  The pre-operational stage, he states . . . i s characterized from the l o g i c a l point of view, by egocentricity; on the one hand, there i s an absence of the desire to find l o g i c a l j u s t i f i c a t i o n for one's statements, and on the other hand, syncretism combines with juxtaposition to produce an excess of subjective and a f f e c t i v e relations at the expense of genuine l o g i c a l implication. (Piaget, 1930, I f , indeed, a f f e c t dominates s o c i a l perception at the  p.  303)  pre-operational  l e v e l , i t would follow that younger children would u t i l i z e agreement to a greater extent than balance when called upon to perform tasks involving psychological consistency  (e.g., tasks requiring prediction of missing  r e l a t i o n s , consistency ratings, e t c . ) .  The rationale for this prediction  i s based upon Zajonc's (1968) review and i t s extension which suggest that  21  Table 3:  Percent of balanced and t r a n s i t i v e responses and explanations in children ranging i n age from 5-12 years (Storm and Knox, 1969).. Transitive Responses  Age  Transitive Explanations  Balanced Responses  Balanced Explanations  5:3- 5:11 years  81.6%  26.2%  59.6%  29.0%  6:2- 7:0  years  93.3%  39.1%  61.2%  30.9%  8:1- 9:0  years  85.0%  70.8%  86.2%  72.8%  9:4-10:11 years  97.9%  85.8%  90.3%  80.9%  11:8-12:10 years  98.3%  77.5%  89.6%  84.0%  22  pleasantness ratings are determined more by agreement than by balance. However, i f the influence of a f f e c t decreases with increasing age, older children would be expected to u t i l i z e balance to a greater extent than agreement i n these tasks.  Of course, i n the pleasantness rating s i t u a t i o n ,  agreement should exert greater influence than balance at a l l age l e v e l s . Data i n accord with this expectation were obtained by Knox and Gutman (1968), i n a study conducted to the 1968 U.S.  i n Bellingham, Washington, immediately  federal e l e c t i o n .  prior  In this study, children i n the  operational age range (9-14 years of age) were asked to predict the Presidential preference of a liked and d i s l i k e d other. based upon balance was  The prediction  that the subjects would perceive the l i k e d other as  preferring the same candidate that they themselves  favored, while the d i s -  l i k e d other would be perceived to prefer a candidate d i f f e r e n t from the subject's own choice.  A tendency to base predictions on agreement would  be indicated i f subjects assigned their own choice to both the liked and to the d i s l i k e d other.  As shown i n Table 4, a greater percentage of subjects  i n a l l age groups responded a desire for or tendency  i n a balanced manner than i n one indicating  toward agreement.  Data i n columns 3 and 4 of  Table 4 suggest also that the tendency to base predictions on agreement  Table 4  decreases as age increases, while u t i l i z a t i o n of balance increases with increasing age, when the task involves prediction  of missing r e l a t i o n s .  (Structures represented i n columns 3 and 4 place balance and agreement i n opposition.)  23  Table 4:  Balance and agreement i n prediction of p o l i t i c a l preferences of a liked and a d i s l i k e d other (Knox and Gutman, 1968).  Predicted Choice of Most Liked Other  Grade  P  V  .0  Same F i r s t Choice as P (Agreement + Balance)  P  V  Predicted Choice of Most Disliked Other .,0  Same Last Choice as P (Disagreement + Imbalance)  P  V  * 0  Saint F i r s t Choice as P (Agreement + Imbalance)  P- - 40  Same Last Choice as P (Disagreement + Balance)  4 (n-37)  67.$%  5.4%  24.3%  54.0%  5 (n=31)  64.5%  12.9%  19.4%  32.3%  6 (n-28)  75.0%  0  14.3%  60.7%  7 (n=27)  55.6%  3.7%  14.8%  59.3%  8 (n=24)  62.5%  8.3%  4.2%  70.8%  9 (n=27)  55.6%  7.4%  7.4%  81.5%  Total = 174  24  Results  o b t a i n e d by Ohashi (1964), however, i n d i c a t e l i t t l e  between agreement and rate triadic  structures  6th g r a d e r s . sociometric  b a l a n c e when c h i l d r e n i n the o p e r a t i o n a l for pleasantness.  They r a t e d  a s e r i e s of P-O-Q  difference  age  range  S u b j e c t s i n Ohashi's study were structures, derived  from a c t u a l  r e l a t i o n s among the c h i l d r e n , f o r p l e a s a n t n e s s on a 7-point  s c a l e r a n g i n g from -3 positivity—i.e.,  to +3.  Ohashi r e p o r t s  a strong  tendency  toward  the more p o s i t i v e r e l a t i o n s c o n t a i n e d i n a s t r u c t u r e ,  h i g h e r the p l e a s a n t n e s s r a t i n g .  However, f u r t h e r i n s p e c t i o n of the  indicates  p r i m a r i l y due  that  t h i s tendency was  That i s , s t r u c t u r e s rated  considerably  There was  containing  more p l e a s a n t than those c o n t a i n i n g  disagreement.  containing  To  a similar  t h a t were b a l a n c e d were r a t e d more p l e a s a n t than those  t h a t were unbalanced.  A t t r a c t i o n , i n o t h e r words, c o n t r i b u t e d  r a t i n g s than e i t h e r agreement or b a l a n c e ; the c o n t r i b u t i o n s and  0 were  negative r e l a t i o n s .  magnitude, to r a t e s t r u c t u r e s  agreement more p l e a s a n t than those c o n t a i n i n g  data  relationship.  p o s i t i v e r e l a t i o n s between P and  a tendency, of l e s s e r  degree, s t r u c t u r e s  to the P/0  the  b a l a n c e were approximately e q u a l to one Ohashi's (1964) f i n d i n g s  of  more to  the  agreement  another.  c o n t r a s t w i t h the  r e s u l t s of most  w i t h a d u l t s which i n d i c a t e t h a t a l t h o u g h a t t r a c t i o n i s an  studies  important  determinant of p l e a s a n t n e s s r a t i n g s i t i s l e s s important than agreement o r balance.  I t i s only  i n the s t u d i e s  a t t r a c t i o n i s found to p l a y very strong  the dominant r o l e .  a t t r a c t i o n e f f e c t s may  Further studies  of Knox (1963) and  i n v e s t i g a t i n g the  be  Among c h i l d r e n , however,  the r u l e r a t h e r  than the  exception.  r e l a t i v e c o n t r i b u t i o n o f a t t r a c t i o n to  the p l e a s a n t n e s s r a t i n g s of c h i l d r e n are r e q u i r e d possibility.  Gutman (1969) t h a t  i n o r d e r to a s s e s s t h i s '  25  Purposes of the Study The results of the studies of Atwood (1969) and of Storm and Knox (1969) which use a prediction procedure to investigate the developmental course of cognitive balance suggest that balance, rather than being a primitive, non-acquired mechanism, emerges with the development of other cognitive s k i l l s such as conservation and s e r i a t i o n between 5-8 years of age.  One implication of this ontogenetic concurrence i s that the a b i l i t y  to reason about s o c i a l relations according to the balance p r i n c i p l e may  be  dependent upon the individual's general l e v e l of cognitive development. Another implication i s that balance, l i k e conservation and s e r i a t i o n , may develop i n stages, as a function of increasing chronological age.  However,  as Braine (1959) has emphasized, the apparent "age of emergence" of various types of reasoning appears to depend, at l e a s t i n part, on the s p e c i f i c experimental procedures u t i l i z e d .  The generality of these findings must  therefore remain uncertain u n t i l they have been confirmed i n d i f f e r e n t experimental settings.  One purpose of the present study then was  to extend  the investigation of the developmental course of cognitive balance beyond the confines of the prediction s i t u a t i o n i n order to determine the generality of the Atwood (1969) and the Storm and Knox (1969) findings. A second, and more specific:  purpose of the present study was  to test  Knox's (1969) contention that the observed age increase i n balance effects between the ages of 5-8 years i n the prediction s i t u a t i o n i s related to an age- or perhaps stage-related s h i f t i n the influence of a f f e c t  (i.e.,  that the s o c i a l perceptions of younger children are determined more by a f f e c t than by considerations of consistency but that with increasing age consistency takes precedence).  26  The l a t t e r purpose i n p a r t i c u l a r was i n the age range 5-12 P-O-X  served by requiring children  to rate hypothetical s o c i a l situations of the  type both for pleasantness and for consistency.  I t was  i f , as Knox suggests, a f f e c t influences the s o c i a l perceptions  assumed that of younger  children more than considerations of consistency there should be d i f f e r e n t i a t i o n between the two  little  types of ratings ( i . e . , situations defined  as pleasant would also be defined as consistent).  On the other hand, i f  there i s a decrease i n the influence of a f f e c t as a function of increasing age,  there should be greater d i f f e r e n t i a t i o n between pleasantness and  consistency among older children.  The influence of a f f e c t would also be  reflected i n the extent to which younger and older children u t i l i z e balance and agreement as a basis for pleasantness and consistency  ratings.  This  follows from the Zajonc (1968) review and i t s extension which indicate that agreement i s most important when the dependent measure relates to affect whereas balance exerts greatest influence when the task relates to psychological  consistency.  A third purpose of the study was  to determine the extent to which  attraction influences the s o c i a l perceptions the majority of previous  of children.  For, although  studies indicate that pleasantness ratings of  adult subjects are determined more by agreement or balance than by a t t r a c t i o n , there i s reason to believe on the basis of Ohashi's (1964) findings that a t t r a c t i o n may  be a more important determinant of the  child's assessments of s o c i a l situations that either agreement or balance. Hypotheses Three s p e c i f i c hypotheses were tested i n the present study. derive from the theory and research reviewed i n the preceding  They  sections.  27  1.  Within the age range 5-12 years, young children i n their ratings  of hypothetical s o c i a l situations w i l l d i f f e r e n t i a t e l i t t l e between pleasantness and consistency.  Relative to the youngest  children, older  children w i l l d i f f e r e n t i a t e more between pleasantness and consistency. Thus, i t i s predicted that as a function of increasing age, correlations between pleasantness and consistency w i l l monotonically decrease across the successive age groups i n the study. 2.  Children at a l l age levels w i l l attach greater weight to agreement  (or disagreement) between P and 0 than to the balance (or imbalance) of the t o t a l s o c i a l s i t u a t i o n when making pleasantness ratings. 3.  The r e l a t i v e importance of agreement and balance on consistency  ratings of s o c i a l situations w i l l vary with chronological age. Younger children w i l l base consistency ratings more on agreement than on balance; balance w i l l exert greater influence than agreement on consistency ratings of older children.  28  CHAPTER I I METHOD Subjects Subjects were 80 children, 20 (10 males, 10 females) from each of the following age groups:  5-6 (mean age 6.0 years), 7-8 (mean age 7.8  years), 9-10 (mean age 9.9 years) and 11-12 (mean age 12.0 years). A l l attended Holy T r i n i t y School i n North Vancouver, B. C,  Subjects i n the  5-6 group were enrolled i n kindergarten, those i n the 7-8 group were i n grade 2.  Subjects i n the 9-10 and 11-12 groups were i n grades 4 and 6  respectively. Overall Design Each c h i l d rated two sets of hypothetical P-O-X structures f o r pleasantness and for consistency. the unit r e l a t i o n s .  The sets d i f f e r e d only i n the form of  Format 1 described P and 0 as partners i n school;  in format 2 they were described as neighbor-playmates.  Testing was  conducted on an i n d i v i d u a l basis, i n two sessions of 15-35 minutes duration, separated by an i n t e r v a l of 6-11 days.  In each session, the subject  rated a set of structures (test s e r i e s ) , performed a conservation task, 3 then rated a second set of structures (retest s e r i e s ) . In order to avoid systematic order e f f e c t s , half of the subjects i n  Two p i l o t studies were conducted p r i o r to the main study. These are described i n Appendix 1. P i l o t Study II i n p a r t i c u l a r , suggested the need for an a c t i v i t y between the test and retest series i n each session. Conservation tasks were selected f o r this purpose i n order to f a c i l i t a t e possible interpretation of results along Piagetian l i n e s and to enable closer comparison of results with those of Atwood (1969). A description of performance, at each age l e v e l , on these tasks i s given i n Chapter III.  29  each age group performed p l e a s a n t n e s s r a t i n g s i n the f i r s t s e s s i o n and c o n s i s t e n c y r a t i n g s i n the second s e s s i o n . remaining s u b j e c t s .  The o r d e r was  r e v e r s e d f o r the  The o r d e r o f format p r e s e n t a t i o n and the o r d e r i n  which -conservation t a s k s ( q u a n t i t y and volume) were a d m i n i s t e r e d a l s o varied.  As  s h o w n  i n F i g u r e 3, format 1 was used f o r the t e s t s e r i e s i n  Figure 3  both s e s s i o n 1 and s e s s i o n 2 f o r h a l f o f -±xi s u b j e c t s i n each age it  s e r v e d as the r e t e s t s e r i e s f o r the o t h e r h a l f o f the s u b j e c t s .  of  the c h i l d r e n were t e s t e d  f o r c o n s e r v a t i o n o f q u a n t i t y i n the  group; Half  first  s e s s i o n and c o n s e r v a t i o n o f volume i n the second; h a l f were t e s t e d f o r c o n s e r v a t i o n o f volume i n the f i r s t in  the second  session.  A c l a s s p a r t y was means of  s e s s i o n and c o n s e r v a t i o n of q u a n t i t y  promised  (and given) to the two youngest groups as a  a r o u s i n g and s u s t a i n i n g m o t i v a t i o n .  The c h i l d r e n were t o l d  that  t h e i r c l a s s would get i t s p a r t y when 20 c h i l d r e n had earned r e d and green tickets.  They were t o l d t h a t the t i c k e t s were earned by h e l p i n g 12 w i t h  some'games."  S u b j e c t s i n the two o l d e r groups were asked to a s s i s t jS  w i t h "some v e r y important experiments." The study was  c r o s s - v a l i d a t e d on a d i f f e r e n t sample, by a second  female E_ i n o r d e r to ensure g e n e r a l i z a b i l i t y o f the f i n d i n g s beyond the s p e c i f i c p o p u l a t i o n s a m p l e d and to reduce the p o s s i b i l i t y were due t o experimenter b i a s . at  that  the r e s u l t s  The c r o s s - v a l i d a t i o n study was  conducted  the same time as the p r i n c i p a l study (June, 1969), u s i n g 32  children,  8 (4 males and 4 females) from each of the f o l l o w i n g age groups:  5-6  30  I. n=5 i n each age group  Pleasantness  Consistency  Test—^-Conservation—•Retest Task  Test  Conservation—>-Retest Task 4-  Fl  II. n=5  Quantity  4  -  4  Test 4  -  4-  Volume  Fl  -jF2  ^Conservation Task -  4  F2  'Conservation Task  >-Retest  4-  Quantity  4-  Fl  Pleasantness *-Retest  -  4-  Volume  F2  Test 4  ^Conservation—•Retest Task -  4  Fl  -  Quantity  4-  F2  Consistency  Pleasantness  Test  Test—^Conservation—>Retest Task  4  F2  Figure 3:  Test  Consistency  Fl  IV. n=5  Hletest  Test—^-Conservation Task  4-  Volume  Fl  Consistency  Pleasantness  F2 III. n=5  F2  ^-Conservation—>-Retest Task -  4  -  Quantity  4-  Fl  4  F2  -  4  -  4-  Volume  Fl  Diagrammatic representation of the experimental design.  31  (mean age 6.0 years), 7-8 (mean age 7.7 years), 9-10 (mean age 10.0 years), and 11-12 (mean age 12.0 years).  A l l attended Dr. R. E. McKechnie School,  a public school of moderate s i z e , i n Vancouver, B. C.  Subjects i n the  5-6 group were enrolled i n kindergarten. Subjects i n the 7-8, 9-10, and 11-12 groups were i n grades 2, A, and 6 respectively.  was b l i n d  as to JEj/s s p e c i f i c purposes and hypotheses. Procedure Two sets of instructions were developed—one  suitable i n language and  content for use with younger children ( i . e . , those i n the 5-6 and 7-8 age groups) and another s e t , i d e n t i c a l i n purpose, but more suited to the interests and a b i l i t i e s of older children ( i . e . , those i n the 9-10 and 11-12 age groups). Procedure for testing younger children (groups 5-6 and 7-8). (a)  Training Session I began with a training period designed to ensure understanding  of the rating procedure.  Six size-graded blocks were placed i n disarray  on the table i n front of the subject. _E said: are a l l d i f f e r e n t s i z e s . carefully how I do i t . "  "See these b l o c k s — t h e y  I'm going to make a stairway out of them.  Watch  E then constructed a stairway (starting with the  smallest block), making sure that the c h i l d watched the procedure.  E  then mixed the blocks and said—"Now, I want you to make a stairway j u s t l i k e the one I made." repeated the demonstration.  If. the c h i l d i n i t i a l l y f a i l e d the task, E E was prepared to discontinue testing i f the  c h i l d f a i l e d to seriate correctly following the second demonstration. There were, however, no cases of f a i l u r e following the second demonstration. E was therefore able to; proceed with a l l subjects to rating scale I which  32  consisted of a 26" x 11%" sheet of white shelf paper on which were drawn six  points, three inches apart.  numbered from 1 through 6.  Intervals delineated by scale points were  As shown i n Figure 4, squares of increasing 4  s i z e p a r a l l e l e d the increase i n scale numbers. as follows:  "I'm  Js introduced the scale  going to build another stairway but this time I'm  to put my blocks on this l i n e .  Figure 4:  going  See the l i t t l e wee box at this end of the  Rating Scale I  l i n e (low end of the s c a l e ) — w e l l , I'm going to put the l i t t l e s t right here on top of the l i t t l e s t  box.  block  I'm going to put this block,  which i s a l i t t l e b i t bigger, i n the next box i n the l i n e . . .I'm  going  to put this block, the biggest block, here at this end of the l i n e i n the biggest box.  See, we have a stairway again."  E then scrambled  blocks and asked the subject to make a stairway on the l i n e .  the  Errors were  The rating scales and training procedures employed i n this study were s p e c i a l l y designed to meet the needs of young, p r e - l i t e r a t e children. They were inspired to some extent by the work of Walster, Berscheid, and Barclay (1967) and Elkind (1964). Walster, Berscheid, and Barclay (1967) investigated post-choiee dissonance reduction i n nursery-school children. They had subjects rate toys, before and a f t e r choosing one, on a scale which consisted of f i v e squares which increased i n s i z e from one square centimeter to f i v e square centimeters. Elkind (1964) used a "stairway" procedure to investigate s e r i a t i o n i n children aged 4-6 years.  33  drawn to the subject's attention and he was scale I was items:  asked to correct them.  Rating  then removed and the child was presented with the following  a p e n c i l , a small p l a s t i c bracelet, a harmonica, a viewmaster, and  a scotty dog. handle i t .  jE drew attention to each item and allowed the c h i l d to  Rating scale II was  then presented.  of nine points, f i v e inches apart.  Scale intervals delineated by these  points were numbered from 1 through 9. the increase i n scale numbers.  Rating scale II consisted  Squares of increasing s i z e p a r a l l e l e d  As shown i n Figure 5, rating scale II was  anchored at the high end by a large red heart and at the low end by a large blue cross.  E said:  "See this l i n e , i t ' s just l i k e the one we made the  Figure 5:  Rating Scale I I  stairway on before, only i t ' s a l i t t l e b i t bigger.  What I want you to do  this time i s to put the toys on the l i n e so the toys make a stairway. The toy you l i k e the very, very b e s t — t h e one you would l i k e to play with most of a l l — s h o u l d go at this end of the l i n e , i n the big box above the red valentine. much.  Now,  The red valentine means that you l i k e the toy very, very  where would you put the toy you l i k e the very, very best?  . . . very good.  Now,  see this blue cross at this end of the line?  blue cross means that you don't l i k e the toy.  The  The toy that you r e a l l y  don't l i k e very much and wouldn't want to play with at a l l should go at  34  this end of the l i n e i n the l i t t l e box by the blue cross. you put the toy you d i s l i k e the most? . . . very good.  Now,  where would  What I want you to  do i s to s t a r t with the toy you l i k e the very, very best and put i t at this end of the l i n e — b y the red valentine. best arid put i t here.  Then take the toy you l i k e next  The l a s t toy, the one you put i n the l i t t l e box by  the blue cross, should be the one you l i k e the l e a s t . two toys the same you can put them i n the same box. toy  game.  Now  I f you l i k e  l e t ' s play the  Show me the toy you l i k e the best—now, where w i l l you put i t ?  When the subject finished, E removed the toys and rating scale and  etc." stated:  "Before we play any more stairway games I'm going to show you  some cards.  Each card t e l l s a story, but the story i s told i n a secret code.  Would you l i k e to learn the secret code?" 1-9.  O.K.?  IS then presented training cards  These cards, described i n Table 5, were designed to teach the c h i l d  Table 5  to  associate a red heart with " l i k e s , " a blue cross with " d i s l i k e s , " and  a box with "something very, very important."  The subject was  then t o l d  that he would be shown some cards containing stories written i n the "secret code."  He was  instructed to examine each card and to t e l l _E the  "story" on i t .  Cards 10-17  (series A) were then presented i n succession.  Any errors or omissions i n " t e l l i n g the story" were drawn to the subject's attention as they occurred. The structures represented i n series A (and i n a l l subsequent series) are shown i n Figure 6.  Structures 1, 2, 7, and 8 are balanced; structures  3, 4, 5, and 6 are imbalanced; a l l relations between persons are r e c i p r o c a l .  35  Table  Card  No.  5:  T r a i n i n g cards  Content o f Card  1-9  _E's  1  explanation  A r e d v a l e n t i n e means t h a t somebody l i k e s s o m e t h i n g . A b l u e c r o s s means t h a t somebody d o e s n ' t l i k e s o m e t h i n g . A b o x means s o m e t h i n g very important.  o  1  /  very,  This c h i l d l i k e s the important thing. See, there i s a r e d valentine.  —->  This c h i l d doesn't l i k e the very important thing. See, there i s a blue cross.  V 7  This c h i l d l i k e s the important thing. See, there i s a r e d valentine.  O  A o  r  X  This c h i l d doesn't l i k e the important thing. See, there is a blue cross. T h e s e two c h i l d r e n l i k e e a c h other. This c h i l d l i k e s this one a n d t h i s c h i l d l i k e s t h i s one. See, there a r e r e d v a l e n t i n e s here and here. T h e s e two c h i l d r e n d o n ' t l i k e each o t h e r . This arrow means t h a t h e d o e s n ' t l i k e him because there i s a blue c r o s s h e r e . T h i s a r r o w means that t h i s c h i l d doesn't l i k e t h i s one b e c a u s e t h e r e i s a blue cross here.  36  Each structure was presented i n p i c t o r i a l form on a separate 3" x 5" card. As shown i n the example i n Figure 6 s t i c k figures were used to indicate persons "P" and "0" and a box was used to symbolize "X," "something very, very important to both children."  The a f f e c t i v e r e l a t i o n s between elements  were indicated by red hearts (likes) and blue crosses ( d i s l i k e s ) .  The  d i r e c t i o n of relations between elements were indicated by black arrows. Separate sets of cards i l l u s t r a t i n g the structures were prepared f o r males and females.  Figure 6:  Structural content of training cards 10-17  Upon completion of series A, card 18 was presented. the s t i c k figure i n the "P" position was c i r c l e d . that you are one of the boys ( g i r l s ) i n the story. with the c i r c l e around him (her).  On this card,  E stated, "Let's pretend You are this boy ( g i r l )  Now, I want you to look at some stories  and whenever you see a boy ( g i r l ) with a c i r c l e around him (her) pretend that boy ( g i r l ) i s you." succession.  Cards 19-26 (series B) were then presented i n  The structures represented i n series B were the same as those  37  i n series A but the presentation sequence was varied and figures i n the "P" position were c i r c l e d .  The subject's task was  to describe the structure  on each card, placing himself i n the "P" p o s i t i o n . (°)  Pleasantness ratings:  test and retest  Pleasantness ratings were made on scale I I I .  Scale I I I contained  f i f t e e n points, f i v e inches apart, anchored at the high end by a smiling face and at the low end by a frowning face (see Figure 7).  Squares of  increasing s i z e again p a r a l l e l e d the increase i n number value of scale intervals.  O  E_ introduced the scale as follows:  I Cl  I I—I  2  3  I T  "Remember when we played  • 4  5  6  Figure 7:  7  8  9  10 11  12  13  14  15  Rating Scale III  the stairway game before, I showed you some toys and you told me which one you l i k e d the very, very best and we put i t i n the biggest box on the line? Then you showed me the toy you liked next best, and we put i t i n the next biggest box.  What I want you to do now,  i s to look at the stories on  these cards and put them i n the boxes on this l i n e .  Put the nicest story,  the one that makes you f e e l very, very happy, i n the biggest box at this end of the l i n e by the picture of the happy, smiling face.  The one that  you l i k e next best, that doesn't make you f e e l quite as happy, should go i n one of these boxes . . . the story that makes you f e e l very, very unhappy should go at this end of the l i n e by the picture of the sad,  38  unhappy face.  O.K.?  When you are a l l finished today y o u ' l l get your red  (or green) ticket for the class party." IS placed cards 27-34 (test series) i n a semi-circle i n front of the c h i l d and ni.id°.  "What I want you to do i s pretend that this c h i l d — t h e one  with the c i r c l e around him (her) i s you and this other c h i l d i s your partner (Format 1).  Do you know what a partner i s ?  Pretend that at the  beginning of the year the teacher picked two c h i l d r e n — y o u boy  and this other  ( g i r l ) i n the card and said that you two were to be partners for the  whole year.  You are to help each other with your school work, s i t together,  and work on s p e c i a l projects together.  Look at a l l the s t o r i e s and ask  youself how happy you would f e e l i f (pointing to structure 5) you l i k e d your partner and he (she) liked you, and you didn't l i k e the box but he (she) liked i t . " '  Or how happy would you f e e l (pointing to structure 4) i f you  didn't l i k e your partner and he (she) didn't l i k e you and you both didn't l i k e what was i n the box.  Put the n i c e s t , happiest story at this end of the  l i n e by the happy, smiling face, then the next nicest story, and the next, u n t i l you get down to the most unpleasant, unhappy story." minimize possible experimenter bias e f f e c t s ,  In order to  went to another part of the  room, with her back to the subject while ratings were being made. The subject was instructed to t e l l IS when he had finished his ratings. When the card numbers and scale positions had been recorded, rating scale III was removed. then conducted.  A test of conservation  (quantity or volume) was  The procedure followed i n administering  these tests was  ^The structures are presented i n diagrammatic form i n the foldout i n Appendix 2. The reader may f i n d i t h e l p f u l to r e f e r to these diagiams throughout the remainder of the paper.  39  as follows: Conservation of quantity. i n height, 2V He was water.  1  The subject was  shown two glasses (4"  i n diameter) f i l l e d to the same l e v e l with colored water.  asked whether or not the two glasses contained the same amount of The water from one glass was  glass (6" i n height, 2 3/4"  then poured into a t a l l e r and wider  i n diameter).  The subject was  asked whether  the two glasses contained more, less, or the same amount of water.  He  was  asked to explain his answer. Conservation of volume. was  In the conservation of volume task, the subject  shown a glass (6" i n height, 2 3/4"  water.  A round piece of P l a s t i c i n e was  l e v e l was marked with a rubber band.  i n diameter) two-thirds f i l l e d with placed i n the glass.  The P l a s t i c i n e was  the water, dried, and r o l l e d into a "sausage."  The new water  then removed from  The subject was  asked  to predict whether the water l e v e l would r i s e more, l e s s , or the same amount i f the "sausage" were placed i n the water.  He was  asked to explain  his answer. Rating scale III was  then placed once more before the c h i l d and  retest series (cards 35-46) was  presented.  12 stated:  "We  are going to  play one more l i n e game before you get your red (or green) t i c k e t . that this i s you (E indicated the figure i n the "P" position) and other c h i l d i s someone who Think how  the  Pretend this  l i v e s near you that you play with (Format 2 ) .  pleasant or happy you would f e e l i f you l i k e d this c h i l d and he  (she) l i k e d you and you both didn't l i k e what was  i n the box  (structure 2 ) .  Think how happy you would f e e l i f you didn't l i k e this c h i l d and he didn't l i k e you and you didn't l i k e what was did (structure 8).  How would you f e e l then?  (she)  i n the box, but your playmate Put the n i c e s t , happiest  40  story here i n the biggest box by the happy, smiling face.  Put the saddest,  most unhappy story i n the l i t t l e s t box by the sad, unhappy face. are a l l finished you w i l l get your t i c k e t . "  When the retest series was  completed, 12 probed b r i e f l y f o r a c l a s s i f i c a t i o n r u l e . "How did you decide where to put the cards?  When you  The c h i l d was asked:  Why did you put these s t o r i e s  here (IS pointed to the low end of the scale) and these stories over here (high end of the scale)?"  Subjects were then given their t i c k e t s , thanked,  and dismissed with the plea not to t e l l their classmates about the task "because i t wouldn't be f a i r . "  Subjects were given a red ticket at the  close of session I and a green t i c k e t at the end of session I I . Session II was conducted  6-11 days after session I.  I f pleasantness  ratings were obtained i n session I, consistency ratings were obtained i n session II and vice versa.  IS introduced session I I as follows:  "Remember  l a s t week, when we played some stairway games and we put toys and s t o r i e s on a l i n e — w e l l , we are going to play some more stairway games today. also going to play some more games with glasses of water today. a l l finished you w i l l get your green ticket and on are going to have the class party.  We are  When we are  (day of week) we  I t should be l o t s of fun . . . Before we  s t a r t the stairway games though, I want to see i f you remember the secret code. Can you t e l l me what i t says on this card?" succession. (c)  Cards 19-26 were then presented i n  IS corrected any mistakes the c h i l d made i n " t e l l i n g the s t o r i e s . "  Consistency ratings:  test and retest  Consistency ratings were made on scale IV.  Like scale I I I , i t  contained f i f t e e n points, f i v e inches apart; squares of increasing size again p a r a l l e l e d the increase i n scale numbers. was anchored at the high end by a large  r  e  d  » " S  a  As shown i n Figure 8, i t n  d  a t  t  h  e  l  o  w  e  n  d  b  y  a  41  red "S" with a black cross over i t .  Subjects were introduced to the task  of rating structures for consistency i n the following manner:  E stated:  i ° i ° i ° i i ° I I I P P I U D I M -4M u  1  2  3  4  5  IJ  U  6  7  Figure 8:  U  8  9  10 11  12  13  14 15  c  O  Rating Scale IV  "Iwantyou to read the s t o r i e s on the cards and decide how much sense they make.  Stories that seem to make sense should go i n the big boxes at the  end of the l i n e by the big red "S." Stories that seem to be "mixed-up" or s i l l y , stories that don't make sense should go i n the l i t t l e boxes at  the end of the l i n e by the S with the black cross over i t .  Now, suppose  I told you that when I was out yesterday I saw a l i t t l e tiny cat eat a great big dog.  Where would you put a story l i k e that?  Right, you would  put i t i n one of the wee boxes by the S with the cross over i t , because i t doesn't make s e n s e — l i t t l e wee cats can't eat big dogs! you that 3 plus 3 i s 8.  Suppose I told  Where would you put a story l i k e that?  a l i t t l e wee box because that doesn't make sense either.  Right—in  3 plus 3 i s 6.  We make 8 by adding 4 plus 4 or 5 plus 3 or 6 plus 2 or 7 plus 1. garten: sense?  Suppose I told you there wera 3 blocks here.  Does that make  No—because there are 6 blocks h e r e — s e e 1, 2, 3 . . . 6.  said didn't make sense.)  (Kinder-  What I  Now, what i f I told you that Peter l i k e s Joe  the very best i n the whole world—but, I know that Joe i s always mean to Peter.  Joe h i t s Peter and pushes him down and kicks him every time  they play together.  Does i t make sense for Peter to l i k e Joe best i n the  42  whole world when Joe i s always mean?  Where would you put this story?  What  i f I told you my best friend was always nice to me—where would you put a story l i k e that?  Well, what I want you to do i s to read the s t o r i e s on  these cards and decide whether or not they make sense and then put them on the l i n e .  I f the story makes l o t s of sense, where would you put i t ? I f  the story seems very s i l l y or mixed up or wrong, where would you put i t ? . . . The s t o r i e s i n the cards are l i k e the ones you saw b e f o r e — t h e y are about the two children who are partners i n school (Format 1). See i n this story (structure 5) you l i k e your partner and he (she) l i k e s you. don't l i k e what's i n the box but he (she) does.  You  Does that make sense?  Or  what about this one (structure 4) . . . you don't l i k e your partner and he (she) doesn't l i k e you.  He (she) doesn't l i k e what's i n the box. You  don't l i k e what's i n the box." When the test items had been rated, scale IV was removed and a conservation task was admiristered. series i n the alternate format was then presented.  A second (retest)  When the r e t e s t series  was completed, 12 probed b r i e f l y for a c l a s s i f i c a t i o n r u l e .  The c h i l d was  asked "How did you decide where to put the cards? . . . Why did you put these stories here (low end of the scale) and these stories over here (high end)?"  The subject was then thanked, given h i s t i c k e t , and dismissed.  Procedure for testing older children (groups 9-10 and 11-12) With the older children, the "secret code" guize was dropped and training cards 10-17 were omitted.  Instead of beginning session I with  practice i n building stairways, E began the session as follows: to show you some cards today.  "I'm going  Each card t e l l s a s t o r y — t h e story i s about  two children and something that i s very, very important to both of them. In some of the s t o r i e s , the two children l i k e each other, i n other s t o r i e s  43  they d i s l i k e each other.  Sometimes both children l i k e or both d i s l i k e the  important thing; i n other stories one c h i l d l i k e s the important thing but the other c h i l d doesn't l i k e i t .  I'm showing these cards to children from  kindergarten right up to grade 6.  Now l i t t l e kids can't read, so I had  to make the stories i n pictures.  Whenever you see a red heart, i t means  . . . ." The introduction to session II also differed f o r older children. E_ stated:  "Remember l a s t week when we did some experiments with c a r d s — w e l l ,  we are going to do some more today. ment with the glasses of water.  We are also going to do another experi-  But, before we s t a r t , I want to make sure  that you remember how to read the cards. these cards? . . . "  Can you t e l l me what i t says on  The remainder of the procedure was i d e n t i c a l to that  used with younger children except that no mention was made of a class party.  44  CHAPTER I I I ASSESSMENT OF THE PROCEDURES Qualitative Observations A l l children i n the 5-6 and 7-8 groups correctly seriated s i x blocks, both o f f and on the " l i n e " ( i . e . , scale I ) .  There were few  occasions when E needed to correct errors or repeat the i n i t i a l demonstration of "stairway" b u i l d i n g .  Subjects indicated, verbally and by placement of  toys on scale I I , that they understood that their task was to arrange things on " l i n e s " i n order of increasing magnitude. Both verbal comments during  training and scale placement of examples  made i t apparent that children at a l l age levels understood that pleasantness ratings were to be based on "how happy I would f e e l i f I were r e a l l y i n the s i t u a t i o n . " consistency  There were s i m i l a r indications that they understood that  ratings were to be made on the basis of how much "sense" the  situations made. It was E/s impression that motivation was high i n a l l age groups, both on the basis of observation  of the children's behaviour during the experiment  and from conversations with the children and their teachers at the conclusion of the study.  An additional indication of high motivation was the willingness  of the children to p a r t i c i p a t e i n the second session. R e l i a b i l i t y of the Ratings Two r e l i a b i l i t y c o e f f i c i e n t s were computed f o r each subject: for pleasantness and one f o r consistency. computed by correlating the subject's  one  Coefficients of each type were  ratings under formats 1 and 2.  Z transformed i n t r a - i n d i v i d u a l c o e f f i c i e n t s were then combined to y i e l d average r e l i a b i l i t y c o e f f i c i e n t s f o r each age group. ^  — — — — — —  These are shown i n  -  Fisher transformations were used i n a l l cases where correlations were averaged or tested for s i g n i f i c a n c e .  45  Table 6.  At a l l age levels r e l i a b i l i t y c o e f f i c i e n t s for pleasantness and  Table 6  and consistency are of s u f f i c i e n t magnitude to warrant the conclusion that the processes underlying the ratings were stable across  formats.  Performance on the Interpolated Conservation Tests Tests of conservation were selected for use as the interpolated a c t i v i t y between ratings made under formats  1 and 2 i n the hope that the  results of these tests might be of aid i n interpreting the ratings and to f a c i l i t a t e comparison of r e s u l t s with those of Atwood (1969).  However, as  shown i n Table 7, although the majority of subjects i n the 5-6 group performed i n a manner consistent with Piagetian theory and research'  Table 7  ( i . e . , 85% f a i l e d to conserve quantity or volume), the performance of In Piagetian theory, a number of cognitive changes are hypothesized to take place at the end of the pre-operational stage, the most important of which i s manifested i n the a c q u i s i t i o n of the "schema of conservation." When a c h i l d i s able to conserve, he r e a l i z e s that certain properties of an object (e.g., quantity, weight) remain constant i n the face of c e r t a i n transformations (e.g., changes i n the object's shape). His thinking i s no longer dominated by physical appearance. However, although an understanding of conservation marks the t r a n s i t i o n from pre-operational to operational functioning, the various types of conservation described by Piaget and his co-workers (Piaget, 1946; Piaget, 1952; Piaget and Inhelder, 1941; Piaget, Inhelder, and Szeminska, 1960) do not develop a l l i n a piece. For example, although conservation of quantity i s generally acquired by age 7-8, conservation of weight does not appear u n t i l 9-10, and conservation of volume i s seldom apparent before 11-12 years. Atwood (1969) and others have used the d i f f e r e n t i a l growth rate of quantity and volume conservation to further subdivide subjects into the Piagetian stages of concrete operational and formal operational functioning.  46  Table 6:  Mean r e l i a b i l i t y  coeffi- ients (Pearson r) for pleasantness  and consistency ratings  Pleasantness  Consistency  5-6  .84  .87  7-8  .93  .92  9-10  .97  .89  11-12  .91  .89  47  Table 7:  % of subjects In each age group showing quantity or volume conservation ( p r i n c i p a l study n^O  i n each age group; cross-  v a l i d a t i o n s-tudy n=8)  Quantity conservation Cross-validation Study  Volume conservation  Age Gp.  Principal Study  Principal Study  Cross-validation Study  5-6 Failure Success  85.0% 15.0%  75.0% 25.0%  85.0% 15.0%  87.5% 12.5%  7-8 Failure Success  15.0% 85.0%  25.0% 75.0%  40.0% 60.0%  62.5% 37.5%  9-10 Failure Success  0 100  %  0 100  %  25.0% 75.0%  12.5% 87.5%  11-12 Failure Success  0 100  %  0 100  %  35.0% 65.0%  12.5% 87.5%  48  children i n the three older groups was  somewhat anomalous.  For example,  i n the p r i n c i p a l study the proportion of volume conservers i n the 7-8 9-10  groups was  9-10  group) for a cognitive a b i l i t y not supposed to be r e l i a b l y  before 11-12  inordinately high (60% i n the 7-8 group and 75% i n the  years of age ( c f . , Piaget and Inhelder, 1941).  children i n the 11-12 task than 9-10  and  found  Furthermore,  group performed no better on the volume conservation  year old children and l i t t l e better than those of 7-8 years.  Thus, i n view of the a t y p i c a l l i t y of the conservation results and the v i r t u a l l y indistinguishable performance of the three older groups, i t d i d not seem f e a s i b l e to c l a s s i f y subjects by conservation l e v e l nor to make any further attempt to r e l a t e these data to the results of the rating task.  49  CHAPTER IV RESULTS AND DISCUSSION OF HYPOTHESES Hypothesis 1 Hypothesis 1 was concerned with the relationship between perceptions of pleasantness and perceptions of consistency i n subjects of varying chronological  age.  Younger children, hypothesized to d i f f e r e n t i a t e  l i t t l e between the pleasantness and the consistency of s o c i a l s i t u a t i o n s , were expected to y i e l d high correlations between pleasantness and consistency ratings.  Older children, hypothesized to d i f f e r e n t i a t e somewhat  more between pleasantness and consistency i n s o c i a l situations were expected to y i e l d lower correlations between the two types of ratings. Separate c o r r e l a t i o n c o e f f i c i e n t s were computed for each subject by correlating h i s responses to the same situations when rating f o r pleasantness and f o r consistency.  I n t r a - i n d i v i d u a l c o r r e l a t i o n c o e f f i c i e n t s were  then averaged f o r each age group. Table 8.  These average correlations are shown i n  Inspection of the table indicates a rather high c o r r e l a t i o n  Table 8  between pleasantness and consistency i n a l l age groups.  A one-way analysis  of variance by age groups of the transformed values of the i n t r a - i n d i v i d u a l correlations indicated no s i g n i f i c a n t between-group effects (F<1).  The  r e s u l t s , indicating a lack of clear d i f f e r e n t i a t i o n between pleasantness and consistency at a l l age l e v e l s , thus f a i l to confirm the hypothesis that d i f f e r e n t i a t i o n would increase between the ages of 5-12 years. 8 These high correlations, while necessary to an inference that children  50  Table 8:  Mean pleasantness-consistency correlations for each age group (Pearson r)  Age Group  r , p (  5-6  .72  7-8  .84  9-40  .79  11-12  .77  51  The absence of evidence of d i f f e r e n t i a t i o n between pleasantness consistency, although predicted i n the case of younger children, was expected  i n the case of older children.  children indicated that they understood  un-  In the t r a i n i n g session, these the meaning of consistency.  predictions tasks (e.g., Atwood, 1969; Knox and Gutman, 1968; Knox, 1969)  and  In  Storm and  children above the age of 7 show a strong tendency to perform  i n accordance with consistency p r i n c i p l e s although younger children appear to respond on a more a f f e c t i v e b a s i s .  Further, there i s evidence that  adults distinguish between pleasantness and consistency when rating hypothetical P-O-X  situations (e.g., Knox, 1963;  Gutman, 1969).  Yet,  the present r e s u l t s indicate that children between the ages of 7-12 do not.  The question that remains i s why?  years  A possible explanation,  r e l a t i n g to the o v e r a l l complexity of the rating task and to the strength of the balance "schema" i n comparison to alternative cognitive biases i s offered i n Chapter VII. Hypotheses 2 and 3 Hypotheses 2 and 3 concerned the r e l a t i v e importance of agreement and balance for pleasantness and consistency ratings.  It was  predicted that  children at a l l age levels would attach greater weight to agreement than to balance when making pleasantness ratings (hypothesis 2) and that younger do not d i f f e r e n t i a t e c l e a r l y between the pleasantness and consistency of t r i a d i c structures are not alone s u f f i c i e n t for this inference. I t i s possible, i n other words, that the children were d i f f e r e n t i a t i n g on some basis masked by the correlations. For example, i t i s conceivable that the children could have u t i l i z e d d i f f e r e n t portions of the raxing scale when assessing pleasantness as compared to consistency. Consequently, each child's d i s t r i b u t i o n of ratings on these two c r i t e r i a was examined. The d i s t r i b u t i o n a l c h a r a c t e r i s t i c s of the pleasantness and consistency ratings were found to be v i r t u a l l y i d e n t i c a l for 90% of the children.  52  children would also base consistency ratings more on agreement than on balance (hypothesis 3).  Balance was expected to exert greater influence  than agreement on consistency ratings of older children (hypothesis 3). The unanticipated high c o r r e l a t i o n between pleasantness and consistency ratings at a l l age levels precluded j o i n t confirmation of these two hypotheses.  Conceivably, however, the data might s t i l l have supported  one or the other h y p o t h e s i s — a p o s s i b i l i t y that warranted cal t e s t s .  further s t a t i s t i -  The s t a t i s t i c a l procedures employed for these tests would also  provide information concerning the r e l a t i v e contribution of agreement, balance, and other s t r u c t u r a l determinants sets of ratings. A between-within  (e.g., attraction) to the two  Thus, analyses of variance were performed on the ratings. design (Winer, 1962, p.320) was used.  Assessed within subjects  were the agreement factor (A) with two l e v e l s , the a t t r a c t i o n factor (B) with two l e v e l s , the format factor (C) with two levels, and a structure factor (0) with two levels that r e f l e c t e d high or low p o s i t i v i t y .  The  age factor (D) with four l e v e l s , was  a between-group factor.  (The o v e r a l l  design i s summarized i n Table 9.)  Zajonc indices were also computed i n  Table 9  order to f a c i l i t a t e comparison of results with those obtained i n previous studies.  Multiple regression equations were f i t t e d to the data of each  subject i n order to examine the r e l a t i v e effects of agreement, balance, and a t t r a c t i o n at the i n d i v i d u a l l e v e l , (a)  Results relevant to hypothesis 2 The results of the analysis of variance of pleasantness ratings are  53  Table 9:  Design of the analyses of variance  Designation  Levels  Factor Label  Structure No.  A  Agreement  1) Agreement 2) Disagreement  1, 2, 3, 4 5, 6, 7, 8  B  Attraction  1) Positive 2) Negative  1, 2, 5, 6 3, 4, 7, 8  C  Formats  1) Partners 2) Neighbor-playmates  a l l structures II tt  D  Age  1) 2) 3) 4)  0  Structures  5-6 years 7-8 years 9-10 years 11-12 years  1) high p o s i t i v i t y 2) low p o s i t i v i t y  a l l structures ti II n  ii  II  ii  1, 3, 5, 8 2, 4, 6, 7  54  summarized i n T a b l e 10.  As shown i n t h i s t a b l e , t h e main - e f f e c t due to  agreement was s i g n i f i c a n t ( F = 7 . 4 2 , d f - 1 , 7 6 ; P < . 0 1 ) .  The p r e s e n c e o f a  s i g n i f i c a n t i n t e r a c t i o n between agreement and a t t r a c t i o n  (F=33.46;  df=l,76;  9 P<.01) i n d i c a t e s t h a t the e f f e c t s o f b a l a n c e were s i g n i f i c a n t  also.  T a b l e 10  Examination of the percent of t o t a l v a r i a b i l i t y  ( i . e . , percent of  total  sum o f s q u a r e s ) accounted f o r by t h e s e two v a r i a b l e s i n d i c a t e s , however, t h a t b a l a n c e e f f e c t s were s l i g h t l y g r e a t e r than t h o s e o f agreement reverse of that p r e d i c t e d i n hypothesis 2 ) .  The p e r c e n t o f  (the  variability  accounted f o r by t h e s e v a r i a b l e s w a s , n e v e r t h e l e s s , e x t r e m e l y s m a l l . ( B a l a n c e accounted f o r o n l y 1.36% of the t o t a l v a r i a b i l i t y ; attributable  t o agreement was .29%.)  S u b j e c t s i n a l l age groups appear  to have based t h e i r r a t i n g s p r i m a r i l y on a t t r a c t i o n . a t t r a c t i o n was 410.09 ( d f = l , 7 6 ; P < . 0 1 ) . o f the t o t a l  variability.*  the amount  The F r a t i o  for  T h i s f a c t o r accounted f o r 45.51%  0  The r e l a t i v e magnitude o f Z a j o n c i n d i c e s f o r b a l a n c e , agreement, and a t t r a c t i o n were c o n s i s t e n t w i t h the f i n d i n g s of the a n a l y s i s o f v a r i a n c e . * B a l a n c e i n v o l v e s the i n t e r a c t i o n o f the l i k i n g r e l a t i o n between P and 0 and the agreement between them c o n c e r n i n g X . B a l a n c e d s t a t e s a r e those i n which P and 0 l i k e one a n o t h e r and a g r e e , o r a l t e r n a t i v e l y , where they d i s l i k e one a n o t h e r and d i s a g r e e . Imbalanced s t a t e s combine l i k i n g w i t h d i s a g r e e m e n t , o r d i s l i k i n g w i t h agreement. ^ I n s p e c t i o n o f T a b l e 10 i n d i c a t e s t h a t e f f e c t s r e l a t i n g to f a c t o r 0 a l s o account f o r a g r e a t e r p r o p o r t i o n of t h e v a r i a b i l i t y t h a n agreement o r balance. ( R e s u l t s and d i s c u s s i o n of f a c t o r 0 a r e p r e s e n t e d I n Chapter V t o g e t h e r w i t h e f f e c t s r e l a t i n g to format d i f f e r e n c e s . ) **Mean p l e a s a n t n e s s and c o n s i s t e n c y r a t i n g s used i n t h e c o m p u t a t i o n o f Z a j o n c i n d i c e s a r e p r e s e n t e d i n Appendix 3 .  55  Table 10:  Summary of analysis of variance of pleasantness ratings  Source of Variation  df  Between Subjects  79  D (Ag e gps) Subj . i n gps  3 76  Within Subjects  MS  F  128.69 39.19  3.28*  % of v a r i a b i l i t y due to significant main effects and i n t e r actions 1.27  1200  A (Agreement) DA A x Subj. i n gps  1 3 76  88.73 40.15 11.95  7.42** 3.36*  .29 .40  B (Attraction) DB B x Subj. i n gps  . 1 3 76  13867.00 21.09 33.81  410.09** .62  45.51  C (Format) DC C x Subj. i n gps  1 3 76  3.94 7.36 3.84  0 (Structure) DO 0 x Subj. i n gps  1 3 76  1123.10 14.52 7.19  156.23** 2.02  3.69  AB (Balance) DAB AB x Subj. i n gps  1 3 76  415.19 9.04 12.41  33.46** .73  1.36  AC DAC AC x Subj. i n gps  1 3 76  21.27 18.78 5.46  AO DAO AO x Subj. i n gps  1 3 76  1590.90 20.15 7.70  BC DBC BC x Subj. i n gps  1 3 76  17.81 18.00 6.24  BO DBO BO x Subj. i n gps  1 3 76  .49 2.96 6.09  .08 .49  CO DCO CO x Subj. i n gps  1 3 76  7.97 2.67 4.80  1.66 .56  1.02 1.91  3.89* 3.44* 206.60** 2.62 2.85 2.89*  .07 .18 5.22  .18  56  Source of Variation  % of v a r i a b i l i t y due to significant main effects and i n t e r actions  df  MS  ABC DABC ABC x Subj. i n gps  1 3 76  31.56 3.06 4.36  7.24** .70  ABO DABC ABO x Subj. i n 8PS  1 3 76  1.19 16.50 5.26  .23 3.14*  Residual Total *p<.05 **p<.01  240 1279  4.68  .11  .16  57  As shown i n Table 11, indices f o r balance were of s l i g h t l y greater magnitude than those f o r agreement, i n a l l groups.  Indices f o r agreement  and balance were small i n comparison to those for a t t r a c t i o n .  The index  Table 11  value of largest magnitude at a l l age levels was that f o r a t t r a c t i o n . The same pattern of results emerged when pleasantness ratings were subjected to multiple regression analysis (see Appendix 4). Average beta c o e f f i c i e n t s f o r balance were greater than those f o r agreement i n a l l groups, and average beta c o e f f i c i e n t s for a t t r a c t i o n were greater than those for balance.  A l l three methods of analysis indicate, i n other words,  that pleasantness ratings were based primarily on the P/0 r e l a t i o n s h i p . The prediction that agreement would exert stronger influence on pleasantness ratings than balance was based on t h e o r e t i c a l considerations outlined i n the Introduction. As previously mentioned, however, strong attraction effects i n pleasantness ratings were obtained by Ohashi (1964) in children, and by Knox (1963) and Gutman (1969) i n adults. results are thus not without precedent.  The present  The results of the present study  (and those of Ohashi, 1964) d i f f e r , however, from those obtained by Knox (1963) and by Gutman (1969) i n the absolute magnitude of a t t r a c t i o n and agreement e f f e c t s .  In Gutman's study, for example, a t t r a c t i o n  accounted 12  for 33.28% of the v a r i a b i l i t y ; agreement accounted f o r 18.41%.  Summaries of the analyses of variance performed are presented i n Appendix 5.  In the  on Gutman's (1969) data  58 Table  11:  Zajonc indices for pleasantness  Effect r a t i o p and o agree Age Group  3-b years 7-8 9-10 11-12  +  pL o  12.72 12.17 13.46 12.66  p and o disagree  -  pL o  4.88 4.02 6.17 5.09  +  pL o  11.38 11.40 11.99 9.61  -  pL o  5.92 5.12 6.37 5.20  4-1 C CD  e CD OJ  U 00  <  1.02 .98 1.07 1.20  a  o  •H 4J  O  CO  U  4-1 4->  <J  2.23 2.58 2.03 2.16  CU  o c c a r-4 td pq  1.15 1.12 1.09 1.22  59  present study, the proportion attributable to a t t r a c t i o n was was  a t t r i b u t a b l e to agreement.  suggests that a t t r a c t i o n may  45.51%;  .29%  The difference i n these percentages  decrease i n importance between childhood  and  adulthood, but that there i s an increase over the same age span i n the importance of agreement, when pleasantness i s the c r i t e r i o n . Within the age range examined In the present study ( i . e . , 5-12  years  of age)  there was  no evidence of i n t e r a c t i o n between age and a t t r a c t i o n  (F<1).  However, the i n t e r a c t i o n between age and agreement was s i g n i f i c a n t  (F=3.36; df=3,76; P<.05).  Inspection  of the means relevant to this i n t e r -  action indicate that differences between agreement and disagreement were greater among older than among younger children (see column 3 of Table 12).  Table 12 The age by agreement i n t e r a c t i o n might perhaps be attributed to the abstract nature of the t h i r d entity i n each t r i a d i c s i t u a t i o n ; that i s , to the p o s s i b i l i t y that younger children may older children i n conceptualizing  "X."  have had greater d i f f i c u l t y than  Such an argument would not, however,  explain the small agreement effects apparent i n Ohashi's (1964) data for Ohashi used three-person structures.  I t seems more l i k e l y that the  age  increase i n agreement effects i s related to an increase i n the a b i l i t y to coordinate information.  (This interpretation w i l l be discussed  i n greater  d e t a i l i n Chapter VII). In sum, (b)  i t i s clear that hypothesis 2 was  not supported by the data,  Results relevant to hypothesis 3 The results of the analysis of variance  summarized i n Table 13.  of consistency  ratings  are  This analysis yielded s i g n i f i c a n t effects f o r  60  T a b l e 12:  Mean p l e a s a n t n e s s r a t i n g s f o r s t r u c t u r e s c o n t a i n i n g agreement and  Age Gp.  Al Agreement  disagreement  A2  A1-A2  Disagreement  1 (5-6 y e a r s )  3.82  8.64  .18  2 (7-8 y e a r s )  8.09  8.26  -.17  3 (9-10 y e a r s )  9.81  9.18  .63  4 (11-12 y e a r s )  8.87  7.40  1.47  61  agreement but  (F=30.38;  df=l,76; P<.01) and b a l a n c e ( F = 3 1 . 1 1 ; df=l,76; P<.01)  i n t e r a c t i o n e f f e c t s between age and agreement  (D x A ) and age and  T a b l e 13  b a l a n c e (D x A x B) were n o t s i g n i f i c a n t . s u p p o r t h y p o t h e s i s 3.  The r e s u l t s , t h e r e f o r e , do not  That i s , i f h y p o t h e s i s 3 were t e n a b l e ,  differences  between agreement and disagreement would have been g r e a t e r i n younger than i n o l d e r c h i l d r e n ; d i f f e r e n c e s between b a l a n c e and imbalance would have been g r e a t e r consistency ment.  i n older children. r a t i n g s to s t r u c t u r e s  Younger c h i l d r e n would have a s s i g n e d 1, 2, 3, and 4 s i n c e a l l c o n t a i n  In o l d e r c h i l d r e n , the p e r c e i v e d  consistency  high  agree-  o f these s t r u c t u r e s  would have been reduced by the imbalance i n s t r u c t u r e s 3 and 4.  Younger  c h i l d r e n would have r a t e d s t r u c t u r e s 5, 6, 7, and 8 as i n c o n s i s t e n t because a l l c o n t a i n disagreement. balance i n structures Inspection  Older c h i l d r e n , responding to the  7 and 8 would have r a t e d  o f mean c o n s i s t e n c y  a l l groups a s s i g n e d  highest  r a t i n g s i n d i c a t e s that subjects i n  r a t i n g s to b a l a n c e d s t r u c t u r e s c o n t a i n i n g  ment ( s t r u c t u r e s 1 and 2 ) . The next h i g h e s t imbalanced s t r u c t u r e s c o n t a i n i n g Structures  them more c o n s i s t e n t .  r a t i n g s were a s s i g n e d  agree-  to  disagreement ( s t r u c t u r e s 5 and 6 ) .  combining b a l a n c e w i t h disagreement (7 and 8 ) and imbalance  w i t h agreement  ( 3 and 4) were r a t e d lowest i n a l l groups.  The s t r u c t u r a l  f e a t u r e which d i s t i n g u i s h e s s t r u c t u r e s 3 , 4 , 7 and 8 from those r a t e d more c o n s i s t e n t i n the presence o f a n e g a t i v e  P/0 bond.  between P and 0, i n o t h e r words, e x e r t e d  greater  r a t i n g s than e i t h e r agreement o r b a l a n c e .  The s i g n o f a t t r a c t i o n  i n f l u e n c e on  consistency  62  Table 1 3 : Summary of analysis of variance of consistency ratings  Source of Variation  df_  Between Subjects  79  MS  J_  6.82**  2.56  30.38** 1.21  3.07  360.33** .26  31.13  (age gps)  3  261.88  Subj. i n gps  79  38.41  D  Within Subjects  1200  A (Agreement) DA A x Subj. i n gps  1 3 76  941.88 37.39  B(Attraction) DB B x Subj. i n gps  1  9537.50  3  6.87  76  26.47  1  1.13 3.27 5.02  C (Format) DC C x Sub  3 76  0 (Structure) DO 0 x Subj. i n gps  3 76  AB (Balance) DAB AB x Subj. i n gps AC DAC AC x Subj. i n gps AO DAO AO x Subj. i n gps BC DBC BC x Subj. i n gps BO DBO BO x Subj. i n gps CO DCO CO x Subj. i n gps  % of v a r i a b i l i t y due to significant main e f f e c t s and i n t e r actions  1  1  31.00  .22 .65  822.40 68.08 12.43  66.19** 5.48**  2.68 .67  31.11** 1.34  1.24  3  380.63 16.41  76  12.24  1  8.78  .74  3  3.41  .29  7G  11.78  1  990.53  96.71**  3.23  3 76  72.70 10.24  7.10**  .71  1  35.78 34.02  3.04  3 76 1  2.89*  .33  12.25** 1.49  .33  11.78  3  102.38 12.46  76  8.36  1 3  2.81 8.14  76  7.68  .00 1.06  63  Source of Variation  df  MS  ABC DABC ABC x Subj. i n gps  1 3 76  4.75 .74 6.30  ABO DABO ABO x Subj. i n gps  1 3 76  113.83 9.78 11.21  240  5.95  Residual Total  *p<.05 **p<. 01  1279  % of v a r i a b i l i t y due to significant main effects and i n t e r actions .75 .12 10.60** .87  .39  64  The F r a t i o f o r a t t r a c t i o n was 360.33 (df=l 76; P<.01>. t  accounted f o r 31.13% of t h e * v a r i a b i l i t y .  This factor  Agreement accounted f o r 3.07% of 13  the v a r i a b i l i t y .  The proportion due to balance was 1.24%.  The same pattern of results with respect to the effects of balance, agreement, and a t t r a c t i o n obtained when the data were analyzed according to Zajonc's (1968) method.  As shown i n Table 14, index values f o r agree-  ment are s l i g h t l y greater than those f o r balance i n a l l but the 7-8 group Table 14  but the most noticeable differences are between balance and agreement on the one hand and a t t r a c t i o n on the other. The pattern was repeated when consistency ratings were subjected to multiple regression analysis.  As shown i n Appendix 4, average beta  c o e f f i c i e n t s for agreement were larger than those f o r balance i n a l l but the 7-8 group.  Average beta c o e f f i c i e n t s f o r a t t r a c t i o n exceed those  for agreement and balance i n a l l groups. thus i n d i c a t e :  A l l three methods of analysis  (a) that agreement exerted s l i g h t l y greater e f f e c t on the  ratings than balance but that a t t r a c t i o n was the primary component of consistency ratings; and (b) that there was no s i g n i f i c a n t i n t e r a c t i o n between age and balance, age and agreement, or age and a t t r a c t i o n . The prediction that balance and agreement would i n t e r a c t with age, 13 Factor 0 effects were also s i g n i f i c a n t (F=66.19; df=l,76; P<.01). The proportion of v a r i a b i l i t y accounted for by factor 0 was 2.68%. An additional 3*23% of the v a r i a b i l i t y i s attributable to the i n t e r a c t i o n between factor 0 and agreement. The contribution of factor 0 to the ratings was thus greater than that of balance.  65  Table 14:  Zajonc Indices for consistency  E f f e c t Ratio p and o agree Age group 5-6 years 7-8 9-10 11-12  +  pL o 11.10 11.80 13.48 13.11  p and o disagree _  pL o 4.72 4.84 7.65 6.08  +  pL o 9.14 9.55 10.46 9.12  §  c o XJ  CD  VJ  e a>  _  pL o 4.32 4.96 6.04 5.48  o co  &  JJ  1.18 1.15 1.28 1.31  2.24 2.18 1.75 1.92  <i...  <u v  c CO  "CO «,  .  1.11 1.16 1.08 1.22  66  i n the case of c o n s i s t e n c y r a t i n g s , was  based  on two  related  assumptions:  (a) t h a t a f f e c t i n f l u e n c e s s o c i a l p e r c e p t i o n s o f younger c h i l d r e n more than c o n s i d e r a t i o n s of c o n s i s t e n c y , and (b) t h a t the i n f l u e n c e o f a f f e c t decreases The  f i r s t assumption  was  supported.  as age i n c r e a s e s .  C h i l d r e n i n the 5-6  group based  c o n s i s t e n c y r a t i n g s on the same s t r u c t u r a l f a c t o r u t i l i z e d ratings—namely,  attraction.  But a t t r a c t i o n was  i n pleasantness  a l s o the primary  of p l e a s a n t n e s s and c o n s i s t e n c y r a t i n g s among o l d e r c h i l d r e n . evidence of an age i n c r e a s e i n u t i l i z a t i o n o f b a l a n c e . thus seem t o suggest  There was  no  The r e s u l t s would  that a f f e c t continues to exert stronger i n f l u e n c e  than b a l a n c e throughout  the 5-12  year p e r i o d .  p r e d i c t i o n s i t u a t i o n ( e . g . , Atwood, 1969; and Knox, 1969)  determinant  R e s u l t s o b t a i n e d i n the  Knox and Gutman, 1968;  Storm  a r e n o t , however, c o n s i s t e n t w i t h t h i s l i n e of r e a s o n i n g .  These l a t t e r s t u d i e s i n d i c a t e t h a t o l d e r c h i l d r e n base p r e d i c t i o n s more on b a l a n c e than on agreement o r a t t r a c t i o n . The d i s c r e p a n c y between r e s u l t s o b t a i n e d i n the r a t i n g s i t u a t i o n i n the p r e d i c t i o n s i t u a t i o n may,  perhaps, be e x p l a i n e d on the b a s i s o f  d i f f e r e n c e s i n task complexity.  Such an e x p l a n a t i o n i s p r o f f e r e d i n  Chapter  and  VII.  Summary The r e s u l t s , i n d i c a t i n g a l a c k of c l e a r d i f f e r e n t i a t i o n between p l e a s a n t n e s s and c o n s i s t e n c y a t a l l age h y p o t h e s i s 1.  levels, failed  The u n i f o r m l y h i g h c o r r e l a t i o n between p l e a s a n t n e s s  c o n s i s t e n c y i n a l l age groups p r e c l u d e d j o i n t 2 and  3.  to c o n f i r m  c o n f i r m a t i o n of  and  hypotheses  C o n c e i v a b l e , however, the d a t a might s t i l l have supported  e i t h e r h y p o t h e s i s 2 or h y p o t h e s i s 3.  The r a t i n g s were t h e r e f o r e s u b j e c t e d  67  to further s t a t i s t i c a l t e s t s .  These tests indicated that balance exerted  s l i g h t l y greater influence than agreement i n a l l age groups when pleasantness was  the c r i t e r i o n (the reverse of that predicted i n hypothesis  2) while i n the case of consistency, agreement exerted s l i g h t l y greater influence than balance.  The effects of balance and agreement were very  small, however, i n comparison to those of a t t r a c t i o n .  Subjects i n a l l  groups appear to have based both pleasantess and consistency ratings primarily on the sign of the P / 0 bond.  68  CHAPTER V ANCILLARY FINDINGS AND 1.  THEIR IMPLICATIONS  E f f e c t s Involving factor 0 (Structure effects) In the discussion of Zajonc indices (pp. 8 - 12) i  f c  w a s  pointed  out  that the eight triads employed i n t h i s study can be ordered into four quadrants by considering the sign of the P/0 bond ( p o s i t i v e or negative) and the presence or absence of agreement between P and 0 with regard X.  to  Theoretically, the two structures assigned to each quadrant should  exert the same e f f e c t on subjects' ratings since both are balanced (or imbalanced), both contain agreement (or disagreement), and both have the same P/0 bond.  The presence of s i g n i f i c a n t main e f f e c t s f o r factor 0 i n  the analysis of variance of pleasantness ratings (F=156.23; df=l,76; P<.01) and consistency  ratings (F=66.19; df=l,76; P<.01) demonstrates, however,  that this i s not the case.  Level 1 structures were rated more pleasant 14  and more consistent than l e v e l 2 structures by a l l age groups. The most parsimonious explanation the number of p o s i t i v e bonds contained of factor 0.  of factor 0 effects i s i n terms of i n the triads included i n each l e v e l  As shown i n the foldout i n Appendix 2, a l l structures i n  l e v e l 1 of factor 0 contain a positive 0/X bond, and a l l structures i n l e v e l 2 of factor 0 contain a negative 0/X bond.  Since each combination  of the other two bonds i s represented i n both l e v e l s , the structures i n l e v e l 1 are more " p o s i t i v e " than those i n l e v e l 2.  Subjects appear to  have preferred the more p o s i t i v e structures. Subsequent analysis of Gutman's (1969) data yielded s i g n i f i c a n t factor 0 e f f e c t s i n the ratings of adult subjects as w e l l . (See Appendix 5.)  69  Structure differences i n p o s i t i v i t y could also account for the s i g n i f i c a n t agreement by structures (A x 0) interaction i n the pleasantness ratings (F=206.60; df=l,76; P<„01) and i n the consistency ratings (F=96.71; df=l,76; P<.01).  The grouping of structures appropriate to this i n t e r -  action i s shown i n Table 15a. C e l l A of the table contains one structure  Table 15  i n which a l l bonds are positive and one i n which two bonds are p o s i t i v e and one negative; c e l l s C and D each contain one structure with two p o s i t i v e bonds and one with a single p o s i t i v e bond; c e l l 3 contains one structure with a single p o s i t i v e bond and one i n which a l l bonds are negative.  As  shown i n Table 15b, the c e l l means for pleasantness are ordered exactly as they should be i f p o s i t i v i t y were the operative factor.  For consistency  (see Table 15c), the most positive c e l l stands out; the other c e l l s are closely grouped.  The c e l l containing the greatest number of p o s i t i v e  relations i s rated highest i n each l e v e l of factor 0.  In l e v e l 1, the  c e l l containing the greatest number of p o s i t i v e r e l a t i o n s contains structures i n which there i s agreement between P and 0.  In l e v e l 2, there are more  positive relations i n the c e l l containing structures with disagreement between P and 0.  The r e s u l t i s that agreement i s more pleasant and more  consistent than disagreement at l e v e l 1 of factor 0, but the reverse tends to be true at l e v e l 2; hence, the i n t e r a c t i o n of factors 0 and  A.  Since a l l that i s required i n order to rate structures i n terms of p o s i t i v i t y i s that one compare the number of p o s i t i v e relations i n the structures, it seems reasonable that p o s i t i v i t y effects should be strong ;  70  Table 15:  Mean pleasantness and consistency ratings relevant to the agreement (A) x structures (0) i n t e r a c t i o n .  01  02  (high p o s i t i v i t y ) p<  \/  A l (Agree)  ±Q  P  f  V  (5)  -J  0  (3)  (1)  \/  -  >0  Pv^  (low p o s i t i v i t y ) p  P««  f  (2)  A  - -»0  V  (8)  _  (4)  *o  P<-  B  - -*0  V  (6)  (7)  C  D  (a) 01  02  01  02  Al  10.95  6.84  Al  10.78  7.41  A2  8.19  8.55  A2  7.30  7.46  X Pleasantness ratings  (b)  X Consistency ratings  (c)  71  i n younger children.  However, since older children are capable of  u t i l i z i n g more complex strategies, one would expect older children to take greater cognizance of the position i n which positive relations occur i n the t r i a d s .  The ratings of older children, i n other words, should be  based more on the arrangement of positive and negative relations i n d i f f e r e n t triads  than on the t o t a l number of positive relations contained.  The s i g n i f i c a n t i n t e r a c t i o n of factor 0 with age i n the consistency ratings (F=5.48; df=3,76; P<.01) i s consistent with t h i s expectation. The interaction with age was  not s i g n i f i c a n t for the pleasantness ratings  (F=2.02; df=3,76; P=.12). The e f f e c t of a t t r a c t i o n — t h e predominant influence on pleasantness and consistency ratings at a l l age levels i s , of course, a p o s i t i v e biasing effect as w e l l , but one s p e c i f i c to the P/0  r e l a t i o n s h i p , which tends to  override the preference for p o s i t i v e bonds elsewhere i n the t r i a d . 2.  Effects involving factor C (Format effects) Main effects due  to format differences were not s i g n i f i c a n t i n either  the analysis of pleasantness ratings (F=1.02; df=l,76; P=.32) or analysis of consistency ratings (F<1). factor C was  Only one i n t e r a c t i o n involving  s i g n i f i c a n t i n the case of consistency—that  a t t r a c t i o n , and  formats.  the  between  age,  In the case of pleasantness, formats were s i g -  n i f i c a n t as a factor i n i n t e r a c t i o n with agreement; age and agreement; age and a t t r a c t i o n ; and agreement and a t t r a c t i o n .  In general, these  interactions suggest that triads involving neighbor-playmates have a more pleasant connotation for younger children than triads involving the more task-oriented  relationship of partners.  Differences  i n these interactions are small, however.  i n the means involved  Effects involving factor C  72  account for less than 1% of the v a r i a b i l i t y i n the data. 3.  Main e f f e c t s of  age  S i g n i f i c a n t main effects of age were obtained both for pleasantness (F=3.28; df=3,76; P<.05) and consistency of 8.73,  8.17,  9.49  and 8.13  (F=6.82; df=3,76; P<,01).  were obtained for each successive age group  when pleasantness ratings were collapsed across structures. average consistency ratings for the four age groups were 7.32, and 8.44.  Means  Corresponding 7.78,  9.40  This would seem to indicate s l i g h t l y d i f f e r e n t use of the  rating scale by subjects of d i f f e r e n t chronological age.*~* Summary An unanticipated with factor 0.  finding i n the r e s u l t s was  the difference associated  In the case of both pleasantness and consistency  ratings  subjects assigned higher ratings to structures i n l e v e l 1 ( i . e . , structures 1, 3, 5, and 8) than to those i n l e v e l 2 (structures 2, 4, 6, and  7).  effect was  attributed  tentatively attributed to p o s i t i v i t y .  That i s , i t was  This  to a preference for structures containing the greatest number of p o s i t i v e relations. There were no s i g n i f i c a n t main effects for formats i n either the pleasantness ratings or the consistency  ratings.  The format factor,  although involved i n several higher order i n t e r a c t i o n s , apparently  contributed  l i t t l e to the ratings.  Pair-wise comparison of age groups according to the Neuman-Keuls procedure (Winer, 1962, pp; 80-85) indicated no s i g n i f i c a n t betweengroup differences either i n the pleasantness ratings or i n the consistency ratings.  73  CHAPTER VI RESULTS OF THE CROSS-VALIDATION STUDY Reliability Average r e l i a b i l i t y c o e f f i c i e n t s for pleasantness and consistency are shown i n Table 16.  As  i  n  the p r i n c i p a l study, r e l i a b i l i t y  the 5-6 group than i n the older groups.  i s lower i n  This i s p a r t i c u l a r l y noticeable  Table 16  in the case of consistency ratings where the average value for the 5-6 group i s .59 while those f o r the other age groups range from .75 to .96. Mann-Whitney U tests indicated, however, that none of the differences between age groups were s i g n i f i c a n t . Comparison across studies (see Tables 6 and 16) indicates that r e l i a b i l i t y was generally lower i n the cross-validation study but the only s i g n i f i c a n t difference between samples was i n the 5-6 group, and then only i n the case of consistency ratings (U=120; z=2.02; P<.04, two-tailed). Hypothesis 1 Hypothesis 1 predicted an increase i n d i f f e r e n t i a t i o n between pleasantness and consistency with increasing age.  The r e s u l t s of the  p r i n c i p a l study f a i l e d to support the hypothesis:  r  groups.  p  r  was high i n a l l  The results of the cross-validation study also f a i l to support  hypothesis 1.  As shown i n Table 17, instead of decreasing with increasing  Table 17  T a b l e 16:  Mean r e l i a b i l i t y  coefficients  (Pearson r  f o r P l e a s a n t n e s s and C o n s i s t e n c y  easentness  Consistency  5-6  .76  .59  7-8  .82  .96  9-10  .89  .75  11-12  .87  .90  75  Table 17:  Mean pleasantness-consistency correlations f o r each age group (Pearson r)  Age  Gp  PC  5-6  .46  7-8  .72  9-10  .78  11-12  .81  76  age, average PC correlations increase with increasing age. Hypotheses 2 and 3 Hypotheses 2 and 3 were concerned with the r e l a t i v e effects of agreement and balance on pleasantness and consistency ratings.  I t was  predicted that subjects i n a l l age groups x<rould u t i l i z e agreement to a greater extent than balance when rating f o r pleasantness (hypothesis 2). Younger subjects were expected to base consistency ratings more on agreement than on balance; balance was expected to exert greater influence than agreement on the consistency ratings of older subjects (hypothesis 3). The results of the p r i n c i p a l study f a i l e d to confirm either hypothesis. Results i n the d i r e c t i o n predicted f o r hypothesis 2 were obtained i n the cross-validation study.  As shown i n Table 18, the F r a t i o f o r agreement  (F=22.19; df=l,28; P<.01) i s s l i g h t l y larger than the F r a t i o f o r balance  Table 18  (F=19.92; df=l,28; P<.01) when pleasantness i s the c r i t e r i o n .  Since there i s  no d i r e c t method of testing for the significance of differences between two values of F, one must turn to the multiple regression analysis f o r a s t a t i s t i c a l test of hypothesis 2.  From a within subject comparison of  regression c o e f f i c i e n t s i n the equations for pleasantness, i t was found that the beta c o e f f i c i e n t s f o r agreement were of greater magnitude than those f o r balance f o r only 50% of the subjects.  This i s clearly not  s i g n i f i c a n t , and thus, as i n the p r i n c i p a l study, leads to r e j e c t i o n of hypothesis 2.  Also, as i n the p r i n c i p a l study, the percent of v a r i a b i l i t y  accounted for by agreement (2.47%) and balance (1.54%) i s small i n  77  Table IQ:  Summary of analysis cf v a r i a n c e of pleasantness ratings  Source of Variation  df  Between Subjects  31  D (Age gps) Subj. i n gps  3 28  Within Subjects  MS  111.46 55.77  F  % of v a r i a b i l i t y due to significant main effects and i n t e r actions  2.00  480  A (Ag reement) DA A x Subj. i n gps  1 3 28  328.32 25.26 14.80  22.19** 1.71  2.47  B (Attraction) DB B x Subj. i n gps  1 3 28  4630.30 52.79 71.32  65.17** .73  35.26  C (Format) DC C x Subj. i n gps.  1 3 28  .50 3.27 5.98  0 (Structures) DO 0 x Subj. i t i gps.  1 3 28  276.12 4.07 6.57  42.01** .62  2.08  AB (Balance) DAB AB x Subj. i n gps.  1 3 28  205.03 12.69 10.30  19.92** 1.23  1.54  AC DAC AC x Subj. i n gps.  1 3 28  .03 3.68 5.81  AO DAO AO x Subj. i n gps.  1 3 28  399.03 45.15 13.59  BC DBC BC x Subj. i n gps.  1 3 28  9.57 3.45 11.50  .83 .73  BO  4.88 2.44 5.10  .96 .48  BO x Subj. i n gps.  1 3 28  CO DCO CO x Subj. i n gps.  1 3 28  18.76 4.62 6.21  3.02 .74  DEO  .08 .55  .01 .63 29.35** 3.32**  3.01 1.02  78  Source of Variation  F  df  MS  ABC DABC ABC x Subj. i n gps  1 3 28  .38 1.25 4.27  .09 .29  ABO DABO ABO x Subj. i n gps  1 3 28  2.82 6.53 2.72  1.04 2.40  Residual  96  5.21  Total  *p<.05 **p<.01  511  % of v a r i a b i l i t y due to significant main e f f e c t s and interactions  79  comparison to that attributable to a t t r a c t i o n (the F r a t i o f o r a t t r a c t i o n was 65.17; this factor accounts f o r 35.26% of the v a r i a b i l i t y ) .  Attraction,  i n other words, was the primary component of pleasantness ratings i n both studies. The analysis of consistency  ratings leads to the same conclusion  concerning hypothesis 3 as the p r i n c i p a l study.  As shown i n Table 19,  Table 19  s i g n i f i c a n t e f f e c t s (P<.01) were obtained for agreement (F=13.48; df=l,28) and balance (F=11.93; df=l,28).  Neither factor interacted with age.  Attraction accounted f o r a greater proportion of the v a r i a b i l i t y (24.33%) than either agreement (3.83%) or balance (.65%). Effects involving factor 0 (Structure effects) The results of the cross-validation study also confirmed the findings of the p r i n c i p a l study with respect to factor 0. and  As shown i n Tables 18  19, s i g n i f i c a n t main effects f o r factor 0 were obtained i n the case  of both pleasantness ratings (F=42;01; df=l,28; P<.01) and ratings (F=18.18; df=l,28; P<.01). 0 and age.  consistency  There was no i n t e r a c t i o n between factor  A l l groups rated l e v e l 1 structures  ( i . e . , structures 1, 3, 5,  and 8) s i g n i f i c a n t l y more pleasant and more consistent than l e v e l 2 structures  ( i . e . , structures 2, 4, 6, and 7).  As i n the p r i n c i p a l study, there was a s i g n i f i c a n t i n t e r a c t i o n between factor 0 and agreement i n the pleasantness ratings (F=29.35; df=l,28; P<.01) and i n the consistency  ratings (F=17.56; df=l,28; P<.01).  Inspection of the relevant c e l l means indicates that l e v e l 1 structures  eo Table 19:  Summary of analysis of variance of  Source of Variation  df  Between Subjects  »*• • •  D (Age gps) Subj. i n gps  3 28  Within Subjects  MS  78.18 62.20  consistency ratings  % of v a r i a b i l i t y due to significant main effects and i n t e r actions  F  1.26  480  A (Agreement) DA A x Subj. i n gps  1 3 28  544.50 18.26 40.38  13.48** .45  3.83  B (Attraction) DB B x Subj. i n gps  1 3 28  3454.90 20.47 66.72  51.78** .31  24.33  C (Format) DC C x Subj. i n gps  1 3 28  46.32 8.26 7.37  6.28* 1.12  0 (Structure) DO 0 x Subj. i n gps  1 3 28  321.95 8.65 17.71  18.18** .49  2.27  AB (Balance) DAB AB x Subj. i n gps  1 3 28  92.82 5.36 7.78  11.93** .69  .65  AC DAC AC x Subj. i n gps  1 3 28  18.76 15.67 9.59  AO DAO AO x Subj. i n gps  1 3 28  255.95 6.97 14.58  BC DBC BC x Subj. i n gps  1 3 28  .13 17.05 19.43  .01 .88  BO DBO  CO DCO CO x Subj. i n gps  1 3 28  94.53 24.27 10.58 1.53  8.94** 2.29  BO x Subj• i n gps  1 3 28  2.69 5.19  .33  1.96 1.63 17.56** .48  .29 .52  1.80  .67  81  Source of Variation  df  MS  F  ABC DABC ABC x Subj. i n gps  1 3 28  1.53 10.02 6.30  .24 1.59  ABO DABO ABO x Subj. i n gps  1 3 28  50.00 28.18 9.72  5.14* 2.90*  Residual  96  8.47  Total *p<.05 **p<.01  511  % of v a r i a b i l i t y due to significant main e f f e c t s and i n t e r actions  .35 .59  82  containing agreement were rated more pleasant and more consistent than those containing disagreement.  Level 2 structures containing disagreement  were rated more pleasant but not more consistent than those containing agreement. Effects involving f a c t o r C (Format effects) In contrast to the p r i n c i p a l study, s i g n i f i c a n t main effects f o r factor C were obtained i n the analysis of consistency ratings (F=6.28; df= 1,28; P<.05).  Inspection of the means indicated that subjects assigned  higher consistency ratings to neighbor-playmate triads.  triads than to partner  There were no format effects i n the case of pleasantness (F<1).  None of the interactions involving factor C were s i g n i f i c a n t for either type of rating. Summary In general, the results of the cross-validation study confirmed the findings of the p r i n c i p a l study: (a) they f a i l e d to provide evidence of an increase i n d i f f e r e n t i a t i o n between pleasantness and consistency as a function of age, (b) they indicated that a t t r a c t i o n was  the primary s t r u c t u r a l component  of pleasantness and consistency ratings at a l l age l e v e l s , (c) they yielded results similar to those of the p r i n c i p a l study with regard to factor 0 (I.e., subjects i n a l l age groups assigned s i g n i f i c a n t l y higher pleasantness and consistency ratings to structures i n l e v e l 1 of factor 0 than to those i n l e v e l 2). The results, i n other words, are not r e s t r i c t e d to the p a r t i c u l a r sample studied nor are they dependent, i n any large measure, on a p a r t i c u l a r 12.  83  CHAPTER VII FURTHER DISCUSSION The results of the present studies indicate that children between the ages of 5-12 years tend to base pleasantness and consistency ratings more on a t t r a c t i o n than on agreement or balance.  The children also  appeared to rely on p o s i t i v i t y to a s l i g h t l y greater extent than on agreement or balance xvhen rating for pleasantness and consistency.  These  r e s u l t s , which d i f f e r from results obtained with children i n the prediction s i t u a t i o n (e;g., Atwood, 1969; Storm and Knox, 1969), may  perhaps be  explained by considering the information processing requirements of the rating task.  In the rating s i t u a t i o n , the c h i l d must not only decide  whether a p a r t i c u l a r structure i s pleasant or unpleasant (consistent or inconsistent) but must evaluate that structure i n comparison to others presented simultaneously (as i n the present study), or i n series (e.g., Ohashi, 1964).  He must hold i n mind f a i r l y complex instructions concerning  the bases and mechanics of rating and also information concerning the s p e c i f i c relationship between the individuals i n the s i t u a t i o n ( i . e . , format information).  The rating s i t u a t i o n thus places considerable demands on the  information processing a b i l i t i e s of the c h i l d .  The children may have based  their ratings f i r s t and foremost on a t t r a c t i o n because i t was way of coping with the complexity of the task.  the simplest  (The c h i l d need focus  attention on only one r e l a t i o n — t h a t between P and 0 — i n order to evaluate the structures i n terms of attraction.)  Evaluation i n terms of agreement  required that the c h i l d simultaneously attend to two r e l a t i o n s — t h o s e between P and X and between 0 and X.  In order to evaluate the structures  in terms 6f balance, the c h i l d must simultaneously attend to three r e l a t i o n s .  84  Processing solely i n terms of a t t r a c t i o n would, however, only enable the c h i l d to dichotomize the structures since four structures contain positive P/0 relations and four contain negative P/0 r e l a t i o n s .  To f u l f i l l  the requirements of the task the children had to devise some means of further d i f f e r e n t i a t i n g between the structures.  I t i s conceivable that  they made an i m p l i c i t count of the number of p o s i t i v e relations contained i n the structures.  In other words, one plausible way to account f o r the  results and, i n c i d e n t a l l y , a way that i s consistent with the overt behaviour of a number of children i n the p i l o t studies i s to suggest that the subjects may f i r s t have separated the structures into those i n which P and 0 l i k e d each other and those i n which they d i s l i k e d each other.  The structures  within each P/0 category may then have been arranged i n terms of the number of positive relations they contained. A two-stage process of this nature would serve to make the task more manageable and i t would account for the ranking of structures i n order of mean pleasantness and consistency ratings (see Table 20), at l e a s t i n the case of the younger children.  Inspection of Table 20 suggests, however,  Table 20  that the older children may have d i f f e r e n t i a t e d between structures more on the basis of a t t r a c t i o n and agreement than on a t t r a c t i o n and p o s i t i v i t y , especially when rating f o r consistency,  A preference f o r agreement over  p o s i t i v i t y would be consistent with Bruner's  (1964) observation that the  complexity of information processing strategies increases with increasing age, since processing i n terms of agreement requires that the subject  8g<.  TABLE 2Q; Rank order of structures i n terms of mean pleasantness and mean consistency ratings.  Age Group A.  5-6  7-8  9-10  11-12  +-+  ++-  ++-  +-+  H  +—  -4+  —H-  -++  -++  -+-  —+  -+-  1-  Pleasantness  Adults (Gutman, 1969)  Highest rating  H  ++-  +-+  +-+  — +  Lowest rating  B.  5-6  7-8  9-10  -H—  +-+  H  +-+  ++-  H  +—  11-12  Consistency  Adults (Gutman, 1969)  Highest rating  +-+  ++-  -++-+  — +  Lowest rating  — +  -+-  -++  86  take cognizance of the position of p o s i t i v e and negative r e l a t i o n s within the structures whereas processing i n terms of p o s i t i v i t y requires only that the subject count the number of positive relations i n the structure. Further studies focusing s p e c i f i c a l l y on agreement and p o s i t i v i t y are, however, required i n order to determine whether there i s a s i g n i f i c a n t change i n the r e l a t i v e effects of these two variables over age.  Studies  focusing on the rating process i t s e l f are, of course, also necessary i n order to determine the t e n a b i l i t y of a two-stage hypothesis. Although  the results of the present studies suggest that u t i l i z a t i o n  of agreement may  increase with increasing age, there was no evidence of an  age increase i n u t i l i z a t i o n of balance.  Balance effects were small i n a l l  groups, both for pleasantness and for consistency.  As previously mentioned  'p. 66) the absence of strong balance effects among older children contrasts with results obtained i n studies i n which children are required to predict missing relations (e.g., Atwood, 1969; Knox and Gutman, 1968; Storm and Knox, 1969).  The discrepancy between results obtained i n the  rating s i t u a t i o n and i n the prediction s i t u a t i o n may, differences i n task complexity.  perhaps, be due to  For example, i n the prediction s i t u a t i o n  the subject i s not required to make comparative judgements between structures; he does not have to consider degrees of pleasantness or degrees of consistency.  Two  relations are presented, he need only supply a positive or  negative sign for the missing t h i r d r e l a t i o n .  The rating s i t u a t i o n , i n  other words, places greater demands on the information processing a b i l i t i e s of the children.  I t i s possible that children r e s t r i c t t h e i r attention  to s p e c i f i c components of the structures as the o v e r a l l complexity of the task increases.  1  87«  Two  studies reported by Singer (1966) support the notion that there  i s a relationship between the information processing requirements of a task and tendencies toward balance. adult  Both studies were conducted with  subjects and both used a prediction task modelled after that of  Morrissette (1958).  16  In these studies, subjects were given p a r t i a l  information concerning the sentiment relations among four persons involved i n an apartment s i t u a t i o n .  Their task was  to predict the remaining relations  and to indicate the degree of tension they would f e e l i n the completed situations.  When subjects were given four r e l a t i o n s (e.g., A  B  C) they  P predicted the remaining  two r e l a t i o n s i n a maximally balanced manner and  reported tension inversely related to the degree of balance. three r e l a t i o n s (e.g., A  B  When g i .en  C) both predictions and tension were more  \  P variable.  With two of s i x r e l a t i o n s given (e.g., A  B  C) there  was  P s i g n i f i c a n t l y less tendency to perceive a balanced system and the degree of balance bore l i t t l e r e l a t i o n to the reported tension. concludes  Singer (1966)  that "these studies show that the motivating effects of i n -  consistency can be v i t i a t e d by  'cognitive f l o o d i n g " ( p . 70). 1  16 In an attempt to test Cartwright and Harary's (1956) formulation of Balance Theory, Morrissette had undergraduate students role-play a move into an apartment. The sentiments among some of the roommates were given. The subject's task was to predict the remaining sentiments and to indicate the degree of tension he would f e e l i n such a s i t u a t i o n . Morrissette found that i n general subjects tended to complete the s i t u a t i o n s i n a balanced manner. He also found a p o s i t i v e correlation between reported tension and the degree of imbalance i n the s i t u a t i o n s .  88  Although d i f f e r e n t i a l task complexity may perhaps account f o r the discrepancy between results obtained with children i n the rating prediction s i t u a t i o n , a task-complexity  vs the  argument cannot be used to explain  the difference between results obtained when adults and children rate structures for consistency.  For example, Knox (1963) and Gutman (1969)  had college students rate hypothetical P-O-X  structures f o r consistency.  In both studies ratings were based more on balance than on agreement or attraction.  Yet i n the present studies, consistency ratings were based  mainly on a t t r a c t i o n .  The difference In consistency ratings of adults and  children could, however, be due to differences i n information processing a b i l i t i e s ( i . e . , differences i n the amount of information that the subject can process simultaneously; i n the f a c i l i t y with which the subject can encode, store, and decode information, and so on). Adults, i n other words, may have less d i f f i c u l t y than children i n coping with the demands of the rating s i t u a t i o n .  They may be more f a c i l e at combining and considering the  three r e l a t i o n s i n each s i t u a t i o n while holding other relevant information i n mind.  The difference  i n consistency ratings of adults and children  could also be due to differences i n the strength of the balance  "schema."  The "schema" or i m p l i c i t code f o r balance may be more firmly established i n adults than i n children.  This could perhaps account f o r the fact that  although adults u t i l i z e balance  to a greater extent than agreement or  a t t r a c t i o n i n the prediction and the rating s i t u a t i o n , strong balance  effects  among children are obtained only i n the easier prediction s i t u a t i o n .  The  balance "schema" i n children may not be of s u f f i c i e n t strength to withstand the competition of a l t e r n a t i v e biases such as a t t r a c t i o n , agreement, and p o s i t i v i t y when the task i s complex.  89  Recommendations f o r future research There do not appear to be any studies investigating the relationship between task complexity and balance i n children.  Such studies should be  conducted, using both rating and prediction tasks.  If the absence of  strong balance effects i n the rating s i t u a t i o n i s , i n f a c t , due to the child's i n a b i l i t y to cope with the information processing requirements of the rating task, balance effects should increase i n strength as task complexity decreases.  By the same token, there should be a decrease i n  balance effects i n the prediction s i t u a t i o n as task complexity increases. There should also be an i n t e r a c t i o n between age, task complexity, and l e v e l of information processing such that complex tasks are d i f f e r e n t i a l l y more d i f f i c u l t for younger children.  Processing i n terms of balance, i n Other  words, should break down at a lower l e v e l of task complexity among younger than among older children. Studies i n which teenage subjects are required to rate structures f o r pleasantness and consistency should also be conducted f o r , although the results of the present study do not indicate an increase i n d i f f e r e n t i a t i o n between the ages of 5-12,  they contrast with results obtained by Knox  (1963) and Gutman (1969) with ratings i n adults. is:  The question that remains  at what point i n ontogenetic time does d i f f e r e n t i a t i o n develop?  When  are children able to c l e a r l y distinguish between what i s pleasant and what i s psychologically consistent when presented with hypothetical s o c i a l s i t u a t i o n s , or, a l t e r n a t i v e l y , when are they able to cope with the requirements of the rating situation? Another method of investigating the developmental  course of d i f f e r -  entiation between pleasantness and consistency would be with an adaptation  90  of the prediction task.  Children of varying chronological age could be  s p e c i f i c a l l y instructed to predict missing relations i n s o c i a l structures so as to maximize pleasantness i n one instance and to maximize consistency i n another. hypotheses  Such a procedure might provide a more sensitive test of the of the present study and further strengthen the interpretation  of results f o r the rating s i t u a t i o n provided above. Comparison of the present results with those of Knox (1963) and Gutman (1969) also raises questions concerning the nature of the rating process i n adults and children, especially as i t concerns consistency. The results of the Knox and the Gutman studies suggest that adults combine agreement and a t t r a c t i o n so as to evaluate the consistency of t r i a d i c structures i n terms of balance.  The present results suggest that children  may focus on the components of balance i n sequence.  That i s , children  appear to evaluate t r i a d i c structures f i r s t and foremost i n terms of attraction.  Attention i s then directed to the presence or absence of  agreement and/or to the t o t a l p o s i t i v i t y of the structures.  In future  studies greater attention should be focused on the rating process i t s e l f .  91  CHAPTER VIII SUMMARY The primary purpose of the present study was  to investigate the  s t r u c t u r a l bases of pleasantness and consistency ratings and to determine the relationship between the two types of judgement i n children ranging i n age from 5-12 years.  An additional purpose of the study was to determine  whether the results of studies by Atwood (1969) and by Storm and Knox (1969) using a prediction procedure to investigate the developmental  course of  cognitive balance would generalize to a d i f f e r e n t dependent measure. s p e c i f i c hypotheses were tested. 1.  Three  I t was predicted that:  i n their ratings of hypothetical s o c i a l situations, young children  would d i f f e r e n t i a t e l i t t l e between pleasantness and consistency. Relative to the youngest children, older children were expected to d i f f e r e n t i a t e more between pleasantness and consistency.  Thus, i t was predicted that as a  function of increasing age, correlations between pleasantness and consistency would monotonically decrease across the successive age groups i n the study. 2.  children at a l l age levels would attach greater weight to agreement  (or disagreement) between P and 0 than to the balance (or imbalance) of the t o t a l s o c i a l s i t u a t i o n , when making pleasantness ratings. 3.  the r e l a t i v e importance of agreement and balance on consistency  ratings of s o c i a l situations would vary with chronological age. Younger c h i l d ren were expected to base consistency ratings more on agreement than on balance; it..was expected that balance would exert greater influence than agreement on consistency ratings of older children. The t e n a b i l i t y of hypothesis 1 was determined by correlating each subject's pleasantness ratings with his consistency ratings.  These i n t r a -  i n d i v i d u a l correlations were then subjected to analysis of variance.  This  92  analysis indicated no s i g n i f i c a n t between-group e f f e c t s — c o r r e l a t i o n s between pleasantness and consistency were uniformly high i n a l l age groups. The r e s u l t s , indicating a lack of clear d i f f e r e n t i a t i o n between pleasantness and consistency at a l l age l e v e l s , thus f a i l e d to confirm hypothesis 1. Although disconfirmation of hypothesis 1 precluded j o i n t confirmation of hypotheses  2 and 3, i t was s t i l l possible that the data might have  supported either hypothesis 2 or hypothesis 3.  The ratings were therefore  subjected to further s t a t i s t i c a l tests ( i . e . , analyses of variance).  These  tests indicated that balance exerted s l i g h t l y greater influence than agreement i n a l l age groups when pleasantness was the c r i t e r i o n (the reverse of that predicted i n hypothesis 2), while i n the case of consistency, agreement exerted s l i g h t l y greater influence than balance.  The effects of balance  and agreement were very small, however, i n comparison to those of a t t r a c t i o n . Subjects i n a l l age groups appear to have based both pleasantness and consistency ratings primarily on the sign of the P/0 bond. Zajonc indices and multiple regression analysis indicated that the results relevant to hypotheses data analysis.  2 and 3 were not dependent on the method of  A cross-validation study conducted concurrently with the  p r i n c i p a l study by an independent  and 'naive" E_ yielded the same pattern  of results with regard to a l l three hypotheses. Although the results of the p r i n c i p a l study and the cross-validation study f a i l e d to y i e l d support f o r the experimental hypotheses  they d i d  confirm one of the major assumptions underlying the hypotheses.  That i s ,  they indicated that affect influences the perceptions of younger children more than considerations of consistency.  The influence of a f f e c t was  reflected i n the subjects' u t i l i z a t i o n of a t t r a c t i o n as a basis for both  93  pleasantness and consistency ratings. The presence of strong a t t r a c t i o n effects i n the pleasantness ratings of children i s consistent with results obtained by Ohashi (1964).  Ohashi  had sixth grade children rate t r i a d i c structures for pleasantness.  His  data also indicate that children base pleasantness ratings more on a t t r a c t i o n than on agreement or balance.  Zajonc indices computed from data obtained  by Knox (1963) and Gutman (1969) indicate a s i m i l a r tendency among adult subjects.  Analysis of other studies with adults by the Zajonc method show  agreement or balance to have contributed s l i g h t l y more to pleasantness ratings than a t t r a c t i o n .  Indices for balance are of greater magnitude  than those f o r agreement or a t t r a c t i o n , on the other hand, i n a l l studies of psychological consistency conducted among adults. favor a t t r a c t i o n or balance, depending on the task.  Studies with children Attraction seems to  influence perceptions of children to a greater extent than agreement or balance when a rating task i s used (e.g., the present study); balance seems to be the more important determinant when the c h i l d i s required to predict missing relations (e.g., Atwood, 1969; Knox and Gutman, 1968; Storm and Knox, 1969). I t was  suggested that differences between adults and children i n the  rating s i t u a t i o n may be due to differences i n information processing a b i l i t i e s and/or to differences i n the strength of the balance "schema." Differences i n results obtained with children i n the rating vs the prediction s i t u a t i o n were tentatively attributed to d i f f e r e n t i a l task complexity.  94  BIBLIOGRAPHY Abelson, R, P., and Rosenberg, M. J . a t t i t u d i n a l cognition. Atwood, G.  Symbolic psychologic:  a model of  Behav. S c i . , 1 9 5 8 , _3, 1 - 1 3 .  A developmental study of cognitive balancing i n hypothetical  three-person systems. Braine, M. D. S.  Child Develop.,  1 9 6 9 , _40,  73-85.  The ontogeny of certain l o g i c a l operations:  formulation examined by nonverbal methods.  Piaget's  Psychol. Monogr., 1 9 5 9 ,  Whole No. 4 7 5 . Bruner, J . S.  The course of cognitive growth.  Amer. Psychol., 1 9 6 4 , 1 9 ,  1-16.  Burnstein, E.  Sources of cognitive bias i n the representation of simple  s o c i a l structures:  Balance, minimal change, p o s i t i v i t y , r e c i p r o c i t y ,  and the respondent's own  attitude.  J . pers. soc. Psychol., 1 9 6 7 , _ 7 ,  36-48.  Cartwright, D., and Harary, F. Heider's theory. Elkind, D.  S t r u c t u r a l balance:  Psychol. Rev.,  a generalization of  1 9 5 6 , 6i3, 2 7 7 - 2 9 3 .  Discrimination, s e r i a t i o n , and numeration of s i z e and  differences i n young children:  Piaget r e p l i c a t i o n study VI.  dimensional J . genetic  Psychol., 1 9 6 4 , 1 0 4 , 2 7 5 - 2 9 6 .  Festinger, L.  A theory of cognitive dissonance.  Evanston, 1 1 1 . :  Row,  Peterson, and Co., 1 9 5 7 . F l a v e l l , J . H.  Developmental psychology  of Jean Piaget.  Princeton, N. J . :  Van Nostrand, 1 9 6 3 . Gutman, G.  Pleasantness, consistency, and tension i n a student population.  Lnpublished, February, 1 9 6 9 . Heider, F.  Attitudes and cognitive organization.  J . Psychol., 1 9 4 6 , 2 1 ,  95  107-112. Heider, F.  The psychology of interpersonal r e l a t i o n s .  New York:  Wiley,  1958. Hershkowitz, A. values:  Interpersonal agreement and disagreement on objects and  a preliminary study.  Unpublished doctoral d i s s e r t a t i o n .  Univ. of Kansas, 1954. Jordan, N.  Behavioral forces that are a function of attitudes and of Hum. Relat., 1953, _6, 273-287.  cognitive organization. Jordan, N.  On cognitive balance.  (Res. Paper P - 1 7 8 ) .  Arlington, Va.:  Institute f ' r Defense Analyses, 1966. Knox, R. E. The components of cognitive balance. dissertation. Knox, R. E.  Unpublished doctoral  Univ. of Oregon, 1963.  Colloquium Address, University of V i c t o r i a , February 1969.  Knox, R. E., and Gutman, G. school children.  P o l i t i c a l s o c i a l i z a t i o n and balance i n American  Unpublished, November, 1968.  McGuire, W. J . Cognitive consistency and attitude change.  J . abnorm. soc.  Psychol., 1960, 6 0 , 345-353. Morrissette, J . An experimental study of the theory of s t r u c t u r a l balance. Hum. Relat., 1958, ll, Neweomb, T. M.  239-254.  An approach to the study of communicative acts.  Psychol.  Rev., 1953, 6 0 , 393-404. Neweomb, T.  Interpersonal balance.  In R. P. Abelson, E. Aronson et a l .  (Eds.), Theories of cognitive consistency:  a sourcebook.  Chicago:  Rand McNally and Co., 1968. Ohashi, M. triad.  Sociometric choice behavior and interpersonal perception i n a Jap. Psychol. Rev., 1964, J>, 7 2 - 8 7 .  96 Osgood, C. E., and Tannenbaum, P. H. prediction of attitude change. Piaget, J . Paul,  The p r i n c i p l e of congruity i n the  Psychol. Rev.,  1955, _62, 42-55.  The child's conception of physical c a u s a l i t y . London:  Kegan  1930.  Piaget, J .  Les notions de mouvement et de vitesse che2 1'enfant.  Presses Universitaires,  1946.  Piaget, J .  Psychology of i n t e l l i g e n c e .  Piaget, J .  The child's conception of number.  Piaget, J . , and Inhelder, B. Neuchatel:  New York: New  Harcourt Brace, York:  1950.  Humanities,  1952.  Le developpement des quantites chez 1'enfant.  Delachaux et N i e s t l e , 1941.  Piaget, J . , Inhelder, B., and Szeminska, A. geometry.  Paris:  New  York:  Basic Books,  The child's conception of  1960.  Price, K., Harburg, E., and Newcomb, T.  Psychological balance i n situations  of negative nterpersonal attitudes. J . pers. soc. Psychol., 1966, J3, !  265-270. Rodrigues, A.  On the d i f f e r e n t i a l effects of some parameters of balance.  J. Psychol., 1965, Rodrigues, A.  61, 241-250.  Tlie psycho-logic of interpersonal r e l a t i o n s .  doctoral d i s s e r t a t i o n . Univ. of C a l i f o r n i a , Los Angeles, Siegel, S.  Non-parametric s t a t i s t i c s .  Singer, J . E.  Motivation for consistency.  consistency: New York:  Academic Press,  s o c i a l perception. s  1966.  McGraw-Hill,  1956.  In S. Feldman (Ed.), Cog  itive  motivational antecedents and behavioral consequences.  Storm, T., and Knox, R. E.  Storm, T.  New York:  Unpublished  1966.  Cross-cultural study of the development of  Unpublished,  and Knox, R. E.  1967.  A c r o s s - c u l t u r a l study of balance  and  97  transitivity.  Unpublished, 1969.  Steiner, I. D., and Spaulding, J . Preference f o r balanced s i t u a t i o n s . Report No. 1 Dept. of Psychol. Univ. of I l l i n o i s , 1966. Walster, E., Berscheid, E., and Barclay, A. M. among modes of dissonance reduction.  A determinant of preference  J . pers. soc. Psychol., 1967, _7_,  211-216. Winer, B. J . S t a t i s t i c a l p r i n c i p l e s i n experimental design.  New York:  McGraw-Hill, 1962. Zajonc, R. B.  Cognitive theories of s o c i a l behaviour.  In G. Lindzey and  B. Aronson (Eds.), Handbook of s o c i a l psychology (Rev. ed.), Cambridge: Addison-Wesley, 1968. Zajonc, R., and Burnstein, E. s o c i a l structures.  The learning of balanced and unbalanced  J . Pers., 1965, 33, 153-163 ( a ) .  Zajonc, R., and Burnstein, E.  Structural balance, r e c i p r o c i t y , and  p o s i t i v i t y as sources of cognitive bias.  J . Pers., 1965, 33, 570-583 (b).  APPENDIX 1 PILOT STUDIES  99  Two p i l o t studies were conducted as f e a s i b i l i t y tests f o r the main study and provided  vehicles f o r developing a suitable procedure f o r young  children. The f i r s t study required pleasantness ratings from subjects aged 4-6 years ( n = l l ) .  Subjects were children of E/s friends and r e l a t i v e s .  The children were tested i n t h e i r homes i n two sessions.  In the f i r s t  session, they were shown how to use the rating scales and then rated two sets of P-O-X structures f o r pleasantness* each set, representing  there were 8 structures i n  the 8 triads shown i n Figure 6.  presented i n p i c t o r i a l form on a separate 3" x 5" card.  Each structure was Stick figures were  used to indicate the two persons P and 0, i n each s i t u a t i o n .  A box was  used to symbolize X, "something very, very important to both c h i l d r e n . " A f f e c t i v e r e l a t i o n s between elements were indicated by red hearts and blue crosses  (dislike).  (like)  The d i r e c t i o n of r e l a t i o n s between elements  was indicated by black arrows.  The subject was instructed to assume the  r o l e of P i n each structure i n the f i r s t set.  He was asked to rate  structures involving two other children i n the second set.  Ratings were  made on a 15-Rpoint s c a l e , 6% feet i n length, anchored at the high end by a smiling face and at the low end by a frowning face. returned  One week l a t e r , JE  to the child's home and asked him to r e c a l l the contents of the  cards and to describe the rating procedure. P i l o t Study II was conducted at the Vancouver Talmud Torah, a Hebrew Day School.  In this, study, children i n grades 2, 4, and 6 (n=16 i n each  grade) rated the two sets of P-0TX structures used i n P i l o t Study I . They rated the structures f i r s t f o r pleasantness and then consistency.  Con-  sistency ratings were made on a 15-polnt scale i d e n t i c a l to that used  100  for pleasantness  ratings i n P i l o t Study I except that t h i s scale was anchored  at the high end by a large red S indicating that situations placed at this end made " l o t s of sense."  At the low end of the scale there was a large  red S covered by a black cross indicating that situations made "very sense."  Testing was conducted i n one session of approximately  duration, prefaced by a training period. during the session, two f o r pleasantness consistency.  little  30 minutes  The subject made four ratings (test and retest) and two f o r  A f i v e minute rest was given between the pleasantness and  consistency rating tasks. In the second p i l o t study, an attempt was made to e s t a b l i s h unit relations between the two persons i n each P-O-X s i t u a t i o n .  This was  accomplished by presenting the various structures i n the following contexts: Format 1: The s t o r i e s that I'm going to show you are about two school children. At the beginning of the year, the teacher told these two children that they were to be partners. They were to help each other with their school work, go to swimming together, work together on s p e c i a l projects, etc. You are one of these c h i l d r e n — t h e one with the black c i r c l e around him (her). Format 2: This time I want you to pretend that you and the other c h i l d are i n I s r a e l . You are staying at a Kibbutz f o r the summer. You and the other c h i l d are the only Canadians on the Kibbutz. You and the other c h i l d sleep i n the same room. You are the child with the black c i r c l e around him (her); the other c h i l d i s your roottcjate. Format 1 was used to present structures i n the test s e r i e s . was used to present the retest s e r i e s .  Format 2  The subject was instructed to view  himself as a participant i n a l l s i t u a t i o n s . The results of the r e c a l l session i n P i l o t Study I indicated that  101  subjects understood the basic elements of the situations and the procedure involved i n making ratings.  Observation of the subjects' behaviour i n  both sessions, however, suggested certain changes i n the training procedure and i n the wording of i n s t r u c t i o n s . training procedure the c h i l d was  For example, at one point i n the  required to s e r i a t e 6-size graded blocks  on a 6-point scale, V% feet i n length. scale p o s i t i o n 1 and a large one was These squares proved confusing  A small square was  drawn below  drawn below scale p o s i t i o n 6.  to the children.  They placed the smallest  block on the small square and the largest block on the large square but were not sure how  to proceed thereafter.  A scale containing 6 squares of  increasing s i z e , placed above the scale numbers, was proved more successful.  rating scales i n the principal^study, i n order  In both p i l o t studies subjects sometimes placed  structure cards half-way between two scale points.  IS had  subject i n order to determine which scale point was  intended.  Further procedural  It  Squares of increasing s i z e were also added to the  pleasantness and consistency to f a c i l i t a t e scoring.  substituted.  to question  the  changes derive s p e c i f i c a l l y from P i l o t Study I I .  Verbal comments by subjects i n P i l o t Study II indicated, f o r example, that retest ratings may  have been influenced to some extent by memory of ratings  made i n the test s e r i e s .  Some form of a c t i v i t y should have been i n t e r -  polated between the test and retest s e r i e s .  Observation-of subjects'  behaviour during P i l o t Study II indicated that motivation toward the end of the session.  This was  tended to wane  p a r t i a l l y a l l e v i a t e d by  shorten-  ing the training procedure and by changing the wording of i n s t r u c t i o n s . Discussion with subjects at the conclusion of the study suggested, however, that a further reduction i n the t o t a l duration of the t r a i n i n g and testing  102  session was necessary.  I t also became apparent during the course of  P i l o t Study II that the order i n which formats were presented should have been varied, as should the order i n which subjects rated for pleasantness and consistency. Ratings obtained i n P i l o t Study I were analyzed according to Zajonc's (1968) method. 1.53,  Index values for a t t r a c t i o n , agreement, and balance were  .99, and 1.11  respectively, indicating that a t t r a c t i o n was the more  important determinant of subjects' pleasantness ratings.  The same result  was obtained i n P i l o t Study II.. As shown i n Table 21, Zajonc indices f o r  Table 21  a t t r a c t i o n exceed those for agreement and balance i n a l l age groups, when computed from pleasantness ratings.  In the case of consistency ratings,  the index value for a t t r a c t i o n was larger than that for agreement or balance in the 2nd grade group; values for agreement were of greatest magnitude among 4th and 6th grade children.  103  Table  1:  Zajonc indices f o r P i l o t Study II  p and o disagree PL 0 PL~0  Attraction  2  12.08  4.34  11.75  3.54  1,0,7  3.03  .97  4  12.88  4.79  10.82  5.43  1.09  2.32  1.17  6  13.36  6.01  11.60  5,91  1.11  2.09  1.09  2  12.88  5.48  9,22  4.25  1.36  2.27  1.17  4  12.66  7.77  6.86  7.46  1,43  1.28  1.38  6  12.66  9.11  3.54  8.02  1.31  1.24  1.17  a) Pleasantness ratings Grade  Balance  p and o agree PL 0 PL 0  Agreement  Effect Ratio  b) Consistency ratings Grade  Grade 2 (n=16), Age range 91-107 months; X age=95 months (7.9 years) Grade 4 (n=16), Age range 114-122 months; X age=118 months (9.8 years) Grade 6 (n=16), 4ge range 136-146 months; X age=i42 months (11.8 years)  APPENDIX 2 DIAGRAMMATIC REPRESENTATION OF THE STRUCTURES INCLUDED IN LEVEL 1 AND LEVEL 2 OF FACTOR 0  L e v e l 1 of Factor O  L e v e l 2 of Factor O  (8)  NOTE:  (7)  Structures 1,2,8, and 7 a r e balanced; the r e m a i n d e r are imbalanced.  APPENDIX 3 MEAN PLEASANTNESS AND CONSISTENCY RATINGS FOR EACH STRUCTURE (BY AGE)  107  ••'fable 22:  Mean pleasantness and consistency ratings for each structure.  Structure No.  Age  Gp.  Pleasantness  Consistency  1  5-6 7-8 3-10 11-12  14.63 15.00 14.83 14.78  14.20 14.45 14.93 14.98  2  5-6 7-8 9-10 11-12  1C.85 9.33 12.08 10.53  8.00 9.15 12.03 11.23  3  5-6 7-8 9-10 11-12  6.75 6.50 8.08 6.98  7.05 6.60 8.10 5.90  4  5-6 7-8 9-10 11-12  3.00 1.53 4.25 3.20  2.38 3.08 7.20 6.25  5  5-6 7-8 9-10 11-12  11.50 10.65 11.83 9.48  9.10 9.60 9.88 9.28  6  5-6 7-8 9-10 11-12  11.25 12.15 12.15 9.73  9.18 9.50 11.03 8.95  7  5-6 7-8 9-10 11-12  6.15 5.00 6.85 5.10  4.38 5.28 5.93 5.43  8  5-6 7-8 9-10 11-12  5.68 5.23 5.88 5.30  4.25 4.63 6.15 5.53  The values i n each c e l l of the above table represent an average of ratings made under formats 1 and 2. A high value indicates that the structure was perceived to be pleasant (or consistent); low values indicate that on the average, Ss perceived the structure to be unpleasant (or inconsistent).  APPENDIX 4 MULTIPLE REGRESSION ANALYSIS OF THE RATINGS  109  (a)  The p r i n c i p a l study In order to rule out the p o s s i b i l i t y that the findings r e l a t i v e to  hypotheses  2 and 3 were an a r t i f a c t of the method of data analysis,  pleasantness and consistency ratings were also subjected to multiple regression analysis.  In this analysis, multiple regression equations  i n standard score form were computed f o r each subject. balance, agreement, and a t t r a c t i o n . criteria.  Predictors were  The subject's own ratings served as  Beta c o e f f i c i e n t s were then averaged for each age group.  average c o e f f i c i e n t s are shown i n Table 23.  These  Inspection of this table  indicates that average beta weights for a t t r a c t i o n exceed those f o r balance  Table 23  and agreement i n a l l groups, both for pleasantness and for consistency. Average beta weights for balance are greater than those for agreement i n the pleasantness equations but i n the consistency equations i n a l l groups but 7-8, average beta weights for agreement exceed those for balance. Coefficients of concordance were calculated i n order to estimate the amount of agreement among subjects i n each age group i n their weighting of the three predictors.  The procedure, described by Siegel ( 1 9 5 6 ) , involves  ranking the beta c o e f f i c i e n t s according to magnitude.  As shown i n Table 24,  W values are s i g n i f i c a n t (P<.01) at a l l age levels for both pleasantness and consistency.  S i g n i f i c a n t values of W may be interpreted as meaning  Table 24  that subjects within a p a r t i c u l a r age group tend to use the three  110  Table 23:  Average regression equations f o r pleasantness and consistency.  Pleasantness  Consistancy  Age  Agreement  Attraction  Balance  Agreement  Attraction  Balance  5-6  .0051  .6457  .1434  .1048  .5413  .0977  7-8  -.0163  .7116  .1010  .1196  .6297  .1257  9-10  .0122  .8167  .1661  .1829  .6543  .1333  11-12  .1665  .65S5  .1784  .2414  .5953  .1955  Table 24: Kendall Coefficients of Concordance  Pleasantness Age Group  Consistency W  S_  5-6  .593  474.00**  .552  451.50**  7-8  .693  554.17**  .483  386.00**  9-10  .773  618.00**  .610  488.00**  11-12  .438  350.00**  .333  266.00**  *p<.05 **p<.01  112  predictors i n s i m i l a r fashion. The frequency of significance of each predictor i s summarized i n Table 25. A t t r a c t i o n was a s i g n i f i c a n t component of pleasantness and consistency ratings f o r the majority of subjects i n a l l groups.  From  70-100% of subjects i n each group assigned p o s i t i v e weighting to a t t r a c t i o n .  Table 25  P o s i t i v e weighting means that i f P and 0 l i k e each other, there i s a tendency of varying i n t e n s i t y , but independent of other predictors, to rate the s i t u a t i o n highly pleasant (or highly consistent). implies the reverse meaning.  Negative weighting  A smaller proportion of subjects (5-40%) i n  each age group also assigned s i g n i f i c a n t weighting to balance and/or agreement.  There i s some i n d i c a t i o n that use of agreement and balance as  a basis for rating increases with increasing age. Overall, the results of the regression analyses lead to the same conclusions with regard to hypotheses and the Zajonc indices.  2 and 3 as the analysis of variance  That i s , the results f a i l to support either  hypothesis. (b)  The cross-validation study Average regression equations computed from ratings obtained i n the  cross-validation study are presented i n Table 26.  The r e l a t i v e weighting  Table 26  of predictors i n these equations i s the same as i n the p r i n c i p a l study  113  Table 2?8  No. of J3s f o r whom predictors are s i g n i f i c a n t components of pleasantness and/or consistency ratings. (n=20 i n each age group.)  Pleasantness Balance Age Gp  +  Agreement  -  +  —  Consistency Attraction  +  —  Balance  +  —  Agreement  Attraction  +  —  +  —  5-6  5  0  0  0  14  0  2  0  2  0  14  0  7-8  4  0  0  1  20  0  2  0  2  0  17  0  9-10  6  0  2  1  20  0  2  0  6  0  18  0  11-12  3  0  5  0  17  0  5  0  8  0  16  0  114 Cross V a l i d a t i o n Study Table  Average regression equations f o r pleasantness and consistency.  Pleasantness  Consistency Attraction  Balance  Balance  Agreement  .4394  .0604  .1446  .4085  .0897  .2160  .5447  .1593  .1872  .4064  .0618  9-10  .1741  .7640  .1132  .2088  .5741  .1165  11-12  .3146  .5412  .2194  .3273  .5855  .1421  Age  Agreement  5-6  .0913  7-8  Attraction  115  i n the case of consistency. In the regression equations for pleasantness, the r e l a t i v e weighting of balance and agreement i s the reverse of that i n the p r i n c i p a l study ( i . e . , beta weights for agreement exceed those for balance).  As mentioned  i n Chapter VI  however, a subject by subject  comparison indicated that beta weights f o r agreement were of greater magnitude than those f o r balance i n only 50% of the subjects.  Thus, the  results of the cross-validation study also f a i l to support hypotheses 2 and 3.  APPENDIX 5 SUMMARY OF ANALYSIS OF VARIANCE OF PLEASANTNESS, CONSISTENCY, AND TENSION RATINGS OBTAINED WITH ADULT SUBJECTS (GUTMAN,  1969)  117  Table 27a:  Summary of analysis of variance of pleasantness ratings.  Source of Variation  df  MS  F  A (Agreement) B (Attraction) AB (Balance) C (Format) AC EC ABC 0 (Structures) AO BO ABO CO ACO BCO ABCO S AS BS ABS CS ACS BCS ABCS OS AOS BCS ABOS COS ACOS BCOS Error  1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 83 83 83 S3 83 83 83 83 83 83 83 83 83 83 83 83  1629.80 2946.50 98.58 .50 .11 .36 3.05 97.50 56.68 13.36 6.30 1.86 .76 9.00 .43 6.60 8.46 11.07 2.72 1.28 1.32 1.31 1.52 1.59 4.92 1.53 1.40  1317.19** 2381.39** 79.68** .41 .09 .29 2.46 78.80** 45.81** 10.80** 5.09* 1.50 .62 7.28** .35 5.33** 6.84** 8.94** 2.20** 1.04 1.07 1.08 1.23 1.29 3.98** 1.24 1.13 .75 .79 .97  Total *p<.05 **p<.01  1343  .92  .97 1.19 1.24  % of v a r i a b i l i t y due to significant main e f f e c t s and inter actions 18.41 33.28 1.11  1.10 .64 .07 .10 6 C 7.93 10.37 2.55  4.61  118  Table  27b; Summary of analysis of variance of consistency ratings.  Source of Variation A (Agreement) B (Attraction) AB (Balance) C (Format) AC BC ABC 0 (Structures) AO BO ABO CO ACO BCO ABCO S AS BS ABS CS ACS BCS ABCS OS AOS BOS ABOS COS ACOS BCOS Error  Total *p<.05 **p<.01  df  MS  83 83 83 83 83 83 83 83 83 83 83 83 83 83 83 83  240.89 718.97 1041.30 .81 .63 .17 1.94 32.50 28.88 57.92 56.27 .09 11.63 2.77 4.41 10.09 7.09 8.82 11.61 2.76 1.77 1.64 1.92 2.29 3.23 3.32 2.56 2.37 2.83 1.90 2.65  1343  F 91.03** 271.68** 393.47** .31 .24 .06 .73 12.28** 10.91** 21.89** 21.26** .03 4.39* 1.05 1.67 3.81** 2.68** 3.33** 4.39** 1.04 .67 .62 .73 .87 1.22 1.25  .97  .90 1.07 .72  % of v a r i a b i l i t y due to significant main effects and i n t e r actions 3.11 9.28 13.44  .42 .37 .75 .73 .15 10.81 7.59 9.45 12.43  119  Table  27c: Summary of  -.ruilysi3  of variance of tension ratings. >  Source of Variation A B AB C AC BC ABC  df  1 1 1 1 1 1  AO BO ABO CO ACO BCO ABCO S AS BS ABS CS ACS BCS ABCS OS AOS BOS ABOS COS ACOS BCOS Error  1 1 1 1 1 1 1 1 1 83 83 83 83 83 83 83 83 83 83 83 83 83 83 83 83  Total  1343  0  MS  2160.40 2731.40 46.50 .19 2.50 .36 3.86 35.36 11.44 15.43 4.53 .96 .00 .03 7.44 9.93 8.97 8.62 2.18 1.83 1.12 1.96 .83 2.30 3.83 1.36 1.74 1.41 1.22 .81 1.32  F  1638.56** 2071,65** 35.27** .14 1.90 .27 2.93 26.32** 8.68** 11.70** 3.43 .73 .00 .02 5.64* 7.53** 6.80** . 6.54** 1.65* 1.39* .90 1.49* .63 1.74** 2.91** 1.03 1.32 1.07 .93 .61  % of v a r i a b i l i t y due to significant main effects and i n t e r actions  23.67 29.93 .51  .39 .13 .17  .08 9.03 8.15 7.83 1.98 1.78 2.09 3.48  

Cite

Citation Scheme:

    

Usage Statistics

Country Views Downloads
United States 5 0
Japan 3 0
China 3 0
Germany 1 2
City Views Downloads
Ashburn 4 0
Beijing 3 0
Tokyo 3 0
Unknown 1 2
Mountain View 1 0

{[{ mDataHeader[type] }]} {[{ month[type] }]} {[{ tData[type] }]}
Download Stats

Share

Embed

Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                        
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            src="{[{embed.src}]}"
                            data-item="{[{embed.item}]}"
                            data-collection="{[{embed.collection}]}"
                            data-metadata="{[{embed.showMetadata}]}"
                            data-width="{[{embed.width}]}"
                            async >
                            </script>
                            </div>
                        
                    
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:
http://iiif.library.ubc.ca/presentation/dsp.831.1-0093329/manifest

Comment

Related Items