Free Play Behaviors in Middle- and Lower-Class Preschoolers: Parten and Piaget Revisited

Abstract

and Lower-Class Preschoolers: Parten and Piaget Revisited. CHILD DEVELOPMENT, 1976, 47, 414-419. 24middle-class and 16 lower-class preschoolers were observed during free play period on 30 consecutiveschool days. The behaviors of the children were coded by cotnbining the social play categories of Partenand the cognitive play schemes of Sniilansky. The results indicated that middle-class preschoolers engagedin significantly less parallel and functional play, and significantly more associative, cooperative, andconstructive play than did their lower-class age mates. Moreover, examination of the combined Parten-Smilansky play scale revealed that middle-class children emitted significantly more associative-constructive and cooperative-dramatic play and significantly less solitary-functional and parallel-functionalplay than lower-class preschoolers. Results also showed that males exhibited significantly more dramatic,solitary functional, and associative-dramatic, and less constructive, solitary-constructive, and parallel-constructive play than their female counterparts. It was suggested that future studies incorporate both theParten and the Smilansky categories of play in investigating the behaviors of preschoolers in a wide varietyof settings.Early observational investigations of chil-dren's free play preferences often focused uponthe formulation of social play hierarchies. Thus, ina now classic study, Parten (1932) discovered thatsocial participation among preschoolers increasedwith the child's age. Parten defined six sequentialsocial participation categories: unoccupied be-havior, solitary play, onlooker behavior, parallelplay, associative play, and cooperative play. Pre-schoolers' modal play preference from 2 to 2%years was solitary play, from 2V4 to 3V4 years wasparallel play, and frorii 3V4 to 4V4 years was associa-tive play.While Parten's social play hierarchy has gen-erally gained acceptance as a measure of thematurity of early peer interaction, it is noteworthythat in the 40 years following her initial investiga-tion, few researchers have attempted to eitherreplicate or extend her original findings. The onerecent replication study (Barnes 1971) revealedthat 3- and 4-year-olds displayed significantlymore unoccupied, solitary, and onlooker activityand significantly less associative and cooperativeplay than Parten found in her original sample.Thus, Barnes has suggested that considerable cau-tion be exercised in using Parten's data in a norma-tive fashion.concerning children's play behaviors stemmedfrom Piaget's (1962) classification of three succes-sive stages according to the degree to which playremains purely sensorimotor or has some bearingon thought itself. Smilansky (1968) elaboratedupon the original Piaget categories and labeledthem as follows: (a) functional play—simple re-petitive muscle movements with or without ob-jects; (b) constructive play—manipulation of ob-jects to construct or to create something; (c)dramatic play—the substitution of an imaginarysituation to satisfy the child's personal wishes andneeds; and (d)games with rules—the acceptance ofprearranged rules and the adjustment to theserules. The four types of play are thought of asdeveloping in a relatively fixed sequence withfunctional play appearing ontogenetically first ininfancy and games with rules last (during concreteoperations). In one recent study, Rubin and Ma-ioni (1975) found that preschoolers engage in sig-nificantly more functional and constructive playthan in either dramatic play or games with rules.Moreover, in support of their belief that play pref-erences are related to cognitive competence, theincidence of dramatic play was found to be posi-tively correlated with spatial relational and clas-sification skills. A negative correlation was foundbetween functional play and the same cognitiveA second major early source of information measures.We would like to express our appreciation to the teachers at the University of Waterloo EarlyChildhood Education Centre for their kind cooperation. We would also like to thank Becky Hansen for herassistance in the data analysis. Requests for reprints should be sent to Kenneth H. Rubin, Department ofPsychology, University of Waterloo, Waterloo, Ontario.