Mother-Child Play: Sequential Interactions and the Relation between Maternal Beliefs and Behaviors

 
CONTINUE READING
Mother-Child Play: Sequential Interactions and
the Relation between Maternal Beliefs and
Behaviors

Amy Melstein Damast
Albert Ehistein College of Medicine

Catherine S. Tamis-LeMonda
New York University

Marc H. Bornstein
National Institute of Child Health and Human Development

      DAMAST, AMY MELSTEIX. T.\Mrs-LEMoxDA, CATHERINE S., and BOHNSTEIX, MARC H .              Mother-Child
      Play: Sequential Interactions and the Relation between Maternal Beliefs and Rehaviors. CHILD
      DEVELOPMENT, 1996, 67, 17.52-1766. This investigation of mother and toddler play had 2 goals.
      The primary goal was to examine the t>'pes of play mothers introduce in direct response to their
      tpddlers' play. A secondary' and exploratory goal was to examine the relation between maternal
      knowledge about child play and actual maternal play behaviors. 50 mothers and their 21-month-
      old toddlers were observed at home duimg free play. Mother and child explorator>', nonsymbolic,
      and symbolic play were coded. Sequential analyses revealed that mothers adjusted their play to
      their children's play level by responding to their children with play that was either at the same
      level or at a higher level than their children's play. Furthermore, mothers who were more
      knowledgeable about early play development more often responded to their children's play by
      introducing higher level play. These flndings suggest that mothers tend to play with their tod-
      dlers in ways that might promote their child's development, and that mothers with more knowl-
      edge about play development provide their children with appropriately challenging play interac-
      tions.

     In Western cultures, play often occurs dren's play? Second, if mothers do respond
in a social setting (Haight & Miller, 1993). differentially to their children's play, does
In the child's first few years, when new cog- mothers' knowledge about early play devel-
nitive and social skills are developing, a opment relate to mothers' actual play with
more knowledgeable social partner, like their children?
mother, is thought to facilitate the develop-
ment of more sophisticated play. Two re-          In the first year, children learn about ob-
lated questions about maternal play behav- jects m their environment through explora-
iors were investigated in the present study, tion—they mouth, look at, and manipulate
First, are mothers sensitive to their chil- them (e.g., Belsky & Most, 1981; Bornstein
dren's play abilities, adjusting their own & Lamb, 1992; Fein, 1981; Nicolich, 1977;
play as a function ofthe level of their chil- Piaget, 1962). Near the end ofthe first year.

            This article is based on a dissertation submitted by the flrst author m partial fulfillment of
      the requirements for the doctoral degree in the Department of Psycholog\' at New York Univer-
      sity. A.M.D. was supported by an N.Y.U. Predoctoral Fellowship; by research grants HD20559,
      HD20807, and MH48915, by an IKTA Fellowship from the National Institute of Child Health
      and Human Development, and by post-doctoral fellowship grant T32HD07384 C T.-L was
      supported by research grants HD20559. HD20807, and MH48915. M.H.B. was supported by
      research grants HD20559 and HD20807. and by a Research Career Development Award
      HD00521 from the National Institute of Child Health and Human Development. A.M.D. thanks
      the members of her dissertation committee, D. Ruble, D. Frye, H. Oster, and J. Daws, for
      their helpful comments and suggestions. We also thank L. Baumwell, L. Cyphers, H. Comes, J.
      Jankowski, R. Kahana-Kalman, S. Kazas, J. McClure, S. Rose, H. Ruff, and C Schmidt for their
      assistance. Requests for reprints should be addressed to Amy Melstem Damast or Catherine
      Tamis-LeMonda, Department of Applied Psychology', New York Universitv, 239 Creene Street-
      5th Floor, New York, NY 10003.
           [Child Development. 1996.67,1752-1766. © 1996 by the Society- for Research in Child Development, Inc.
           All rights resented. 6o09-3920/96/6704-0026S01 00]
Damast, Tamis-LeMonda, and Bornstein                     1753

children begin to engage in nonsymbolic            first and second years (Belsky, Goode, &
play as they examine the unique qualities          Most, 1980), as well as greater nonsymbolic
and functions of objects (e.g., pushing but-       play (and less simple exploration) during
tons on toy phones, stacking blocks, nesting       the toddler years (Fiese, 1990). Maternal
cups). During the second year, play more of-       involvement and turn-taking behaviors are
ten takes on a nonliteral "as if" quality, as      negatively associated with simple explora-
children incorporate difFerent forms of sym-       tion during toddlerhood (Fiese, 1990). In
bolic play into their repertoires (see Belsky      addition, Tamis-LeMonda and Bornstein
& Most, 1981; Bornstein & O'Reilly, 1993;          (1991) found that mothers' own play behav-
Fein, 1981; Fenson, Kagan, Kearsley, & Zel-        iors (i.e., gestural and verbal solicitations
azo, 1976; Fenson & Ramsev, 1980;                  and demonstrations of play activities) with
McCune-Nicolich, 1981; Nicohch, 1977;              their children relate to their children's play
 Schaefer, Gitlin, & Sandgrund, 1991; Tamis-       behaviors: Mothers who engage in more
LeMotida & Bornstein, 1995; Tamis-                 symbolic play with their 13- and 20-month-
 LeMonda, Damast, & Bornstein, 1994; Un-            oids have toddlers who engage in more sym-
 gerer, Zelazo, Kearsley, & O'Leary, 1981;         bolic play at these ages.
 Watson & Fischer, 1977).
                                                         The literature on maternal respon-
     \'\'hiile children's play increases in so-    siveness and child cognitive development
phistication with age, at any one age chil-        implies that contingency in maternal inter-
dren exhibit different levels of play in differ-   active behaviors might also be central to
ent social contexts (e.g., Fiese, 1990;            children's play behaviors (e.g., Bakeman,
O'Connell & Bretherton, 1984). One possi-          Adamson, Brown, & Eldridge, 1989; Baum-
ble explanation for this obsen'ation can be        well. Tamis-LeMonda, & Bornstein, 1996;
found in Vygotsky's theory of the zone of          Beckwith & Cohen, 1989; Bornstein, 1989;
proximal development Vygotsky (1978) sug-          Bornstein & Tamis-LeMonda, 1989; Born-
gested that children may function between          stein, Tamis-LeMonda, et al., 1992; Olson,
two levels of development—their actual de-         Bates, & Bayles, 1984; Olson, Bayles, &
velopmental level and their potential devel-       Bates, 1986). This literature demonstrates
opmental level (i.e., the level at which they      that it is not merely the Frequency of positive
can function iii collaboration with a more ex-     maternal interactive behaviors that Facili-
perienced partner). Vygotsky and others            tates children's development, but rather
(e.g., RogofF, 1990; RogofF & Wertsch, 1984;       both the contingency (i.e., the temporal and
Wertsch, 1984, 1985; Zukow, 1986) suggest           sequential relation) and the appropriateness
that more advanced partners structure inter-        (i.e., the content, sophistication, or topic) oF
actions that exceed children's actual level         mothers' behaviors in the context oF the on-
and approach children's potential level,            going interaction. In other words, maternal
thereby helping to advance children's actual       behaviors have a positive influence on chil-
level of development. For example, O'Con-           dren's development when those behaviors
nell and Bretherton (1984) demonstrated             are both appropriate to and contingent on
increases in children's play diversity (a           children's behaviors.
measure fqund to relate to later play
sophistication and intelligence) during col-
labora;tive play with mother as compared to             With respect to the appropriateness oF
solitary play. Similarly, Fiese (1990) showed      maternal play behaviors, developmental
tl^at toddlers spent a greater percentage of       studies suggest tliat the level and the Fre-
their time in nonsjTnbolic and symbolic            quency oF mothers' play relate to their chil-
play, and a lower percentage of their time in      dren's age and ability (Belsky et al., 1980;
exploratory piay, during mother-child play         Brooks-Gunn & Lewis, 1984^ Fiese, 1990;
than during play alone.                            Tamis-LeMonda & Bornstein, 1991). As a
                                                   group, mothers tend to suggest more sophis-
     O'Connell and Bretherton {1984) re-           ticated play to children when their children
ported, that it is not simply a mother's pres-     are older and more capable oF this type oF
ence, but her active participation, that ac-       play (Tamis-LeMonda & Bornstein, 1991).
counts for differences in children's play          In addition, mothers tend to support their
across social contexts. Maternal physical at-      children's autonomy by offering fewer play
tention-focusing behaviors during interac-         suggestions as children grow older (Belsky
tion have been found to predict greater ex-        et al., 1980; Fiese, 1990). Thus, it appears
ploratory competence (i.e., a weighted             that mothers are generally sensitive to their
composite measure of exploration, nonsym-          children's play abilities in that they change
bolic play, and symbolic play) during the          their behaviors appropriately with child age.
1754      Child Development
     Little is known, however, about the con-         eventful perinatal histories and were
tingency (i.e., responsiveness) of maternal           healthy and free of any known develop-
play to children's play. Are mothers ad-              mental delays during the course ofthe study.
justing their play responsively to the indi-          They were all from middle- and upper-
vidual play behaviors of their children (i.e.,        middle-class intact families (M = 57.7, SD
on an episode-by-episode basis)? The first            = 7.6, on the Hollingshead Four Factor In-
goal of this investigation was to address this        dex of Social Status, 1975; Gottfried, 1985),
Qiiestion. Our expectation was that mothers           with 90% of mothers hiiving completed 4 or
woiild respond appropriately and contin-              more years of college. At the 21-month visit,
gently to their children's play on an episode-        45 of the 50 were only children, one had a.
by-episode basis.                                     lO-month-old sibling, and four had a sibling
                                                      younger than 4months.
     Within the context of group sensitivity
to child play behaviors, we expected to find          Play at Home
variability in the extent to which individual              Procedures.—Dyads were visited m
mothers adjust their play to their children's         their homjas at a time convenient to mother
play. The second, exploratory goal ofthis in-         and child and when only the two would be
vestigation was, therefore, to examine                present. Data from a 10-min mother-child
whether rnothers' ability to respond sensi-           free-play session, part ofa larger home visit,
tively on an episode-by-episode basis to              are the focus ofthe present report.^ Mother
their children's play is, at least in part, ex-       and child were filmed while playing with a
plained by their knowledge ofthe progres-             standard set of toys that included cups,
sive nature of children's play. It has been           plates, spoons, teapot and cover, doll, baby
suggested that mothers' knowledge about               blanket, baby battle, bus with removable
eertain developing abilities in children re-          peopie, sponge, telephone, blocks, and a set
lates to their behaviors with their children          of nesti:ng cups. The toy set represented ob-
(e,g., Goodnow, 1988; Goodnow & Collins,              jects With whieh all children were familiar
1990; Hunt & Paraskevopoulos. lOSO; Ko-               in their everyday routines and which lend
phanska, 198|0; McGltlicuddy-DeLisi, 1982;            tiiemselves to a variety of age-appropriate
Miller, 198S'. Sigel, McGiliicuddy-DeLisi, &          actMties. Each mother was instructed to sit
Goodnow, 1992; Tamis-LeMonda et al.,                  with her child and behave as she normally
1!J94). In a comprehensive review of paren-           would when she and her child played. In
tal knowledge of development. Miller (1988)           additioii, mothers were asked to ignore the
note^ that parents are gene .rally accurate in        expenmenter''s presence to the extent possi-
their knowledge of developmental mile-                ble. Mothers were told that they could use
stones, and tliat variatioi:i among parents in        any or all of the toys provided, but that the
the aepuraey of their knowledge is pre-               chiM's own ttoys should not be included in
dictive (albeit modestly) of their behaviors          die play session.
wifh their children. Accordingly, we as-
sessed the a^ccuracy of mothers' knowledge                  Coding.—Mother play and child play
ofthe relative difficultj,' of'vrarious chilcl play   were fioded from Tiddeotapes using an object-
acts and asked whether this knowledige ac-            centered, event-sampling procedure. Both
pounts for variation among mothers in their           partaers' play was coded using; the mutually
responsive play behaviors.                            eiclusive codes of exploration, nonsymbolic
                                                      play, and symbolic play; during times when
                                                      chiiireri were not engaged in one of these
Method                                                three types of play for 2 or more seconds,
                                                      theywere assigned a code af off-task. Mater-
Participants                                          nal play could be verbal (e.g., saying "drink
     Fifty mothers (M age = 33 years, SD =            some tea"), gestural (e.g., pretending to pour
3.4) and their firstborn, 21-month-oId chil-          from a teapot into a cup and handing the cup
dren (Mage = 651 days, SD = 8.5; 26 males,            to tlae qhild to drink), or a combination ofthe
24 females) participated in this study. Fami-                (eig., saying "let's drink tea" and pre-
lies were recruited from a pediatric unit in               l      to (feink from a teaeup). The onset
a large urban university hospital and from                              i occurred with the start ofa
Ioeal obstebic groups. All children had un-

            1 The entire home visit consisted of four videotaped sessions occumng in a standardized
       order- 10 min ofthe child playing alone with a standard set of toys, 10 mm of mother-child tree
       play with another set of toys, 10-15 min of experimenter-child play using a third set of toys, and
       45 min of naturalistic interaction (during which mdtfeers were free to do whatever they pleased
       with or without their childien).
Damast, Tamis-LeMonda, and Bornstein                    1755
mother's vocalization (e.g., "Call daddy on       and mothers' play separately: percent agree-
the phone") in the case of verbal or verbal-      ment was 84% for both child and mother
gestural play, or with the start of her gesture   play events. For play acts on which both
(e.g., handing an object to her child, pointing   coders agreed, reliabilities were calculated
to an object) or demonstration (e.g., pushing     for the level of child play (i.e., for all ehild
buttons on a telephone) in the case of ges-       acts, whether or not mother responded), the
tural play. Children's play was always based      level of maternal play, and the level of the
on gestures accompanied by visual regard          child act that preceded maternal play in the
and could occur with or without vocaliza-         sequence (i.e., only child acts to which
tions. For example, if a child pretended to       mother responded). Cohen's kappa was used
drink from a teacup, the child's action was       as a conservative measure of reliability
coded as a single, self-directed symbolic act,    (Bakeman & Gottman, 1986). Kappas were
whether or not the child said "drink tea."        good to excellent for each code: K = .67 for
Children's play was coded from the onset of       child play level (range = .62 to .76), K = .77
the child's regard. Children's voadizations       for maternal play level (range = .58 to .95),
in the absence of gestures were not coded         and K. = .66 for preceding child level (range
as play to avoid confounding symbolic play         = .54 to .73).
ability with language ability.
                                                  Maternal Knowledge of Play
     For both mother and child, play events             Mothers' knowledge of child play was
terminated when the player looked at an-           assessed using a questionnaire mailed to
other object, engaged in a thematically unre-     them 1 week prior to the home visit and col-
lated action vvith the same object, or tumed       lected by the experimenter at the comple-
attention from the toys altogether for a pe-      tion of the visit. The play activities on the
riod of 2 or more seconds. As mentioned,           questionnaire represented one example of
when a child directed attention away from          each ofthe levels of a hypothesized 24-level
the toys for 2 or more seconds, the behavior       play scale (Tamis-LeMonda et al., 1994; Ap-
was coded as off-task.                             pendix A). This 24-level play scale was de-
     Based on the above mother and child          veloped by integrating previousfindingsex-
play codes, sequences of play acts were ob-       tant in the literature in order to create a more
tained by assigning each maternal play act a       comprehensive description of early develop-
code indicating the child's activity' prior to    mental progressions in play (see Belsky &
or at the onset of the act. For example, if        Most, 1981; Fenson et al., 1976; Fenson &
mother s^id "talk on the phone" after the         Ramsey, 1980; McCune-Nicolich, 1981; Ni-
child manipuiated the phone, the mother's         colich, 1977; Tamis-LeMonda & Bornstein,
play was coded as symbolic play preceded           1991; Ungerer et al., 1981; Watson &
by child exploration. Likewise, if mother put     Fischer, 1977). Seven important transitions
the lid on the teapot while the child was         were incorporated into the scale: (1) the
"drinldng" from a teacup, the motlier's play      transition from simple exploration to non-
was coded as nonsymbolic play preceded by         symbolic play; (2) the transition from non-
a child symbolic act.^ Due to the sequential      symbolic play to sy-mbolic play; (3) the tran-
nature of interaction, coders made one pass       sition from seif-directed pretense to
thrpugh the videotapes, coding both mother        other-directed pretense; (4) the transition
and toddler play togettier. Coders were en-       from animate-directed pretense to inani-
couraged to stop and replay segments of in-       mate-directed pretense; (5) the transition
teraction to maximize coding accuracy.            from single acts of pretense to sequenced
                                                  acts of pretense; (6) the transition from
     Interrater agreement.—Interrater reli-       agentive pretense to vicarious pretense; and
abilities were obtained by having two coders      (7) the transitionfromliteral pretense to sub-
independently score ail mother and child          stitutive pretense. Moreover, there is some
play behaviors on 10% ofthe sampie for the        consensus that eertain transitions occur be-
entire 10-min session. A.s a first step to as-    fore others; for example, the transition from
sessing interrater agreement, we examined         self-directed to other-directed play is ex-
how often coders agreed on the identifica-        pected to occur before the transition from
tion of an event, calculated for children's       literal to substitutive play, on account ofthe

           This code system does not imply that the child is always the "initiator" and the mother
     always the "respondent" in the interaction. Sequences provide no information conceming which
     partner established the topic focus; instead, they describe only how a mother responded to a
     particular behavior on the part ofthe child.
1756     Child Development
more complex cognitive demands of substi-           play and specific patterns of maternal re-
tutive play (i.e., substituting one object for      sponsive play were related to one another.
another requires decontextualization from
the object, whereas other-directed play does             For each set of analyses, dyads that did
not). Additionally, the scale incorporates the      not have complete data on all pertinent vari-
notion that combinations of components of           ables were not included. For example, in
play are more difficult to master than are sin-     ANOVAs investigating matenial nonsym-
gle components of play. For exainpie, pre-          bolic play, mothers who did not engage in
tending to feed a doH with a block (Le., a          nonsymbolic play were necessarily ex-
combination of pretense toward an inani-            cluded. As a consequence, reported results
mate otiier and substitutive pretense) is           are based on samples of 42 to 50 dyads. Prior
jmore difficult than pretending to feed a doll      to analysis, data were examined for extreme
with a toy bottle (i.e., a single act of pretense   bivariate outliers using scatterplots. Cook's
toward an inanimate other),                         D, Stud.enti2ed Deleted Residual, and Le-
                                                    verage statistics. Four dya:ds whose individ-
     preliminary' analyses support the hy-          ual data altered the significance ofthe corre-
pothesized ordering of these 24 levels of           lations (either positively or negatively)
play (Tamis-LeMonda, Albright, Damast,              between maternal knowledge about play
Fox, & Bornstein, 1995). In a sample of 108         arid ma^ternal play behaviors were excluded
21-month-old cMtdren (including the 50              from tfte correlation analyses (Judd &
children in the present study) playing alone        MeClelland, 1989). Finally, because the
with the same standard set of toys, the pres-       AMO'teAs reVealed no consistent differences
ence or absence of each ofthe 24 play levels        in the base rates of maternal and child play
waS coded. The play levels were then rank-          by chiid gerider, and there were no gender
ordered based ofli Uie percentage of children       differetiees in any ofthe sequential analyses,
engaging in each level. The linearity ofthe         results are presented for the entire sample.
hypothesized scale was tested by comparing
these rank-orderings with the ordering in           Descriptive Analyses of Base Rates of
the proposed scale using Kendairs tau. The          Mother and Child Play
tau was signifiGant, tau = .72, p < .001, indi-          The mean frequencies and proportions
CE^ting substaiitial agreement between the          (number of maternal play acts at each level
hypothesized play scale and the order ob-           of play divided by the total number of mater-
served during toddlers' solitary play.              nal plaj' acts, calculated separately for each
                                                    mpthelr) of tlie three levels pf maternal play
      Examples ofthe 2,4 play levels were pre-      are presented in Table 1. A within-subjects
seiited in random order on the question-            rejaeatfed-measures ANOVA identified sig-
naire, (The labels of the play levels that ap-      nificMt differences in the base rates (i.e.,
pear in Appendix A were not provided.)              frequebiCies) of the differerit play levels, F(2,
Mothers were asfed to rarik the 24 different        98) = lCP.94,p < .001. Specifically, mothers
play activities iix order of increasing diffi-      pidtflptfid more symbolic play than either
cillti,', wfith 1 being the easiest and 24 the      expiorkton or nonsymbolic play, is(49) =
most difficult                                      Xi.64 aitd 10.67, respectively, ps < .001.
                                                    They showed no difference in their rates of
                                                    esplofation and nonsynabolic play, f(49) <
Results                                             I. (Ptfesnits for all descriptive analyses were
     Results are discussed in the following         identicsil when transfortned proportions
order. f;irst, base rates of the three levels of    were used as the dependent measure.)
maternal play, the four levels of child play,
and the rates with which these child levels              The mean frequencies and proportions
precedecl maternal play are presented and           (nurnt>er of child play acts- at each play level
analvzed using analyses of variance. Second,        diyi&d by the total rmmber of child play
analyses of the sequences of mother and             acts, paJeuiated separately for each child) of
child piay behaviors are reported. The goal         the three childi play levels and off-task epi-
of these analyses was to describe dependen-         sodes are also presented in Table 1. These
cies of maternal play on preceding child            rates rgpi-esent the tdtal number of child acts
play. Finally, relations between maternal           diujdg Ilie play session (iiot just the subset
knowledge of child play development and             of chiilcl acts to which mcfjhars responded). A
observed maternal play are reported. For            within-subjects repeated-fioeasures ANOVA
these analyses, measures of the accuracy of                 e differences in rates, F(3, 147) =
materiial knowledge of progressions in child                 , p < .001^ Child(eii engaged in explo-
Damast, Tamis-LeMonda, and Bornstein                       1757
                                                  TABLE 1
                  BASE R \ T E S OF MATERNAL PL.4IY AND C H I L D ACTIVITIES BY LEVEL

                      Level of Play             Mean             Range                SD
                  Mother:
                   Exploratory                 9.5 (.20)     0-24 (.00-.46)         5.0 (.10)
                   Nons>'mbolic                8.8 (.18)     0-27 (.00-.52)         6.6 (.12)
                   Symbolic                   30.8 (.63)     4-63 (.14-.93)        12.7 (.16)
                  Child:
                   Explorator\'               35 9 (.43)     13-51    (.24-.59)     8.4   (.08)
                   Nonsymbolic                11.8 (.14)      0-30    f.00-.28)     7.4   (.08)
                   Symbolic                   18.2 (.22)      5-43    (.07-.40)     7.5   (.08)
                   OfF-task                   18 2 (.22)     10-.32   (.11-.35)     5.4   ( 05)
                      NOTE —Proportions in parentheses.

ration more often than they were off-task or               levels of child play, F(3, 147) = 42.14, p <
engaged in nonsymbolic or symbolic play,                   .001. Specifically, greater numbers of mater-
ts(49) = 13.50, 17.89, and 11.60, ps < .001,               nal play responses were offered when chil-
respectively. (This may occur because chil-                dren were either off-task or exploring than
dren often orient to and explore objects be-               when they were engaged in either nonsym-
fore engaging in nonsymbolic and symbolic                  bolic or symbolic play, ^(49) = 9.67, p <
play.) In addition, children were off-task or              .001. Additionally, mothers responded more
engaged in symbolic play more frequently                   frequently when children were engaged in
than they engaged in nonsymbolic play,                     exploration than when they were off-task or
ts(49) = 5.04 and 3.93, ps < .001, respec-                 engaged in nonsymbolic or symbolic play,
tively. There was no difference in the rates               ts{49) = 4.12, 11.17, and 6.05, respectively,
at which children were off-task or engaged                 ps < .001. Mothers also responded more fre-
in symbolic play, f(49) < 1.                               quently when children were off-task than
                                                           when they were eagaged in nonsymbolic or
     The mean frequencies and proportions                  symbolic play, ts{49) = 8.30 and 2.72, ps <
(number of maternal piay acts in response to               .001 and .01, respectively. Lastly, mothers
each level of child play divided by the total              responded more frequently when children
number of maternal play acts, calculated                   were engaged in symbolic play than when
separately for each dyad) of mothers' play                 they were engaged in nonsymbolic play,
responses to each level of child play are pre-             t(49) = 3.95, p < 001.
sented in Table 2. Maternal play responses
were collapsed across level of maternal play               The Sequential Nature of Mother-Child
in order to question what mothers were re-                 Play Interactions
sponding to in their children's play, exclu-                    Mother-child play sequences were ana-
sive of how they were responding. A within-                lyzed by first calculating transitional proba-
subjects     repeated-measures       ANOVA                 bilities for all combinations of the three lev-
indicated a difference in the number of                    els of maternal play (i.e., exploration,
times mothers responded to the different                   nonsymbolic play, symbolic play) and four

                                                 TABLE 2
                     M.\TEHNAL PLAY BY LEVEL OF PRECEDING CHILD PL^Y ACT

                  Level of Preceding
                    Child Play Act             Mean              Range                SD
                 Off-task"                . 12.6   (.29)     3-28     (.12-.53)    5.5    (.10)
                 Exploratory                17 0   (.38)     .3-39    (.15-.61)    7.7    (.10)
                 Nonsymbolic                 56    (.12)     0-16     (.00-33)     4.4    (.08)
                 Symbolic                    9.7   (.22)     0-25     (.00-..50)   6.1    (.12)
                      NOTE —Proportions in parentheses
                      " OfF-task episodes iire included as child play acts because they, like
                 child play acts, represent children's attention at the time mothers choose
                 to o£fer play suggestions.
1758     Chad Development
levels of preceding child play (i.e., off-taslc,     X 4 (preceding child play level) within-
exploration, norisymbolic play, symbolic            subjects repeated-measures ANOVA to com-
play), using the formulas and procedures de-        pare the difference from expected probabil-
tailed by Bakeman and Gottman (1986).               ity' ofthe differen.t sequences. This ANOVA
These transitional probabilities were then          indicated ^ significant interaction of mater-
converted to z scores, and the z scores were        nal play level x preceding child play level,
used as dependent variables in subsequent           F(6, 258) = 33.73, p < .001. These results
analyses. Z scores control for the base rates       are presented in Table 4.
of both the "given" and "target" behaviors.
Table 3 provides descriptive data of the fre-            Figure 1 displays z scores for each child
quencies of the 12 child play—mother play           play-mather plag sequence. When compar-
sequences.                                          ing the four maternal exploratory z scores to
                                                    cbaiice, maternal exploratory play was pre-
     Z scores were used to examine differ-          ceded by child ofF-task episodes signifi-
ences between t t e observed rates ofthe se-        cantly more often than would be expected
queiices and their expected rates given the         by chance, f(49) = 8.43, p < .001, and it was
observed base rate probabilities of the indi-       preceded by child exploration, nonsymbolic
vidual "given" and "target" events (i.e.,           play, and symbolic play less often than
Bakeman & Gottman's, 1986, "first-order"            would be expected by chance,
Damast, Tamis-LeMonda, and Bornstein 1759
                                                       TABLE 3
                            MATERNAL PLAY BY LEVEL AXD BY LEVEL OF PRECEDING
                                               CHILD PLAY

                           Level of Maternal Play and
                          Level of Preceding Child Play          Mean      Range     SD
                       Maternal exploratory play;
                        Child off-task    '                       4.4      0-13      3.1
                        Child exploration                         2.8      0-8       2.3
                        Child nonsjnnbolie play                   0.8      0-.5      1.2
                        Child symbolic play                       0.8      0-4       1.1
                       Maternal nonsymbolic play.
                        Child off-task                            2.1      0-7       17
                        Child exploration                         3.3      0-16      34
                        Child nonsymbolic play                    2.2      0-9       2.5
                        Child symbolic play                       0.6      0-3       0.8
                       Maternal symbolic play:
                        Child off-task                            6.2      1-16      3.6
                        Child exploration                        10.6      2-23      .5.2
                        Child Bonsymbolic play                    2.6      0-10      2.5
                        Child symbolic play                       83       0-24      5.6

                        TABLE 4                               ration at a rate expected by chance, f(49) =
  DiFFEREXCES IN Z SCORES RY M A T E F X A L                  1.32, N.S. Gonsequently, when comparing
     LEVEL AXD BY PRECEDING CHILD LEVEL                       the four maternal symbolic play z scores to
                                                              each other, child s\TTibolic play increased
   Maternal Play Level                                        the likelihood of maternal symbolic play sig-
    and Comparison of                                         nificantly more than did child off-task epi-
  Preceding Child Level           n*        t         p       sodes, exploration, and nonsymbolic play,
                                                              t(48) = 9.71, t{4o) = 9.57, and t(48) = 6.31,
Exploratory:                                                  respectively, ps < .001.
  O-T vs. EXP                     49       7.30     .001
  O-T vs. NS                      45       6.60     .001           In summary, sequential analyses re-
  O-T vs. SYM                     48       8.64     .001      vealed noteworthy patterns with respect to
  EXP vs. NS                      45     -0.49                both increases and decreases in the likeli-
  EXP vs. SYM                     48       180     .078       hood of certain levels of maternal play given
  NS vs. SYM                      45       2.97    .005
Nonsymbolic:                                                  certain levels of child play. When children
  O-T vs. EXP                     49     -1.42                were off-task or engaged in exploration, the
  O-Tvs. NS                       45     -4.80     .001       greatest increase in probability (although
  O-T vs. SYM                     48       4.35    .001       not always a significant increase above the
  EXPvs. NS                       45     -3.86     .001       rate expected by chance) occurred at a ma-
  EXP vs. SYM                     48       5.92    .001       ternal play level one level higher than chil-
  NS vs. SYM                      45       9.37    .001       dren's play level (i.e., mothers suggested ex-
SjTnbolic:                                                    ploration when children were off-task, and
  O-T vs. EXP                     50    -4.95      .001       nonsymbolic plav when children were ex-
  O-Tvs. NS                       46    -1.45                 ploring). In contrast, when children engaged
  O-T vs. SYM                     49    -9.71      .001
  EXP vs. NS                      46      4.23     .001       in nonsymbolic and symbolic play, the great-
  EXP vs. SYM                     49    -6.31      .001       est increase in probability occurred at a ma-
  NS vs. SYM                      46    -9.57      .001       ternal play level that matched the preceding
                                                              child activity (i.e., mothers suggested non-
     NOTE —O-T off-task; EXP. exploraticn, NS: non-           symbolic play when children engaged in
symbolic play, SYM- symbolic play                             nonsymbolic pla}/, and symbolic play when
     ^ n for each analysis vanes as a result of tbe absence
of certain sequences in lndiwdual dyads                       children engaged in symbolic play). Mater-
                                                              nal play that was either lower or much
                                                              higher (i.e., three levels higher) than chil-
                                                              dren's ongoing play decreased in likelihood.
                                                                  Exploratory analyses were conducted
                                                              next to determine whether mothers were re-
                                                              sponding to their children's nonsymbolic
1760      Child Development
              z~score
       1.50 r

       1.00

       0.50

       0.00

      -0.50

      -1.00

      -1.50
                        Exploratory                 Nonsymboiic                       Symbolic
                                             Maternal Play Level

                ^B     Child Off-task                             Child Exploratory
                i    -I Chiid Nonsymbolic                         Child Symbolic

                              • p < 06, " p < 01, " • p < 001
                              FIG 1.—Maternal play level by child play level

and symbolic play with play that was lower               of these three categories of response (i.e.,
than, equal to, or higher than their children's          lower than, equal to, or higher than the pre-
preceding play act within the nons^inbolic               ceding child play level), within nonsymbohc
and symbolic play categories (see Appendix               and symbolic play separately, were summed
A). Specifically, for each maternal nonsym-              and used in subsequent analyses.
bolic play act that followed a child nonsym-
bolic play act, and for each maternal sym-                    Within-subjects       repeated-measures
bolic play act that followed a child symbolic            ANOVAs were used to identify differences
play act, both maternal play and preceding               in the frequencies of lower, equal, or higher
child play were coded using the levels ofthe             maternal nonsymbolic and symbolic play be-
24-level play scale. Mothers' play was then              haviors separately. For both maternal non-
recoded as being lower than, equal to, or                symbolic and symbolic play, there were sig-
higher than the level of the preceding child             nificant main effects, F(2, 98) = 3.61, p <
play. (Interrater reliabilities, using the cod-          .05, and F(2, 98) = 34.50, p < .001, respec-
ings of the same five dyads as above, were               tively. Mothers were more likely to suggest
again good to excellent: K = 1.00 for non-               nonsymbolic or symbolic play at levels
symbolic levels and K = .63 for symbolic                 equal to or higher than the level of their chil-
levels.) For example, if a child stirred with            dren's preceding nonsymbolic or symbolic
a toy spoon in a cup and the mother stirred              play, rather than lower than the preceding
with another spoon in another cup, she was               level of child play, ts(49) = 2.44 and 2.25,
matching the child's overall level of sym-               ps < .05, and ts(49) = 6.95 and 8.97, ps <
bolic play as well as the more specific sym-             .001, respectively. There were no significant
bolic act of self-directed single act pretense.          differences in the rates at which mothers
On the other hand, if a child stirred with a             suggested nonsymbolic or symbolic play at
toy spoon in a cup and the mother suggested              the same level as their children's preceding
that the child "feed the dolly," the mother              nonsymbolic or symbolic play or at a higher
was matching the child's overall level of                level than the preceding child play level,
symbolic play, but vvas prompting the child              ts{4Q) < 1.
to a higher level within symbolic play by                Relations between Maternal Knowledge
suggesting that the child include an inani-              about Play and Maternal Play
mate other in a sequence of two different                     Maternal play summary measure.—To
pretense schemes. The frequencies of each                examine relations between maternal knowl-
Damast, Tamis-LeMonda, and Bornstein                     1761
edge of play and maternal play, a siagle mea-       sequenced acts of pretense, literal pretense
sure of maternal play was created. The goal         as easier than substitutive pretense, and
was to create a measure of play to index            agentive pretense as easier than vicarious
mothers' attempts to "scaffold" their chil-         pretense. (All p values for z scores ranged
dren's play (i.e., attempts to help tlieir chil-    from .05 to .001.)
dren reach their potential level of perfor-
mance). For each dyad, the mean of the                   A measure of the overall accuracy of
transitional probabilities of seven play se-        each mother's rankings was generated by
quences was calculated; these sequences             correlating her ranking of the 24 play acts
are represented in Appendix B.* Sequences           with the hypothesized play scale. As a
were selected to represent instances in             group, individual mothers' rankings of these
which mothers responded with play at levels         acts correlated highly with the empirical
that were higher than the level of the pre-         scale, mean r = ,69, SD = .16. There was,
ceding child activity. Thus, higher values on       however, substantial individual variation in
this measure indicate a greater tendency of         these correlations, with scores ranging from
mothers to engage in more advanced levels           .09 to .91.
of play relative to their children's own play.           Correlation between maternal knowl-
The theoretical basis of this measure was           edge about play and. maternal play sum-
Kuhn's (1972) finding that children are moti-       mary measure.— To examine the relation
vated to perform at advanced stages (in Pia-        between maternal knowledge of play and
getian tasks) when they witnessed behaviors         observed maternal play, mothers' accuracy
performed at stages higher than Ilieir own          in ranking play was correlated with the ma-
stage, As a group, mothers offered 45% (SD          ternal play summary measure. Mothers
= 7%) of their play responses at levels             whose rankings of play corresponded more
higher than the preceding child play level.         closely with the hypothesized scale re-
                                                    sponded to their children's play with play
      Accuracy of maternal knowledge of             that was higher in level than the preceding
 play progressions.—Because the task of             child play level, n41) = .33, p < .05. In addi-
 ranking 24 different child play acts might be      tion, although four of the seven sequences
 difficult for mothers, the first set of analyses   included in this score involved mothers'
 investigated how accurate mothers, as a            suggesting symbolic play to their children,
 group, were in their knowledge of early play       maternal knowledge of play development
 development (see also Tamis-LeMonda et             was unrelated to mothers' overall frequency
al., 1994). Correlation analyses revealed           of suggesting symbolic play, r(41) = .10,
 strong agreement between both ithe mean            N.S. Thus, maternal knowledge about play
and modal maternal rankings ofthe play acts         is related to the specific coordination of the
and the hypothesized play scale, rs = .88 for       levels of mothers' and children's play, rather
both analyses. Further analyses explored the        than mothers' overall level of play.
accuracy of mothers' knowledge ofthe tran-
sitions in play overviewed above. Specifi-
cally, the percentage of mothers ranking the        Discussion
more difficult play act as higher than the less           The present study investigated the level
difficult act (e.g., the percentage of mothers      and contingency of maternal play on chil-
ranking other-directed pretense higher than         dren's play, as well as the possibility that the
self-directed pretense) was compared to             accuracy of mothers' rankings of the diffi-
chance using z scores. These analyses again         culty of child play acts accounts, in part, for
revealed that mothers are accurate in their         the content of mothers' play. Results indi-
knowledge of play development. Mothers              cated that mothers are sensitive to the level
ranked exploratory play as easier ilian non-        of their children's play, adjusting the level
symbolic play, and nonsymbolic play as eas-         of their own play to their children's play in
ier th^n symbolic play. Moreover, within            both a macroanalytic manner (i.e., consider-
symbolic play, mothers ranked pretending            ing base rate frequencies and unconditional
toward self as easier than pretending toward        probabilities) and a microanalytic manner
Other, single acts of pretense as easier than       (i.e., considering episode-by-episode se-

             Transitional probabilities were used, rather than frequencies or unconditional probabili-
      ties, because they control for tlie base rate of the child act. This was considered necessary
      because certain levels of toddler play force mothers' prompts to he higher than child play. For
      example, when children are off-+ask, any maternal prompt is necessarily higher than children's
      play.
1762      Ghild Development
quences and transitional prohabilities). In         Data from the present study advance our un-
addition, mothers who are more knowledge-           derstanding ofthis dynam^ic by demonstrat-
able about play development tend to suggest         ing that mothers also adjust the frequency of
play to their children at a more sophisticated      their suggestions on an episode-by-episode
level than their children's ongoing play.           basis at any one age. That is, they provide
                                                    more play suggestions when their children
            evaluating the generalizability of      are engaged in "less cornpetent" (i.e., off-
these findings, it may be important to con-         task and simple exploratory) behaviors, and
sider charact:eristics of the sample particular     they provide fewer play suggestions when
to this investiga^tion. Specifically, the moth-     their children are engaged in "more compe-
ers who participated in this study were             tent" (i.e., nonsymbolic and symbolic play)
somewhat older, more educated, and of a             behaviors.
higher socioeconomic background. Maternal
knowledge of and sensitivity to play may dif-            From a microanalytic perspective,
fer in younger mothers, less educated moth-         mothers appear to adjust the level of their
ers, arid mothers from other socioeconomic          play contingent on their children's play,
or cultural backgrounds.                            thereby functioning withiri their children's
                                                    "?:One of proximal deyelopment" (ZPD; Ro-
     From a macroanalytie perspective,              goff & Wertsch, 1984; Vygotsky, 1978).
mothers are sensitive to the level of sophisti-     When children are engaged in less sophisti-
cation of their ehildren's play acflvities, and     cated activities, mothers are likely to re-
they adjust the frequency and the level of          spond with play at one level higher than
their play responses based on the level of          their children's level. Mothers attempt to fo-
their children's play. Across the play session      cus their children's attention to the toys
as a whole^ motliers suggest play to their          \¥hen their children are oJEf-task, and (though
children most frequently at the overall level       not significantly above chance levels) to en-
of their children's play; both toddlers and         courage their children to engage in nonsym-
motliers engage in more symbolic play than          bolie play when they are involved in simple
noasj'mbolic play. In addition, mothers offer       objed manipulation. Thus, at these times
significantly more play resporises when their       mothtirs appear to use their play to "chan-
children are engaged in less sophisticated          nel" their children's behavior toward focus
activities (i.e., when off-task or engaged in       on toys and more sophisticated levels of jalay
px|pibration). In contrast, when their chil-        (Hodapp, Goldfield, & Boyatzis, 1^84;
dren are engaged in aiore sophisticated ac-         TroutiTian, Hazen, & Cook, 1992). When
tivities (i.e., in norisyriibolic or symbolic       children are engaged in more sophisticated
play), mothers offer fewer play suggestions.        play, mothers are more likely to match their
                                                    oVvn play to the level of their children's play,
     These findings concur with other evi-          or to engage in higher levels of play within
dence of parental sensitivity to develop-           the broader categories of nonsymbolic and
mental level and parental support of auton-         sjTuboIic play. This strategj^^ enables moth-
omy in ehildren. As described by Macceby            ers to elaborate on their children's behaviors
(1984), parents generally appear to be sensi-       at the same level.
tiye to the developmental abilities of their
ciiiidren, adjusting their own actions, over             Decreases in the likelihood of certain
time, to accommodiafe their children's              maternal play behaviors in response to cer-
emerging a(bij:it;ies. Maccoby suggested that       tain toddler behaviors also support the no-
paretafaJ nijenitoring of ehildren's nidment-       tion that mothers generally function within
by-moment; behaviors is firequent during in-^       their children's "zone of proximal develop-
fancy and tfjdidlerhodd/ Specific exi3«nE>les of    ment " Decreases in the likelihood of mater-
parents' ad|iistiHg their behavior across time      nal exploiatory play when preceded by child
are (Siyident: in a variety of parentirigd0m:ains   nonsvmbolie and child symbolic play, and
(Bornstein. Tal, et al., 1992; lEeckhausen,         decreases in the likelihood of maternal non-
1987; O'Cdnneti & Brethertbii, }Mk Pawjnt           symbolic play when preceded by child sym-
Moss GoBselin:; & St;-ta«.rtent: 1993; logofi;      bolic play, indicate that mothers tend wot to
Ellis, & Gardner, i§M; Tamis-LeiMonda &             respoiid to their children's play with play
Bornstein, 199J). For instance, mothers tend        that is lower in level (i.e., below their chil-
to show decreases iin particular interactions       dren's ZPD). The same pattem holds true
and increases in others as their children           when ea;a.minin.g trends within nonsymbolic
 grow dlder^attd imore co!rapeteirt(e.f., Belsky    and within symbolic play; mothers tend to
 et al.. I980;'fieirnstein & Taniis-LeMonda,        make suggestions that are equal or higher in
                                                    level, rtot lower in level, than their chil-
 199b; Rogoff, Malkin, & Gilbride, 1984).
Damast, Tamis-LeMonda, and Bornstein                   1763
dren's play. A decrease in the likelihood of        likely to respond to their children's play
maternal exploratory play when preceded by-         with play that is more sophisticated. The ef^
child exploration suggests that motlbers tend       feet size of this correlation is modest and
not to orient their children to a toy when          comparable to relations between parental
their children are already examining it, or         knowledge and behavior in other domains
shift their children's focus from one toy to        reported in the literature (Miller, 1988). On
another. In addition, although mothers en-          a macroanalytic level, no relation emerged
gaged in more symbolic play than nonsym-            between maternal knowledge about play
bolic play across the play session, they did        and overall frequency of maternal symbolic
not tend to suggest symbolic play in re-            play. Thus, mothers who are more knowl-
sponse to their children's nonsymbolic play         edgeable about early play development re-
or off-task behaviors. Overall, these results       spond sensitively on an episode-by-episode
might suggest that mothers' play behaviors          basis, rather than simply offering more sym-
with their children serve to scaHbld their          bolic play suggestions across the entire ses-
children's play on an episode-by-episode ba-        sion, irrespective of tiieir children's current
sis and across the entire play session. Fur-        play level. These differences in sensitivity
thermore, the data suggest that children's          between macroanalytic and microanalytic
dyadic play behaviors guide their mothers'          analyses of knowledge and scaffolding be-
participation in play interactions; mothers         haviors reinforce the value of considering
appear to use their children's play cues (e.g.,     the sequential nature of interactions (see
object focus, actions) as a framework for their     also Lewis & Lee-Painter, 1974; Tingley &
own play behaviors with their children.             Golden, 1992).
      Beyond the empirical contribution, the             Research investigating parental cogni-
present investigation expands the extant            tions has tended to explore relations be-
play literature methodologically. The use of        tween cognitions and developmental out-
sequential analyses allowed for a more de-          comes in children (Miller, 1988); these
tailed examination of an important aspect of        relations have frequently been examined
sensitive maternal interactive behavior—            with little regard to the parental behaviors
contingency. It is evident that motliers' sen-      that might mediate them, although the need
sitivity to their children's developmental          for such data has been discussed (Hunt &
level is both global (i.e., aeross an entire play   Paraskevopoulos, 1980; Kindermann & Skin-
session) and specific (i.e., on an episode-by-      ner, 1989). Furthermore, parental cognitions
episode basis). It also became evident,             are ofiten investigated in relation to reports
through these analyses, that matches be-            of parental behaviors (either self-reports or
tween the levels of children's and mothers'         e.xperimenter-completed reports) rather than
play observed in the literature (e.g., Tamis-       to direct observations of parental interac-
LeMonda & Bornstein, 1991) are at least par-        tions (Sigel et al., 1992, Stevens, 1984). Fi-
tially detennined by mothers' contingent,           nally, when parental knowledge about de-
sensitive responding.                               velopment is related to parenting behavior,
                                                    researchers have typically assessed (1)
     Analyses of mothers' knowledge of play         knowledge of general milestones and devel-
revealed that mothers, as a group, are knowl-       opment (e.g., the KIDI, MacPhee, 1981; the
edgeable about development in eaily play.           KEID, Stevens, 1984) and/or (2) more gen-
Nonetheless, interesting and in.Formative           eral parenting stj'les £tnd practices (e.g., au-
variation among mothers does exist. From a          thoritative/authoritarian parenting; Kochan-
microanalytic perspective, mothers' knowl-          ska, 1990). By relating parental cognitions in
edge ofthe progressive nature of play relates       the domain of play to observed parental be-
to episode-by-episode adjustments in their          haviors in the same domain, the present
play with their children. Mothers who are           study takes a step toward elucidating speci-
more knpwledgeable about the relative dif-          Rcity in the interplay betvveen maternal cog-
ficult>' of various toddler play acts are more      nitions and actual mother-child interactions.
1764 Ghild Development
 Appendix A
                                               Empirical Play Scale

                         Level                                       Action on Mothers' Questionnaire
Exploration:
   (1) Mouthing                                              (1) Suck block.
   (2) Simple rtlanipulatidn                                 (2) Hold spoon and look at it.
Nonsymbolic play:
   (3) Unitary functional                                    (3)   Turn wheel on toy car.
   (4) Inappropriate ci^mbinations                           (4)   Put toy dish on car.
   (o) Combinatioiis based on perception                     (5)   Stack toy plates.
   (6) Combinations based on function                        (6)   Put toy lid on teapot.
Symbolic play:
   (7) Self                                                  (7)   Feed self with toy spoon
  (8) Agentive animate                                       (8)   Wash mom with toy sponge.
  (9) Agentive inanimate                                     (9)   Rock doll.
 (10) Sequenced self                                        (10)   Stir in toy cup and eat from toy spoon.
 (11) Sequenced agentive animate                            (11)   Pour into toy cup from toy teapot and feed
                                                                   inom.
   (12)   Sequenced agentive inanimate                      (12)   Cover doll with blanket and pat to sleep.
   (13)   Vicarious                                         (13)   Make; doll wave "hi."
   (14)   Self substitution                                 (14)   Use block as sponge and wash own face.
   (15)   Agentive animate substitution                     (15)   Put toy plate on mom's head as hat.
   (16)   Agentive inanimate substitution                   (16)   Use spoon as brush and brush doll's hair.
   (17)   Sequenced vicarious                               (17)   Make stuffed bear walk to toy car and
                                                                   drive away.
  (18) Sequenced self substitution                          (IS)   Stir in toy pot with comb as spoon and eat
                                                                   from comb.
  (19) Sequenced agentive animate substitution              (19)   Feed mom with block and wipe her mouth
                                                                   with t()y sponge.
  (20) Sequenced agentive inanimate substi-                 (20)   Wash doll with block as sponge and dry
       Itution                                                     vvith towel.
  (21) Vicarious substitution                               (21)   Jifake toy person drive away in nesting cup
                                                                   as car.
  (22) Sequenced vicarious substitution                     (22)   Fut toy bib on doll as coat and make her
                                                                   walk,
  (23) Self-removed                                         (23)   Make one doll kiss another doll.
  (24) Emotive                                              (24)   Make doll fall down and cry.

Appendix B                                                   References
  Sequences in Maternal Play Summary Score                  Bakeman, R., Adamson, L. B., Brown, J. V., & El-
                                                                dridge, M. E. (1989). Can early interaction
 Mother Suggests                When Child Is                  predict? How and how much? In M. H.
                                                                Bornstein & N. A. Krasnegor (Eds.), Stability
(1) Exploration           (1) Off-task                          and continuity in mental developjnent: Be-
(2) Nonsymbolic           (2) Off-task                          havioral and biological perspectives (pp.
    play                                                       235-248). New York: Erlbaum.
(3) Nonsymbolic           (3) Exploring
    play                                                    Bakeman, R., & Gottman, J. M. (1986). Observing
(4) Symbolic play         (4) OfF-task                         interaction: An introduction to sequential
(5) Symbolic play         (5) Exploring                        analysis. New York: Cambridge University
(6) Symbolic play         (6) Engaged in nonsym-               Press.
                              bolic play                    Baumwell, L., Tamis-LeMonda, C. S., & Born-
(7) Symbolic play         (7) Engaged in symbolic              stein, M. H. (1996). Styles of maternal inter-
                               l                               action and toddler language comprehension.
    ' when a 21-month-old engaged m symbolic play, a           The importance of maternal sensitivity. Manu-
mother who suggests symbolic play is suggesting play at        script under review.
an appropriately sophisticated level given the age of the   Beckwith, L., & Cohen, S. E. (1989). Maternal re-
child and is considered to be scafFoldmg her child's play      iponsiveness with preterm infants and later
                                                               competency. In M. H. Bornstein (Ed.), Mater-
                                                               nal responsiveness. Characteristics and con-
                                                               sequences (pp. 75-87). San Francisco.
                                                               Jossey-Bass.
Damast, Tamis-LeMonda, and Bornstein                        1765
Belsky, J., Goode, M. K., & Most, R. K. (1980).        Goodnow, J. J. (1988). Parents' ideas, actions, and
     Maternal stimulation and infant explorator>'          feelings: Models and methods from develop-
     competence: Cross-sectional, correlabonal,            mental and social psychology. ChildDevelop-
     and experimental analyses. Child Develop-             ment, 59, 296-320.
     ment, 51, 1168-1178.                              Goodnow, J. J., & Collins, W. A. (1990). Develop-
Belsky, J., & Most, R. K. (1981). From exploration         ment according to parents: The nature,
     to play: A cross-sectional study of infant free       sources, and consequences of parents' ideas.
     play behavior. Developmental Psychology,              Hillsdale, NJ: Erlbaum.
     17, 630-639.                                      Gottfried, A W. (1985) Measures of socioeco-
Bornstein, M. H. (Ed). (1989). Maternal respon-            nomic status in child development research:
     siveness: Characteristics and consequences.           Data and recommendations. Merrill-Palmer
     San Francisco: Jossey-Bass.                           Quarterly, 31, 85-92.
Bornstein, M. H., & Lamb, M. E. (1992). Develop-       Haight, W. L., & Miller, P. J. (1993). Pretending
     ment in infancy: An introductiott (3d ed.).           at home. Albany, NY: SUNY.
     New York: McGraw-Hill.                            Heckhausen, J. (1987). Balancing for weaknesses
Bornstein, M. H., & O'Reilly, A. W. (Eds.). (1993).        and challenging developmental potential. A
     The role of play in the development of                longitudinal study of mother-infant dyads m
     thought. San Francisco: Jossey-Bass.                  apprenticeship interactions. Developmental
Bornstein, M. H., Tal, J., Rahn, C , Calperin,             Psychology, 23, 762-770.
     C. "Z.f Pecheux, M.-G., Lamour, M., Azuma,        Hodapp, R. M., Goldfleld, E. C & Boyatzis, C. J.
     H., Toda, S., Ogino, M., & Tamis-LeMonda,              (1984). The use and effectiveness of maternal
     C. S. (1992). Functional analysis of the con-          scaffolding in mother-infant games. Child De-
     tents of maternal speech to infants of 5 and 13       velopment, 55, 772-781.
     months in four cultures: Argentina, France,       Hollmgshead, A. B. (1975). The Four-Factor In-
     Japan, and the United States. Developm.ental           dex of Social Status Unpublished manu-
     Psychology, 28, 593-603.                               script, Yale University.
Bornstein, M. H., & Tamis-LeMonda, C. S. (1989).       Hunt, J. McV., & Paraskevopoulos, J. (1980). Chil-
     Maternal responsiveness and cognitive devel-           dren's psychological development as a func-
     opment in children. In M. H. Bornstein (Ed.),         tion ofthe inaccuracy of their mother's knowl-
     Maternal responsiveness: Characteristics and          edge of their Eibilities. Journal of Genetic
     con.sequences (pp. 49-61). San Francisco: Jos-        Psychology, 136, 285-298.
     sey-Bass.                                         Judd, C. M., & McClelland, G. H. (1989). Data
Bornstein, M. H., & Tamis-LeMonda, C S. (1990).            analysis: A model-comparison approach (pp
    Activities and interactions of mothers and             214-237). New York. Harcourt Brace Jova-
     their children in the flrst six months of life:       novich.
     Covariation, stability, continuity', correspon-   Kindermann, T., & Skinner, E. (1989). Mothers'
     dence, and prediction. Child Development.             perceptions of children's progress in develop-
    61, 1206-1217.                                         mental ta.sks: Organizers of contingencies in
Bornstein, M. H., Tamis-LeMonda, C. S., Tal, J.,           everyday interactions. Paper presented at the
     Ludemann^ P., Toda, S., Rahn, C. W., Pe-              meeting of the Societ\' for Research m Child
    cheux, M., Azuma, H., & Vardi, D. (1992). Ma-           Development, E[ansas City, MO.
    temal responsiveness to infants in three soci-     Kochanska, G. (1990). Maternal beliefs as long-
    eties: The United States, France^, and Japan.          term predictors of mother-child interaction
    Child Development, 63, 808-821.                        and report. Child Developmsnt, 61, 1934-
Brooks-Gunn, J., & Lewis, M. (1984). Maternal re-           1943.
    sponsivity in interactions with handicapped        Kuhn, D. (1972). Mechanisms of change in the de-
    infants. Child Development, 55, 782-793.               velopment of cognitive structures. Child De-
Fein, G. G. (1981). Pretend play In childhood: An          velopment, 43, 833-844.
    integrative review. Child Development 52           Lewis, M., & Lee-Painter, S. (1974). An interac-
     1095-1118.                                            tional approach to the mother-infant dyad. In
Fenson, L , Kagan, J., Kearsley, R. B., & Zelazo,          M. Lewis & L. A. Rosenblum (Eds.), The ef-
    p. R. (1976). The developmental progression           fect ofthe mfant on its caregiver (pp. 21-48).
    of manipulative play in the flrst two years.           New York: Wiley.
    Child Development, 47, 232-236.                    Maccoby, E. E. (1984) Socialization and devel-
Fenspa, L., & Ramsey, D. S. (1980). Decentration           opmental change. Child Development, 55,
    and integration of the child's play in the sec-        317-328.
    ond year. Child Development, 51^ 171-178.          MacPhee, D. (1981), Manual: Knowledge of In-
Fiese, B. H. (1990). Playful relationships: A con-        fant Development Inventory. Unpublished
    textual analysis of mother-toddler interaction        manuscript. University of North Carolina.
    and symbolic play. Child Development, 61,          McCune-Nicolich, L. (1981). Toward symbolic
    1648-1656.                                             functioning: Structure of early pretend games
You can also read