197

JOURNAL OF APPLIED BEHAVIOR ANALYSIS 2004, 37, 197–207 NUMBER 2 (SUMMER 2004)

INCREASING PRETEND TOY PLAY OF TODDLERS WITHDISABILITIES IN AN INCLUSIVE SETTING

CYNTHIA F. DICARLO

LOUISIANA STATE UNIVERSITY HEALTH SCIENCES CENTER

AND

DENNIS H. REID

CAROLINA BEHAVIOR ANALYSIS AND SUPPORT CENTER,

MORGANTON, NORTH CAROLINA

We evaluated a program for increasing pretend toy play of 2-year-old children withdisabilities in an inclusive classroom. Classroom personnel implemented the program,which involved choices of classroom centers containing toys that tend to occasion pretendplay in toddlers without disabilities, along with prompting and praise. Increases occurredin independent pretend-play rates among all 5 participating toddlers. Results are discussedregarding the importance of promoting toy play of very young children with disabilitiesthat is similar to the type of play of their nondisabled peers, and the need to identifycritical program components that are applicable in inclusive settings.

DESCRIPTORS: pretend play, young children with disabilities

A continuing concern in the provision ofservices for young children with disabilitiesis ensuring that individuals develop impor-tant skills while participating in inclusiveprograms. Participation of young childrenwith disabilities in programs with typicallydeveloping children is a major part of rec-ommended practices in early intervention(Sandall, McLean, Santos, & Smith, 2000)and is endorsed by the Council for Excep-tional Children and the National Associa-tion for the Education of Young Children(Sandall, McLean, & Smith, 2000). Despitethe emphasis on inclusive services, questionsremain regarding how those services shouldbe provided to allow sufficient learning op-portunities for young children with disabil-ities (DiCarlo, Reid, & Stricklin, 2003;Hauser-Cram, Bronson, & Upshur, 1993).

One particular area of concern in inclu-

We thank Sarintha Stricklin for her administrativesupport of the research and Patricia Snyder for hercomments on an earlier draft of the manuscript.

Requests for reprints should be sent to Dennis H.Reid, Carolina Behavior Analysis and Support Center,P.O. Box 425, Morganton, North Carolina 28680.

sive early-intervention settings is toy play.The importance of toy play as a form ofenvironmental engagement among youngchildren is well accepted (Weinberger &Starkey, 1994; Wolery & Werts, 1994). It isalso well recognized that young childrenwith disabilities often engage in less toy playthan their typically developing peers (Blasco,Bailey, & Burchinal, 1993), and frequentlyneed special interventions to increase toyplay (Blasco et al.; Reid, DiCarlo, Schepis,Hawkins, & Stricklin, 2003).

When considering means of increasingtoy play among young children with dis-abilities, attention is warranted regarding thespecific type of play that is targeted. Toy playamong typically developing children pro-gresses from simple to more advanced typesof play (McCabe, Jenkins, Mills, Dale, &Cole, 1999). However, children with dis-abilities frequently fail to engage in more ad-vanced or complex toy play that is charac-teristic of their nondisabled peers (McCabeet al.; Sigafoos, Roberts-Pennell, & Graves,1999). Failure to progress to more complex

198 CYNTHIA F. DICARLO and DENNIS H. REID

toy play can reduce learning opportunitiesand detrimentally affect overall skill devel-opment (Weinberger & Starkey, 1994).Consequently, when investigating means ofincreasing toy play among young childrenwith disabilities, it is beneficial to focus ontoy play that is commensurate with the typeor complexity of play of their same-age non-disabled peers (cf. McCabe et al.).

One specific type of play that is commonamong typically developing young childrenis pretend play with toys (Casby, 1991). Pre-tend toy play is referred to in a number ofways, including make-believe (Weinberger& Starkey, 1994), symbolic (Schrader,1990), and dramatic (Howe, Moller, Cham-bers, & Petrakos, 1993) play. In general, thistype of play involves using a toy to simulatea real-life action or situation, such as pre-tending to feed a doll (Howe et al.). Pretendplay is considered important in the devel-opmental process and is related to the de-velopment of a variety of adaptive skills(Schrader, 1990; Weinberger & Starkey,1994). Correspondingly, promotion of pre-tend play is a common goal of curricula andeducational programs for young children(Goldstein & Cisar, 1992; Sigafoos et al.,1999). Despite the recognized importance ofpretend toy play, few empirical studies haveinvestigated ways of promoting this type ofplay among young children (e.g., 2-year-olds) with disabilities (Howe et al., 1993;McCabe et al., 1999), and especially amongyoung children with disabilities in inclusiveearly-intervention settings (cf. Reinhartsen,Garfinkle, & Wolery, 2002).

The purpose of this investigation was toevaluate a method of increasing pretend toyplay among 2-year-old children with dis-abilities in an inclusive classroom. In accor-dance with recommended practices in earlyintervention (McWilliam, 2000), and to po-tentially enhance the practical applicabilityof the procedures (Schepis, Reid, Ownbey,& Parsons, 2001), all intervention compo-

nents were designed to be incorporated intothe existing classroom routine by regularclassroom staff.

METHOD

Setting and ParticipantsThe setting was an inclusive classroom

that served children between the ages of 15and 36 months. Two groups of children at-tended the classroom. One group attendedfor a half day on Mondays and Wednesdays,and one group attended for a half day onTuesdays and Thursdays. Each group includ-ed 12 children, half of whom had disabilitiesand half of whom did not. Each child withdisabilities met eligibility criteria for specialeducation services according to Part C of theIndividuals with Disabilities Education Actof 1997 (Shelden & Rush, 2001).

The classroom was staffed by a certifiedearly-intervention special educator, a speech-language pathologist, an occupational ther-apist, and a teacher assistant, and was orga-nized into interest centers, such as manipu-lative materials, dolls, kitchen, computer,and sand play. The investigation was con-ducted in the kitchen and doll centers,which were equipped with toys that are con-sidered to occasion pretend-play actionsamong children without disabilities (seeHowe et al., 1993, and Behavior Definitions,below). In addition, most of the pretend toyplay by the typically developing 2-year-oldchildren in the classroom was observed tooccur in these two interest centers.

At the outset of the study, 3 children withdisabilities between the ages of 26 and 30months participated, 2 of whom attendedthe classroom on Mondays and Wednesdays.The 3rd child attended on Tuesdays andThursdays. Based on the Early InterventionDevelopmental Profile (Rogers &D’Eugenio, 1981), Sally functioned cogni-tively in the 18-month range, Nate in the16-month range, and Kirk in the 22-month

199INCREASING PRETEND TOY PLAY

range. Adaptively, the children functionedbetween 9 and 19 months below chronolog-ical age level. Each child was ambulatory,and each required assistance from staff tocomplete self-care routines (Nate and Kirkreceived nutritional intake through gastros-tomy tubes). Each child vocalized infre-quently for apparent communication pur-poses (only Kirk occasionally used actualwords). Nate and Kirk used a voice outputcommunication device when specificallyprompted by a staff person, and Sally useda few manual signs when prompted.

While the intervention phase of the in-vestigation was in effect with the 3 targetchildren with disabilities, 2 additional chil-dren with disabilities were added to the in-vestigation for replication purposes. Jill at-tended the classroom on Tuesdays andThursdays, and Chas attended on Mondaysand Wednesdays. Jill was 27 months of ageand functioned cognitively in the 16-monthrange. Chas was 29 months old and func-tioned cognitively in the 14-month range.Both Jill and Chas were ambulatory.

The 5 children were selected for the in-vestigation based on staff reports and pre-baseline observations that indicated their fre-quencies of pretend toy play were below thatof their nondisabled peers. When the targettoddlers did play with toys, most of the toyplay was less advanced than pretend play,usually involving functional play with simplecause–effect toys (cf. Sigafoos et al., 1999).

Behavior Definitions and Observation System

The target behavior was independent pre-tend-play actions. Based on definitions inprevious research (Casby, 1991; McCabe etal., 1999; Weinberger & Starkey, 1994), pre-tend play was defined as at least a single-stepaction that appeared to imitate a real-life sit-uation involving objects that correspondedto the toys used in the action. Single-stepplay, involving one specific action with a toy,is characteristic of the type of pretend play

frequently observed among 2-year-old typi-cally developing children (Howe et al.,1993). Examples of single-step pretend toyplay included a child appearing to pour adrink into a cup (e.g., tilting a toy pitcherdown toward a toy cup), stirring a toy spoonin a toy bowl, talking on a toy telephone,and feeding a doll by placing a toy spoon tothe doll’s mouth. To be scored as indepen-dent, each pretend-play action had to occurwithout any prompt within the preceding 5s. Staff prompts could be verbal (e.g., ‘‘brushthe doll’s hair’’), modeling (e.g., a staff mem-ber showing how to pretend to scoop foodfrom a bowl to a plate), or physical assis-tance in the form of partially or totally guid-ing the child through a pretend-play action.Separate occurrences of pretend play wererecorded if the child completed a pretend-play action with one toy and then completeda pretend-play action with another toy. Sep-arate occurrences also were recorded if theplay involved only one toy and pretend playwith that toy was terminated and then re-initiated after at least 5 s. Observers also re-corded each toy used for a pretend-play ac-tion by each child during each observationsession, as well as each choice of toys pro-vided by a staff person to a child, andwhether the child made a choice in responseto this staff action (see Baseline, below).

Pretend play was observed continuouslythroughout 10-min free-play periods. Ob-servations occurred for only 1 toddler at atime, and each toddler was observed a max-imum of twice per classroom day. For reli-ability purposes, the periods were dividedinto 1-min intervals. During the 10-min ob-servation sessions, a child’s behavior was ob-served only while he or she was present inone of the two centers that contained thepretend-play toys.

Interobserver agreement checks were con-ducted for all 5 children and during bothexperimental conditions on 17% of all ob-servations. Interobserver agreement was cal-

200 CYNTHIA F. DICARLO and DENNIS H. REID

culated on a minute-by-minute basis by di-viding the smaller number of recordings ofindependent pretend-play actions of one ob-server by the larger number of the other ob-server and multiplying by 100%. If both ob-servers recorded zero occurrences for a par-ticular minute of observation, then 100%agreement was recorded for that minute.The resulting minute-by-minute figures werethen averaged for a given session. Interob-server agreement for independent pretendplay averaged 87% (range, 55% to 100%)per session. The lower range was due to asession in which one observer recorded oneoccurrence of pretend play and one observerrecorded zero for that minute (resulting inan agreement of 0%).

We calculated interobserver agreement forthe number of different toys played with byeach child by dividing the smaller numberof toys recorded by one observer during asession by the larger number recorded by theother observer and multiplying by 100%.Interobserver agreement averaged 95%(range, 75% to 100%). Interobserver agree-ment averaged 94% (range, 0% to 100%)for number of choices provided by a staffperson to a child and 94% (range, 0% to100%) for number of choices actually madeby a child.

Experimental Conditions

Baseline. Baseline involved the routinefree-play time in the classroom. Childrenwere free to move about the classroom andplay with toys of their choosing. As part ofthe usual classroom routine, staff attemptedto prompt toy play by providing a choice ofinterest centers when they observed that achild was not engaged in play. The choiceprocess involved a choice board that includ-ed picture and object symbols that repre-sented two of the various interest centers.The staff member typically presented theboard to a child who was not playing andasked him or her to tell in which of the two

centers on the board he or she would like toplay, while simultaneously naming the centeroptions. Following a choice, the staff mem-ber escorted the toddler to the chosen center.Presentation of choices in this manner wasnot provided on a systematic or set schedule,but averaged approximately once per childduring a given 10-min free-play period. Thetwo specific interest centers represented onthe choice board also varied nonsystemati-cally and included all of the interest centersin the classroom. Otherwise, during base-line, staff generally attempted to supportchild engagement in activities through, forexample, instructions and prompting, assist-ed children with self-care needs, and provid-ed supervision for all toddlers in the class-room. As with use of the choice board, therewas no consistent system for prompting toyplay (e.g., different staff used different typesof prompts and prompt sequences).

Responsive teaching program. The teachingprogram was designed to be responsive totwo issues related to pretend toy play thatbecame apparent during baseline observa-tions. Specifically, during baseline (a) thechildren did not visit the two interest centersin which the pretend-play toys were availableconsistently, and (b) when the children werein a target center, they frequently did notengage in a pretend-play action. Hence, theresponsive teaching program was designed toincrease the amount of time the childrenspent in the target centers and to increasetheir pretend-play actions. The teaching pro-gram also included a child-directed compo-nent, as recommended in early-interventionservices (Haney & Cavallaro, 1996), to pro-vide repeated choices of interest centers thatcontained pretend-play toys.

Providing a choice of the two interest cen-ters represented the first component of theresponsive teaching program. At the begin-ning of a free-play session, a staff memberprovided a choice of the two centers usingthe choice board. The staff member pre-

201INCREASING PRETEND TOY PLAY

sented the choice board to the child andasked him or her which center he or shewould like while naming both centers (e.g.,‘‘Do you want to play in the kitchen or withthe dolls?’’). The staff person ensured thatthe child looked at each object on the boardand waited 5 to 7 s for a response. Eachchild had been repeatedly exposed to thisprocedure prior to and during baseline. Afterthe child looked at each object on the board,the next object that the child either touched,pointed to, or looked at was considered thechild’s choice response. If the child onlylooked at an object, the staff member guidedthe child’s hand to touch the correspondingobject. If the child did not make a choice,the staff member arbitrarily selected an in-terest center and guided the child’s hand totouch a corresponding object. The child wasthen escorted to the chosen center (choseneither by the child or by the staff person).Once in the selected center, the staff mem-ber waited 10 s for the child to touch a toy.If the child did not touch a toy within 10s, the staff member initiated toy play byguiding the child’s hand to touch a toy. Staffinitiation of toy play in this manner oc-curred only when the child entered an in-terest center immediately after the choiceboard had been presented.

Following toy contact, initiated either bythe child independently or prompted by thestaff person, least-to-most assistive promptswere used to assist the child in completinga pretend-play action with the toy if thechild did not complete such action indepen-dently within 5 s of touching the toy (if thechild completed a pretend-play action in-dependently after touching a toy, the staffperson provided no assistance but praisedthe child’s actions). The least-to-most assis-tive prompts began with a verbal promptand then proceeded to a modeling prompt,and then to a physical prompt as necessary.Following completion of the prompted pre-

tend-play action, the staff member praisedthe child’s pretend play.

If at any time during the observed free-play session the child left a target interestcenter, the process of providing a choice ofthe two centers using the choice board wasrepeated. Throughout the responsive teach-ing condition, the choice board was pre-sented an average of 2.4 times per sessionfor Sally, 1.5 for Nate, 5.8 for Kirk, 1.8 forJill, and 3.0 for Chas. Sally made an inde-pendent choice on 75% of the presentations,Nate on 95%, Kirk on 77%, Jill on 100%,and Chas on 80%. Choices included bothof the interest centers.

In summary, the responsive teaching pro-gram consisted of three components. First, achoice of the two interest centers that con-tained pretend-play toys was provided at thebeginning of a free-play session and when achild left one of the centers during the ses-sion. Second, once the child was in a targetcenter following the choice-board presenta-tion, a staff member prompted toy play ifthe child did not touch a toy within 10 s ofentering the center. Third, if the child com-pleted a pretend-play action following con-tact with a toy, the child was praised for theaction; if the child did not independentlycomplete a pretend-play action within 5 s,the child was prompted to engage in pretendplay with the toy and praise was then pro-vided.

Experimental Design

The experimental design for the first 3children with disabilities was a multiplebaseline across the 3 toddlers. Two concur-rent AB designs were then used with the 2additional children.

Comparison Observations with TypicallyDeveloping Children

To evaluate the degree to which the levelof pretend toy play among the children withdisabilities approximated the level of such

202 CYNTHIA F. DICARLO and DENNIS H. REID

play among their nondisabled peers, com-parison observations of the pretend toy playof 3 nondisabled children in the classroom(age range, 19 to 25 months) were made.The observations were conducted in thesame manner using the same behavior defi-nitions and observers as with the childrenwith disabilities. Interobserver agreementwas assessed on one occasion, resulting in70% agreement for independent pretendplay and 80% for number of different toyswith which each child played. The 3 typi-cally developing children—Zen, Hal, andBill—were observed on 10, 8, and 8 occa-sions, respectively. Observations occurredduring the time period that encompassed thebaseline and responsive teaching conditionsfor the first 3 children with disabilities.

RESULTS

As indicated in Figure 1, during baseline,the first 3 target 2-year-olds were observedto be in the doll and kitchen interest centersinconsistently. When the children were inthese centers, low rates of independent pre-tend play were observed. For Sally, Nate, andKirk, independent pretend-play actions av-eraged 0.05 per minute (range, 0 to 0.2),0.13 (range, 0 to 0.5), and 0, respectively.During the responsive teaching program, all3 children visited the interest centers everysession, and increases in rate of independentpretend play occurred for each child. Pre-tend play increased to averages of 0.6 perminute (range, 0.1 to 1.3), 1.0 (range, 0 to2.3), and 0.1 (range, 0 to 0.7), for Sally,Nate, and Kirk, respectively, although theincrease occurred very gradually for Kirk.Similar results occurred for Jill and Chas(Figure 2). For Jill, independent pretendplay increased from an average of 0.06 perminute (range, 0 to 0.5) during baselinewhen she was in the target centers to 0.6(range, 0.3 to 1.0) during the program, al-though there appeared to be a decreasing

trend. For Chas, pretend play increased froma baseline average of 0.2 per minute (range,0 to 0.5) to 1.4 (range, 0.8 to 2.4) duringthe responsive teaching program.

In addition to the changes in pretend playdescribed above, the number of differenttoys each of the 5 children played with dur-ing the responsive teaching program in-creased relative to baseline. The averagenumber of toys played with during baselinefor Sally, Nate, Kirk, Jill, and Chas were 1.3,0.6, 1.9, 3.5, and 4, respectively (data notshown in figures). During the program therespective averages increased to 9.3, 6.6, 6.4,8, and 7.8.

Comparison Observations With TypicallyDeveloping Children

Across the 3 children without disabilities,independent pretend-play actions averaged0.6 per minute. (Mean rate for Zen was 0.9with a range of 0 to 2; for Hal, it was 0.5with a range of 0 to 1.4; for Bill, it was 0.4with a range of 0 to 0.8.) These results in-dicate that during baseline, each of the chil-dren with disabilities had a lower averagerate of independent pretend-play actionsthan the average for each of the childrenwithout disabilities. In contrast, during theresponsive teaching program, all but 1 of thechildren with disabilities (Kirk) engaged inindependent pretend play at a rate similar toor higher than the rate for the children with-out disabilities. Also, during the responsiveteaching program, the children with disabil-ities played with at least as many differenttoys on average across sessions as did thechildren without disabilities (averages forZen, Hal, and Bill were 6.1, 2.4, and 4.2,respectively).

DISCUSSION

Results indicated that the responsiveteaching program was accompanied by in-creases in rate of independent pretend toy

203INCREASING PRETEND TOY PLAY

Figure 1. Number of independent pretend toy-play actions per minute for each of the first 3 participantsduring each observation session in both experimental conditions (the open circles indicate when a child didnot enter a target center throughout the observed free-play period).

204 CYNTHIA F. DICARLO and DENNIS H. REID

Figure 2. Number of independent pretend toy-play actions per minute for each of the 2 later participantsduring each observation session in both experimental conditions.

play each time the program was implement-ed with 2-year-old children with disabilities.On average, the increase brought the 5 chil-dren with disabilities to pretend-play ratessimilar to those of children without disabil-ities, with one exception (Kirk). In accor-dance with recommended practices in earlyintervention (Schepis et al., 2001), regularstaff members carried out all components ofthe responsive teaching program during rou-tine classroom activities. Such featuresshould enhance the practicality of imple-mentation in terms of other staff being ableto incorporate the program procedures intotheir classroom routines (McWilliam, 2000).

Also in accordance with recommended prac-tices (Reinhartsen et al., 2002), the programincluded a child-directed component by pro-viding the children with choices of interestcenters (albeit restricted to two) and thenhonoring their choices by allowing them toplay in the chosen center. Incorporating achild-directed component into staff-imple-mented procedures may make the programmore acceptable in early-intervention set-tings that espouse a child-directed philoso-phy (Ivory & McCollum, 1999).

Although toy play of the children withdisabilities increased to levels similar to thoseof their peers without disabilities, it should

205INCREASING PRETEND TOY PLAY

be noted that the type of play observedamong the two groups was not necessarilyidentical. As indicated previously, single-steppretend-play actions are commonly observedamong typically developing 2-year-olds(Howe et al., 1993). However, typically de-veloping toddlers may also engage in moreadvanced pretend play with toys that in-volves various multiple-step actions. Thistype of play was not monitored as part ofthis investigation because of the focus onsingle-step actions, which appeared to be areasonable goal for the toddlers with dis-abilities given the infrequency of their pre-tend toy play. Also, duration measures of in-stances of pretend play were not addressed.Duration of pretend-play actions may alsovary across children with and without dis-abilities, and may relate to differences inquality of pretend play. Future research iswarranted to determine methods of evalu-ating and perhaps increasing pretend toyplay among very young children with dis-abilities that targets multiple-step actionsand duration of play to compare more pre-cisely their quality of play to that of theirnondisabled peers.

One seemingly noteworthy aspect of theresults is that the observed increases in pre-tend toy play occurred among very youngchildren with disabilities in an inclusive set-ting. Past research on toy play has focusedon 3- to 5-year-olds in preschools. To ourknowledge, research specifically on methodsof increasing pretend toy play among 2-year-olds in inclusive settings prior to enteringpreschool classrooms has not been reported,although the need for such research has beenacknowledged (Reinhartsen et al., 2002).The results offer encouragement for inter-vening with children at quite young ages ininclusive settings to help advance their levelof play.

The increases in pretend-play actions withtoys also are noteworthy in that these resultssuggest that pretend toy play may have be-

come reinforcing for these children. How-ever, the exact mechanism underlying the in-creases in independent pretend play cannotbe determined from the current investiga-tion. For example, it may have been thatstaff praise alone was sufficiently reinforcingto increase pretend toy play. Given the lowfrequencies of pretend toy play (as well asinconsistent presence in classroom interestcenters with pretend-play toys) during base-line, however, it seems unlikely that the chil-dren would have experienced sufficient re-inforcement from praise to account for theamount of behavior change observed. Alter-natively, increased familiarity with the toysthat likely resulted from staff prompting pre-tend toy play may have been responsible forthe observed increases (e.g., Hanley, Iwata,Lindberg, & Conners, 2003). Limitingchoices of interest centers during the sessionsto the two that were equipped with toys de-signed for pretend play may also have con-tributed to the increases. That is, limitingchoices to toys considered to be related topretend play may have increased contactwith these types of toys. Another variablethat may affect pretend play is preferencesfor specific toys that tend to occasion thistype of play. In the current investigation,preferences for specific pretend-play toyswere not addressed.

Additional research is warranted to iden-tify which components of the teaching pro-gram might be responsible for increases inpretend play, as well as the impact of pref-erences across pretend-play toys. Nonethe-less, given the lack of pretend play among2-year-olds with disabilities relative to thistype of play among typically developing 2-year-olds, it is important to demonstrateways to increase pretend toy play (Sigafooset al., 1999). Demonstrations are particular-ly needed in inclusive settings that are be-coming more common for young childrenwith disabilities. If effects of the programevaluated here are replicable in other inclu-

206 CYNTHIA F. DICARLO and DENNIS H. REID

sive settings, then more refined investiga-tions could delineate separate program com-ponents that may or may not be necessaryfor the program’s overall efficacy. Such re-search could provide useful information tohelp young children with disabilities engagein toy play at a level similar to that of theirtypically developing peers.

REFERENCES

Blasco, P. M., Bailey, D. B., & Burchinal, M. A.(1993). Dimensions of mastery in same-age andmixed-age integrated classrooms. Early ChildhoodResearch Quarterly, 8, 193–206.

Casby, M. W. (1991). Symbolic play I: A develop-mental framework. Transdisciplinary Journal, 1,219–231.

DiCarlo, C. F., Reid, D. H., & Stricklin, S. B. (2003).Increasing toy play among toddlers with multipledisabilities in an inclusive classroom: A more-to-less, child-directed intervention continuum. Re-search in Developmental Disabilities, 24, 195–209.

Goldstein, H., & Cisar, C. L. (1992). Promoting in-teraction during sociodramatic play: Teachingscripts to typical preschoolers and classmates withdisabilities. Journal of Applied Behavior Analysis,25, 265–280.

Haney, M., & Cavallaro, C. C. (1996). Using eco-logical assessment in daily program planning forchildren with disabilities in typical preschool set-tings. Topics in Early Childhood Special Education,16, 66–81.

Hanley, G. P., Iwata, B. A., Lindberg, J. S., & Con-ners, J. (2003). Response-restriction analysis: I.Assessment of activity preferences. Journal of Ap-plied Behavior Analysis, 36, 47–58.

Hauser-Cram, P., Bronson, M. B., & Upshur, C. C.(1993). The effects of the classroom environmenton the social and mastery behavior of preschoolchildren with disabilities. Early Childhood ResearchQuarterly, 8, 479–497.

Howe, N., Moller, L., Chambers, B., & Petrakos, H.(1993). The ecology of dramatic play centers andchildren’s social and cognitive play. Early Child-hood Research Quarterly, 8, 235–251.

Ivory, J. J., & McCollum, J. A. (1999). Effects ofsocial and isolate toys on social play in an inclusivesetting. Journal of Special Education, 32, 238–243.

McCabe, J. R., Jenkins, J. R., Mills, P. E., Dale, P. S.,& Cole, K. N. (1999). Effects of group compo-sition, materials, and developmental level on playin preschool children with disabilities. Journal ofEarly Intervention, 22, 164–178.

McWilliam, R. A. (2000). Recommended practicesin interdisciplinary models. In S. Sandall, M. E.

McLean, & B. J. Smith (Eds.), DEC recommendedpractices in early intervention/early childhood specialeducation (pp. 47–54). Longmont, CO: SoprisWest.

Reid, D. H., DiCarlo, C. F., Schepis, M. M., Haw-kins, J., & Stricklin, S. B. (2003). Observationalassessment of toy preferences among young chil-dren with disabilities in inclusive settings: Effi-ciency analysis and comparison with staff opinion.Behavior Modification, 27, 233–250.

Reinhartsen, D. B., Garfinkle, A. N., & Wolery, M.(2002). Engagement with toys in two-year-oldchildren with autism: Teacher selection versuschild choice. Research & Practice for Persons withSevere Disabilities, 27, 175–187.

Rogers, S. J., & D’Eugenio, D. B. (1981). Develop-mental programming for infants and young chil-dren: Assessment and application. In D. S. Schafer& M. S. Moersch (Eds.), Early intervention devel-opmental profile (rev. ed., pp. 23–28). Ann Arbor:University of Michigan Press.

Sandall, S., McLean, M. E., Santos, R. M., & Smith,B. J. (2000). DEC’s new recommended practices:The context for change. In S. Sandall, M. E. Mc-Lean, & B. J. Smith (Eds.), DEC recommendedpractices in early intervention/early childhood specialeducation (pp. 5–13). Longmont, CO: SoprisWest.

Sandall, S., McLean, M. E., & Smith, B. J. (Eds.).(2000). DEC recommended practices in early inter-vention/early childhood special education. Long-mont, CO: Sopris West.

Schepis, M. M., Reid, D. H., Ownbey, J., & Parsons,M. B. (2001). Training support staff to embedteaching within natural routines of young childrenwith disabilities in an inclusive preschool. Journalof Applied Behavior Analysis, 34, 313–328.

Schrader, C. T. (1990). Symbolic play as a curriculartool for early literacy development. Early Child-hood Research Quarterly, 5, 79–103.

Shelden, M. I., & Rush, D. (2001). The ten mythsabout providing early intervention services in nat-ural environments. Infants and Young Children,14, 1–13.

Sigafoos, J., Roberts-Pennell, D., & Graves, D.(1999). Longitudinal assessment of play andadaptive behavior in young children with devel-opmental disabilities. Research in DevelopmentalDisabilities, 20, 147–162.

Weinberger, L. A., & Starkey, P. (1994). Pretend playby African American children in Head Start. EarlyChildhood Research Quarterly, 9, 327–343.

Wolery, M., & Werts, M. G. (1994). Current practicewith young children who have disabilities: Place-ment, assessment, and instruction issues. Focus onExceptional Children, 26, 1–26.

Received March 12, 2003Final acceptance February 17, 2004Action Editor, Linda Cooper-Brown

207INCREASING PRETEND TOY PLAY

STUDY QUESTIONS

1. According to the authors, what is pretend toy play and why is it an important behavior?

2. How was pretend play defined and measured?

3. What features of the interobserver agreement procedures and results suggest that pretendplay may have been difficult to observe in a reliable manner?

4. Briefly describe the three components of the responsive teaching program.

5. Summarize the three general characteristics of the results.

6. What aspects of the observation procedures must be considered when comparing rates ofpretend play exhibited by the participants and their nondisabled peers?

7. Describe the general types of interventions (independent variables) that may have accountedfor observed changes in the participants’ pretend-play behavior.

8. How might one determine the behavioral benefits of teaching children to engage in pretendplay?

Questions prepared by Sarah Bloom and Stephen North, University of Florida