Ïã¸ÛÁùºÏ²Ê¼´Ê±¿ª½±

Special Needs (Ambulatory and Verbal) & No Developmental Concerns: (Bundy, Nelson, Metzger, & Bingaman, 2001; n = 77 children (42 males, 35 females); age range = 22 - 118 months; mean age = 63.8 months; n = 68; no developmental concerns, n=9; special needs (ambulatory and verbal))

• ToP Research Version 1/Study Group 1: All 12 raters¡¯ scoring fit the measurement model. 96% of ratings were expected by the Rasch model, indicating interrater reliability after rater training

Special Education & No Developmental Concerns: (Bundy et al., 2001; n = 124 children (69 males, 55 females); age range = 15 - 118 months; mean age = 55 months; enrolled in special education, n = 16; no developmental concerns, n = 108)

• ToP Research Version 2/Study Group 2: All 26 raters¡¯ scoring fit the measurement model. 96% of ratings were expected by the Rasch model

Able-bodied; Physical Disability; Cerebral Palsy, Spina Bifida; Below-knee Amputation from Birth Anomaly; Hemiplegia from CVA; Stickler¡¯s Syndrome: (Harkness & Bundy, 2001; n = 50; age range = 2 to 12 years; mean age = 83.8 months (7 years); able-bodied children, n = 25; children with physical disabilities, n = 25; cerebral palsy, n =14; spina bifida, n = 7; below-knee amputation from birth anomaly, n = 2; hemiplegia from CVA, n = 1; Stickler¡¯s Syndrome, n = 1)

• ToP Version 3: Data from all raters (100%) matched predictions of a Rasch model, so authors concluded high interrater reliability

No Delays; Cerebral Palsy; Hemiplegia; Diplegia; Quadriplegia; Mild, Moderate and Severe Developmental Impairments: (Okimoto, Bundy, & Hanzlik, 2000, p.78; n = 38 (26 males, 12 females; age range 3 - 32 months); n = 19 children with no delays (chronological age range 3 - 18 months); n = 19 children with cerebral palsy (chronological age range 8 - 32 months); mean chronological age of children with no delays = 9 months; mean chronological age of children with cerebral palsy = 18 months; n = 5 children with hemiplegia; n = 3 children with diplegia; n =11 children with quadriplegia; n = 7 children with mild developmental impairments; n = 4 children with moderate developmental impairments; n = 6 children with severe developmental impairments)

ToP Version 3: Data from all raters conformed to the expectations of the Rasch model: of 1,296 scores assigned (24 items x 54 tests), 2.5% were unexpected (22 for the item ¡°Shares¡±; 7 for the item ¡°Decides¡±).

Predictive Validity

No Developmental Concerns; Special Needs (Ambulatory and Verbal): (Bundy et al., 2001; n = 89 children (49 males, 40 females); age range = 15 months - 118 months; n = 80; no developmental concerns, n = 9; special needs (ambulatory and verbal)

ToP Research Version 3/Study 3: Adequate predictive validity: The ToP and Children¡¯s Playfulness Scale (CPS) correlate with a Pearson Product Moment coefficient of r = .46 (p = .0001)

Construct Validity/Convergent

No developmental concerns; Special Needs (Ambulatory and Verbal):(Bundy et al., 2001)

• ToP Research Version 3/Study 3: Adequate correlation was found between the ToP and the Children¡¯s Playfulness Scale (r = .46)

Children with no delays; children with cerebral palsy and developmental delays: (Okimoto et al., 2000; n = 38 (26 males, 12 females); age range 3 - 32 months; n = 19 children with no delays (chronological age range 3 - 18 months; mean chronological age = 9 months); n = 19 children with cerebral palsy and mild to severe developmental impairments (chronological age range 8 - 32 months; mean chronological age = 18 months)

ToP Version 3: Goodness-of-fit for items measuring relative playfulness and difficulty of items for typically developing young children and young children with cerebral palsy conformed to expectations of Rasch measurement model: t values ranged from ¡Ü 2 to ¡Ý -2 and MnSq values ranged from .6 to 1.4; paired t test indicated that mean score of typically developing children was significantly higher than mean score of matched counterparts with cerebral palsy and developmental delays (t = 3.938; p = .0005)

Item Response

Special Needs (Ambulatory and Verbal) & No Developmental Concerns:

• ToP Research Version 1/Study 1: 93% of items fit the measurement model (4 of 60 did not fit), suggesting playfulness is unidimensional. A 95% fit was desired but not met (Bundy et al., 2001)

Special Education & No Developmental Concerns:

• ToP Research Version 2/Study 2: 93% items fit the measurement model (5 of 68 did not fit), suggesting unidimensional playfulness  (Bundy et al., 2001)

Typically Developing; Autism; Developmentally Delayed; Cerebral Palsy; Abused; Other; Missing Data: (Bundy, Waugh, & Brentnall, 2008, p. 138-139; n = 265; age range = 15 months - 12 years; mean age = 57 months; typically developing, n = 78; autism, n = 73; developmentally delayed, n = 47; cerebral palsy, n = 14;, abused, n = 10; other, n = 29; missing data, n = 14)

• ToP Version 4: ¡°Because this study was performed to illustrate a process for test development rather than to present a test for use in practice, we employed a generous cutoff criterion of infit mean square statistics of 2.00 or less . . .Goodness of Fit mean square statistics from all 29 ToP items were within the limits set for the study. Data from 92% of the participants (n = 244) were within the limits.¡±

Person Response

Special Needs (Ambulatory and Verbal) & No Developmental Concerns:  (Bundy et al., 2001)

• ToP Research Version 1/Study 1: 10% of subjects (8) did not fit the measurement model; the percentage of children who did not fit the model was higher among children with special needs (44%, 4 children) than among typically developing children (5%, 4 children)

Special Education & No Developmental Concerns: (Bundy et al. 2001)

• ToP Research Version 2/Study 2: 98% of children¡¯s response patterns fit the measurement model

Able-bodied; Physical Disability; Cerebral Palsy, Spina Bifida; Below-knee Amputation from Birth Anomaly; Hemiplegia from CVA; Stickler¡¯s Syndrome: (Harkness & Bundy, 2001)

ToP Version 3: The ToP¡¯s item hierarchy may differ for children with and without physical disabilities. Person-response validity for children with physical disabilities did not meet the 95% fit to the Rasch model (only 92% fit). Most unexpected ratings stemmed from 4 items. Future research is needed on how the environment impacts playfulness scores

Typically Developing Children: (Brentnall, Bundy, & Scott Key, 2008, p.138; n = 20 (8 males, 12 females; age range 2 years, 8 months - 6 years; mean age 48 months; standard deviation 9 months; typically developing children from six childcare centers in Toronto)

Moderate test-retest reliability: interclass correlations between scores for Time 1 and Time 2 from each observation period: first 15 minutes: ICC = 0.668; 95% confidence interval = 0.331 to 0.584; p (2-tailed) = ©‚.001; last 15 minutes: ICC = 0.408; 95% confidence interval = -0.30 to 0.715; p (two-tailed) = .033; entire 30 minutes: ICC = 0.033; 95% confidence interval = -0.405 to 0.460; p (two-tailed) = .433

Typically Developing Children: (Brentnall, Bundy, & Scott Key, 2008)

• Unspecified ToP Version: Statistically significant differences between pre and post 6-week non-specific play materials intervention means (Z = -1.94; p = .025, one tailed; Cohen¡¯s d effect size of .55 (95% confidence interval, -0.08 to 1.19 with the approximate test mean at 0)) (Brentnall et al., 2008; n = 20; age range = 5 - 7 years)
• Unspecified ToP Version: Statistically significant differences between pre- and post-intervention means of a 6-week non-specific play materials intervention, (Z = -1.94; p = .025, one tailed; Cohen¡¯s d effect size of .55 (95% confidence interval, -0.08 to 1.19 with the approximate test mean at 0) (Bundy, Luckett, Naughton, Tranter, Wyver, Ragen, ... Spies, 2008; n = 20; age range = 3 5 - 7 years)

Interrater Reliability

Developmental Disabilities; No Developmental Disabilities: (Hamm, 2006, p. 92; n = 40; age range = 6 - 38 months)

• ToP Version 4: ¡°To determine inter-rater reliability, rater calibrations were examined for fit to the measurement model. Both raters fit the measurement model for the ToP . . .indicating that the scales were used in a consistent manner. For this sample, 100% of the ratings assigned conformed to the Rasch model supporting reliability of the scale when scored by trained raters¡±

Interrater & Intrarater Reliability

Autism; Pervasive Developmental Disorder:  (O¡¯Brien, Coker, Lynn, Suppinger, Pearigen, Rabon ¡­ Ward, 2000, p. 43; n = 4 children with diagnosis of autism; n = 2 children with dual diagnosis of pervasive developmental disorder; n = 2 children receiving medication; n = 3 children receiving additional occupational therapy services; range of age = four to six years; mean age = 55 months

Unspecified ToP Version:¡°Intrarater and interrater reliability was established for the seven student researchers in the use of the Test of Playfulness (TOP)¡±

Concurrent Validity

Autism: (Muys, Roger, & Bundy, 2006; n = 24 (17 males, 7 females); age range = 41 - 86 months; mean age = 63.4 months)

• ToP Version 4: Excellent concurrent validity, comparing the ToP, Version 4, with the Children¡¯s Playfulness Scale (CPS) with Pearson correlation coefficients, both for unstructured play (r = .814, p < .01) and structured play (r = .799, p < .01)

Convergent Validity

Sensory Processing Disorder: (Bundy, Shia, Qi, & Miller, 2007, p. 204; n = 20 with deficits in SPD (4 girls, 16 boys); age range = 4.4 years to 9.8 years, mean age = 6.9; n = 20 typically developing (9 girls, 11 boys); age range = 4.7 years - 11.7 years; mean age = 7.5 years)

• Unspecified ToP Version: ¡°The Rasch analysis provided two pairs of goodness-of-fit statistics for each child...data from 30% of the children with SPD (n = 6) feel outside the acceptable range¡±

Autism: (Muys et al., 2006)

• ToP Version 4: Excellent convergent validity with the Children¡¯s Playfulness Scale for unstructured and structured play settings (r = .814, .799 respectively)

Item Response

Developmental Disabilities; No Developmental Disabilities: (Hamm, 2006, p. 92)

• ToP Version 4: ¡°All items (100%) fit the model supporting that the ToP reflects a unidimensional construct of playfulness.¡±

Person-response

Developmental Disabilities; No Developmental Disabilities: (Hamm, 2006, p. 92)

• ToP Version 4: ¡°Data from 100% of the participants and raters conformed to the Rasch model.¡±

Typically Developing: (Bundy et al., 2008; n = 20; age range = 3 5 - 7 years)

• Unspecified ToP Version: Statistically significant differences between pre- and post-intervention means of a 6-week non-specific play materials intervention, (Z = -1.94; p = .025, one tailed; Cohen¡¯s d effect size of .55 (95% confidence interval, -0.08 to 1.19 with the approximate test mean at 0)

Sensory Processing Disorder: (Bundy et al., 2007; n = 20 (16 boys, 4 girls); age range = 4.4 - 9.8 years; mean age = 6.9 years, SD = 1.6 years; all children had sensory modulation dysfunction, some also had dyspraxia)

• Unspecified ToP Version: ToP scores pre and post a 20-session sensory-integration intervention were not statistically significantly different between a pre- and post-intervention measurements (p = 0.43, 0.41, 0.21)

Developmental/Communication/Autism Disorders:  (O¡¯Brien et al., 2000; n = 8 (TX, n = 4, CO, n = 4); age range = 4 - 6 years; mean age = 55 months)

Unspecified ToP Version: Two children in the control group and two in the treatment group demonstrated a significant increase in ToP scores pre- and post a 10-week OT intervention designed to increase playfulness; no significant changes in scores for the remaining children

Cerebral Palsy: (Reid, 2004; n = 13 (male = 7; female = 6); wheelchair users = 7; ambulatory with and without assistive devices = 6; age range = 8 - 12 years; mean age = 10 years, 5 months; SD = 1 year, 3 months)

Unspecified ToP Version: means and standard deviations from ToP subscales for 3 of 8 randomly selected video-recorded sessions of 13 participants in 12 virtual reality environments during an intervention designed to stimulate playfulness demonstrated playfulness in 4 categories (moderate levels of playfulness in motivation, control and framing; lower than moderate scores in suspension from reality)