A Systematic Review on Clinimetric Properties of Play Instruments for Occupational Therapy Practice

Table 1

Characteristic and psychometric reporting of individual studies.

Author	Year	Instrument	Objective	Study design	Country	Participant	Rater	Finding
Dender & Stagnitti [108]	2017	IPPS	To explore the content and cultural validity for social aspect of the instrument	Qualitative	Australia	6 pairs of indigenous children (i.e., 12 children); 14 community elders and mothers	—	The extension instrument is culturally accepted and nonjudgmental.
Golchin et al. [109]	2017	ChIPPA	To establish the reliabilities, content, and cross-cultural validity of the translated Persian version of the instrument	Cross-sectional (validity); Cohort (reliabilities)	Iran	5 occupational therapists; 31 typical children	2 researchers	Internal consistency is . Reliability is excellent for intrarater (), interrater () and moderate to strong for test-retest (). Content validity is strong ().
Stagnitti & Lewis [129]	2015	ChIPPA	To investigate the predictive validity of the instrument on semantic organization and narrative retelling skills using SAOLA	Cross-sectional	Australia	48 typical and at risk of learning difficulty children	3 examiners	The instruments predicted 23.8% of semantic organization and 18.2% of narrative retelling skills.
Dender & Stagnitti [107]	2011	I-ChIPPA	To investigate the cultural appropriateness of the adapted instrument and its reliability	Qualitative; Cross-sectional	Australia	23 indigenous Australian children (i.e., 12 pairs)	4 indigenous children	Cultural adaptation is satisfactory. The toys were found to be gender-neutral (). Overall, interrater reliability on toy use is moderate ().
Pfeifer et al. [120]	2011	ChIPPA	To establish the cross-cultural validity and reliability of the translated Portuguese version of the instrument	Cross-sectional	Brazil	14 typical children	1 occupational therapy student and 1 supervisor	Validity is established where the play material and duration are appropriate with the Brazilian context. Intrarater reliability is good (). Interrater reliability is moderate ().
McAloney & Stagnitti [117]	2009	ChIPPA	To investigate the concurrent validity of the instrument	Cross-sectional	Australia	53 typical children	1 researcher	Significant negative correlation was found between play and social.
Uren & Stagnitti [128]	2009	ChIPPA	To investigate the construct validity of the instrument	Cross-sectional	Australia	41 children of typical or minor disabilities	5 teachers	There is probable evidence on construct validity of the instrument Penn Interactive Peer Play Scale (PIPPS) and Leuven Involvement Scale for Young Children (LIS-YC) where several components were significantly moderately correlated.
Swindells & Stagnitti [127]	2006	ChIPPA	To investigate the construct validity of the instrument	Cross-sectional	Australia	35 typical children	2 researchers	Interrater reliability is strong (). There is probable evidence on construct validity of the instrument with Vineland Social-Emotional Early Childhood Scales; overall not significant but certain aspects were found significantly correlated.
Stagnitti & Unsworth [124]	2004	ChIPPA	To establish test-retest reliability of the instrument	Longitudinal	Australia	38 typical and developmental delay children	1 researcher	Test-retest reliability is moderate to strong ().
Stagnitti et al. [125]	2000	ChIPPA	To ascertain the discriminant validity and interrater reliability of the instrument	Cross-sectional	Australia	82 typical and preacademic problem children	3 occupational therapists	Interrater reliability is excellent (). Discriminant validity is established ()
Sposito et al. [123]	2019	Knox PPS	To verify the reliabilities of the Brazilian version of the instrument	Cross-sectional	Brazil	135 typical children	2 undergraduate occupational therapy students	Overall, the internal consistency is good (). Overall intrarater reliability () is reported to be moderate to excellent and interrater reliability () is moderate.
Pacciulio et al. [119]	2010	Knox PPS	To investigate the reliability and repeatability of the Brazilian version	Cohort	Brazil	18 typical children	2 examiners (one is the researcher; no further detail)	Strong intrarater correlation between the two occasions (). Strong interrater correlation between the two examiners ().
Lee & Hinojosa [116]	2010	Knox PPS	To establish the interrater and concurrent validity of the revised version of the instrument	Cross-sectional	United States of America	61 children with autism	2 researchers	Interrater reliability is excellent () and construct validity with VABS is moderate (, ).
Jankovich et al. [112]	2008	Knox PPS	To establish the interrater and construct validity of the revised version of the instrument	Cross-sectional	United States of America	38 typically developing children	2 occupational therapy students	Interrater agreement is high (81.8%–100%). Higher agreement was achieved on observation of older than younger children. Construct validity showed higher agreement between chronological and average play age for older than younger children.
Harrison & Keilhofner [111]	1986	Knox PPS	To determine the interrater and test-retest reliability and validity of the original instrument	Cross-sectional (interrater; concurrent validity); Longitudinal (test-retest)	United States of America	60 disabled preschool children	3 observers (detail not mentioned)	Overall interrater reliability is substantial (). Overall test-retest correlation is strong (). Concurrent validity indicates that the instrument correlates moderately with Parten’s Social Play Hierarchy () and Lunzer’s Scale on Organization of Play Behavior (). The instrument correlated moderately with age (r = 0.01–0.91) for disabled children but strongly with typical children.
Bledsoe & Sheperd [102]	1982	Knox PPS	To determine the inter-rater, test-retest reliability and validity of the revised instrument	Cross-sectional (inter-rater; concurrent validity); Longitudinal (test-retest)	United States of America	90 typical children	2 researchers cum observers	Overall, the inter-rater and test-retest yielded satisfactory correlation.; Concurrent validity indicates that the instrument correlates moderately with Parten’s Social Play Hierarchy and Lunzer’s Scale on Organization of Play Behavior.; The construct validity indicates that the instrument is correlated strongly with age.
McDonald & Vigen [118]	2012	McDonald Play Inventory	To examine the content, construct and discriminative, validity, internal consistency, and test-retest reliability of the instrument	Cross-sectional (validities, internal consistency); Longitudinal (test-retest)	United States of America	124 children; 17 parents	Self/proxy-rating; 7 children (test-retest)	Content validity is overall moderately correlated between items. Construct validity found that the instrument can discriminate between typical and disabled children. Concurrent validity between parent-child rating has low to moderate correlation (). Test-retest was strongly correlated () between two time points. Internal consistency: .
Schneider & Rosenblum [122]	2014	My Child’s Play	To describes the development, reliability, and validity of the instrument	Cross-sectional	Israel	334 mothers	—	Concurrent validity with Parent as a Teacher Inventory is fair (; ). Factor analysis established construct validity (). Gender (girls>boys) and age were significantly different in score. Internal consistency: .
Lautamo & Heikkilä [113]	2011	PAGS	To investigate the interrater reliability of the instrument	Cross-sectional	Finland	78 typical and atypical children	12 professionals (teachers, occupational therapist, physiotherapist)	MFR on expected agreement (44.1%) and the observed agreement (50.8%) with Rasch kappa of 0.12.
Lautamo et al. [115]	2011	PAGS	To evaluate the validity of the instrument for use with children with language impairment over typical children	Cross-sectional	Finland	156 typical and language impairment children	Proxy-rating (teachers, special education teachers, nurses, physiotherapist, occupational therapist)	The analysis found significant difference between the two groups, but 80% of the items are considered stable.
Lautamo et al. [114]	2005	PAGS	To determine the construct validity of the instrument	Cross-sectional	Finland	93 typical and atypical children	Proxy-rating (teachers, special education teachers, nurses, occupational therapist)	The construct validity of the instrument is established by internal scale validity, and person response validity achieved strong goodness of fit value.
Behnke & Fetkovich [101]	1984	Play History Interview	To determine reliability in terms of interrater and test-retest and validity of the Play History Interview	Cross-sectional (interrater; concurrent validity); Longitudinal (test-retest)	United States of America	30 parents with nondisabled or disabled children	2 researchers cum raters	Concurrent validity with Minnesota Child Development Inventory is overall moderate to strong.; Known-group validity is able to discriminate between disabled and nondisabled children ().; Interrater reliability is moderate to strong while test-retest has fair to strong correlation.
Sturgess & Ziviani [126]	1995	Playform	To explore the consistency on rating the instrument between three groups of rater	Cross-sectional	Australia	13 children; 13 parents; 1 teacher	—	Qualitatively, the rating between the three groups is relatively similar; parents scored slightly more positive than the children, but teachers are the most positive.
Bundy et al. [106]	2009	T-TUM	To investigate the translatability of the instrument to practice known as T-TUM (ToP+TOES Unifying Measure)	Cross-sectional	United States of America	265 atypical children	—	At least 92% of the outcomes were within the limit for goodness of fit. The reliability enhanced to for T-TUM.
Brentnall et al. [103]	2008	ToP	To evaluate the validity of instrument rating over different lengths and point of time	Cross-sectional	United States of America	20 typical children	3 researchers cum raters	Different time points have no significantly different observation outcome () but significantly different than longer observation time () but provide no added information. Longer observation time has poorer test-retest value () compared to shorter time ().
Rigby & Gaik [121]	2007	ToP	To investigate the stability of the instruments over three different settings	Cohort	United States of America	16 children with cerebral palsy	1 researcher	The score showed significant difference across the three settings (i.e., home, community, and school) (). The children are most playful at home and least playful at school.
Hamm [110]	2006	ToP + TOES	To examine the validity and reliability of the instruments with children with and without disabilities	Cross-sectional	United States of America	40 children with and without disabilities	2 trained raters	Interrater agreement is 100%. Item response validity is 100%, and internal scale validity is 100%. There is less playfulness but higher correlation of the instrument with children with disabilities than without disabilities.
Bronson & Bundy [104]	2001	ToP + TOES	To evaluate the validity of the two instruments	Cross-sectional	United States of America	160 children with and without disabilities	10 raters (not specified)	The reliability is acceptable: . TOES construct validity is acceptable (94% fit). The environment (i.e., TOES) is correlated significantly with playfulness (i.e., ToP) (; ). The TOES has significant difference between typical and disabled children (; ).
Bundy et al. [105]	2001	ToP	To investigate the construct and concurrent validity and interrater reliability of the instrument	Cross-sectional	United States of America	124 children (typical and special education) in total	26 occupational therapists	Construct validity explained 93% of the items unidimensional construct on playfulness. Concurrent validity with Children’s Playfulness Scale was found to be moderate (; ). Interrater reliability achieved 96% consensus.
Okimoto et al. [130]	1999	ToP	To investigate the reliability and validity of the instrument	Cross-sectional	United States of America	54 videotaped mother-CP-child dyad	3 occupational therapists	The reliability is 97.5% fit within the acceptable range. The instrument was found to be sensitive to change.

ChIPPA: Child-Initiated Pretend Play Assessment; I-ChIPPA: Indigenous ChIPPA; IPPS: Indigenous Play Partner Scale; Knox PPS: Revised Knox Preschool Play Scale; PAGS: Play Assessment for Group Setting; ToP: Test of Playfulness; TOES: Test of Environmental Supportiveness; T-TUM: ToP-TOES Unifying Measure.