Literature search results

From our database search, we identified a total of 529 articles, including 95 articles from PubMed, 103 from EMBASE, 156 from Web of Science, 48 from CINAHL, and 127 from MEDLINE. The search was performed until January 31, 2023, without restriction of early publication time.

All identified articles were imported to EndNote, and 424 duplicates were removed. The titles and abstracts of the remaining 105 articles were screened, and 68 irrelevant articles were excluded, resulting in 37 additional articles. Then, two articles were excluded due to unavailability of the full text (conference abstracts), and 35 were assessed for eligibility. We further excluded 13 articles for the following reasons: three articles were reviews [15, 34, 35], one was a dissertation [36], one study did not investigate the measurement properties of PDMS-2 [37], and eight studies used PDMS-2 to assess other scales [2, 38,39,40,41,42,43,44]. Finally, 22 articles were included in our assessment. The detailed selection process and number of articles in each step are shown in Fig. 1.

Fig. 1

Flow diagram of the article selection according to the PRISMA

Characteristics of the included studies

The characteristics of the included articles are shown in Table 1. The studies were intercontinental, mainly from Europe [21, 45,46,47,48,49,50], followed by Asia [20, 22, 51,52,53,54,55] and North America [14, 17,18,19, 56, 57]. Specifically, six studies were from the USA [14, 17,18,19, 56, 57]; four studies were from Taiwan, China [51,52,53,54]; three, Portugal [21, 45, 46]; two, Brazil [58, 59]; two, South Korea [20, 55]; one, Belgium [47]; Spain [50]; the Netherlands [48]; Iran [22]; and the UK [49]. The participants in these studies were both normal [14, 19,20,21, 45, 46] and exceptional [17, 18, 22, 47,48,49,50,51,52,53,54,55,56,57,58,59] children. Exceptional children were identified as having various disabilities, such as developmental delays [17, 47, 48, 51, 52, 56, 57], premature birth [49, 58, 59] and neurological diseases [18, 50, 53,54,55]. The age of the children ranged from 0 to 6 years.

Table 1 Basic characteristics of the included articlesSynthesis of evidence for the measurement properties of PDMS-2

The overall assessment of the PDMS-2 measurement properties and the corresponding quality of evidence for each measurement property are shown in Table 2. The detailed quality of evidence data are provided in the supplementary material (Table S3).

Table 2 Summary of the findingsContent validity

Of the 22 included articles, only one study methodologically assessed the content validity of the PDMS-2 standard recommended by the COSMIN [59]. The study systematically assessed the content validity of the PDMS-2 by interviewing experts in the field and judged the relevance and comprehensiveness of the scale. The overall rating of the results for content validity was found to be sufficient, and the quality of evidence was moderate. Since this study did not report comprehensibility, it was not possible to judge the overall rating of comprehensibility (Table 2).

Structural validity

Four of the 22 included articles assessed the bifactor structural validity of the PDMS-2 by classical test theory (CTT) [14, 21, 45, 59]. The overall rating of the results for structural validity was found to be sufficient. The quality of evidence was high, and all studies were judged as very good (Table 2).

Internal consistency

Two studies examined the unidimensionality of the PDMS-2 subscales through Rasch analysis and indicated that most items on the scale met the unidimensionality requirement [52, 58]. Four of the 22 included articles assessed the internal consistency of the PDMS-2 [21, 45, 50, 54]. The Cronbach’s alpha values for the internal consistency of PDMS-2 were 0.999 (Reflex), 0.86–0.999 (Stationary), 0.89–0.999 (Locomotion), 0.87–0.991 (Manipulation), 0.76–0.999 (Grasping) and 0.89–0.999 (Visual–Motor). The overall rating was sufficient, and the quality of evidence of all included studies was high for internal consistency (Table 2).

Cross-cultural validity/measurement invariance

Of the 22 included articles, only one study assessed the cross-cultural validity of the PDMS-2 [46]. However, the methodology used in this study did not meet the COSMIN methodological requirements.

Reliability

Ten studies assessed the reliability of the PDMS-2 [14, 20, 22, 45, 48, 50, 53,54,55, 59]. According to the COSMIN manual [30], these studies can be divided into test-retest reliability, interrater reliability and intrarater reliability.

Eight studies assessed the test-retest reliability of the PDMS-2 [14, 22, 45, 48, 53,54,55, 59]. These studies mainly used the intraclass correlation coefficient (ICC) [22, 45, 53,