The current study examined the psychometric properties of PAT(0-3). According to the results, the Cronbach's α values of PAT(0-3), Education part and Healthcare part are all above 0.70, suggesting high reliability, making it suitable for clinical settings. [38][39].The test-retest reliability exhibited large values for PAT(0-3) and its subscales, illustrating that the scale was stable and the temporal consistency was very good[36]. Moreover, PAT(0-3) scores varied in different maternal education levels and maternal ages, and were related to the child development indicators, which represent that it has satisfied external validity.
The internal consistency of the health and feeding subscale did not meet the satisfactory criteria, which refers that content and context of the subscale should be well defined[40]. The results from the CNCPQ study also showed the low internal consistency coefficients of health and feeding subscales(0.53 and 0.48)[12], which implies that it is not easy to acquire ideal items in the health and feeding parts within the NCF. In contrast, the result from an evaluation scale for nurturing and nursing ability of diabetic mothers showed that all the subscales' Cronbach's α coefficients were above 0.6[41] , inferring that subscales of a comprehensive scale could have ideal internal consistency when applied in the special population. The CFA results show that the scale did not align well with the theoretical framework, possibly due to the complexity of the scale contents and assessment methods. This scale evaluates parental parenting knowledge, beliefs as well as parenting behaviors through both interview and observation methods. Observational scales, such as HOME and PCI scale based the Barnard model, faced challenges in achieving ideal CFA results until certain items were removed[42][43][44].
Education part and Healthcare part in PAT(0-3) assessed distinct parenting skills and illustrated different psychometric properties. Given that health and education partnership is a systematic project, the development of PAT (0-3) requires a deep integration of evidence-based educational perspectives and social medicine concepts[45], rather than a simple combination of biomedical and behavioral training views. To enhance the validation of PAT (0-3), it may be necessary to adopt rational construction methods akin to the HOME and PCI scales, enriching concept subscales and reclassifying items based on theoretical connections rather than relying solely on statistical analyses. Since Hajjar commented that validity research involves a dynamic process that requires examinations and revised practices in a never-ending feedback loop[35], PAT(0-3) needs to be revised in a never-ending way.
According to the results, Education part and Healthcare part tend to predict child mental and physical development outcomes respectively. The positive correlations showed a good concurrent validity of different PAT(0-3) parts and verified that positive parenting aspects can contribute to different positive developmental outcomes. As child growth and development outcomes stem from complex relationships encompassing biological, behavioral, social, and environmental conditions[46], and researches on relationships of comprehensive nurturing care measurements and child development outcomes are insufficient, it is necessary to discuss the two parts’ predictive properties separately. Since positive relationship was found between education aspect in PAT(0-3) and child neuropsychological development, education aspect mainly referring to responsive caregiving and early learning can predict child metal development. Undoubtably the results were accepted by researchers worldwide, as previous studies have highlighted the significance of responsive parent-child relationships and parental support for early learning in t promoting early child neuropsychological development[47][48]. Meanwhile, the healthcare component within PAT(0-3), which encompasses health, nutrition, and safety, failed to predict the metal development of children, but it did predict their physical growth. Studies by Morrison and colleagues reported that parents’ healthcare knowledge, attitudes, and behaviors affected child health outcomes across the domains of disease prevention, acute illness care, and chronic illness care[49]. Additionally, Cheng’s review suggested that low parent health literacy was related to a range of pediatric health risks[50]. Irene and colleagues suggested that the feeding scale could identify 80% of infants and toddlers with feeding disorders. These studies collectively indicated that poor parents’ children healthcare knowledge and practice could influenced child health, hampering child growth[51].
Maternal education level was found to have significant positive associations with the PAT(0-3) total score. It can be attributed to the fact that caregivers with high education have a high capacity for understanding and applying information. Glick reported that parents without a high school education had 8.5 times the adjusted odds of low health literacy compared to those who were more educated[52]. Ip and colleagues argued that parents with lower education level often had less knowledge towards parent-child interactions[13]. These findings suggest the need to promote parenting among underprivileged families.
The score of PAT(0-3) increased with maternal age. While certain prior studies did not identify the relationship[11], it is commonly believed that parenting ability would improve with maternal age, similar to the results from Xie’s study[53]. This improvement in parenting ability with increasing maternal age can be attributed to the fact that older mothers often bring prior child-rearing experience to their parenting practices[54].
The score of Education part in PAT(0-3) was positively related to child age probably due to the improvement of maternal parenting experiences with child age. Martens and colleagues reported that the parenting quality of home environment improved from infants of 3 month to 12 month[55]. However, the score of Healthcare part did not significantly increase with the age of children, consistent with findings reported by Yin in China[33].
This study has several strengths. Firstly, PAT(0-3) addresses an existing gap in measurement of early parenting combining observations with face-to-face interviews specifically aligning with the nurturing care framework. This approach allows for a more objective and comprehensive evaluation[15]. Secondly, the evaluation of PAT(0-3) was conducted among caregivers in different geographical areas of China, demonstrating ideal internal consistency, test-retest reliability, and external validity. . Additionally, some indicators of construct validity were acceptable, suggesting that the instrument can be used for early parenting measurement in clinics of China to further improve caregivers’ parenting skills, behaviors, family nurturing environment and children’s early development. Lastly, PAT(0-3) included education and health parts, which offers diverse psychometric properties and facilitates a comprehensive assessment of early parenting practices.
There were several limitations of this study. Firstly, the items of the subscale of health and feeding exhibited less than ideal internal consistency, suggesting a need for modifications in these areas. Secondly, while the construct validity was partly acceptable, further efforts are required to enhance its overall construct validity. Thirdly, the participants were exclusively selected from urban areas and mostly had high education level, thus the generalizability of the study findings among caregivers with low educational levels might be limited, and the sample could not be viewed as representative of the general population.