Parent supervision attributes profile questionnaire (PSAPQ) for young children: psychometric properties of the Chinese version

In China, thousands of children die from unintentional injury each year; the incidence rate of injury is from 19.4 to 64.3% and is the leading cause of mortality for children between the ages one and 14 [1]. Unintentional injury is a major cause of hospitalization and death for children in China and elsewhere; moreover, it is a heavy burden for society [2‐4]. Most of the injuries are preventable since studies show that inadequate supervision is an important contributing factor to child injury [5‐10]. Appropriate supervision could prevent children from attempting unsafe activities and assist them to accomplish tasks successfully [11]. When parents overestimated their children’s coping skills for unsafe environments that often put the children in danger; on the contrary, if parents underestimated the coping skills, they could become more protective and obstruct their children’s development of essential skills for safety and autonomy [12]. It is difficult to obtain the data regarding parental supervision; to date, methods including naturalistic observation, self-report about supervision, and event participation monitoring methods were used, all of which are labor and time consuming [5‐9, 13‐18].

Morrongiello [15, 16] developed the Parent Supervision Attributes Profile Questionnaire (PSAPQ) to assess parental protectiveness and supervision of children from parental interaction with the observational measures. The PSAPQ has 29 items, including four subscales: protectiveness, supervision beliefs, tolerance for children’s risk taking, and belief in fate. The psychometric properties of PSAPQ were tested by using 192 parents who had children aged 2–5 years old [15]. The results showed that the subscales were representative of four constructs by using confirmatory factor analysis, and the internal consistency (Cronbach’s α) was good for the subscales: protectiveness (α = .78), supervision beliefs (α = .77), tolerance for children’s risk taking(α = .79), and belief in fate (α = .78). In addition, the test-retest reliability for the four subscales were all acceptable (r = .72–.80, p < .001). Intercorrelations was significant between parents’ self-reported PSAPQ and the naturalistic observation by researchers. Further, convergent and discriminant validity were tested, and all indicated good construct validity. Correlations among the four subscales were within the acceptable range, for example, protectiveness and tolerance for children’s risk taking (r = − .37), protectiveness and belief in fate (r = − .13), and supervision beliefs and belief in fate (r = − .21); the results showed the factors were significantly distinct in different constructs. The PSAPQ has been widely used to assess supervision and children’s risks of unintentional injury in different studies with other children’s age groups, and all showed good psychometric properties [13‐17].

In China, there is a specific type of family structure not found elsewhere in the world, which is called left-behind children. In this type of family structure, the parents are away from their hometown for work while their children stay in the hometown with grandparents. Prior studies revealed that single-parent children and left-behind children’s families are more vulnerable than the two-parent families for unintentional injury because of a lack of supervision [19‐21]. However, controversial findings pointed out that caregivers who spent more time taking care of their children couldn’t decrease the unintentional injury; instead, the quality of supervision matters [20, 22]. Therefore, a properly validated instrument to measure the supervision of children among Chinese primary caregivers is important and necessary. Currently, there is no available measurement of parental supervision in Chinese language. Thus, the aims of this study were to translate and adapt the PSAPQ into Chinese (C-PSAPQ), and further to test the reliability, validity, and factor structure of the C-PSAPQ in Chinese primary caregivers of 3–6 years old children.

A total of 296 caregiver-child dyads (Table 1) participated in this study, with a mean age of 59.1 (SD = 5.4) for caregivers and 5.3 (SD = 1.4) for children (163 boys and 133 girls). The majority of the study participants came from nuclear families (62.8%), with an average annual income of 50,000 yuan (55.7%) which is above the national average in China. More than half of the caregivers were college educated (65.5%), and were grandparents (63.9%). The frequency of injury in the past year was 113 (38.2%) never had any injury, 38 (12.8%) had one, 40 (13.5%) had twice, and 105 (35.5%) had more than three times. The average of severity of injury was 1.3 (SD = 0.8, ranged from 0 to 10).

The total score for the C-PSAPQ was 101.36 (SD = 13.77); for the four subscales were: 34.30 + 5.02 (protectiveness), 33.71 + 5.56 (supervision beliefs), 28.42 + 5.61 (tolerance for children’s risk taking), and 4.96 + 2.25 (belief in fate). Compared to the findings from Petrass’ study [13], the family caregivers in the current study scored significantly higher in protectiveness (M = 19.3; t = 51.38, p < .01), supervision beliefs (M = 18.3; t = 47.67, p < .01), and tolerance for children’s risk taking (M = 22; t = 19.67, p < .01). However, the mean score of belief in fate was significantly lower than Petrass’ results (M = 13.1; t = − 62.16, p < .01).

The purpose of this study was to adapt the C-PSAPQ and assess the psychometric properties of this scale among the Chinese population. Findings from this study show that the C-PSAPQ is linguistically relevant to Chinese family caregivers for young children. As we translated the questionnaire across culture, no item had great modification to the Chinese cultural content. The C-PSAPQ may covers protectiveness, supervision beliefs, tolerance for children’s risk taking, and belief in fate.

The C-PSAPQ had acceptable internal consistency reliability, all above .80, which is in accordance with previous studies of this scale [11, 14]. The Cronbach’s α for the two subscales, supervision beliefs (α = .64) and belief in fate (α = .61), was less than previous reports (Supervision = .77, Fate = .78), but was still acceptable [25]. This might be related to the differences between Eastern and Western parents in differing socioeconomic, cultural, and religious factors, which affect the perception of the ability to care for children and attendant dangers. Recently, a few published studies suggest cultural differences may contribute to race differences in injury of the children [26, 28, 37]. The western culture encourages child risk-taking, adventures and impulsion. Many caregivers in the Western families expressed self-confidence in supervision and a belief that the safety of the children is a matter of luck or fate [6, 14, 37]. Some parents also believe that childhood injuries are normal and that they do not have the ability to prevent their children from injury, so they might pay less attention than parents who had higher self-confidence in controlling for injuries [38, 39]. However, the caregivers in Latin American and Asian families do not encourage their children’s risk-taking behaviors in order to protect them from unintentional injury [40, 41]. In China, it is quite easy for Chinese parents to blame themselves if the child has an injury because most Chinese parents have no religious faith; they believe that if the injuries happen they result from the parent’s negligent care for the children [42]. In the current study, almost all primary caregivers (97%) had no religious faith; most of them think they are masters of themselves instead of victims of destiny. Various researches showed parental education related to pediatric injury: the degree of supervision varied across different educational levels. Those caregivers with high educational levels accurately supervised to assess the risk level and to cope with the risk [14, 41]. In the current study, there are 194 (65.5%) caregivers with at least college level education, and the mean score of belief in fate was 4.96 which was lower than Petrass’ results [13]. This indicated that most caregivers had the confidence to supervise and keep their children safe. On the contrary, the scores of protectiveness, supervision beliefs and tolerance for children’s risk taking in this study are higher than Petrass’ study [13]. The results show that caregivers were inclined to pay more attention to children in order to avoid injuries. Although the mean score and the range for the extended version of the PSAPQ have been reported [13, 15, 34], to date, normative scores for the PSAPQ have not been published. The norms and cut-off scores for the PSAPQ should be further explored to identify children who had higher risk of injury as a result of lack of parental supervision.

Multiple methods were used in this study to assess the validity of the C-PSAPQ. The results from the CFA indicate that this instrument includes four dimensions just as Morrongiello originally proposed [34]; however, the loading to each factor was different when compared to the original version (Table 3). Although CFA confirmed the C-PSAPQ included the same four factors as Petrass’ study [13], the two subscales, protectiveness and supervision beliefs, are highly correlated with each other, which was also established in a study in Portuguese [14]. Since the exploratory factor analysis (EFA) was not performed in this study, further study should include EFA since the constructs might be comprehended differently in different cultures. Regarding the concurrent validity, although the C-PSAPQ scores were significantly correlated with the naturalistic observations, the correlation was relatively low, which indicates concurrent validity should be further explored.

To our best knowledge, this is the first study to test the psychometric evaluation of the C-PSAPQ in Chinese. Although we found sound psychometric properties, some limitations need to be considered. First, the purpose of this study was to test the psychometric properties of the C-PSAPQ; however, the convergent/discriminant validity in the constructs of supervision beliefs and protectiveness were ambiguous in this study. Therefore, further validation is needed, particularly, the study participants were from four kindergartens in the city of Daqing; thus, the findings might not be presented to other areas of China. It is necessary to increase a large sample size in a future study and to employ psychometric test of the C-PSAPQ in both urban and rural areas. Second, the majority of the caregivers’ educational levels in the present study was above college (65.5%) and most of the primary caregivers are grandparents (63.9%). The different types of supervision between grandparents and parents should be further assessed since the parenting style could be different between generations. In urban China, the three-generation family is very common; children are cared for by both their parents and grandparents who all pay much attention to the children in order to avoid injuries [43]. In contrast, the left-behind children were solely cared for by the grandparents; therefore, the injury prevalence rate might be different across the types of family structures and warrant further study. Third, although the concurrent validity had a statistical significance, the correlation was low in this large sample size, which either indicated the self-report questionnaire could not replace the observation method to predict unintentional injuries or it is problematic when comparing the association of the assessments between parents and the researchers. Fourthly, the participants of this study were all physically and mentally healthy children, and behavioral problems were excluded; the study could not provide a full explanation regarding parental supervision of children who would need additional attention. Further study should examine the characteristics of parental supervision and behaviors in the group of children with ADHD.

Parental supervision is important for preventing child injury, but, to date, there is no reliable questionnaire to measure parental supervision in China. This study has shown that the C-PSAPQ has acceptable reliability, construct validity, concurrent validity, but not convergent/discriminant validity. Due to the unacceptable convergent/discriminant validity (supervision beliefs and protectiveness) of the C-PSAPQ, an exploratory factor analysis is needed to ensure that the same trait is measured by its indicators in different cultures in a further study.