Feeling Like a Professional May Drive How Childcare Trainees Rate Their Own Caregiving Skills

Among young people moving from vocational classrooms into real childcare settings in southern China, those who identify strongly with the caregiving profession tend to rate their own caregiving performance more favorably — and much of that link appears to run through how engaged they feel at work. That is the central finding of a new cross-sectional study published in Frontiers in Psychology, which surveyed 318 work-integrated childcare trainees and early-career practitioners across Guangdong Province. The researchers, Chen Chunhua of Zhanjiang Preschool Education College and He Yuda of Azman Hashim International Business School at the University of Technology Malaysia, report that professional identity showed a strong total association with self-rated caregiving performance, and that work engagement accounted for roughly a third of that association.

The study addresses a workforce segment that is expanding rapidly in China: individuals entering childcare practice through vocational education pathways. In what Chinese policy discourse calls integrated childcare, care and early education for children under three — and in some settings up to age six — are delivered within a single institutional framework rather than being split between custodial care and preschool education. Practitioners in these settings combine health monitoring, safety protection, feeding and hygiene routines, responsive interaction, play and developmental support, and communication with families. The role differs from conventional preschool teaching in its stronger emphasis on physical care, health surveillance, and infant-toddler routines.

Participants were recruited through 11 vocational colleges in Guangzhou, Shenzhen, Foshan, Dongguan, and Zhanjiang, which maintained placement agreements with 50 childcare institutions, including stand-alone childcare centers, childcare classes attached to kindergartens, community-based inclusive childcare points, and health-education integrated institutions. Questionnaires were distributed between 2 March and 29 April 2026 through institutional and placement communication channels. Of 352 returned questionnaires, 34 were excluded for reasons including invariant responding, excessive missingness, very short completion times, and duplicate submissions, leaving an analytic sample of 318 — a retention rate of 90.3 percent. Because the survey circulated through open placement channels, the researchers could not determine how many people saw the invitation, so no conventional response rate could be computed.

The sample was overwhelmingly female — 276 of 318 participants, or 86.8 percent — and young, with 56.3 percent aged 18 to 20. Most were majoring in infant and childcare, preschool education, or early education, and 62.3 percent were on internships, with another 27.0 percent on employment-linked placements and 10.7 percent already employed in early-career practice. The authors deliberately describe their participants as work-integrated trainees and early-career practitioners rather than established professionals, noting that the label of practitioner would overstate the occupational position of most of the sample. This boundary position — between education and employment — is precisely what makes the population theoretically interesting.

The theoretical framework combined two perspectives. Social identity theory holds that people derive part of their self-concept from membership in valued groups, and that this identification shapes motivation and behavior. Job demands–resources theory describes how personal and job resources predict engagement, which in turn predicts performance-related outcomes. The authors argue that professional identity qualifies as a personal resource because it supplies goal-relevant meaning that is stable across the varied and often mundane tasks of caregiving, allowing routine work to be construed as progress toward becoming a competent member of the occupation rather than as a set of discrete obligations. Combining the two theories yields a prediction neither makes alone: that the identity–performance association should operate substantially through engagement, even where occupational membership is provisional — a condition under which social identity theory alone would predict a weak effect.

A crucial conceptual clarification runs through the paper. Care quality in early childhood settings is conventionally assessed by trained external observers using standardized instruments such as the Classroom Assessment Scoring System or the Infant/Toddler Environment Rating Scale. Self-appraisal cannot substitute for such observation, because practitioners are poorly positioned to evaluate their own interactional quality and because self-ratings are subject to leniency and self-enhancement. The outcome measured here is therefore self-rated caregiving performance — the practitioner’s own appraisal of how well they apply caregiving knowledge, execute caregiving skills, and support children’s development and safety — not an indicator of the care children actually experience. The authors are explicit that their findings speak to how practitioners construe their own work, not to what children experience.

Statistically, the study was unusually thorough for a survey of this kind. The three-factor measurement model showed good fit, with a comparative fit index of 0.954 and a root mean square error of approximation of 0.039. Merged alternatives — in which two of the three constructs were collapsed — fitted significantly worse, providing direct evidence that professional identity, work engagement, and self-rated caregiving performance are empirically distinguishable rather than artifacts of a single undifferentiated self-evaluation factor. In the structural model, professional identity was positively associated with self-rated caregiving performance with a total standardized coefficient of 0.694, a direct coefficient of 0.465, and an indirect association through work engagement of 0.229, with a bias-corrected 95 percent confidence interval of 0.128 to 0.339. Professional identity alone accounted for just over half the variance in work engagement, and the two predictors jointly accounted for 53.2 percent of the variance in self-rated caregiving performance.

The robustness checks were extensive. Because participants were nested within 11 colleges and 50 institutions, the researchers computed cluster-robust standard errors and re-estimated the indirect association in a college-level cluster bootstrap; the adjusted estimate of 0.229 with a confidence interval of 0.120 to 0.353 still excluded zero. A Monte Carlo simulation with 3,000 replications showed empirical power of 0.990 for the indirect association at the achieved sample size. Adding twelve covariates — gender, age band, certificate status, major, practice status, and city — moved no coefficient by more than 0.02. Practice-status groups did not differ significantly on any construct, though the authors caution that these are failures to detect differences rather than demonstrations of equivalence, particularly given the small employed subgroup of 34 participants.

The authors themselves flag the study’s limits with unusual candor. The design is cross-sectional and single-source: all constructs were self-reported by the same individuals at one point in time, so the results describe a covariance structure compatible with the proposed ordering but cannot establish temporal order or causation. Reverse and reciprocal orderings remain plausible — trainees who believe they perform competently may develop stronger identification as a consequence, and longitudinal evidence in teacher populations suggests identity and engagement influence each other over time. Shared self-evaluative disposition is another alternative: respondents with generally favorable self-perceptions may rate themselves highly on everything. Unmeasured common causes such as placement quality, supervisory support, and prior childcare experience could also raise all three constructs together. Discriminant validity evidence was mixed, with average variance extracted below the conventional 0.50 threshold for all three constructs, and the Fornell–Larcker criterion not satisfied for the two pairs involving professional identity.

The effect sizes also deserve scrutiny. The total association of 0.694 is substantially larger than typical meta-analytic estimates for engagement–performance relationships, which cluster around 0.30 to 0.45 with self-rated criteria. The authors attribute this partly to both variables being self-appraisals collected in one instrument, and partly to latent-variable estimation correcting for measurement error; the raw composite-score correlation of 0.601 sits closer to the published range. They also note that work engagement’s incremental contribution is more modest than its statistical significance suggests, since the indirect association accounts for only about a third of the total, leaving a substantial direct association that may reflect mechanisms not modeled here, such as feedback-seeking, receptivity to mentoring, or the tendency to interpret difficult tasks as professional development rather than burden.

Perhaps the most striking theoretical result is that the identity–engagement association was as strong in this boundary population as in studies of established employees, despite the provisional occupational membership and limited task autonomy that characterize placement work. Social identity theory would predict attenuation under such conditions. That the association was not attenuated suggests, the authors propose, that anticipated rather than achieved membership may function as a psychological resource — a possibility that longitudinal work could test directly. The contribution is best characterized, they write, as a boundary test rather than a new theory: evidence that a sequence documented in settled workforces also appears in a population where its theoretical preconditions are weaker.

For practice, the implications are framed as candidates for intervention research rather than proven practices. Structured orientation, mentoring relationships, guided reflection on the social value of caregiving, and exposure to experienced practitioners are plausible ways to cultivate professional identity during work-integrated training, while placement design features such as workload calibrated to trainee capacity, regular supervisory feedback, and meaningful task allocation may support engagement. But the authors attach a firm caution: because the outcome is self-appraised, interventions that raise identity and engagement might raise self-rated performance without altering observed caregiving quality. Future evaluations, they argue, should include observational or supervisor-rated criteria rather than relying on practitioner self-report alone — and longitudinal, multi-source designs are the clearest next step for determining whether identity genuinely precedes engagement, or whether the two simply rise and fall together.

Subject of Research: Associations among professional identity, work engagement, and self-rated caregiving performance among childcare trainees in China

Article Title: Professional identity and self-rated caregiving performance among work-integrated childcare trainees and early-career practitioners: an indirect association through work engagement

Article References: Chunhua, C., & Yuda, H. (2026). Professional identity and self-rated caregiving performance among work-integrated childcare trainees and early-career practitioners: an indirect association through work engagement. Frontiers in Psychology, 17, 1919887. https://doi.org/10.3389/fpsyg.2026.1919887

Image Credits: AI Generated

DOI: 10.3389/fpsyg.2026.1919887

Keywords: childcare workforce, professional identity, work engagement, self-rated job performance, vocational education, work-integrated learning, structural equation modeling, Guangdong, early childhood care, organizational psychology, cross-sectional survey, China