Contemporary Chinese vocational interest scale in the digital age: development and validation

The study addresses the need to update vocational interest typologies and assessment tools in light of cultural factors and rapid changes in the occupational landscape driven by the fourth industrial revolution. While Holland’s RIASEC model has been influential, evidence suggests cultural variability in interest structures and limitations of legacy instruments in covering emerging digital-era occupations. The purpose is to define contemporary Chinese vocational interest types and to develop and validate a localized scale suitable for early adolescents (middle school students) to inform educational tracking (vocational vs. regular high school) and future career planning.

The paper reviews Holland’s RIASEC model and its global applications, noting predictive validity for academic and career outcomes. It highlights cross-cultural considerations showing mixed fit of RIASEC in Chinese contexts and calls to align interest types with evolving occupational categories. Prior Chinese studies have suggested additional or re-labeled types (e.g., Technological/Operational, Expressive, Powerful, Natural/Ecological/Biotic, Adventurous). Synthesizing this literature and contemporary job taxonomies in China, the authors propose nine types: Artistic, Natural, Enterprising, Conventional, Technological, Investigative, Powerful, Social, and Operational. Artistic integrates prior 'Expressive' elements; Natural combines 'Natural' and 'Adventurous' emphases; Enterprising and Powerful are separated to reflect business vs. political orientations; Technological is introduced to capture digitalization-related interests.

Design and procedure: A multi-step scale development process following best practices (item generation, expert review, item analysis, EFA for item reduction within subscales, and CFA in an independent sample). Participants: 1,332 middle school students from Shandong Province, China; after data quality exclusions, N=1,255 (51.08% male, 47.58% female; grades 7–8; mean age 12.75 years, SD=0.73). Random split into Sample A (N=628; Mage=12.75, 47% girls) for item analysis and EFA, and Sample B (N=627; Mage=12.76, 48% girls) for CFA and convergent validity. Item generation: Initial pool built from prior RIASEC and Chinese interest scales and updated using the 2022 General Code of Occupational Classification (with emphasis on Technological items). Obsolete items were removed or modernized (e.g., replacing outdated hands-on items), and new digital-age items were added (e.g., understanding how AI works). Expert review: Two psychometricians and six psychology students evaluated item clarity, suitability, and type classification; 135 items (15 per type) retained with consensus. Measures: Present scale—135 dichotomous items (1=yes, 2=no) across nine proposed types (A, N, E, C, T, I, P, S, O). External measure—SDS Form E (60 dichotomous items; six subscales R, I, A, S, E, C) for convergent validity. Item analysis (Sample A): For each subscale, high vs. low groups compared via independent-samples t-tests; items with nonsignificant discrimination (p>0.05) flagged for deletion. Deleted items: Artistic item 10; Conventional items 7 and 14; Operational items 4, 5, and 15; resulting in 129 items. Exploratory factor analysis (Sample A): Separate EFAs per subscale using principal axis factoring with one factor forced to test unidimensionality. Adequacy: KMO values—Artistic .82, Natural .87, Enterprising .87, Conventional .80, Technological .91, Investigative .85, Social .85, Powerful .84, Operational .76; Bartlett’s tests significant (ps<.001). Item retention criteria: loadings > .30 and minimum three items per factor. Final retained scale: 70 items across nine subscales—Artistic (8), Natural (8), Enterprising (8), Conventional (7), Technological (9), Investigative (7), Powerful (8), Social (7), Operational (8). All loadings ≥ .34. Internal consistency (Sample A): Cronbach’s alpha per subscale ranged from .70 to .86. Confirmatory factor analysis (Sample B): Separate CFAs per subscale in Mplus with categorical indicators; model fit generally acceptable to good across subscales with RMSEA approximately .048–.097, SRMR .040–.120, CFI .951–.989, TLI .937–.985. Reliability (Sample B): Alphas ranged from .67 to .87 across subscales. Convergent validity: Correlations between present subscales and SDS subscales showed theoretically consistent associations, including strong links between corresponding constructs (e.g., present Artistic with SDS Artistic; present Investigative with SDS Investigative; present Social with SDS Social). Operational correlated strongly with SDS Realistic, supporting construct alignment. Technological and Investigative showed notable association with SDS Investigative, reflecting cognitive/analytic overlap.

• Item pool refinement: From 135 items to 129 post item-discrimination screening; EFA retained 70 items across nine subscales (A=8, N=8, E=8, C=7, T=9, I=7, P=8, S=7, O=8). - Factor adequacy (Sample A): KMO indices—A .82; N .87; E .87; C .80; T .91; I .85; S .85; P .84; O .76; Bartlett’s ps<.001. All retained items loaded > .34 on their intended factor. - Reliability: Sample A alphas ranged .70–.86; Sample B alphas: Artistic .73; Natural .72; Enterprising .81; Conventional .73; Technological .87; Investigative .68; Powerful .71; Social .68; Operational .67. - CFA fit (Sample B) by subscale: Example indices—Artistic RMSEA=.048 (90% CI .031–.064), SRMR=.041, CFI=.984, TLI=.978; Technological RMSEA=.079 (.066–.093), SRMR=.040, CFI=.976, TLI=.969; Operational RMSEA=.059 (.045–.073), SRMR=.060, CFI=.951, TLI=.938. Most subscales demonstrated acceptable-to-excellent fit (CFI/TLI ~.95–.99; RMSEA ~.05–.08). - Convergent validity with SDS (Sample B, n=627): Present Artistic with SDS Artistic r=.485 (p<.01); Present Investigative with SDS Investigative r=.533 (p<.01); Present Social with SDS Social r=.460 (p<.01); Present Enterprising with SDS Enterprising r=.306 (p<.01); Operational with SDS Realistic r=.577 (p<.01). Present Technological also correlated with SDS Investigative r=.533 (p<.01). Overall pattern supports construct validity of the new subscales and their mapping to established interest dimensions. - Conceptual contributions: Empirical support for nine vocational interest types in contemporary China, including differentiation between Enterprising (wealth/business oriented) and Powerful (power/politics oriented), a combined Natural type (nature/adventure), and a novel Technological type reflecting digitalization.

Findings indicate that vocational interests among Chinese early adolescents can be reliably organized into nine dimensions that align with cultural-contextual realities and the digital transformation of work. The scale demonstrated coherent internal structure and acceptable psychometrics, with EFAs and CFAs supporting unidimensional subscales. Convergent validity with SDS confirms that several new subscales map onto established constructs (e.g., Operational to Realistic; Artistic, Investigative, Social, Enterprising to corresponding SDS types), while also capturing distinct domains not covered by RIASEC (Technological, Powerful, Natural). The differentiation between Enterprising (market/wealth orientation) and Powerful (political/power orientation) reflects salient value distinctions in China’s sociocultural landscape. The Technological dimension captures interests tied to emerging digital occupations, addressing gaps in legacy instruments. Practically, the instrument supports educational subject selection under China’s reformed high-school entrance framework and informs early career exploration and counseling. The results thus address the research aim by clarifying updated interest types for contemporary China and providing a validated measurement tool suited to the digital age.

This study defined nine contemporary vocational interest types in China and developed a corresponding adolescent scale with sound psychometric properties. By modernizing item content and introducing culturally and contextually relevant dimensions (Technological, Powerful, Natural, Operational), the instrument extends beyond classical RIASEC to better reflect current work environments. The validated scale can aid educational tracking and career guidance for middle school students. Future research should: (1) examine generalizability across age groups and populations; (2) conduct longitudinal studies to track stability and change in interests; and (3) expand convergent and predictive validity evidence with academic and career outcomes.

• Sample limited to middle school students (ages ~11–14) from a single province; generalizability to other age groups and regions remains to be tested. - Cross-sectional design precludes conclusions about developmental change; longitudinal validation is needed. - Convergent validity assessed primarily with SDS; additional external criteria (academic performance, job satisfaction/performance, career outcomes) should be incorporated to strengthen validity evidence.