Job Skills and Occupations: PRISMA 2020 Systematic Literature Review Screening and Analysis Dataset

Abstract

This dataset contains the complete screening, eligibility assessment, relevance scoring, thematic classification, and methodological synthesis used in a systematic literature review on job skills and occupations. The review follows the PRISMA 2020 framework and is based on records retrieved from Scopus using the search query TITLE-ABS-KEY("Job Skill") AND TITLE-ABS-KEY("occupation"), with English-language and open-access criteria documented in the screening protocol. A total of 41 records were identified and screened. No duplicate records were identified and no records were removed through automation tools. Two records were excluded during the initial screening stage, resulting in 39 reports assessed for eligibility. Eight additional reports were excluded during eligibility assessment, leaving 31 studies for the final synthesis, corresponding to an inclusion rate of 75.6%. Each record was evaluated using a five-dimension relevance rubric with a maximum score of 10 points: C1 – job-skill construct centrality; C2 – occupational anchoring; C3 – empirical or methodological rigour; C4 – labour-market or workforce outcome relevance; and C5 – transferability to computational skill-occupation modelling. The workbook also documents the inclusion and exclusion criteria and provides explicit reasons for each inclusion or exclusion decision. The 31 included studies were organised into six thematic clusters: (1) Occupational Skill Taxonomies and Measurement Instruments; (2) Skill Mismatch, Overeducation and Credential Dynamics; (3) Computational and AI-Driven Skill-Occupation Matching; (4) Technological Change, Automation and Green Transitions; (5) Education, Training and Employability Pathways; and (6) Worker Heterogeneity, Mobility and Labour-Market Equity. Five methodological families were identified across the included studies: econometric and statistical modelling; machine learning and natural language processing on text data; network and graph analytics; survey and psychometric analysis; and framework, curriculum and competency design. The dataset is intended to support reproducibility, secondary analysis, evidence synthesis, bibliometric research, labour-market studies, occupational skill modelling, artificial intelligence-based skill matching, workforce development research, and future systematic literature reviews.