Digital Transformation and Athlete Human Capital Configuration: A Configurational Perspective on Value Creation

README Dataset for: "Digital Transformation and Athlete Human Capital Configuration: A Configurational Perspective on Value Creation in Elite Sport" ============================================================= 1. CASE SELECTION ============================================================= This study examines Chinese track and field athletes across three competitive tiers based on the human capital development stage theory: formation stage (early-career), accumulation stage (national-level), and maturity stage (international elite). Screening Procedure: - Initial pool: 622 athlete cases identified from four sources: (a) Chinese athletes ranked in the top 100 of each event on World Athletics (as of February 23, 2025); (b) Chinese athletes who broke world or continental records between 2019 and 2025; (c) Chinese Athletics Association's 2024 Paris Olympics roster; (d) China's roster for the 2024 World U20 Championships. - Cross-validation: Athletes appearing in multiple sources were counted once. - Stratified sampling: Three tiers with differentiated criteria: - Top tier (n = 8): Olympic/World Championship medalists OR world top-50 ranked OR record-breakers, with ≥1 million social media followers or ≥5 million topic views. - Middle tier (n = 11): National Championships medalists OR Paris Olympics participants OR world top-100 ranked, with≥100K followers or ≥10 million topic views. - Early-career tier (n = 9): U20/U18 medalists OR national ranking 7–16, with some data supplemented via interviews. - Expert validation: 3 track and field coaches independently assessed tier classification (inter-rater agreement: 89.3%). Final sample: 28 athletes (male: 13; female: 15), covering sprinting, jumping, throwing, and race walking events. ============================================================= 2. VARIABLE DEFINITIONS AND DATA SOURCES ============================================================= 2.1 Condition Variables (1) Achievement Capital Operationalized as competitive performance level, measured by medals and placements at major international and domestic championships (Olympic Games, World Championships, Asian Games, National Games, National Championships, U18/U20 events), with Olympic participation without a medal recognized as a distinct achievement level. (2) Social Capital Operationalized as online visibility and engagement, measured by verified followers on Weibo and Douyin, total engagement metrics (likes, comments, shares), and topic view counts. (3) Institutional Capital Assessed on two dimensions: (a) Institutional Support: Existence of government policies supporting market-based investment in athlete development (e.g., dedicated funding programs, subsidy policies, streamlined market access). (b) Policy Suitability: Availability of local government policies supporting sports technology enterprises,innovation platforms, and digital transformation initiatives. Composite score = MAX(Institutional Support, Policy Suitability). (4) Technology Application Measured through multiple channels: use of AI-assisted training systems or biomechanical analysis technologies, data-driven rehabilitation and intelligent wearable devices, and access to comprehensive scientific and medical support systems (including national team training programs with integrated scientific support infrastructure). Sources include General Administration of Sport reports, local government media, sports forums, athlete interviews, and self-media content. (5) Digital Commercialization Ability Assessed on two dimensions: (a) Brand Endorsement: Number of verified brand endorsements, identified through keyword searches in official media and brand websites. (b) Social Media Operations: Professional team management, content update frequency, engagement style, and presence of virtual avatar/digital persona. 2.2 Outcome Variable Human Capital Value Realization, assessed on two dimensions: (a) Income: Estimated from competition bonuses, salary, and commercial endorsements. Top-tier income estimated from media reports and Forbes athlete rankings; middle-tier from national salary benchmarks and endorsement data; early-career tier supplemented via interview data. (b) Career Development: Career transition capability and academic capital accumulation, assessed through publicly available career records and interview data. 2.3 Calibration Procedures Achievement Capital: 4-level ordinal scale (1, 0.67, 0.33, 0) based on highest competitive achievement level: (1) international medalist or Asian record holder → 1; (2) Asian Games medalist or National Games gold medalist → 0.67; (3) National Championships gold medalist, U18/U20 medalist, or Olympic selected competitor → 0.33; (4) none of the above → 0.Social Capital: Direct calibration in fsQCA using total follower count (million) on verified Weibo and Douyin accounts. Anchors: full membership = 2.0M, crossover = 0.42M, full non-membership = 0.023M. Technology Application: 3-item checklist (AI-assisted training or biomechanical analysis; data-driven rehabilitation and intelligent wearable devices; comprehensive scientific and medical support systems). Scored 0–1: all 3 met = 1, any 2 = 0.67, any 1 = 0.33, none = 0. National team training programs with integrated scientific support infrastructure satisfy criterion (3). Digital Commercialization Ability: 3-item checklist (virtual/AI persona; endorsement volume: >8 brands = high, 3–8 = medium, <3 = low; professional social media management). Scored 0–1. Institutional Capital: MAX(Institutional Support, Policy Suitability). Each sub-dimension scored 0–1 based on a three-item policy checklist. Income: Direct calibration in fsQCA. Income midpoints derived from estimated annual ranges. Anchors: 25th percentile = ¥262,500, median = ¥375,000, 75th percentile = ¥1,150,000. Career Development: Qualitative assignment based on evidence of (a) diversified post-athletic roles, (b) academic degree pursuit, or (c) limited career planning. Scores:0, 0.33, 0.67, 1. Human Capital Value Realization: Direct calibration in fsQCA using income midpoint as primary input with the same anchors as Income, holistically incorporating career development observations. 2.4 Reproducibility: NCA Analysis Code The NCA analysis was conducted using R package "NCA" (Dul et al., 2020). The input data file "NCA.csv" has the following column order: | Column | Variable | |:------:|----------| | 1 | HC Value (outcome variable) | | 2 | Achievement Capital | | 3 | Social Capital | | 4 | Technology Application | | 5 | Digital Commercialization | | 6 | Institutional Capital | Necessity Analysis (one condition at a time): ```r library(NCA) data <- read.csv("NCA.csv", header = TRUE) # Column 2 (Achievement Capital) vs Column 1 (HC Value) model1 <- nca_analysis(data, 2, 1, ceilings = "cr_fdh", test.rep = 10000) nca_output(model1, test = TRUE) # Column 3 (Social Capital) vs Column 1 (HC Value) model2 <- nca_analysis(data, 3, 1, ceilings = "cr_fdh", test.rep = 10000) nca_output(model2, test = TRUE) # Column 4 (Technology Application) vs Column 1 (HC Value) model3 <- nca_analysis(data, 4, 1, ceilings = "cr_fdh", test.rep = 10000) nca_output(model3, test = TRUE) # Column 5 (Digital Commercialization) vs Column 1 (HC Value) model4 <- nca_analysis(data, 5, 1, ceilings = "cr_fdh", test.rep = 10000) nca_output(model4, test = TRUE) # Column 6 (Institutional Capital) vs Column 1 (HC Value) model5 <- nca_analysis(data, 6, 1, ceilings = "cr_fdh", test.rep = 10000) nca_output(model5, test = TRUE) Bottleneck Analysis (all conditions simultaneously): # Columns 2-6 (all conditions) vs Column 1 (HC Value) model_bn <- nca_analysis(data, c(2:6), 1, test.rep = 10000) nca_output(model_bn, bottlenecks = TRUE) =============================================================3. DATA COLLECTION TIMELINE Secondary data collection: February 2025 – April 2025Interview data collection: May 3–5, 2025Data analysis: May 2025 – July 2025 =============================================================4. INTERVIEW DATA Interviews were conducted with early-career tier athletes whose publicly available data were insufficient for variable calibration. Participant ID Interview Date Duration Mode INT001 May 3, 2025 52:20 WeChat Voice INT002 May 4, 2025 19:55 WeChat Voice INT003 May 4, 2025 47:57 WeChat Video INT004 May 5, 2025 37:24 WeChat Voice Mean duration: 39.4 minutes (SD = 14.1) All participants provided written informed consent. Interviews were audio-recorded with participant permission. Interview data were anonymized and used to supplement secondary data for calibrating Technology Application, Digital Commercialization,Institutional Capital (both sub-dimensions), Career Development, and Income conditions for early-career athletes with insufficient publicly available data. =============================================================5. FILE INDEX File Access Table_S1_Variable_Definitions.xlsx Public Table_S2_Raw_Data_Anonymized.xlsx Public Table_S3_Calibrated_Data.xlsx Public Table_S4_NCA_Results.xlsx Public Table_S5_fsQCA_Truth_Table.xlsx Public Table_S6_fsQCA_Solutions.xlsx Public Table_S7_Data_Source_Tracker.xlsx Restricted Document_S1_Interview_Protocol.pdf Public Document_S2_Thematic_Analysis.pdf Public File descriptions: Table_S1_Variable_Definitions.xlsxVariable operationalization, data sources, and calibration anchors for all conditions and outcome. Table_S2_Raw_Data_Anonymized.xlsx. Raw scores for 24 public athletes (interview participant data excluded to protect privacy). Table_S3_Calibrated_Data.xlsx Fuzzy-set calibrated scores for fsQCA analysis (all 28 cases). Table_S4_NCA_Results.xlsxNCA bottleneck tables, effect sizes, and necessity test results. Table_S5_fsQCA_Truth_Table.xlsxTruth tables for high and non-high HC Value Realization(main analysis and robustness check). Table_S6_fsQCA_Solutions.xlsxComplex, parsimonious, and intermediate solutions; core vs. peripheral condition identification; robustness check. Table_S7_Data_Source_Tracker.xlsx [Available upon request]Detailed data source log for each athlete and each variable(with URLs and collection dates). Contains case-levelinformation that could potentially identify interview participants. Available upon reasonable request to the corresponding author for academic verification purposes only. Raw interview transcripts and audio recordings are not included. Document_S1_Interview_Protocol.pdfSemi-structured interview guide. Document_S2_Thematic_Analysis.pdfThematic analysis coding framework and key findings. =============================================================6. DATA RESTRICTIONS Table_S7 (Data Source Tracker) is not included in this public dataset because it contains case-level data provenance that could potentially identify interview participants. It is available upon reasonable request to the corresponding author for academic verification purposes only. Raw interview transcripts and audio recordings are not included in this dataset to protect participant privacy. =============================================================7. RELIABILITY AND VALIDITY Inter-rater reliability: κ = 0.83 Convergent validity: r = 0.45–0.67 among related variables All calibration thresholds verified through expert consultation and multiple sets of sensitivity analyses.