Analysis Data and Codebook for “Power-Base Classification and Interfirm Trust: Reward Power, Separability, and Configurational Co-Presence”
Description
This dataset accompanies the study “Power-Base Classification and Interfirm Trust: Reward Power, Separability, and Configurational Co-Presence” The study examines how dependence and different power bases are associated with interfirm trust in relationships between Korean manufacturing SMEs and their primary foreign customers. The repository contains the de-identified, analysis-only data used in the study, together with a detailed codebook and README file. The structural equation modeling (CB-SEM) analysis uses 166 firms, while the fuzzy-set qualitative comparative analysis (fsQCA) uses 163 complete cases. The released dataset includes the 37 survey items used in the reported analyses, covering focal-firm dependence, perceived partner dependence, non-mediated power, reward power, coercive power, trust, relationship quality, partnership behavior, competitive advantage, and financial performance. To protect confidentiality and limit the release to information necessary for research verification and re-analysis, the repository does not include original respondent identifiers, contact information, firm-level demographic variables, or survey items not used in the reported analyses. Newly generated anonymous case identifiers are provided. Construct definitions, item codes, English renderings, measurement sources, and variable mappings are documented in the accompanying codebook and supplementary material.
Files
Steps to reproduce
Use Power_Trust_Analysis_Data.xlsx as the analysis dataset and Power_Trust_Codebook.xlsx for variable definitions and item mappings. The SEM analysis uses 166 cases. The fsQCA analysis uses the 163 complete cases provided in the fsQCA_163 sheet. Construct the study variables using the item groupings documented in the codebook and Supplementary Appendix A: focal-firm dependence (MFD: SA4–SA5), perceived partner dependence (PPD: EA1–EA3), non-mediated power (NMP: GA1–GA3 and GB1–GB3), reward power (RWP: GC1–GC3), coercive power (CP: GD1–GD4), trust (TR: PC2, PC3, PC4, PC7), relationship quality (RQ: RC3–RC5, RF1, RF3), partnership behavior (PS: PD1–PD3), competitive advantage (CA: P7–P10), and financial performance (FP: P4–P6). For the covariance-based SEM analysis, estimate the measurement models using robust maximum likelihood (MLR) with full-information maximum likelihood for missing item responses. The retained power specification separates non-mediated power, reward power, and coercive power. In the structural model, MFD, PPD, NMP, RWP, and CP are entered as standardized unit-weighted composites, while TR, RQ, PS, CA, and FP are modeled as latent constructs. A correlated residual is specified between RF1 and RF3. Indirect effects are evaluated using 5,000 percentile bootstrap resamples. For fsQCA, use the 163 complete cases. The five conditions are PPD, MFD, NMP, RWP, and CP, with trust (TR) as the outcome. Calibrate construct scores using the direct method with the 10th, 50th, and 90th percentiles as the full non-membership, crossover, and full membership anchors, respectively. For memberships falling exactly at 0.5, assign 0.501 when the raw score is above the crossover and 0.499 otherwise. Construct the truth table using a frequency threshold of 3 cases, a raw consistency threshold of .80, and a PRI consistency threshold of .70. Interpret the intermediate solution as the primary solution. Robustness checks reported in the Supplementary Material use alternative calibration anchors (5/50/95 and 15/50/85), stricter consistency thresholds, a frequency threshold of 2, alternative directional expectations, alternative treatment of exact crossover memberships, and a two-item reward-power measure excluding GC3. The analyses were conducted in R using lavaan for SEM and the QCA package for configurational analysis. Exact numerical replication requires implementing the specifications described above; analysis scripts are not included in the present repository.
Institutions
- Chungnam National UniversityDaejeon, Daejeon