Combining PLS-SEM, NCA, and fsQCA: Three Logics, One Coherent Design
PLS-SEM, NCA, and fsQCA are not three ways to test the same hypothesis. They describe net effects, necessary conditions, and sufficient configurations, respectively.
A sound integrated study prespecifies each method’s question, keeps concepts consistent without equating scores, calibrates fsQCA transparently, and uses findings complementarily rather than by majority vote.
Three methods address three research questions
- PLS-SEM
Relationships and predictive contribution
- NCA
Necessary conditions for an outcome level
- fsQCA
Sufficient configurations of conditions
Conceptual diagram. Integrate through questions and measurement, not mechanical cross-confirmation.
Source: Lê Minh · Conceptual illustration, not software output.1. Three Methods, Three Logics
A variable may have a small net effect, be necessary for a certain outcome level, or appear only in some sufficient configurations. These results need not contradict one another (Dul, 2016; Ragin, 2008).
| Method | Logic | Main output |
|---|---|---|
| PLS-SEM | Symmetric/net effect | Path coefficients, R², prediction |
| NCA | Necessary but not sufficient | Effect size, ceiling, bottleneck |
| fsQCA | Set-theoretic/sufficient configurations and asymmetry | Consistency, coverage, recipes |
2. Designing Integration from Research Questions
Map the questions
Assign each question to net-effect, necessity, or sufficiency logic.
Define constructs
Maintain consistent substantive definitions and specify the scores used in each analysis.
Run PLS-SEM
Assess the measurement model, structural model, and prediction according to the objective.
Run NCA
Assess necessity hypotheses, effect sizes, permutation tests, and bottlenecks.
Calibrate fsQCA
Disclose anchors for full membership, crossover, and full non-membership.
Compare findings
Explain complementarity, asymmetry, and limitations; do not count significant methods as votes.
3. fsQCA Calibration and Robustness Assessment
Latent scores are not automatically fuzzy-set membership scores. Calibration requires substantive meaning or a justified distributional basis; do not choose crossover points merely to optimize solutions.
Report necessity analysis, truth tables, frequency and consistency thresholds, PRI, solution consistency/coverage, and robustness under other reasonable choices.
fsQCA identifies set relations in observational data. Causal claims still depend on theory, timing, design, and excluding alternative explanations.
4. A Template for Integrating Results
Example report
PLS-SEM indicated that [X] had a [ ] net effect on Y. NCA indicated that [Z] was a necessary condition [degree], d = [ ]. fsQCA identified [number] sufficient configurations, with solution consistency [ ] and coverage [ ]. These three evidence layers were interpreted as average effects, necessity constraints, and configurational pathways, respectively.
- Each method has its own question.
- Concepts are consistent and transformations are disclosed.
- NCA includes permutation tests and conditional bottlenecks.
- fsQCA includes anchors, thresholds, and PRI.
- Each method has robustness checks.
- Do not conflate net effects with necessity/sufficiency.
Conclusion
The strength of a PLS-SEM–NCA–fsQCA design lies in the complementarity of its three logics. It is convincing only when each method addresses the appropriate question and findings are integrated through reasoning rather than superficial agreement.
References
- Dul, J. (2016). Necessary Condition Analysis. Organizational Research Methods, 19(1), 10–52.https://doi.org/10.1177/1094428115584005
- Ragin, C. C. (2008). Redesigning Social Inquiry: Fuzzy Sets and Beyond. University of Chicago Press.https://press.uchicago.edu/ucp/books/book/chicago/R/bo5973952.html
Academic note: Before integration, check each analysis’s assumptions and quality separately; the three methods cannot compensate for one another.