PLS-SEM and fsQCA in Green Purchase Research: Lessons from Vietnam's FMCG Industry

1. Why is this article worth reading?

For students of quantitative research, the article offers a relatively clear example of combining variable-oriented and configurational analysis within one consumer-behavior study.

Many theses and manuscripts use PLS-SEM to test a causal model and then add fsQCA as an extension. Not every study needs fsQCA, however. It adds value when the research question concerns multiple causal pathways, combinations of conditions, or possible asymmetry between high and low outcomes.

Dinh Van Hoang and colleagues examine consumers' green purchase intention for FMCG products in an emerging market. The topic matters in practice because firms invest in packaging, eco-labels, green advertising and sustainability messages. The academic question extends beyond whether a firm engages in green marketing: do consumers notice it, trust it, and translate that perception into purchase intention?

2. FMCG and green consumption

FMCG stands for Fast-Moving Consumer Goods: products with short consumption cycles, frequent purchases, generally modest spending per purchase and often rapid purchase decisions. These features distinguish green purchasing in FMCG from high-value purchases such as electric vehicles, household appliances or sustainable tourism services.

FMCG consumers may spend little time evaluating a brand's environmental claims. Decisions often involve limited information, constrained budgets, established shopping habits and strong influence from family, friends, social media or distribution channels. Green marketing in FMCG therefore faces two major obstacles: low attention and high skepticism.

The Vietnamese context makes this topic particularly relevant because rapid consumption coexists with substantial price sensitivity. Consumers may appreciate green messages without buying a product that costs more, is harder to find, or appears to involve greenwashing: environmental communication that exceeds the brand's actual environmental practices.

In FMCG, perceiving a brand as green and believing that it genuinely honors environmental commitments are different cognitive states. The gap between them is an important contribution of the article.

3. Research question and theoretical foundation

The article examines how consumers develop green purchase intention from a perceived green marketing strategy. Environmental consciousness and social influence are antecedents. Green brand image and green trust are two important mediating mechanisms.

The main theoretical foundation can be understood through signaling theory. In markets with information asymmetry, firms signal environmental responsibility through green labels, packaging, advertising, certification and sustainability messages. Consumers cannot observe the entire supply chain directly, so they interpret these signals to evaluate the brand.

A signal does not automatically produce behavior. It may make a brand appear greener, but without trust its effect on purchase intention may be weak or nonsignificant. The article therefore distinguishes green brand image from green trust rather than combining them into an ambiguous concept of green perception.

4. PLS-SEM and fsQCA: complementary causal logics

PLS-SEM and fsQCA address different questions. PLS-SEM is variable-oriented: it estimates relationships between latent variables, tests hypotheses, evaluates measurement reliability and validity, and then examines the statistical significance of structural paths. Its central question is how one variable changes as another increases, given the other variables included in the model.

fsQCA is a set-theoretic, configurational approach. Rather than asking how much X affects Y, it asks which combination of conditions is sufficient for a high level of Y. It is particularly relevant when several different pathways can lead to the same outcome, a phenomenon called equifinality.

Combining the methods is justified when the discussion integrates their findings. Simply placing PLS-SEM and fsQCA alongside one another can produce many techniques but little contribution. By contrast, PLS-SEM can describe average effects while fsQCA identifies configurational pathways, allowing the methods to complement each other persuasively.

5. Research design and analytical workflow

The article uses a sequential mixed-method design. Its qualitative phase develops and adapts the model to the study context. The main quantitative phase analyzes 275 valid responses with PLS-SEM and fsQCA. This is a reasonable choice when the goal is both to test a theoretical model and to explore configurations of conditions associated with high green purchase intention.

PLS-SEM assessment should proceed from the measurement model to the structural model. Researchers need to examine indicator reliability, internal consistency, convergent validity and discriminant validity before interpreting path coefficients. Inadequate measurement undermines structural interpretations of causal relationships.

In fsQCA, Likert responses are not analyzed simply as raw scores. They require calibration into fuzzy-set membership scores between 0 and 1. The article uses three anchors based on the five-point scale and the data distribution, then examines necessary and sufficient conditions. Beginners often give calibration too little attention, although it directly affects the final configurations.

6. Reading the PLS-SEM results

The article's PLS-SEM results indicate substantial explanatory power for green purchase intention. Adjusted R2 for GPI is approximately 0.501, meaning the model explains about half of its variance in the survey sample. This is not a low level in consumer-behavior research, but R2 does not establish causality in the sense of a controlled experiment.

The supported relationships include EC → GMS, SI → GMS, GMS → GBI, GMS → GPI, GMS → GT, GT → GPI and GBI → GT. GBI → GPI and mediation through GBI in GMS → GPI are not supported. By contrast, green trust has significant direct and mediating roles.

• Supported: EC → GMS: consumers with greater environmental consciousness tend to recognize and evaluate green marketing strategies more positively.
• Supported: SI → GMS: social influence is a strong antecedent, consistent with consumers being influenced by communities, reference groups and social media.
• Supported: GMS → GBI, GMS → GT and GMS → GPI: perceived green marketing strategy affects brand image, trust and purchase intention.
• Not supported: GBI → GPI: green brand image has no statistically significant direct effect on green purchase intention in this model.
• Supported: GT → GPI and GMS → GT → GPI: green trust is more important than green brand image as a mechanism translating green marketing into purchase intention.
• Additional relationship: GBI → GT: green brand image still matters because it helps build trust, even without directly increasing purchase intention.

Statistical nonsignificance does not mean a concept is useless. Here, green brand image does not directly lead to purchase intention, but it is associated with green trust. A suitable interpretation is that a green image may be an initial cognitive layer, yet may be insufficient to translate into purchase intention without trust.

7. Reading the fsQCA results

The fsQCA results add another interpretive layer. GMS, GT and GBI all have consistency above 0.8 in the necessity analysis, but do not exceed the 0.9 threshold for strict necessity. A cautious description is that these conditions are nearly necessary or highly important for high GPI, rather than claiming that each condition is absolutely necessary on its own.

For sufficiency, the article reports two main configurations for high GPI. The first is the presence of GMS; the second combines GT and GBI. Solution coverage is 0.980 and solution consistency is 0.841. The reported configurations therefore cover most high-outcome cases and have an acceptable level of consistency under an fsQCA interpretation.

8. Green trust and green brand image: the main lesson

The most instructive finding is not that green marketing has a positive effect, which is relatively predictable. It is the distinction between green brand image and green trust. The former has no significant direct effect on green purchase intention, whereas the latter has direct and mediating effects.

The finding can be understood at three levels. First comes recognition of a brand's green image. Second comes belief that the brand's environmental commitments are trustworthy. Third comes intended behavior: willingness to buy. The first level does not automatically lead to the third; the second makes that translation more convincing.

This is plausible in FMCG, where consumers encounter many green claims: environmentally friendly packaging, eco-labels, plastic-reduction messages and low-emission commitments. As these claims become widespread, a green image may lose its ability to differentiate brands unless supported by credible evidence, consistent experiences or third-party confirmation.

For thesis and article writers, an unexpected finding is not necessarily a weakness when it is explained in a suitable theoretical and practical context. It can make the contribution clearer than a model that merely reconfirms familiar relationships.

9. Common misunderstandings when combining PLS-SEM and fsQCA

Many doctoral manuscripts struggle because of misunderstanding the methods' roles rather than incorrect software use. Avoid the following errors when planning a similar study.

10. Practical checklist

Use this checklist before deciding to combine PLS-SEM and fsQCA in marketing, management, consumer behavior or education research.

• Does the research question have two levels? One concerns effects between variables; the other concerns combinations of conditions or multiple pathways to an outcome.
• Does the theory support configurational reasoning? Do not add fsQCA if the theory anticipates only a simple linear relationship.
• Has measurement been adequately assessed? Evaluate loadings, reliability, AVE and HTMT before interpreting the structural model.
• Is fsQCA calibration justified? Anchors should reflect theory, measurement scales, the data distribution or accepted practice in the field.
• Do you distinguish necessary from sufficient conditions? A condition may be important without meeting the threshold for strict necessity.
• Does the discussion integrate both methods? Show what fsQCA adds to PLS-SEM rather than repeating the same conclusion.
• Have you reported the study's limitations? Cross-sectional data, self-reports and a specific context all limit generalizability.

11. Key terminology

• PLS-SEM: Partial Least Squares Structural Equation Modeling. A variance-based approach to structural equation modeling, commonly used with latent-variable models and prediction-oriented research.
• fsQCA: Fuzzy-set Qualitative Comparative Analysis. A fuzzy-set method for identifying necessary conditions, sufficient conditions and causal configurations.
• Green trust: Consumers' belief that a brand reliably fulfills its environmental commitments.
• Green brand image: Consumers' associations and perceptions of a brand's green or environmentally friendly image.
• Calibration: The process of converting raw data into set-membership scores in fsQCA, usually between 0 and 1.
• Consistency: The extent to which a condition or configuration is consistent with an outcome under set-theoretic reasoning.
• Coverage: The extent to which a condition or configuration covers the outcome cases to be explained.
• Equifinality: The possibility that different causal pathways lead to the same outcome.