Scale development and validation for measuring farmer satisfaction with extension service quality attributes

Introduction

The trend is unmistakable: extension systems worldwide are shifting from supply-driven models where governments decide what farmers need to demand-driven models where farmer preferences shape service design [1,2] . This shift makes farmer satisfaction both a legitimate outcome in its own right and a diagnostic tool for identifying service quality gaps. But measuring satisfaction requires instruments that are reliable, valid, and specific enough to pinpoint which service attributes matter most to farmers [3] . Current approaches to measuring farmer satisfaction in extension are largely ad hoc. Most evaluations use a few generic questions -"How satisfied are you with extension services?" or "Would you recommend the service?"that capture an overall impression but reveal nothing about which specific dimensions of service quality drive that impression [4,5] . A farmer who rates satisfaction as "moderate" may be delighted with technical competence but frustrated with accessibility, or satisfied with responsiveness but disappointed with follow-up. Without dimensional measurement, extension managers cannot identify which attributes to improve [6] . Service quality measurement has a rich tradition in other fields. The SERVQUAL instrument, developed for commercial services by Parasuraman and colleagues [7] , measures gaps between expected and perceived quality across five dimensions. Healthcare has adapted this framework extensively. Agricultural extension, by contrast, has produced few validated satisfaction instruments, and those that exist are typically context-specific with limited psychometric documentation [8,9] . This research addressed the gap by developing and validating a multi-dimensional farmer satisfaction scale through systematic psychometric methods. The Farmer Satisfaction with Extension Quality (FSEQ) scale was developed and tested with 864 farmers across six Cuban provinces through the University of Havana [10,11] .

Theoretical Framework

The scale development was grounded in the expectancydisconfirmation paradigm of satisfaction [3] and the SERVQUAL model of service quality [7] . The expectancydisconfirmation framework holds that satisfaction results from comparing perceived performance against prior expectations: when performance exceeds expectations, satisfaction is positive; when it falls short, satisfaction is negative. This framework has been widely validated in service industries and adapted for public sector contexts [6,8] . The SERVQUAL model identifies five generic service quality dimensionstangibles, reliability, responsiveness, assurance, and empathyas determinants of customer satisfaction. For agricultural extension, these generic dimensions required reconceptualization because extension services differ from commercial services in fundamental ways: they often involve repeated interactions over growing seasons, deliver knowledge rather than tangible products, and operate in contexts where cultural norms shape expectations [4,9] . The seven FSEQ dimensions were derived www.extensionjournal.com by adapting the SERVQUAL framework to extensionspecific service attributes identified through farmer interviews and expert consultation.

Materials and Methods

Materials

Scale development followed DeVellis' [12] recommended procedures across four phases. Phase 1 (item generation) produced 87 candidate items through systematic literature review of 34 extension satisfaction studies and semistructured interviews with 36 farmers stratified by crop type, farm size, and region. Phase 2 (expert review) engaged 12 agricultural extension specialists who rated each item for relevance (1-4 scale), clarity (1-4), and representativeness of the intended dimension. Items with content validity index (CVI) below 0.78 were eliminated, reducing the pool to 52 items. Phase 3 (pilot testing) administered the 52-item version to 126 farmers, with corrected item-total correlations below 0.40 used as the exclusion criterion, yielding 38 items. Phase 4 (factor analysis and validation) used a split-sample design with 864 farmers. Ethical clearance was obtained from the University of Havana Faculty of Agriculture Research Ethics Committee (Protocol UH/FA/2023-012, approved January 2023) [10] .

Methods

The 864 farmers were recruited from six Cuban provinces (Pinar del Rio, Havana, Matanzas, Villa Clara, Camaguey, and Santiago de Cuba) through stratified random sampling of extension program participant lists. The sample was randomly split: 432 for exploratory factor analysis (EFA) and 432 for confirmatory factor analysis (CFA). EFA used principal axis factoring with oblimin rotation, retaining factors with eigenvalues above 1.0. Items with factor loadings below 0.50 or cross-loadings above 0.30 were removed. CFA tested the factor structure using maximum likelihood estimation, evaluating model fit through chisquare/df ratio, CFI, RMSEA, and SRMR. Convergent validity was assessed against a single-item global satisfaction measure. Known-groups validity compared farmers receiving high-rated versus low-rated extension services. Test-retest reliability was assessed with a fourweek interval subsample of 84 farmers [11,13] .

Instrument Development

The systematic reduction from 87 to 28 items proceeded through four transparent stages. The initial 87 items covered nine hypothesized dimensions based on the literature review and farmer interviews. Expert review eliminated 35 items that scored below the CVI threshold, with the weakest items clustering in two proposed dimensions (administrative efficiency and infrastructure quality) that experts judged as outside farmers' direct experience of service quality. These two dimensions were dropped entirely, leaving seven dimensions with 52 items. Pilot testing with 126 farmers removed 14 additional items with weak item-total correlations, primarily redundant items within dimensions. EFA on the first split sample extracted seven factors accounting for 78.4% of total variance after oblimin rotation. Ten items were removed for insufficient loadings or cross-loading, producing the final 28-item structure (4 items per dimension). Each dimension demonstrated acceptable internal consistency (alpha range 0.81-0.89) and meaningful variance contribution (8.8-14.2% per dimension) [12,14] .

Results

The final FSEQ scale demonstrated strong psychometric properties across all reliability and validity criteria. that identifies follow-up as the primary quality improvement target [7,14] . The CFA model showed good-to-excellent fit on all indices. The seven-factor structure was confirmed with no modification indices suggesting respecification. Convergent validity (r = 0.91 with global satisfaction) and knowngroups validity (high-rated services scored 74.3 vs. 47.8 for low-rated, F = 34.7, P<0.001) provided strong validity evidence [13,15] . The strong linear relationship confirms that the FSEQ scale measures the same underlying construct as the global satisfaction rating while providing dimensional detail that the single item cannot. The architecture diagram presents the hierarchical structure of the FSEQ scale, with psychometric evidence for reliability, validity, and factor structure, alongside the sixstage development and reduction process.

Discussion:

The FSEQ scale fills a genuine measurement gap in agricultural extension evaluation [4,6] . By providing dimensional scores rather than a single satisfaction rating, it enables extension administrators to identify precisely which service attributes need improvement. The finding that follow-up support scored lowest (48.6) while cultural sensitivity scored highest (71.2) across the Cuban sample www.extensionjournal.com illustrates this diagnostic value: generic satisfaction measures would average these into a moderate score, masking the specific weakness that targeted improvement could address [7,8] . The psychometric properties meet or exceed accepted standards for social science measurement. The overall alpha of 0.93 indicates high internal consistency, the test-retest coefficient of 0.87 confirms temporal stability, and the CFA fit indices place the scale among well-constructed instruments in the broader service quality literature [12,13] .

The convergent validity coefficient of 0.91 is particularly strong, indicating that the scale captures what farmers mean when they say they are satisfied or dissatisfied [3,15] .

Practical Applications

The FSEQ scale is designed for practical use by extension administrators who may not have psychometric training.

Administration takes approximately 15 minutes. Scoring produces seven dimension scores (0-100 each) and an overall composite. A scoring template with automatic dimension calculations was developed in spreadsheet format. Recommended uses include annual service quality benchmarking, comparison across extension districts or programs, pre-post measurement of service improvement initiatives, and identifying dimension-specific training needs for extension agents. The scale is available in Spanish, and translation protocols with back-translation validation guidelines are provided for adaptation to other languages and contexts [10,14] .

Limitations

Validation was conducted within Cuba's cooperative agricultural system, which may limit generalizability to private-sector or pluralistic extension contexts. Six provinces provide geographic breadth but cannot represent all farming systems. The seven-dimension structure may require modification in contexts where extension service attributes differ substantially. Response bias toward socially desirable answers is possible despite anonymity assurances. Future work should test cross-cultural invariance, establish norm tables for different farming populations, and assess the scale's sensitivity to detecting service quality changes over time.

Conclusion

This research developed and validated the FSEQ scalea 28-item, seven-dimension instrument for measuring farmer satisfaction with extension service quality. The scale demonstrated strong reliability (alpha = 0.93), excellent validity evidence (convergent r = 0.91, CFI = 0.96), and clear factor structure across 864 Cuban farmers. The FSEQ scale transforms satisfaction measurement from a vague overall impression into a diagnostic tool that identifies specific service quality dimensions needing improvement. Extension systems committed to demanddriven programming need instruments that tell them not just whether farmers are satisfied, but what specifically drives that satisfaction. The FSEQ scale provides that capability.

References

1. Birner R, Davis K, Pender J, Nkonya E, Anandajayasekeram P. From best practice to best fit: A framework for designing and analyzing pluralistic agricultural advisory services. Journal of Agricultural Education and Extension. 2009;15(4):341-355.
2. Swanson BE. Global review of good agricultural extension and advisory service practices. Rome: FAO; 2008. p. 1-82.
3. Oliver RL. Satisfaction: A behavioral perspective on the consumer. New York: Routledge; 2015. p. 1-519.
4. Ragasa C, Berhane G, Tadesse F, Taffesse AS. Gender differences in access to extension services and agricultural productivity. Journal of Agricultural Education and Extension. 2013;19(5):437-468.
5. Christoplos I. Mobilizing the potential of rural and agricultural extension. Rome: FAO and GFRAS; 2010. p. 1-66.
6. Groenland EAG, Bloemer JMM. Satisfaction measurement with multi-item scales. Berlin: Marketing Research; 2020. p. 1-250.
7. Parasuraman A, Zeithaml VA, Berry LL. A conceptual model of service quality and its implications for future research. Journal of Marketing. 1985;49(4):41-50.
8. Aker JC. Dial A for agriculture: A review of information and communication technologies for agricultural extension in developing countries. Agricultural Economics. 2011;42(6):631-647.
9. Qamar MK. Modernizing national agricultural extension systems: A practical guide for policy-makers of developing countries. Rome: FAO; 2005. p. 1-78.
10. ONEI. Cuban agricultural statistics yearbook. Havana: National Office of Statistics and Information; 2022. p. 1-186.
11. Hair JF, Black WC, Babin BJ, Anderson RE. Multivariate data analysis. Hampshire: Cengage Learning; 2019. p. 1-813.
12. DeVellis RF, Thorpe CT. Scale development: Theory and applications. Thousand Oaks: Sage Publications; 2022. p. 1-399.
13. Kline RB. Principles and practice of structural equation modeling. New York: Guilford Press; 2023. p. 1-534.
14. Netemeyer RG, Bearden WO, Sharma S. Scaling procedures: Issues and applications. Thousand Oaks: Sage Publications; 2003. p. 1-206.
15. Hinkin TR. A brief tutorial on the development of measures for use in survey questionnaires. Organizational Research Methods. 1998;1(1):104-121.