Repertory Grid Technique and Honey’s Content Analysis: gathering and processing qualitative data

Abstract

Purpose

This study aims to demonstrate a rigorous approach to applying the Repertory Grid Technique (RGT) and Honey’s Content Analysis (HCA) to obtain and process qualitative data through structured interviews.

Design/methodology/approach

An illustrative case study using the OpenRepGrid package from the open-source software R facilitates a deeper understanding of these techniques. The study subjects were employees of a corporate charter company.

Findings

The RGT enables the identification of key attributes as perceived by interviewees regarding the phenomenon, whereas HCA clarifies how these attributes impact the desired analysis outcome. The presented case study identified constructs related to the client–supplier relationship and their impact on service performance from the provider’s perspective.

Research limitations/implications

This study illustrates the use of qualitative methods based on an interpretative naturalistic approach to rigorously and systematically capture interviewees’ perspectives.

Practical implications

The combination of RGT and HCA can be a valuable tool for management studies by allowing controlled researcher interference in empirical investigations. In addition, the data-driven selection of constructs by interviewees can lead to the emergence of novel theories.

Social implications

Using diverse methodologies enables researchers to address complex managerial challenges that often surpass the capabilities of conventional analysis methods.

Originality/value

The proposed methodology offers a robust understanding of phenomena from the interviewees’ perspectives. Consequently, this study highlights the potential of these techniques for theoretical and empirical research in the field of administration.

Citation

Bastos Carvalhais Barroso, M., Martins, R.S. and Freitas, J.S. (2024), "Repertory Grid Technique and Honey’s Content Analysis: gathering and processing qualitative data", RAUSP Management Journal, Vol. 59 No. 4, pp. 386-401. https://doi.org/10.1108/RAUSP-11-2022-0249

1. Introduction

We focus on qualitative research as an interpretive naturalistic approach that emphasizes the qualities of the investigated entities rather than their quantities (Gephart, 2004). This approach studies phenomena in their natural environment, concentrating on the meanings attributed to these phenomena by the social actors involved. This type of research is crucial as it provides insights that are challenging to obtain through quantitative methods, such as understanding social processes and evidence of human interactions and meanings. Therefore, more qualitative studies are essential for expanding knowledge across various management fields (Goffin, Raja, Claes, Szwejczewski, & Martinez, 2012).

This study uses a methodology designed to capture individuals’ subjective interpretations of a particular phenomenon. Structured interviews, typically conducted through questionnaires and models, have limitations regarding discovering new insights and the risk of contamination by the interviewer’s preconceived perspectives. Conversely, informal interviews lack focus and consistency in the interviewees’ responses. Consequently, despite management researchers advocating for the discovery of appropriate ways to handle tacit knowledge, there is a dearth of techniques to manage this type of knowledge effectively (Jankowicz, 2001). According to Goffin et al. (2012), interviews are challenging due to the ambiguity of the subject matter, leading to superficial answers to direct questions.

Expanding the use of robust qualitative methods is necessary to obtain reliable information through interviews. This article addresses the research question:

RQ1.

How can one capture an interviewee’s construction of meanings within a domain of knowledge in a robust and structured manner?

To answer this question, we present an application of the Repertory Grid Technique (RGT) and Honey’s Content Analysis (HCA) to understand interorganizational relationships between clients and providers of corporate charter services from the service provider’s perspective.

The article is structured as follows: Section 2 provides a brief overview of RGT and HCA, exploring these techniques within the context of qualitative methods. We emphasize that the article aims to demonstrate a technique rather than engage in a paradigmatic discussion. Section 3 presents the application of the suggested methodology, investigating the relationships between clients and suppliers. We explain the study’s objectives, the application of RGT and the step-by-step process of HCA. Section 4 discusses how rigor can be enhanced in qualitative studies and the potential of the method, categorized as formal qualitative, to contribute to management studies by introducing a more traditional approach to interpretivism. We also highlight the method’s constraints within the defined scope. Section 5 summarizes the implications for management researchers.

2. The selected methods

The proposed met6hodology integrates the RGT with HCA, forming a qualitative approach centered on clarifying the meanings individuals attribute to the studied phenomenon. This application in management studies clarifies variables within their contextual frameworks, addressing complex phenomena unsuitable for surveys and experiments (Gil, 2002). The emergence of qualitative methods, often as a critique of quantitative approaches, underscores a diverse research practice that proves its value independently, encompassing both exemplary and subpar research (Flick, 2009).

2.1 Repertory Grid Technique

Developed by George Kelly in his seminal 1955 work, “The Psychology of Personal Constructs,” the RGT is grounded in the theory of personal construct psychology. This theory asserts that the evidence is not based on objective facts but on individuals’ interpretations of these facts. Consequently, RGT is applicable across various domains (Jankowicz, 2001). According to Goffin et al. (2012), RGT effectively clarifies constructs when existing literature fails to do so, making it particularly suitable for exploratory studies where constructs remain ambiguous.

One fundamental assumption of RGT is constructive alternativism, positing that individuals interpret the same phenomenon differently and a person’s interpretation may vary over time (Jankowicz, 2004). This aligns with the epistemology of interpretivism, or social constructivism. As Creswell and Poth (2018) explained, interpretivism involves individuals seeking to understand their world by developing subjective meanings from their experiences. Interpretive researchers aim to describe and comprehend the meanings held by group members and the implications of differing interpretations in social interactions (Gephart, 2004).

Jankowicz (2004) defined RGT as a set of rating scales using the individual’s constructs as the basis for evaluation. This classification procedure captures direct descriptions of personal perceptions of reality, using the individual’s language. RGT interviews, a structured method, accurately reflect participants’ perspectives without interviewer bias. Although interviews can explore complex phenomena, direct questions often yield superficial responses due to topic ambiguity. Thus, RGT’s indirect approach allows for a deeper understanding (Goffin et al., 2012).

Occupations involving skill-based and cognitive tasks, such as strategic decision-making by managers, rely on experience, intuition and subjective judgment. By mapping mental representations, RGT effectively makes tacit knowledge explicit (Jankowicz, 2001). This approach allows researchers to investigate the complexity of viewpoints without imposing restrictions. Moreover, it emphasizes relying on participants’ perspectives, with meaning emerging through interpersonal interactions (Creswell & Poth, 2018). Thus, interpretive research builds social science concepts inductively, grounded in social actors’ concepts (Gephart, 2004).

Smith (1980) was a pioneer in applying RGT in administration, advocating for its use in management due to its ability to structure and influence managerial thought processes. Easterby‐Smith, Thorpe, and Holman (1996) noted that this technique has gained traction in management development and research. Its use in administration captures subjective perceptions and cognitive processes, offering a deeper understanding of phenomena through individual experiences and viewpoints. Rogers and Ryals (2007) further explored RGT’s strengths and weaknesses in understanding complex business relationships, demonstrating its effectiveness in examining nuanced topics and decision-making mechanisms. Goffin et al. (2012) highlighted its application in supply chain management, noting that RGT provides insights that conventional methods often miss, enriching qualitative research with quantitative analyses. Ultimately, RGT’s capacity to capture subjective perceptions and cognitive processes makes it invaluable for generating deeper insights into individual experiences and perspectives within administrative studies.

RGT consists of four components: topic, elements, constructs and ratings (Figure 1). The topic defines the specific area of interest. Elements are specific, comparable items used to elicit constructs related to the topic. Constructs – basic units of description and analysis formed by opposing poles – are identified by systematically comparing elements. For example, “reliable” versus “unreliable” becomes meaningful only when the implied contrast is understood (Jankowicz, 2001). Ratings involve evaluating each element within each construct, capturing the interviewee’s perspective on each element (Jankowicz, 2004). In addition, RGT considers that standard constructs or “general summaries” encapsulate the interviewee’s overall view on the topic.

In the example (Figure 1), the topic is interorganizational relationships, with corporate charter clients as elements. Three clients, “3,” “7” and “8”, are chosen for comparison. The interviewer asks, “What do two of these elements have in common that differentiates them from the third?” The respondent might say clients “7” and “8” are more reliable than client “3.” If “reliable” is deemed a positive attribute, it is recorded on the left side of the grid, whereas its opposite, “uneven demands,” is noted on the right. Clients are then rated on a scale from 1 (reliable) to 6 (uneven demands). This process continues until no further constructs are obtained. Finally, the standard constructs are explained, and the interviewee rates each client accordingly.

2.2 Honey’s Content Analysis

HCA is applied when multiple grids are developed, each containing constructs identified by individual interviewees. Aggregating these varied constructs to examine the similar meanings within the grids while maintaining the unique perspectives of each interviewee to the greatest extent possible is essential (Jankowicz, 2004). The constructs are first systematically categorized to achieve this aggregation:

The categorization of constructs provides the most important technique for analyzing more than one grid, being used to indicate how a group, or sample of people, construe some topic of common interest and, by implication, to say something about how individual members of the group compare with each other (Jankowicz, 2004, p. 88).

The constructs provided by all respondents are systematically grouped and categorized based on the meanings they convey. These categories are derived from the constructs themselves using the bootstrapping technique. By identifying similarities and differences among the constructs, the meanings are aggregated into similar or distinct categories (Jankowicz, 2004). After this process, a reliability test is conducted, validating the findings with each respondent or involving a second evaluator to ensure consensus.

This methodology allows for aggregating diverse meanings expressed in a set of grids into general statements about the sample. However, traditional content analysis does not incorporate the use of element classifications available in the original grids. To address this limitation, Honey (1979) developed a technique that aggregates various constructs from a sample and enables the use of individual meanings conveyed by each respondent’s evaluations (Jankowicz, 2004).

According to Jankowicz (2004), HCA assigns two indices to each construct, reflecting the degree to which the classifications of the specific construct align with the overall construct classifications. The first index is the similarity score. Consequently, the similarity score interprets the extent to which each construct matches the standard construct. This score is obtained by transforming each sum of differences between the construct and the overall construct into percentages. The calculation is performed according to the equation below. SD is the sum of the differences between the element ratings of the two analyzed constructs, N is the Likert point scale used and E is the number of elements:

100− SD(N−1)× E × 100

The second index, H-I-L (high, intermediate and low-level), indicates that individuals vary in their typical similarity scores. The respondents’ different ranges of similarity scores are associated with specific personal metrics. Accordingly, Honey’s procedure acknowledges that similarity scores are relative. In addition to recording their actual percentage values, it identifies whether they fall within each individual’s high, intermediate or low ranges (Jankowicz, 2004). For every interviewee, the similarities between constructs are categorized into three levels. For instance, for a respondent whose similarity values range between 80% and 95%, a construct with 80% similarity would be classified as low level. Conversely, for an interviewee with similarity values ranging from 40% to 80%, a construct with 80% similarity would be classified as high level.

According to Rojon, McDowall, and Saunders (2018), categories encompassing high- and intermediate-level constructs are retained in forming an initial model as individual participants deem them significant. Categories with more low-level than high-level constructions are discarded, as participants do not strongly associate them with the topic.

This methodology preserves individual richness by aggregating personal constructs at the group level within a constructivist social structure. Honey’s (1979) approach facilitates aggregation, providing a flexible and multifunctional methodology. This technique aligns with the theory of personal construct psychology, as the constructs elicited by individuals result from social interactions (Rojon et al., 2018). Finally, the general findings must be validated through interviews to ensure they accurately reflect the interviewees’ views.

2.3 Repertory Grid Technique and Honey’s Content Analysis in qualitative studies

According to Wright (2017), significant attention has been directed toward grid design, administration, information acquisition and data analysis. Nevertheless, a comprehensive understanding of the underlying driving force behind the method, rooted in Kelly’s (1969) personal construct theory, is essential for its proper utilization. Personal construct psychology fundamentally deals with “future anticipation” (Wright, 2017). As Kelly (1955, p. 34) articulated, “it is the future which tantalizes man, not the past. Always he reaches out to the future through the window of the present.” Kelly further elaborates on his theory through 11 associated “corollaries,” all grounded in constructs – ways in which we anticipate, perceive, interpret and make sense of our lived experiences. It is crucial to note that a defining characteristic of people’s constructions is their bipolarity; Kelly strongly believed that we understand the world by comparing similar and dissimilar things (Wright, 2016). Therefore, by understanding a person’s construct system, we can better anticipate how they perceive the world and the direction of their behavioral actions (Wright, 2017).

Thus, the foundational theory of the RGT, known as Kelly’s personal construct theory (1969), aligns with constructivism. According to Guba and Lincoln (1994), paradigms consist of a set of fundamental beliefs representing a worldview that defines, for its adherents, the nature of the world, their place within it and the spectrum of possible relationships with that world and its components. Within paradigms, these authors present constructivism, which aims to understand and reconstruct individuals’ initial constructions, seeking consensus while remaining open to new interpretations as information and sophistication increase.

One contentious debate within the organizational studies field pertains to using “templates” in qualitative research. Gioia et al. (2022) convened various proponents of this debate. Some argue that if templates “codify best practices and conventions for a particular qualitative method” (Harley & Cornelissen, 2020), then the issue behind their inappropriate use may stem from the initial phase of systematizing the tacit knowledge developed by early users. Conversely, others (e.g. Mees-Buss, Welch, & Piekkari, 2020), aligned with the hermeneutic tradition, criticize approaches like the Gioia Method for providing only a theorized and abstract version of informants’ expressions. For this group, the use of templates is restricted to interpreting qualitative research acceptable to the neo-positivist mainstream. However, the qualitative community generally concurs that templates are neither inherently good nor bad; their significance lies in their research application (Gioia et al., 2022).

Thus, we acknowledge the limitations of combining RGT and HCA as adherents of the naturalistic interpretative paradigm. As Mees-Buss et al. (2020) highlighted, social research, particularly interpretative qualitative research, faces the challenge of addressing the subjectivity of the social world without compromising its scientific integrity. The underlying premise of the naturalistic approach is that researchers can capture the subjective meaning participants attribute to their actions, faithfully reproducing their experiences as narrated by themselves (Gubrium & Holstein, 1997). Guba and Lincoln (1982) explained that this perspective does not deny the mutual influence between researchers and the researched; instead, it suggests that researchers actively explore the perceptions resulting from this inherent interaction. To manage subjectivity appropriately, researchers must carefully handle and process data to identify abstract categories that align with participants’ actions and words. Consequently, theory develops from the data rather than being imposed on them.

Alternative methods within the same conceptual framework, specifically those oriented toward eliciting and analyzing cognitive mental models, share the goal of maintaining formality in subjective analysis. Consequently, Kelly’s theory aligns with other methodologies grounded in a similar theoretical standpoint. For instance, Policy-Capturing, an experimental technique capable of providing powerful insights on the cognitive bases of decision-making processes (Nokes & Hodgkinson, 2017), exemplifies this. Notably, the theory distinguishes between idiographic dimensions, which seek to identify unique aspects of individuals’ decision policies and mental representations, and nomothetic dimensions, which aim to identify “general trends” in decision policies across the sample (Aiman-Smith, Scullen, & Barr, 2002).

This article does not intend to foster a paradigmatic discussion but rather to demonstrate a technique from the perspective of actionable knowledge. According to Jarzabkowski and Wilson (2006), these are knowledge artifacts that research offers to managers in practice. A practical perspective draws on philosophical traditions like American pragmatism and Aristotelian perspectives on action through practical reasoning and engagement. It has become more prominent in social sciences as a “practical turn.” Fundamentally, a pragmatic viewpoint entails using knowledge to address specific requirements of a given scenario. Thus, to understand actionable knowledge, we should not only examine theories available for use but also, more pertinently, evaluate whether the artifacts resulting from these theories are applicable (Jarzabkowski & Wilson, 2006).

3. Application of the methodology

3.1 Research design

This case study uses a specific methodology to address the following research question:

RQ1.

Which interorganizational governance mechanisms are associated with the technical performance of corporate charter services from the service provider’s perspective?

By deepening the understanding of interorganizational governance mechanisms and their impact on performance within the empirical field of corporate charter services, this study contributes to the existing literature on service production. In addition, it enhances comprehension from the supplier company’s perspective within the dyad, which possesses technical knowledge of the service. Simultaneously, the client determines the interorganizational governance mechanisms, i.e. the criteria for executing the transportation. The RGT was used to explore how interviewees implicitly differentiated governance mechanisms. Moreover, to identify the “general trends” in decision policies across the sample by accessing the subject’s cognitive frameworks, RGT was combined with HCA.

3.2 Sample selection

The sample for applying the RGT consisted of employees from the charter company selected for interviews. Consequently, a case study strategy was adopted. According to Gil (2002), a case study involves an in-depth examination of one or a few subjects to achieve comprehensive knowledge. It also facilitates the investigation of contemporary phenomena within their context, mainly when the boundaries between facts and context are not well-defined (Yin, 2001).

The case selection was carried out theoretically, meaning it was chosen for its particular suitability in elucidating and expanding the relationships among constructs (Eisenhardt & Graebner, 2007). Specifically relevant criteria that made this case compelling include the large number of charter contracts, the average tenure of boundary-spanning employees, and the direct involvement of top management in interorganizational relationships. Over the past five years, the company has provided corporate charter services to ten firms, with four active charter clients at the time of the study. The substantial volume of charter contracts and the company’s expertise in the sector facilitate the comparison of governance instruments adopted by various clients from the perspective of service experts.

The topic selected for the RGT application was interorganizational relationships. The researcher focused on the last ten charter service contracts executed for different client companies. The number of clients aligns with Jankowicz’s (2004) recommendation of using 5–12 elements, as this range allows for effective case comparison. The first sample consisted of five individuals, all boundary spanners within the company directly involved in the customer relationship management process. These individuals were responsible for the last ten charter services performed by the company and were chosen due to their extensive tenure with the company.

A second sample of ten additional employees was included to incorporate perspectives from various company areas. These employees had indirect relationships with management activities related to at least five of the company’s most recent charter contracts. The selection of five elements was deemed necessary to enable a proper distinction between cases (e.g. Jankowicz, 2004). Only those with relevant experience in the subject matter were selected to ensure the interviews were conducted with knowledgeable individuals.

3.3 Data collection

Data collection took place in a designated room within the company’s case unit facility, which was reserved to minimize distractions during the technique’s application. Fifteen interviews were conducted, each lasting between 60 and 105 min, leading to a total of 20 h.

Before applying the RGT, the procedure was briefly explained to the interviewees using a standard presentation prepared by the researcher. The explanation emphasized that the RGT aims to understand how the interviewee perceives and compares the elements systematically, noting that there are no right or wrong answers (Jankowicz, 2004).

The procedure began after the interviewees understood the technique’s purpose and mechanics. Three elements were randomly selected, and the interviewee was asked to compare them. An attribute describing one end of the grid was written on one side. The interviewee was then asked to identify the opposite of this attribute to establish a contrast. Based on the interviewee’s responses, the researcher organized the construct with the positive aspect on the left and the negative aspect on the right. The left side of the grid represented the 1 end of the scale, whereas the right side represented the 6 end of the scale. According to the construct, the respondent assigned scores ranging from 1 to 6 for each element.

This process continued until as many constructs as possible were identified, with the intention of obtaining between 8 and 12 constructs per interviewee (Jankowicz, 2004). In addition, a standard construct such as “in general – allows service satisfaction” and “in general – hinders service satisfaction” was considered. It was explained that these overarching constructs represent the interviewer’s perspective on the overall technical performance of the service for the respective client, referred to as service satisfaction within the business context.

The outcome of the data collection is a set of grids, each reflecting the constructs identified by individual interviewees. The data captured in these grids is represented both in words and numbers. Words explicitly communicate the constructs, whereas numbers quantify the elements according to the various constructs within the grid (Jankowicz, 2004).

3.4 Data analysis

3.4.1 Stage 1: Individual decision-maker focus.

The initial data analysis phase used methods designed to explore relationships within individual grids. These procedures aimed to identify the factors and patterns that service providers use to interpret the governance instruments defined by the client company. The analysis involved calculating correlations between the constructs within each grid using the OpenRepGrid R software package. This approach facilitated the identification of critical factors that service providers consider when interpreting governance instruments and allowed for recognizing patterns and differences between the governance frameworks of various organizations.

Kendall’s and Spearman’s correlations are nonparametric methods commonly used in applied sciences. This study selected Kendall’s correlation due to its robustness and slightly greater efficiency compared to Spearman’s correlation, making it the preferred estimator for this analysis (Croux & Dehon, 2010).

The correlation between constructs was calculated for the analysis of individual grids. By examining these correlations, the researcher could identify similarities in how interviewees discussed the elements, thereby making the interviewees’ tacit knowledge explicit making it possible to interpret their perspectives (Jankowicz, 2004). This correlation analysis helped identify patterns in the governance frameworks of different organizations. Specifically, the analysis of correlations between constructs listed by each boundary spanner provided insights into the relationships between variables as reported by the same interviewee. A Kendall’s correlation greater than 75% was used as the criterion to denote a high-intensity relationship. Table 1 illustrates an example of constructs with nonobvious relationships.

3.4.2 Stage 2: General trends across the sample focus.

The analysis of multiple grids enabled the identification of primary factors considered by experts in the field, providing insights into the frequency of reference to governance instruments. This phase involved examining patterns across these grids to identify similarities in the perceptions of interorganizational governance mechanisms among employees within the same organization.

Initially, the analysis focused on the ratings of the overarching construct. Statistical measures, including mean, range and standard deviation, were calculated from the aggregated scores. Higher scores given by suppliers indicated poorer performance of the charter service. Clients were then ranked in ascending order based on their average overall technical performance construct scores. Objective characteristics differentiating clients with higher and lower scores were investigated. Based on documentation reviewed for each service contract and client, two main aspects that impact the technical performance of the service emerged: the client company’s field of activity and location. Services provided to specific sectors or within certain locations exhibited comparatively lower technical quality as assessed by the supplier.

Respondents’ constructs were grouped into categories through a coding process. This involved a detailed interpretation of each interviewee’s explanations of their constructs. Constructs were analyzed to identify commonalities in meaning, and categories were refined iteratively untiln each category represented a distinct significance. Individual confirmation sessions were conducted with each interviewee to address biases and discrepancies. The 15 interviews produced 205 constructs, categorized into 28 final groups.

The frequency of mention served as a measure of importance. Categories of constructs mentioned by more than 25% of the total sample (including all respondents) and sample A1 (boundary spanners only) were highlighted. Table 2 shows a segment of the resulting data.

Applying these criteria, we identified 13 critical attributes in the interorganizational relationship from the supplier’s perspective. These key attributes were further analyzed using HCA. The most frequently cited construct category was “Respect for the Supplier” (mentioned by 73% of respondents), followed by “Learning and Improvement,” “Bureaucracy” and “Joint Problem-Solving” (each cited by 60% of respondents).

HCA facilitated the identification of critical factors considered by service experts in understanding governance instruments. The analysis used the construct “in general: high satisfaction with the service – in general: low satisfaction with the service” to compare different grids. Similarities between constructs were calculated following Jankowicz’s (2004) methodology (see Table 3). This index measures similarity by summing the differences between elicited and provided constructs. The H-I-L index was assigned to each elicited construct, indicating its similarity score and impact on service satisfaction.

HCA highlights the relationship between elicited constructs and the overall construct (Raja et al., 2013). The H-I-L index is used to categorize constructs as high (H), medium (I) or low (L) similarity. The general similarity score for each category is calculated by assessing the percentage of constructs at each level. Key attributes significantly impacting service performance included “Specific Investment” (H index 50%), “Financial Default” (H index 43%), “Mediated Power” (H index 43%) and “Service Location” (H index 40%).

3.5 Results

The findings indicate that, from the respondents’ perspective, interorganizational governance mechanisms significantly impact the technical performance of the provided service. Among the constructs elicited by respondents, categories such as respect for the supplier, learning and improvement, bureaucracy, and joint problem-solving emerged as the most frequently cited. These factors were identified as critical in differentiating corporate charter clients and, consequently, in shaping the interorganizational governance mechanisms stipulated in the contracts. For example, the substantial emphasis on respect for the supplier highlights variations in how boundary spanners from client companies interact with supplier employees. Sometimes, the dominant party exerts authority over the supplier based on the contractual relationship (Huo, Flynn, & Zhao, 2017). This authority is frequently perceived as arrogance, aggressiveness, disrespect and rudeness in the studied interpersonal dynamics.

Respondents evaluated each relationship based on a service performance metric, which was considered as an aggregation of quality, cost, speed, flexibility and reliability of transportation. HCA revealed that specific investment, financial delinquency, mediated power and service location significantly impact performance from the respondents’ perspective. For instance, service location and specific investment – associated with service characteristics and the specific transportation needs of the client company – are critical determinants of service performance. Service location emphasizes the importance of road infrastructure and the cities where the service is delivered, while specific investment pertains to the variability and characteristics of the vehicles used in the service. These variables are therefore fundamental to charter transportation operations and crucial to understanding technical performance from the supply perspective.

The analysis also revealed that service providers prioritize relational governance mechanisms over contractual ones when interpreting client-defined governance factors. This finding challenges the theoretical expectation that suppliers place greater emphasis on contractual instruments compared to relational ones (Um & Oh, 2020). Given the substantial power of client companies, often global enterprises, and the limited agility of the Brazilian legal dispute system, suppliers may not rely exclusively on formal contracts to manage uncertainties and sustain ongoing relationships. Moreover, as noted by Alvim (1984), charter contracts are generally drafted by the client’s legal department and focus primarily on the service provider’s obligations.

4. Rigor, contributions and limitations

Sophisticated methods can enhance research rigor; however, demonstrating rigor is more crucial than the specific methods used (Goffin et al., 2012). The method itself, serving merely as a template, does not inherently provide rigor to scientific work. Its significance lies in its application within the research context (Gioia et al., 2022). While the proposed method may facilitate methodological rigor, the research quality depends on various factors.

When evaluating management research, specific criteria are implicitly or explicitly used (Savall, Zardet, Bonnet, & Peron, 2008). These criteria establish a boundary to prevent low-quality, unreliable or illegitimate work from entering the mainstream. However, Symon, Cassell, and Johnson (2018) highlighted the inherently precarious nature of evaluation within the management field. Despite its frequent portrayal as a straightforward application of consensus-based benchmarks, the evaluation process is fraught with epistemological ambiguities.

Interpretivism in research methodology involves several critical criteria for establishing an internally reflexive audit trail. These criteria include credibility, dependability, confirmability, validity and the potential for transferability or logical inference (Symon et al., 2018). Pratt (2009) underscored the importance of incorporating fundamental elements within the methods section to ensure high-quality research. These elements include justifying the research, assessing its contribution to theory, and providing context-specific rationale for the chosen unit of analysis. In addition, data should be presented clearly and coherently, demonstrating how findings are derived from the collected data. Using organized data to construct the research narrative and modeling successful qualitative work styles by established authors are also recommended to achieve consistency in qualitative research publishing.

Guba and Lincoln (1989) proposed criteria for assessing qualitative research, emphasizing confirmability, credibility, transferability and dependability. Riege (2003) provided insight into each criterion. Confirmability examines the logical and unbiased developments from data interpretations, ensuring the integrity of findings by objectively linking data to conclusions. Credibility refers to the extent to which findings are verified by interviewees or peers, acknowledging multiple interpretations of reality. Transferability assesses the generalizability of findings across different contexts based on the studied context. Dependability requires a detailed description of all research stages to ensure transparency.

The RGT is valuable for exploring ambiguous constructs. To enhance internal validity and credibility, parallel coding by two researchers or cross-validation with interviewers is crucial for maintaining consistency in data interpretation. The RGT provides a structured framework that can be applied across various cohorts. The combination of RGT and HCA significantly strengthens research through structured data collection and analysis.

However, the RGT and HCA are not without limitations. According to Alexander, Van Loggerenberg, Lotriet, and Phahlamohlaka (2010), the RGT is most suitable for research involving participants with substantial practical experience due to the method’s specific requirements. The method’s applicability is contingent upon having a comparable amount of elements, which may restrict its use in cases of excess or scarcity of elements. In addition, the elements must be concrete and practical examples, avoiding abstract entities. Constructs must be bipolar, excluding nonbipolar constructs from the method. In the case of HCA, handling a large number of interviewees with diverse experiences may result in challenges in aggregating constructs across different participants. Furthermore, while HCA retains interviewees’ constructs, some data intricacies may be lost during the aggregation and interpretation of results.

Another limitation of the proposed method is the paradox of sensemaking. Despite moving away from the structure-functionalist approach of traditional cognitive theories, there is an attempt to establish ostensibly ‘objective’ knowledge about sensemaking. This endeavor relies heavily on micro/interpretive methods for data collection and uses systematic comparison approaches to understand the meanings individuals attribute to their experiences. The goal is to uncover patterns and consistent connections within sensemaking processes. However, while valuing subjective and socially constructed human experiences, researchers strive to detach from these experiences and render them objective. As Allard-Poesi (2005) notes, this approach may lead researchers to contradict their conceptualization of sensemaking and risk diluting, if not losing, the essence of sensemaking they aim to clarify.

5. Conclusion

The primary objective of this study has been to develop a systematic approach for researchers to effectively capture and analyze an interviewee’s construction of meaning within their specialized domain. To achieve this, the article presents a robust methodology that integrates the RGT and HCA for collecting and analyzing data from structured interviews. This methodology emphasizes a qualitative investigative approach combined with structured interpretation and data analysis, making it replicable. Adopting this method significantly enhances and supplements knowledge dissemination practices in fields dealing with inherently ambiguous phenomena rooted in social interactions and subjective meanings.

The application of RGT enables the identification of critical attributes emphasized by interviewees concerning the specified topic. HCA further clarifies the influence of these attributes over different interviewees regarding the established construct. This methodological approach provides comprehensive insights into the nuances of each attribute and identifies those with the most significant impact on the overarching metric, specifically focusing on service performance within the given context. However, it is essential to note that while RGT and HCA effectively uncover implicit rules of decision-makers, they do not represent an exhaustive list of methods.

From a pragmatic perspective, the study emphasizes the use of knowledge to address specific scenarios and offer practical insights for managerial practice. The proposed methods are highly applicable to various issues, with the central focus being the comparison of different objective elements and understanding the interviewee’s perspective. This study’s contribution is to present a methodology that involves structured interviews, which addresses the limitations associated with interviewer bias in questionnaire-based approaches. It also mitigates challenges related to informal interviews, such as lack of focus and the absence of standardized patterns across interview sessions. In contrast, HCA provides a protocol for comprehensive analysis of these structured interviews.

Considering the dichotomy between transferability and replicability, the strength of this methodology lies in its capacity for widespread application across diverse contexts, rooted in its commitment to understanding individual constructs. Exploring similar methodologies from other disciplines within management studies offers promising opportunities for gaining insights into cognition and decision-making processes, expanding the tools available for effectively understanding and addressing managerial challenges.

For instance, the empirical study provides valuable managerial insights applicable to companies engaged in service-oriented interorganizational relationships, particularly in contexts characterized by significant power imbalances between clients and suppliers, where the former holds a dominant position. Achieving superior service performance relies on fostering enduring, forward-thinking relationships. For the service recipient, demonstrating a genuine interest in understanding the service’s nuances and the constraints the supplier faces is crucial. This proactive approach reduces the risk of imposing demands that could negatively impact service quality.

In summary, this study advocates for the practical application of methodologies aligned with a naturalistic interpretative paradigm to clarify the tacit criteria used by decision-makers through structured interviews. The expanded and theoretically grounded use of these methods facilitates controlled researcher intervention in empirical inquiries and allows for the organic emergence of novel theories and insights from the data. This approach advances scholarly understanding and has significant implications for social contexts, potentially reshaping managerial strategies and enhancing decision-making processes within organizational frameworks.

Figures

Figure 1.

Repertory grid example

Table 1.

Correlation between constructs

Positive pole	Negative pole	Correlation
Easy to handle	Difficult communication with the client	97%
Extra requests are decided quickly	Extra requests are time-consuming

Source: Authors’ own work

Table 2.

Categories of listed constructs

Category	Description	Frequency of mention
		A1 + A2 (%)	A1 (%)
Respect for the supplier	The client has respectful interpersonal relationships with supplier employees	73	60
Learning and improvement	Continuous improvement of service driven by the customer	60	80
Bureaucracy	Complexity and slowness of processes	60	20
Joint problem solving	The customer seeks to solve problems jointly with the supplier	60	40
Excessive demand	The customer presents excessive rigidity, constant complaints and a lack of flexibility	53	0
Supplier autonomy	The customer has a dialogue with the supplier without performing excessive impositions	47	40
Location of the service	Availability of infrastructure and services in the region	47	60

Source: Authors’ own work

Table 3.

Honey’s analysis example

Category	Positive pole	Negative pole	Similarity (%)	H-I-L
Respect for the supplier	Easy to handle, educated manager	Arrogant and difficult-to-handle manager	56	H
Learning and improvement	The client brings growth (learning) to the supplier	The customer does not add value (knowledge) to the supplier	30	L
Learning and improvement	Demanding customer	The customer does not demand anything	28	L
Management knowledge	Management-skilled leader	The leader has no management capacity	26	L
Ambiguity of the contract	Clear contract without implied clauses	Difficult contractual interpretation	46	I
Ambiguity of the contract	Complex traffic service	Easy traffic service (standard route)	34	L
Customer availability	Easy to handle	Difficult communication with the client	42	I
Mediated power	The customer understands the supplier	The customer pressures the supplier excessively	70	H
Mediated power	Stability in supplier evaluation	Evaluation of unstable supplier, with constant changes	54	I

Source: Authors’ own work

Author contributions: All authors contributed to the study conception and design. Material preparation, data collection and analysis were performed by Marina Bastos Carvalhais Barroso. The first draft of the manuscript was written by Marina Bastos Carvalhais Barroso and all authors commented on previous versions of the manuscript. All authors read and approved the final manuscript.