Methodological concerns in case-based research in industrial engineering: revisiting the challenges towards further recommendations

Abstract Paper aims This paper addresses difficulties among the Brazilian scholarly community in industrial engineering (IE) when conducting case-based research. It also provides further recommendations to increase methodological rigour. Originality The paper contributes to the practice of case research by providing a historical perspective of research methodology in Brazil and offering guidance to improve case research adoption as well as the methodological rigour. Research method The main challenges when conducting case research were first identified through a literature review. Then, an exploratory survey with Brazilian scholars was conducted to identify challenges perceived by those. Recommendations are then provided, especially regarding the data analysis stage. The recommendations are discussed in the light of the existing literature and based on authors’ experience in conducting qualitative research. Main findings Difficulties when conducting case research identified by scholars can be classified in three ‘Aquila’s hells’: (i) weak theoretical background, (ii) careless case study design/planning; and (iii) fragile/uncertain data analysis. Suggestions to improve the data analysis process consist of building a narrative, data reduction, improving coding, etc. Improving validity is also necessary. Implications for theory and practice The recommendations are especially meaningful to early-stage researchers and provide guidance to improve robustness when conducting case research.


Introduction
Operations management (OM) is a domain that encompasses several other sub-areas.OM is a blend of different academic disciplines and practical fields of application; one of those fields is industrial engineering -IE (Slack et al., 2004).Whatever the purpose of investigation within OM, choosing an adequate research design and approach is a sine qua non condition to achieving robust, rigorous, and reliable results and conclusions.
The research design aims to lead the researcher to address questions of a specific research problem (Saunders et al., 2012).A suitable and consistently organized research design ensures rigour, reliability, and a comprehensive answer to the research question being addressed.Among a wide range of methodological approaches, case-based research has been pointed out as one of the most popular in OM (Filippini, 1997;Filippini & Voss, 1997;Scudder & Hill, 1998;Pannirselvam et al., 1999;Voss et al., 2002;Barratt et al., 2011), being widely adopted among the OM scholarly community over the last decades (Sousa-Zomer et al., 2022).

Literature background on case-based research
Case research is a methodological approach that focuses on understanding the dynamics present within specific settings, studying phenomena in their contexts rather than independent of context.Case research is the method that uses cases studies as its basis", i.e. case study is the analysis element.A pertinent definition is provided by Leonard- Barton (1990) when considering case study as a historical investigation of a past or current phenomenon, in which the context is relevant.Hence, the adoption of case studies can provide a Production, 33, e20220095, 2023 | DOI: 10.1590/0103-6513.202200953/16 description, generate, test, and/or refine theory (Eisenhardt, 1989).It is a compelling theory-building method, a sound approach to developing new models or concepts that can help industrial engineers, OM researchers, and practitioners understand or deal with a situation (Childe, 2011).Over the past years, scholars have recognized that qualitative research approaches, including case studies, can play a significant role in advancing OM theory and practice, which is still weak in theory-building efforts, especially at the grand theory level (Soltani et al., 2014).
The required rigor when doing research is understood as a fundamental element for an adequate adoption of any research method (Hatchuel, 2009).Moreover, methodological rigor contributes to the validity of the research and, thus, its recognition as a serious and well-conducted study.By considering the imperative need for rigor when doing investigations, researchers should justify and make their decisions understandable and unmistakable when planning and conducting research.The concern of adopting a suitable research approach has motivated several publications not only to dedicate to presenting and diagnosing research methods (e.g., Berto & Nakano, 2000, 2014) but also to offer recommendations (e.g., Cauchick Miguel, 2007;Mello et al., 2012;Cauchick Miguel & Dresch, 2018).
Internationally and particularly regarding the Brazilian IE context, one defining milestone that brought researchers' attention to the most common research methods in OM was a special issue of the International Journal of Operations Management (IJOPM).Four relevant manuscripts in research methodology were published discussing approaches such as action research (Coughlan & Coghlan, 2002), case research (Voss et al., 2002), modelling and simulation (Bertrand & Fransoo, 2002), and survey (Forza, 2002).Among those methodological research approaches, case research has consistently been one of the most powerful research methods (Voss et al., 2002).When case studies are well conducted, those enable an in-depth comprehension of a phenomenon.Relevant issues when carrying out case research are discussed next.

Relevant issues when choosing case research
The choice of case study as the appropriate research approach to explore a specific phenomenon should be made by considering a range of aspects.Undoubtedly, one of the most challenging research decisions is to select the methodological research method among various alternatives usually available.Indeed, this is also true when deciding on case-based research.More broadly, the unit of analysis, the case study, is a history of a past or current phenomenon drawn from multiple sources of evidence, including data from on-site observation, interviews, secondary data, among others (Leonard- Barton, 1990).As added by the previous cited author, any fact relevant to the stream of events describing the phenomenon is a potential datum in a case study since the context is essential.
First, it is crucial to consider that if the phenomenon under consideration requires an interpretive philosophy, i.e., the researcher needs to make sense of the subjective and socially constructed meanings expressed about the phenomenon being studied, then a qualitative research design should be considered (Saunders et al., 2012).'Then, in the realm of qualitative research, numerous research strategies can be contemplated such as action research, case-based research, and others (Saunders et al., 2012).The case study is a valuable research strategy to emphasize the rich, real-world context in which the phenomena occur (Eisenhardt & Graebner, 2007).As a theory-building approach deeply embedded in rich empirical data, building theory from case studies will likely produce an accurate and testable theory (Eisenhardt & Graebner, 2007).The case-based approach is appropriate for exploring answers to 'why ', 'what' and 'how' types of questions (Saunders et al., 2012).For this reason, a case study is a practical research approach for both exploratory and explanatory research goals (Saunders et al., 2012).
As mentioned earlier, when applying case studies, the researcher will have to consider multiple data sources such as interviews, in loco observations, documentary data, etc. to build the 'whole picture of the phenomenon under analysis.Consequently, the multiple data sources must be 'triangulated' when analyzing the data.Triangulation refers to using different data collection techniques within a study to ensure that the data is showing is what the researcher has identified in the data analysis (Saunders et al., 2012;Hussein, 2009), to develop a chain of evidence (Carter et al., 2014).
Once case studies have been chosen as the most suitable research approach to explore a question or phenomenon, a range of other aspects needs to be defined.Yin (1994) points out that choices should be made concerning using single or multiple cases and holistic cases versus embedded cases.As additional definition of single and multiple case is offered by Coombs (2022), mostly based on Creswell & Poth (2018).Nevertheless, the reader should bear in mind that it is a simplistic source and more in-depth discussion would be required.Single cases should be considered when the case represents a critical case or a unique or extreme case (Yin, 1994).The selected case may provide the opportunity to observe and analyze a phenomenon few have considered.The critical aspect of single case selection is ensuring that the choice is the most suitable to answer the research question at hand and meet the study's objectives (Yin, 1994;Saunders et al., 2012).
On the other hand, a case study strategy can also incorporate multiple cases, focusing on identifying whether the findings can be replicated across cases (Yin, 1994).The cases will then be chosen on the basis that similar results are predicted, termed by Yin (1994) as literal replication or based on a contextual factor that is different across the cases.In this situation, the impact of the difference across the case is predicted by the researcher, and when it is realized, it is termed theoretical replication (Saunders et al., 2012).When the study starts with a range of predictions and theoretical propositions on which the predictions are based, the study adopts a deductive approach aiming to test the theory.Therefore, a multiple case study approach is chosen to address one of the two forms of replication.The second dimension Yin (1994) proposed refers to the unit of analysis.If the research is concerned with an entire organization, the unit of analysis is considered holistic (Yin, 1994;Saunders et al., 2012).The case is considered embedded if the research is concerned with an organization's group, department or sub-unit.
Besides the decisions on the units of analysis, a case-study design involves the definition of other elements, including the case study boundaries and analytical techniques, which are part of the stage' plan the case', which are elaborated next.

Planning and conducting case research
There are a couple of issues that should be taken into consideration when planning case research.Firstly, the research question (and/or objectives) should be developed based on the literature and research gaps.Secondly, the researcher should critically consider the methodological fit, i.e., if case-based approach is a suitable research approach to tackle the research problem.Table 1 elaborates on the alignment between the broad research purpose and methodological approach and Table 2 summarizes some examples of sources according to the purpose of case research.
Thirdly, a robust plan should be developed when adopting case research, i.e., efforts and time must be put before thinking in going to the field.Figure 1 summarizes the main stages of case research and its overall design and execution.More details on the steps to plan and conduct case research can be found elsewhere (Cauchick Miguel, 2007).
Table 3 provides a brief description of each case research step in Figure 1 and highlights some relevant issues to be considered when designing and conducting case research.
Table 4 was adapted from Corrêa (1992) and provides a valuable structure to address some of the questions researchers face when making choices and developing the research design.The table compares different approaches regarding the research requirements and characteristics.Table 3. Case research stages and their description, and relevant issues to take into account.As shown in Table 4, each methodological approach has specific characteristics.Case research combines a range of distinct characteristics but conducting case research with adequate methodological rigor is not a trivial task.The following section discusses some challenges when conducting case research identified in the literature and highlighted by researchers among the Brazilian IE community.

The context of research methodology and case research in the Brazilian IE scholarly
The movement towards improving the adoption of more robust research methodologies and increasing methodological rigor in Brazilian IE academics is relatively recent compared to other countries.This movement was triggered by a few scholars in the middle 90s in the Department of Production Engineering at the University of São Paulo (USP).Later on, other scholars from different institutions across the country were involved with the subject.When Professor Henrique Correa (now in the Rollins College at Florida in the USA) returned from a doctorate obtained in Warwick, he proposed to the head of the department that research methodology should be introduced.As Professor Correa highlighted (Correa, 2020): He [the head of department] suggested that I should lead a course on research methodology […] but I replied that I would organize research seminars to discuss research design and methods with master and doctorate students based on the material I was collecting and studied during my doctorate degree.Participants in this seminar included research students such as Roberto Martins, João Turrioni, José Paulo Fusco, and many others from other educational institutions.
The research methodology seminars were conducted from 1993 to 1995 until Professor Correa left the University of São Paulo in 1996.In 1997, Professor Afonso Fleury created a course on research methodology that is still being delivered today.The seminars and the course have educated several research students through the decades.As a reference for this time, one of the first theses with a proper chapter on research methodology is the one by Martins (1999).Later on, he was involved in the efforts to improve research methodology within the industrial engineering community.Figure 2 illustrates the macro phases in the progress of research methodology adoption and improvement in Brazil.This representation was developed by the authors based on main milestones, checking facts with one of the main contributors involved, as well as the accumulated experience of one of the person in this movement.
Case research is less employed among the North American operations management community, but is a widely applied approach in Europe (Drejer et al., 2000) as well as in the Brazilian IE scenario (e.g., see Berto & Nakano, 2014;Cauchick Miguel & Dresch, 2018).To the best of our knowledge, no equivalent study has been published in the past decade as the data in Figure 3.However, the figure illustrates this point by offering an example of the adoption of case research in a major domestic IE journal (anonymized).
A recent search for case research published in the Production Journal, for instance, has shown that the number of case research has been growing since the 2000s.The growing number of case studies in recognized national and international academic outlets reflects the efforts among the operations management community, both nationally (e.g., Cauchick Miguel, 2007) as well as internationally (e.g., Voss et al., 2002) in supporting improvements and methodological rigor in qualitative research, as qualitative research methodologies are essential for new theory development and evolvement of the field.Nevertheless, improvements in methodological rigor among the Brazilian IE scholarly community are needed.

Research design
The interdependently objectives of this paper was to address the existing difficulties among the Brazilian scholarly community in IE when conducting case-based research in addition to offer recommendations to mitigate those difficulties.The development of this work was then divided into three steps: (i) a literature search; (ii) an exploratory survey; and (iii) recommendations to address the difficulties.Those are described in more details next.

Literature search
Two main steps were followed in reviewing the literature on case research: (i) selection of the main methodological articles on case research approach published in the last 15 years or so (circa 30 publications), and (ii) a search in one of the main operations management journals publishing case-based research, in order to complement the previously mentioned set of articles.
First, a literature search was conducted to identify the main challenges when conducting case studies raised in the main case research papers.However, this literature review was not meant to be exhaustive of all publications on case study methodology that could have been retrieved.Instead, the review aimed to be representative of the most relevant articles (well-cited) on case research methodology published in peer-reviewed journals that have addressed challenges when conducting case-based research.The selection of articles helps navigate the subject and identify issues that need further attention in the IE context.
The search was carried out in the Scopus and Web of Science databases, as they are some of the most prominent ones for OM research (Thomé et al., 2016).The search was performed in the titles, abstracts, and keywords by using the following terms: case study, case research, case-based approach, combined with the terms operations management, research methodology, research methods, and challenges.This set of keywords was chosen based on other relevant publications covering the domains investigated in this study (e.g., Eisenhardt, 1989;Voss et al., 2002, and others).After reading the titles and abstracts, papers that were not aligned with the research purpose (identification of challenges and difficulties) were discarded, resulting in 28 articles for final analysis.More details of the selection and analysis of the papers can be found in Cauchick- Miguel et al. (2019).
Afterwards, a complementary search for methodological case research papers was conducted in the International Journal of Operations and Production Management (IJOPM), one of the most well-known journals in OM and which publishes a high number of qualitative research.Only two articles were found that have not been identified in the previous search.

Exploratory survey
The second step was to conduct an exploratory survey with Brazilian researchers mostly.This survey type may be carried out when the purpose is to gain preliminary insight on a subject (Forza, 2002).The author emphasised that exploratory surveys are relevant and extensively adopted in OM.Actually, this kind of survey does not necessarily demand a theoretical model, research hypotheses, non-probabilistic sample size, pre-test, or minimum response rate [for further details refer to Forza (2002), p.188-189].Thus, to guide this step this work followed Forza's recommendations.
A sample of all authors' contacts of IE academics was used.Seventy experienced researchers were consulted.They were from a range of institutions (a purposive sample), who are usually interested in case research in OM.It considered a single open-ended question: what are the main difficulties when conducting case studies?In addition, the researchers were asked to list "up to three difficulties when engaging in case studies" they considered being the most limiting (difficulties) in adopting the case study as a research approach.The question was sent by email with subsequent two follow up messages.The results enable to raise the difficulties and from the analysis elements to mitigate them.
After 10 weeks of sending the email (response rate about 51%), with 116 statements (in total) were obtained, read, and grouped by an affinity diagram (Figure 4).From those, 8 statements had no direct relation to what had been requested, i.e., they were could not be connected difficulties associated with case research and, thus, were discarded.This resulted in 108 statements for further analysis.According to their contents, the statements were assigned to each stage of the step-by-step framework for case study previously shown in Figure 1.More details on the survey design can be found in Cauchick-Miguel & Sousa (2018); Cauchick- Miguel et al. (2019).
The identified difficulties provided the basis for raising recommendations to address them and enhance the quality of the outputs of case research and theory-building process in OM.The recommendations were elaborated based on previous literature on case research.Those were particularly relevant concerning the data analysis stage, which was identified as one of the most critical stages when conducting case-based research, from the results of data collection and analysis (Cauchick-Miguel & Sousa, 2018).

Results: case research difficulties and recommendations for addressing them
A range of difficulties when conducting case study was identified from the literature search, as summarized in Table 5.Those are related to the required theoretical background, research design, data collection and analysis, and the research report in the case research stages of Figure 1.
Concerning the survey performed with researchers in the Brazilian IE scholarly to identify the main difficulties faced when engaging in case research, 116 aspects were obtained, as mentioned in the previous section.Those represent struggles when conducting case research.Those were clustered and organized following the stages of case research, as shown in Figure 1, apart from other categories of difficulties that emerged.Table 6 summarizes the identified difficulties.
As shown in Table 6, the difficulties raised by the researchers are not necessarily the same identified in the academic literature, as the ones related to data analysis, for example.Additionally, difficulties related to the pilot test step were not reported, possibly because this step is rarely performed by those who took part in the survey.At follows, the identified issues can be summarized in some categories and provide the basis to suggest improvements that may improve the methodological rigour when conducting case research.After reading, recording, and organizing the statements provided by survey participants, a content analysis was performed.Then, the main difficulties presented in the previous section, were further classified into three 'Aquila's hells', namely: • Weak theoretical background, e.g., previous literature analysis that does not lead to a research opportunity or that identifies research gaps, as already identified by previous scholars as an essential step in any research effort (e.g., Eisenhardt & Graebner, 2007;Barratt et al., 2011;Cauchick Miguel & Dresch, 2018); • Careless case study design/planning (Gerring, 2004;Eisenhardt & Graebner, 2007;Baškarada, 2014), e.g., selection of the unit of analysis; and • Fragile/uncertain data analysis, e.g., definition of criteria for data analysis, description of the data analysis procedures, coding and analyzing the data in such a way that the research provides new insights instead of just a description of the data (Lu & Shulman, 2008;Barratt et al., 2011).
When revisiting previous studies that offer recommendations, those were related to the following actions and decisions in the case research (Cauchick Miguel, 2007;Cauchick Miguel & Dresch, 2018): (i) development of a theoretical background (e.g., identify research gaps, constructs, contradictions, etc.); (ii) define the type of research (e.g., exploratory, explanatory, etc.); (iii) planning the research (e.g., consider the types of validity; see section 6); (iv) establish criteria for case selection (e.g., access is essential but it is not enough to justify a choice); (v) data collection (e.g., develop a robust research protocol, take multiple sources of evidence into account); and (vi) data analysis (e.g., organize data, triangulate the data, code the data, identify patterns, cross-analysis in multiple cases, etc.); and (vii) correlate findings to existing literature and theory in order to move forward (i.e., create a new theory, test, or extend/refine it).Indeed, all of them are relevant to case research.Nevertheless, the data analysis stage was identified in the survey as one of the main constraints in the context of the Brazilian industrial engineering scholars.Thus, the recommendations here mainly addressed concerns regarding the data analysis stage.To illustrate that, a couple of examples of quotes from two respondents were: (i) Difficulty of analyzing the data, and this can lead to a text that is too descriptive, and (ii) Difficulty of discussing the results and confronting them with the existing theory associated to the phenomenon or the investigated research problem.
As can be seen, the quotes are somewhat simple, suggesting that this stage of case research is still weakly addressed.This is concerning as the data analysis is a critical step in the case research and an appropriate step for generating new findings and theory development.Quote (i) is related to usual reported results on case-based research within the Brazilian IE scholar, i.e., the results are too descriptive and less analytical.Actually, the 'how' description of data analysis is usually neglected in case research when the outlets from the Brazilian IE researcher are examined.The reports are often a storyline of what was collected in the field, and the methods usually focus on describing the data sources and collection procedures, with limited attention given to explaining 'how' the data has been triangulated and analyzed, indicating a poor use of data analysis strategies and techniques, e.g.coding.
Data analysis can range from a shallow description to a theoretical interpretation of data and facts.After collecting data in the field (considering multiple sources of evidence, as an example in Table 7), data reduction must be carried out (i.e., not 'all' data should be included in the analysis or the report).
The analysis should only consider data that is narrowly related to the research objectives and constructs.A narrative should be produced as early as possible.Actually, a priori theorization is essential to frame the research design (Ketokivi & Choi, 2014), allowing a deeper understanding of the narrative in the light of the literature, for instance.It is recommended to type up field notes as soon as possible both to maximize recall and to facilitate follow-up and filling of gaps in the data.
Interviews should be conducted by at least two of the authors, in order to enhance the information gathering's reliability (Dubé & Paré, 2003), especially for data analysis.In addition, idiosyncratic responses should be disregarded in the interest of focusing on dominant patterns among interviewees (Tortorella et al., 2021).
If interviews were audio-recorded, information should be fully transcribed and subsequently analyzed qualitatively and discussed by the authors.Summaries should be then consolidated after reaching a consensus on the main findings (Miles & Huberman, 1994;Miles et al., 2014).The transcriptions should be also made as shortly after as possible, for instance, to remember details of the interview environment such as interviewees' reactions.The same procedure is true for field observations.Observations, however, have their potential drawbacks.A researcher may give meaning to a situation based on observation without checking out that meaning with participants (Corbin & Strauss, 2008).Thus, it is important to triangulate the different data sources, so the findings of the case study will be supported by multiple sources of evidence.Other information like secondary data should also be considered to support the analysis in addition to internal documents that the researchers managed to have access.All of this data documentation produces a case narrative made up of the transcriptions of notes, all data sources, and ideas and insights.Narrative accuracy may be enhanced by letting key informers assess draft reports.
As mentioned earlier, data reduction of the raw data is necessary.After the data collection, the first step is to convert these data into text files organized in a case study database.However, just transcribing the data and building a narrative considering all sources of evidence is not enough.According to Yin (1994), data analysis involves examining, categorizing, tabulating, testing or recombining evidence to produce findings that contribute to existing literature.It is important to highlight that any scientific research should contribute to the scholarly literature, and this contribution should be evident from the research findings.
The researcher can start the analysis process by playing with the data.Some options to start manipulating the data include (Miles & Huberman, 1994): (i) developing a matrix of categories and placing the evidence within the established categories; (ii) creating data displays such as flowcharts and other graphics, (iii) tabulating the frequency of different events, and (iv) using a temporal scheme to describe the events.Table 8 presents an example of display organizing the collected data for a modularity study.If multiple cases are investigated, a display summarizing the data for each case should be developed, and a cross-case analysis should be performed afterwards to identify convergent and divergent aspects across the cases and data sources.A display is an example of a data organizing technique that can help to identify patterns and relationships in the data and to develop a clear chain of evidence that will lead to robust findings supported by the data.A clear chain of evidence is one of the conditions to increase the validity of the research.Preliminary analyses after manipulating the data can then support moving towards an analytical strategy.According to Yin (1994), the researcher can use four different analytical strategies in the data analysis process: • Relying on theoretical propositions: if theoretical propositions were identified in the literature review and the objectives and design of the case study were defined based on such propositions, those should be the focus and guide the data analysis; • Working the data from the 'ground up': this strategy is the opposite of working with propositions; instead of analyzing the data with a preliminary view of the propositions, this strategy consists of finding new insights and developing concepts from the data, i.e., inductively; • Developing a case description: this strategy consists of organizing the case study data according to some descriptive framework; and • Analyzing rival explanations: this strategy defines and tests rival explanations and can be applied with the three previous strategies.
A detailed description of each of those strategies can be found in Yin (1994).The point is that the researcher should be aware that a procedure should be followed for the data analysis, and a strategy should be adopted, depending on the purpose of the study.
Additionally, it is important to highlight that coding the data is a common practice when employing analytical strategies.Coding the data is, actually, the first step in the data reduction process (Sousa, 2005).It consists of highlighting parts of the texts and developing codes that might represent pre-defined categories (e.g., when adopting a deductive approach and, for example, working with pre-established propositions) or that will support the development of new themes and concepts (i.e., when working on the data from the 'ground up' or conducting an inductive analysis).The codes can be seen as 'blocks' that represent the data related to what has been explored, related to either the research question or the constructs identified in the literature.Table 9 shows an example of codes and their meaning for a previous work looking at modularity.To facilitate the assembly of the truck body, they are mounted upside down (...) there are two lines in parallel, one that assembles the body and the other that then turns the body into its normal position and completes the assembly of the truck (fits the cabin, banks, etc.).The door is mounted in the cab and its components installed in it, without removing it from the cab, as there is no space for a sub-assembly of the doors outside the cab.
The company had a project for the instrument panel to be assembled by the supplier, but it didn't work out, as the sub-suppliers of panel components would charge a higher value for the parts than the value they sell to the company
starting point for case research is the literature, that should be mapped before designing the cases and going to the field -Review the literature to identify constructs that will be tested in the field -Develop a framework that represents the constructs and their relations -Develop research questions and/or objectives Design and plan the cases A careful design and a detailed planning make case research more robust and enable to minimize possible research limitations in the next steps -Establish robust criteria for selecting the cases and unit of analysis (define it accurately) -Contact interviewers as early as possible and plan ahead the schedule for data collection -Develop a research protocol for data collection (research questions, interviewees' profile, support documents, etc.) -Define previously 'how-to' analyze data -Consider a pilot test for debugging instruments for data collection Conduct pilot test A pilot trial helps to mitigate possible misunderstandings and mistakes to be avoid in the 'real' data collection and analysis -Check constructs, data collection procedures, and evidence sources at preliminary level -Enable to examine data quality and assess analysis procedures -Provide a bit of experience for researcher less familiar with case research Data collection After contacting the informants' and have their consent, all data collection instruments should be applied for gathering data from the units of analysis -Check the quality of evidence sources and their contents -Employ multiple sources of evidence as well as data collection techniques -Get and record data using multiple approaches, if available -Reduce/mitigate the influence of the researcher(s) Data analysis Establish a comprehensive and robust data analysis by employing appropriate strategies and techniques to interpret data from all sources of evidence -Use of support tools and software (e.g., Atlas.ti,NVivo, VOS Viewer, etc.) -Produce an overall narrative and apply data reduction -Employ content analysis and develop ways to display data for further analysis -Establish causal relationship among constructs Generate report A strong report provides evidence that the research is rigorous and well-designed and conducted -Provide a detailed presentation of methodological design and procedures employed for collecting and analyzing data -Show the research protocol and other research supporting documents -Draw managerial and/or theoretical implications Source: developed based on Voss et al. (2002) and Cauchick Miguel (2007).

Figure 2 .
Figure 2. Milestones in IE research methodology.

Figure 4 .
Figure 4. Affinity diagram by grouping the statements (photo by the first author).Note: colors represent a main category (step in Figure 1) where the title is showed by the post-it; white statements have multi-steps.

Table 2 .
References on case study considering its intent with regards to theory advancement.

Table 4 .
Design choices when comparing case research with other approaches.

Table 5 .
Some of the difficulties when conducting case-based research.
Characterization of the case study in terms of its timing and type Production, 33, e20220095, 2023 | DOI: 10.1590/0103-6513.20220095

Table 7 .
Data collection details from various sources of evidence in multiple cases.

Table 8 .
Example of a display for organizing collected data and constructs.