Convergent interviewing: a tool forstrategic investigationWil Williams* and Duncan LewisUniversity of Glamorgan Business School, UK

! This paper explores the application of a relatively under-reported qualitative researchmethod known as convergent interviewing.

! Specifically it examines the practical relevance of an adapted version of this techniquein practitioner-orientated strategic management research.

Copyright © 2005 John Wiley & Sons, Ltd.

Copyright © 2005 John Wiley & Sons, Ltd. Strategic Change, June–July 2005

p. 433) is one instance of an inductiveapproach.

A feature of strategicmanagement research is

the complexity of thesituations being examined

Convergent interviewing has emerged as aqualitative technique that attempts to addressresearch topics that lack theoretical underpin-ning (Dick, 1990). It is also argued that theconvergent interviewing technique is aninductive, flexible, evolving research instru-ment (Dick, 1990). Whilst the academic fieldof strategic management cannot be accused oftheoretical shortcomings, there might besome tangible benefits to be derived fromadoption of a research instrument that specif-ically offers the researcher or consultantenhanced flexibility to aid the understandingof complex situations. We therefore offer thispaper as an innovative application of this tech-nique in terms of consultancy and strategicmanagement research. What follows is pre-sented in three parts. First, we outline the

Strat. Change 14: 219–229 (2005)Published online in Wiley InterScience(www.interscience.wiley.com). DOI: 10.1002/jsc.719 Strategic Change

* Correspondence to: Wil Williams, Strategy Department,Welsh Institute for Competitive Advantage, University ofGlamorgan Business School, Trefforest, Rhondda CynonTaff CF37 1DL, UK.E-mail: wwillial@glam.ac.uk

Introduction

One of the problems with many consultancyor academic research projects in the strategyfield is that sometimes the client does notalways know exactly what they wantresearched (Reige and Nair, 2004). Even if theclient has an idea of the potential outputs andoutcomes at the outset of a study, these are likely to evolve as the research develops,unexpected findings emerge or circumstanceschange (Rao and Perry, 2003). A feature ofstrategic management research is the com-plexity of the situations being examined(Stacey, 1996; Beinhocker, 1997; Courtney et al., 1997; Pascale et al., 2000; Kurtz andSnowden, 2003). Strategic issues are often notthe product of one factor but the result of amyriad of interconnected dependent and inde-pendent variables (Beinhocker, 1997). In thesecircumstances researchers often have to adoptan inductive model where theory evolves fromresearch undertaken. Grounded theory, ofwhich convergent interviewing is an example(after Glaser and Strauss, 1967; in Burns, 2000,

220 Wil Williams and Duncan Lewis

broad principles of convergent interviewingas a methodology. Second, we illustrate theapplication of the technique for both strategicmanagement in academic and consultancyoutputs. Third, we offer concluding thoughtson the applicability of the technique for strate-gic management researchers.

Convergent interviewing explained

A chosen methodology is informed primarilyby the context of the study. However, it is alsoessential that a researcher’s ontological andepistemological positions are recognized, asare the potential influences of such positions(Burrell and Morgan, 1979; Easterby-Smith et al., 1994; Hughes and Sharrock, 1997). Theimportance of fitting the research methodol-ogy and methods to the appropriate contexthas been identified by a number of writersincluding, for example, Silverman (2001) andFlick (2002).

We contend that much, if not all, organiza-tional strategic research takes place in a socio-economic context and therefore it often has toadopt an essentially qualitative approach. Thisdoes not discount the importance of quantita-tive research in this field, but it acknowledgesthe importance of subjective elements thatcontribute to strategic decisions. Qualitativeresearch is often associated with terms suchas: soft; flexible; subjective; political; casestudy; speculative and grounded (Halfpenny,1979; in Silverman, 2001). Abercrombie et al.(2000) simply define qualitative analysis as the study of the human condition. It is often,but not always, contrasted with quantitativeresearch traditions and a positivist epistemol-ogy (Abercrombie et al., 2000, p. 284).Silverman (2000) distinguishes between quali-tative techniques that are data-gathering andthose that are analytical in nature. Qualitativeapproaches can address both inductive,grounded research and deductive, exploratorystudy (Gummesson, 2000). It is broadly agreed(Fields, 1988) that within the variousapproaches to data gathering in qualitativeresearch the following examples of keyresearch techniques can be identified:

observation; interviews; focus groups; ques-tionnaires; case studies (including ethno-graphic studies and action research); oralhistories/stories; documentary review (includ-ing textual analysis) and visual analysis.

To these qualitative techniques, analyticalqualitative techniques such as content analysisand discourse analysis (Fields, 1988) can beadded. In some situations different terminol-ogy is applied to similar techniques dependingon the perspective of the researcher or writer.Silverman (2000) identifies four main cate-gories of qualitative research: observation;analysing texts and documents; interviews;recording and transcribing. The last categoryshould not really be considered as a stand-alone method, but instead might be viewed asa means of operationalizing participant obser-vation or interviews. Convergent interview-ing, as its title suggests, falls into the third ofSilverman’s categories of qualitative research— interviews, but also incorporates elementsof the first category, observation.

Dick (1990) coined the term ‘convergent (orconvergence) interviewing’ in the 1980s. AsReige and Nair (2004, pp. 74–75) indicate,convergent interviewing is highly suitable forexploratory, inductive research. Dick (1990, p.2) describes convergent interviewing as ‘Away of collecting qualitative informationabout people’s attitudes and beliefs throughthe use of interviews’. Dick (1990) alsodescribes it as an action research technique,although he also recognizes convergent inter-viewing as a modus operandi that can be usedfor ‘pilot research’ as well as a generalist toolin the armoury of the qualitative researcher.Convergent interviewing has also beendescribed by Rao and Perry (2003) as an ‘in-depth interviewing technique with a struc-tured data analysis process’ (p. 237).

Dick (1990) distinguishes between ‘content’and ‘process’ in convergent interviewing. Heviews the content of convergent interviewingas unstructured but the process as semi-structured. The structure in the process isderived from an embedded, ordered route ofdesign and analysis. He outlines a cyclical pro-cedure: design; data collection; analysis and

Copyright © 2005 John Wiley & Sons, Ltd. Strategic Change, June–July 2005

interpretation, that he claims is common to allforms of qualitative research, including con-vergent interviewing. However, certain formsof qualitative research utilize multiple cyclesthat include: redesign; data collection; rein-terpretations; redesign; etc. (see Figure 1).Using this method, the research process con-tinues until each element of succeeding datadoes not alter the interpretation of the situa-tion. It might therefore be thought of as areductionist iterative process.

The convergent interviewing process issummarized in Table 1. Dick (1990, p. 1)believes that interviews should be conductedon a one-to-one basis. Where a number ofinterviewers are involved, they shouldcompare findings. This addresses a major crit-

icism of qualitative techniques, namely, inter-viewer bias. Where a qualitative piece ofresearch is the product of an individual’s viewson a particular subject, that person’s findingsmight be coloured by any number of factors(Dick, 1990, p. 4). Initially, interviews areunstructured, but as the number of interviewsand the richness of data improves, questionscan become more focused to expose key find-ings. This allows the researcher to gain adeeper understanding of a situation. Conver-gence can also occur naturally within an inter-view as an interviewer explores particularissues as they emerge. Interviewers will tendto filter information as they interview, tryingto identify areas of agreement and disagree-ment between interviewees. In a sense con-vergent interviewing legitimizes this naturaltendency by making it an explicit part of theresearch strategy. In a strategic context thismight revolve, for example, around issuessuch as differences or agreement about stake-holder understanding or strategic intent.

Dick describes convergent interviewing ascyclical in nature (see Figure 2), in that itallows refinement of the question, the answersand the method applied (Dick, 1990, p. 2). Hedescribes the process of convergent inter-viewing as a series of ‘successive approxima-tions’ (Dick, 1990, p. 3).

Convergent interviewing 221

Copyright © 2005 John Wiley & Sons, Ltd. Strategic Change, June–July 2005

Redesign

Reinterpretations

DataCollectionAnalysis

DataCollectionAnalysis

Convergence

Divergence

Design

Interpretation

Iterations

Figure 1. Convergent interviewing process.Source: After Dick (1990, p. 3).

Table 1. The convergent interviewing process

1. The research topic is identified2. Key themes for interviews are identified by

(a) Reference to the literature(b) Reference to key informants — the clients or

the research team3. Interviewees are identified — normally experts in

the area of research4. Interview takes place5. Key issues are identified 6. Interview schedule is amended in light of interview

findings7. Interviews are concluded when convergent themes

are identified

Source: Nair and Riege (1995); in Rao and Perry (2003,p. 237).

Refine theQuestion

Refine theAnswerRefine the

Method

Figure 2. Convergent interviewing process of succes-sive approximations.Source: Dick (1990, p. 3).

222 Wil Williams and Duncan Lewis

It is important for the strategic managementresearcher or consultant to distinguishbetween what Dick describes as ‘content’ (i.e.the topic being researched) and the ‘process’(i.e. convergent interviewing), as these cancome together over the course of a researchproject. He states that it is important that theresearcher keeps their interpretations openfor as long as possible. To this end the inter-view should begin with as open a question aspossible, only later moving on to specificissues. Dick (1990) suggests that the initialquestions should almost be content-free. Laterquestions should be used to remove dis-crepancies and ambiguities. Thus, within astructured data analysis process, the initialinterview content (the questions) should beunstructured.The structured process could beopen to adjustment through the course of aresearch project, as the target populationmight be expanded. For example, where it isidentified that there has been inconsistency inviews about elements in the strategic decision-making processes in an organization, inter-viewers can focus their questioning to exploreareas of discrepancy. Continuing this example,different individuals or groups in organizationsmay have diverse perspectives on sources ofstrategic direction in an organization. By focus-ing on this element in the process (i.e. ques-tions) in successive interviews, areas of clearconvergence or divergence can be established.It is argued that because of the complexity ofstrategic situations this process is essential.

Dick (1990) describes the structure of con-vergent interviewing as a dialectic process,where disagreements and agreements are usedto identify patterns of convergence and diver-gence. This is self-evident in the process, butacknowledging the fact is important whenconsidering the method. He states, ‘Conver-gence and discrepancy provide ways of iden-tifying the needle of informative data in thehaystack of irrelevancy’ (Dick, 1990, p. 10).It is important to maintain a healthy respectfor both patterns of data, as they could beequally significant. Dick specifically proposesthat, in each interview, the researcher activelytries to disprove emerging explanations of the

data. He claims that if this technique is usedsystematically it will provide the study withobjectivity when refining subjective data(Dick, 1990, p. 11). In a strategic context, achief executive’s view, if divergent from per-spectives of other levels in the organization,might indicate a fundamental failure at thestrategic apex of an organization. If strategyshould be examined from multiple perspec-tives as Mintzberg et al. (1998) would argue,different stakeholders may have very differentyet valid perspectives on a strategic issue.

A series of convergent interviews in atypical study would be continuously modifiedwith refinement of the content and process.According to Reige and Nair (2004), inter-viewing will terminate ‘when a stable patternof clear agreements or disagreementsemerges between all or most of the intervie-wees, and where different opinions andbeliefs can be explained’ (p. 75). The devel-opment of a progressive report, which can beamended after each interview, is a usefulvehicle for production of research findings. Toallow a strategic assessment to be made andstrategic change to be managed it is importantthat some ‘firm’ foundations of evidence andstructure are established. Convergent inter-viewing allows the strategy researcher toestablish some basis for further examination orchange.

Validity and reliability

Issues of validity and reliability cannot beignored when utilizing qualitative techniques(Silverman, 2000, p. 1) and convergent interviewing is no exception. Reliability isdescribed by Hammersley (1992) as ‘thedegree of consistency with which instancesare assigned to the same category by dif-ferent observers or by the same observer ondifferent occasions’ (p. 67). Furthermore,Silverman warns of the dangers of this form of measurement of reliability as a differentcontext might elicit an alternative finding(2001, p. 225). Kirk and Miller (1986, p. 10)identify that reliability can be achievedthrough consistency in procedure. It is here

Copyright © 2005 John Wiley & Sons, Ltd. Strategic Change, June–July 2005

that convergent interviewing proves its worthas a method that offers the strategic manage-ment researcher a clear and established pro-cedural basis with which to address potentialproblems of reliability.

Issues of validity in convergent interview-ing, as they do in all forms of qualitativeresearch, centre on ‘anecdotalism’ (Silverman,2000, p. 10). Qualitative research is often crit-icized for being merely ‘anecdotal’ or ‘illustra-tive’ and conducted in casual or unstructuredways (Mason, 2002, p. 1). Kirk and Miller iden-tified two types of ‘anecdotal error’ in terms ofvalidity of qualitative data. A type 1 erroroccurs when a researcher believes a statementis true when it is not, whereas a type 2 erroris the opposite, when a statement is rejectedalthough it is true (Kirk and Miller, 1986, pp.29–30). Rao and Perry (2003, pp. 240–241)identify three types of validity measures:namely, construct, internal and external, thatcan and should be applied to research that utilizes convergent interviewing. Constructvalidity is established through practicalapproaches to the subject (Emory and Cooper,1991; in Rao and Perry, 2003, p. 240). Inter-nal validity refers to the casual relationshipbetween variables (Zikmund, 2000; in Rao andPerry, 2003, p. 240). Finally, external validityrelates to the generalizability of the study.

It must be noted that the notion of validitycomes from quantitative research and there-fore, in certain respects, the concept of valid-ity might not be applicable to the same extentwhen considering convergent interviewing.Convergent interviewing, as with most formsof qualitative research, generates largeamounts of data which makes analysis both dif-ficult and complex. Turner (1983) suggests a‘solution’ to this problem by ensuring the datais firmly grounded in theory where inter-relationships between data and theory can be clearly established (Turner, 1983, p. 333).However, this is not always the case whenusing convergent interviewing. Convergentinterviewing allows researchers to identifytheory through inductive, rather than deduc-tive logic. The process explicit in convergentinterviewing assists the researcher in finding

their way through the mass of data producedby the process to identify the key issues per-tinent to the subject being studied. It is herethat the technique of convergent interviewingmight prove most useful to the strategyresearcher and consultant. The complexity ofstrategic processes, often with multiple inter-pretations by stakeholders, requires a flexibleresearch framework in respect to the bound-aries of the research content, whilst maintain-ing some structure to the process.

The complexity of strategicprocesses requires a

flexible researchframework

Advantages and disadvantages ofconvergent interviewing

Rao and Perry identify three main benefits ofconvergent interviewing (2003, p. 238): first,as a method for quickly converging on keyissues in an area of emergent research; second,as an efficient mechanism for data analysisafter each interview; third, as a method for rec-ognizing the end-point for research. Both Raoand Perry (2003, p. 237) and Dick (1990, p. 3)clearly identify that the main advantage ofconvergent interviewing compared to in-depth interviews is its ability to refine contentand the process of the interview and to focus-in on broad research issues. Rao and Perry(2003, pp. 237–239) critique convergent inter-viewing against in-depth interviewing, caseresearch and focus groups and clearly establishthe main strength of convergent interviewingis its progressive nature, that allows a researchtheory or theme to emerge from successiveiterations of the interview process. This facetof convergent interviewing is invaluable whenresearching strategic issues.

Woodward identifies limitations of the con-vergent interviewing method (Woodward,1996; in Rao and Perry, 2003, pp. 239–240).

Convergent interviewing 223

Copyright © 2005 John Wiley & Sons, Ltd. Strategic Change, June–July 2005

224 Wil Williams and Duncan Lewis

First he, like Dick (1990), highlights the poten-tial for interviewer bias. Second, Woodwardbelieves that researchers need to have priorknowledge of the research subject so that theycan contribute to the ‘. . . meaningful infor-mation to the exploratory research . . .’,however, this may once again result inresearcher bias. Reige and Nair identify priorknowledge as a key issue in the use of the tech-nique (2004, p. 76), whilst Woodward (1996)has concerns that convergent interviewing on its own may not be sufficient in terms of validity to provide results that can be generalized.

Despite these criticisms we maintain that asa technique, convergent interviewing hasmuch to offer the strategic managementresearcher. Faced with the complex inter-twining of variables from multiple sources, thestrategy researcher must take into accountcopious perspectives quickly, analyse themwith integrity and offer potential solutions andobservations.To this end, we decided to adoptthe technique of convergent interviewing totwo research projects with the aim of under-taking some formal evaluation and observa-tion. It is to these examples we now turn.

The research projects

The convergent interviewing method wasapplied in two UK public sector commis-sioned research projects. The aim of the firstproject was to evaluate the provision of ser-vices by a public-sector funded developmentagency. The aim of the second project was thedevelopment of a community regenerationstrategy for a county region. Both projectstook place in highly politicized environmentswith multiple stakeholders. A key reason forthe research team’s involvement in both pro-jects was their perceived objectivity as bothclient organizations were some geographic dis-tance from the research team’s base. All themembers of the research team had knowledgeof the localities and their socio-economicunderpinnings, which was important in under-standing the individual complexities of eachproject.

Essentially the same methodology andmethod were applied in both projects. Theonly difference was that in the second project,the development of a community regenerationstrategy for a county region, a larger researchteam of six interviewers was utilized, com-pared to only four interviewers for the firstproject. In both cases a project director liaisedwith the clients and oversaw the direction of the project, although all the members of the research teams were actively involved inthe development and consideration of themethod.

In both projects the clients identified theresearch aims but did not specify the method-ology, methods, or provide any indication asto the themes that they wished to see emergefrom the data-gathering processes. Both acad-emic and grey literature (mainly in the form ofreports and minutes of meetings) were con-sulted by the research teams. Lengthy discus-sions were also held with both sets of clients,after which key themes were identified,agreed upon and noted by the research team.This information provided interviewers withbackground to the issues that they mightencounter in sessions with key respondents.However, the researchers were always mindfulof the very ‘emergent’nature of strategic issueson the ground.

The research team and the client jointlyidentified key respondents. Deciding on thesampling frame is a critical element in the con-vergent interviewing process. ‘In qualitativeresearch like this [convergent interviewing],the sampling method is purposeful ratherthan random’ (Patton, 1990; in Rao and Perry,2003, p. 242). Care was taken to ensure that awide spread of views was canvassed. Impor-tantly, the research team identified ‘politicallyneutral’ respondents who were interviewed atthe outset of the process in an attempt toavoid initial potential bias. Additional infor-mants were identified during the process ofdata gathering. In addition to key respondentinterviews the process utilized focus groups.Focus groups were used because of the natureof some of the stakeholder groupings whoseviews needed to be gained. Resource con-

Copyright © 2005 John Wiley & Sons, Ltd. Strategic Change, June–July 2005

straints and research timings also influencedthe selection of focus groups as an appropri-ate data-gathering mechanism. It is importantto note that the use of focus groups is a departure from the conventional convergentinterviewing technique; however, the samedialectic process was applied, where dis-agreements and agreements are used to iden-tify patterns of convergence (agreement) anddivergence (disagreement) in both focusgroups and one-to-one interviews.

In both cases the research teams underwentthorough briefings on the process and contentthroughout the projects. This was essential toensure that the teams were fully conversantwith the process and content being adopted.This is a necessary prerequisite for convergentinterviewing, as a ‘rogue’ interviewer mightinvalidate findings.The focus groups and inter-views with key respondents were conductedin as free-following a manner as was possibleunder such conditions.This follows the adviceof Silverman (2001) who argued that ques-tions should be as open-ended as possible toensure ‘authenticity’ (p. 13). In both projectsthe lead questions were framed as openly aspossible. For example:

Can you tell us about the performance ofXXXX organization?

Can you tell us what you think of the com-munity regeneration strategy in XXXX?

The initial project involved three focus groupsand nine key respondent interviews. Thesecond project consisted of seven focusgroups and fifteen interviews. All focus groupsand interviews were digitally recorded andtranscribed. Usually three members of theresearch team would attend a focus group(two interviewers and a scribe) and twowould conduct an interview. The interviewersacted, as Dick suggests, independently of eachother but in parallel (Dick, 1990). The sameopen-ended questioning approach adopted in the interviews, was adopted in the focusgroups. As themes surfaced, interviewersexplored emerging aspects of the study.

A strict method of reporting findings fromeach focus group and interview was estab-lished. This involved interviewers makingoutline notes of key issues that came out ofmeetings. Notification of these issues wouldthen be passed to the project manager whohad an overview of the research being under-taken. This allowed relatively easy identifica-tion of emergent themes and convergent anddivergent thoughts on those themes fromrespondents. Focus groups are particularlyuseful where multi-perspectives are requiredbut the researcher has limited resources,although, it must be noted that focus groupshave inherent dangers in terms of groupdynamics. Regular research team meetingsallowed interviewers to discuss findings andfor the project directors to amend the processand content with the full commitment andunderstanding of the research team. Tran-scripts of meetings were only used for confir-mation of assertions or to quote exemplarstatements about particular issues.

As the number of focus groups and inter-views increased, an iterative review of theprocess and content was conducted, whichallowed the research team to focus on key con-vergent or divergent themes until they weresatisfied with the validity and reliability oftheir observations. In both projects, draftreports were produced and discussed with theclients, which also included some of the keystakeholder groupings. Feedback and com-ments were received before the final draft wasdelivered to the client.

Findings and discussion

Both projects were well received by theclients and stakeholder participants. Clientfeedback suggests that our method of conver-gent interviewing did indeed identify the mainissues and represented the ‘reality’ of the situ-ations being researched. We do not makeclaims that either project provided completesolutions for the clients, but that the tech-nique of convergent interviewing was effec-tive in providing a progressive mechanism forreduction of uncertainty (Philips and Pugh,

Convergent interviewing 225

Copyright © 2005 John Wiley & Sons, Ltd. Strategic Change, June–July 2005

226 Wil Williams and Duncan Lewis

1994; in Reige and Nair, 2004, p. 73), whichis a key facet for any strategic research project.

As experienced researchers, the researchteam involved in both studies were able tomake comparisons with other recently usedmethods and found the convergent interview-ing technique useful for a number of reasons.First, the structure of the process gave somecertainty in projects that were nebulous andpotentially without boundaries. Convergentinterviewing provided a rationale for theresearch teams to establish boundaries fortheir research around the convergent or diver-gent themes. Second, the process also allowedfor limited resources to be used efficiently andeffectively. Focus groups and interviews hadtangible objectives and means of reportingfindings. Linked with regular research teammeetings, this allowed bias to be identified andan element of objectivity in the study.The con-vergent and divergent views of researchers inthese meetings were an invaluable element inthe research process, allowing for creative andemergent interpretative discussion that never-theless retained objectivity as a result of themethod.

There has been some discussion in the lit-erature (Reige and Nair, 2004; Dick, 1990)about the use of divergent information. Ini-tially, Dick suggests that it should be dis-carded. However, as Reige and Nair indicate,divergent information, in particular circum-stances, should be utilized (Reige and Nair,2004, p. 78). These studies highlighted thisperfectly, where individual respondent’s diver-gent views opened up avenues for furtherresearch by the team and produced someuseful and illuminating findings.

Woodward’s concerns (Woodward, 1996; inRao and Perry, 2003, pp. 239–240) regardingthe convergent interviewing technique wererecognized by the research team before theresearch design phase and during interviewtraining. However, as already illustrated, themain issue of bias was felt to have beenaddressed by the use of convergent inter-viewing, particularly with the use of focusgroups and interviews. By gaining multipleconvergent or divergent interpretations of

views, the research team could be confidentin the validity and reliability of their findings.

The danger of bias through prior knowledgeof the subject was overcome to an extent byattempting to ensure that one member of theinterviewing team had little prior knowledgeof the situation, organization or individualinterviewee’s perspective. Obviously, as theresearch on both projects progressed thisbecame more difficult. However, in the earlyphases of these projects this objectivity wasessential as certain research team membershad been intimately involved in discussionwith the client and examination of client documentation.

The three types of validity measure identi-fied by Rao and Perry (2003), namely con-struct, internal and external validity, wereaddressed by using a number of techniques.Construct validity was achieved through triangulation of focus groups, interview questions and primarily through the verynature of the convergent interview process.Asdescribed, the interview content and processwas re-evaluated and re-designed through theinterview programme. Internal validity was adifficult standard to meet. However, the con-vergent interviewing process relies on casualrelationship identification. This was achievedthrough detailed and rigorous analysis of thedata by the project team. Woodward’s (1996)concerns must be recognized, that convergentinterviewing on its own may not be sufficientin terms of external validity to provide resultsthat can be generalized. Both sets of findingswere presented as generalizable only to the sit-uation being examined and as such should notbe held up as models for ‘similar’ situations soas to establish ‘theory’. However, the repeti-tive nature of the process might be consideredas a form of generalization of the findings. Reli-ability in the form of consistency in this studywas achieved using a structured process ofinterviews. The structured process of record-ing, writing and interpreting data also addedto the method’s reliability. The use of multipleinterviewers working in parallel in the designand conduct of the interviews also con-tributed to the reliability of the interview data.

Copyright © 2005 John Wiley & Sons, Ltd. Strategic Change, June–July 2005

The research teams acted as a ‘steering group’(a term used by Dick, 1990) in the design,analysis and evaluation of the focus group andinterview content and process. As Rao andPerry (2003) identify, validity and reliability inqualitative research comes through forms ofcross-checking and convergent interviewing isa method where cross-checking of data is afundamental feature of the process. The use ofmultiple interviewers, it is argued, is an addi-tional source of validity and reliability in theprocess.

The holistic nature of strategic managementis such that its study invariably involves a mul-titude of perspectives, factors and agents. Theboundaries of any strategic research projectare transient and, as such, can rarely be firmlyset when designing the method. The contentof strategic research is often a moveable butrich feast that requires structure in its processso as to allow some foundation for findings to emerge. In the two cases outlined above itwas found that convergent interviewing was a valuable tool in researching this type of complexity.

Concluding remarks

For projects where there is little or no acade-mic theory, where it is unclear which body oftheory can be brought to bear, or where thereis little precedent in terms of established prac-titioner routine, convergent interviewing is ahighly applicable tool in qualitative strategicresearch. The projects and processesdescribed here deviate from the establishedpractice as detailed by writers such as Dick(1990) and Rao and Perry (2003) largelythrough the use of focus groups. However, wewould argue that the essential characteristicsthat underpin convergent interviewing make

Convergent interviewing isa highly applicable tool in

qualitative strategicresearch

it a sufficiently powerful method in the typeof situation described here.

The apparent success of the projects, evi-denced by the positive reactions of the clientsand key stakeholders, reaffirmed the utility ofconvergent interviewing in inductive, albeit‘commercial’, multi-variant research, i.e.research that may have a number of academicand non-academic foci. Dick (1990) offers forany experienced researcher a valuable obser-vation, ‘The best time to design the data-collection procedures is at the end of theprogram’ (p. 7). Convergent interviewing, toan extent, addresses this issue. Where the areaof research does not clearly sit within one par-ticular field of study and therefore the researchquestion or methodology and method may not be at all clear, or the researcher cannotconduct a pilot study, a multi-cyclicalapproach to research, such as convergentinterviewing, might be more appropriate. Weargue strongly, based on the experience of thecases outlined above, that convergent inter-viewing should find its place in the toolkit ofthe strategic management analyst and consul-tant, if only because the very nature of strate-gic analysis provides a complexity of terrainunseen by functional specialists.

Research into strategic issues lends itself toinductive qualitative methods such as conver-gent interviewing. The use of multiple andexperienced researchers was invaluable ingathering, but more importantly evaluatingand synthesizing data. This allowed the find-ings to be presented with some confidence interms of their validity and reliability. Conver-gent interviewing was particularly useful foridentifying the ‘boundaries’ and therefore keythemes, which should be applied to studieswith amorphous margins.

The removal of discrepancies in interviewresponses is where a main advantage of con-vergent interviewing is gained. Dick (1990)describes content analysis of interview tran-scripts as ‘masochism’ (p. 9), and all of thoseexperienced researchers involved in thestudies described here were glad to haveavoided such an approach. To that end,through the convergent interviewing process,

Convergent interviewing 227

Copyright © 2005 John Wiley & Sons, Ltd. Strategic Change, June–July 2005

228 Wil Williams and Duncan Lewis

the minutiae of transcript evaluation and syn-thesis and the comparatively time-consumingprocess of content analysis was not onlyavoided but justifiably dispensed with.

Whilst Dick sees convergent interviewing aspart of the repertoire of the action researcher,it is felt that it is worthy of consideration byall qualitative academic and practitionerresearchers. In the field of strategic manage-ment research, where variables are often manyand unclear, convergent interviewing mightjust assist the analyst and consultant. Accept-ing that the use of focus groups might be crit-icized by the convergent interviewer purist,we would contend that they were a necessaryand valuable addition to the technique. Any-thing that can contribute significantly to thereliability and validity of a study’s findings is atleast worthy of careful consideration and atbest laudable of practical application.

Biographical notes

Wil Williams is a Principal Lecturer in the University of Glamorgan Business School. Wilis Head of the Strategy Field and a leadingmember of the Welsh Institute for CompetitiveAdvantage.

Duncan Lewis is The ACAS Professor ofWorkplace Futures and Head of the Depart-ment of Strategy.

References

Abercrombie N, Hill S, Turner B. 2000. ThePenguin Dictionary of Sociology. Penguin:London.

Beinhocker E. 1997. Strategy at the edge of chaos.The McKinsey Quarterly (1): 22–40.

Burns RB. 2000. Introduction to ResearchMethods. Sage Publications: London.

Burrell G, Morgan G. 1979. Sociological Paradigmsand Organisational Analysis: Elements of the Sociology of Corporate Life. Heinemann:London.

Courtney H, Kirkland J, Viguerie SP. 1997. Strategyunder uncertainty. Harvard Business Review75(6): 67–81.

Dick R. 1990. Convergent Interviewing. Inter-change: Brisbane.

Easterby-Smith M, Thorpe R, Lowe A. 1994. Man-agement Research — An Introduction. SagePublications: London.

Fields EE. 1988. Qualitative content analysis of tele-vision news: systematic techniques. QualitativeSociology 11(3): 183–193.

Flick U. 2002. An Introduction to QualitativeResearch. Sage Publications: London.

Gummesson E. 2000. Qualitative ResearchMethods in Management Research. Sage Publications: London.

Hammersley M. 1992. What’s Wrong With Ethnog-raphy? Methodological Explorations.Longmans:London.

Hughes J, Sharrock W. 1997. The Philosophy ofSocial Research. Longmans: London.

Kirk J, Miller M. 1986. Reliability and Validity in Qualitative Research. Sage Publications:London.

Kurtz CF, Snowden DJ. 2003. The new dynamics ofstrategy: sense-making in a complex and com-plicated world. IBM Systems Journal 42(3):462–484.

Marshall C, Rossman GB. 1995. Designing Quali-tative Research. Sage Publications: London.

Mason J. 2002. Qualitative Researching. Sage Publications: London.

Mintzberg H, Ahlstrand B, Lampel J. 1998. StrategySafari: A Guided Tour Through the Wilds ofStrategic Management. Financial Times/PrenticeHall: London.

Pascale R, Millemann M, Gioja L. 2000. Surfing theEdge of Chaos. Textre Publishing: London.

Patton MQ. 2002. Qualitative Research and Eval-uation Methods. Sage Publications: London.

Rao S, Perry C. 2003. Convergent interviewing tobuild a theory in under-researched areas: princi-ples and an example investigation of internetusage in inter-firm relationships. QualitativeMarket Research: An International Journal6(4): 236–247.

Reige AM, Nair G. 2004.The diversity of convergentinterviewing applications for early researchersand postgraduate students. The MarketingReview 4: 73–85.

Rocco TS. 2003. Shaping the future: writing up themethod on qualitative studies. Human ResourceDevelopment 14(3): 343–349.

Silverman D. 2000. Doing Qualitative Research.Sage Publications: London.

Copyright © 2005 John Wiley & Sons, Ltd. Strategic Change, June–July 2005

Silverman D. 2001. Interpreting Qualitative Data:Methods for Analysing Talk, Text and Interac-tion. Sage Publications: London.

Stacey RD. 1996. Complexity and Creativity inOrganisations. Berrett-Koehler Publications: SanFrancisco.

Turner BA. 1983. The use of grounded theory forthe qualitative analysis of organisational behav-iour. Journal of Management Studies 20(3):333–348.

Convergent interviewing 229

Copyright © 2005 John Wiley & Sons, Ltd. Strategic Change, June–July 2005