IR 1

A critical reviewof interventionist research

Vicki BaardDepartment of Accounting and Finance, Faculty of Business and Economics,

Macquarie University, Sydney, Australia

Abstract

Purpose – The purpose of this paper is to further develop the epistemological base of interventionistresearch (IR) as a valid accounting and management research methodology, through the identificationof intervention theory and an IR framework derived from social sciences. Moreover, this paper seeks tocontribute to empirical knowledge of IR through a critical review of limited empirical evidence relatingto intervention theory and the extant IR frameworks derived from action research.

Design/methodology/approach – Texts and academic journal papers that judiciously reviewintervention theory, intervention research frameworks were identified systematically; along withempirical research addressing theoretical and methodological deficiencies of IR and, providingevidence to inform practical considerations when undertaking IR.

Findings – The key findings include rare empirical evidence addressing theoretical shortcomingsand application of intervention theory, an IR framework derived from social sciences with extremelylimited use in accounting and management research, deficiencies in action research orientedframeworks labelled as alternative forms of IR, an alternate perspective to positivistic validity andreliability issues and other practical considerations to facilitate the conducting of IR.

Originality/value – The novelty of this paper lies in the diminution of the fragmented nature of IRthat undermines its scientific value through the identification of an intervention theory and IRframework experiencing extremely limited use in accounting and management research, with theexception of a cross-disciplinary (management accounting and information systems) doctoral study,optimising IR utilisation with greater degrees of validity and reliability and, finally, a proposedalternative research design for utilisation in IR.

Keywords Action research, Research methods

Paper type Literature review

IntroductionRecently, interventionist research (IR) has entertained renewed interest as a researchmethodology in accounting and management. Given the practical knowledge outcomes,in contrast to purely theoretical knowledge development (KD) focused research;organisations and practitioners grapple with finding the relevance of the theoreticalliterature to solving problems. Chartered Institute of Management Accountants (CIMA)called for IR proposals in 2006 to promote the usefulness of management accounting fororganisations using this approach. Lukka (2006) explained why CIMA embarked on thisresearch initiative, through rhetoric embracing how to make management accounting“findings of value to practice” and that “interventionist research aims to narrow the gapbetween practice and academic theory”. It seems that CIMA calling for IR was a timelyone; there is a dearth of IR in management accounting. Jonsson and Lukka (2007)concurs, save a few Finnish studies (Kasanen et al., 1993; Lukka, 2007; Labro andTuomela, 2003; Malmi et al., 2004; Labro et al., 2005) which packaged IR as aconstructive research approach (CRA). Other management accounting studies using anaction research (AR) approach included Seal et al. (1999), Arnaboldi and Azzone (2004)

The current issue and full text archive of this journal is available at

www.emeraldinsight.com/1176-6093.htm

A critical reviewof interventionist

research

13

Qualitative Research in Accounting &Management

Vol. 7 No. 1, 2010pp. 13-45

q Emerald Group Publishing Limited1176-6093

DOI 10.1108/11766091011034262

and Andriessen (2007)[1]. Jonsson and Lukka (2007) and van Aken (2004) also refer todesign-based research (DBR) as design science; Jonsson and Lukka indicate that designscience is an alternative form of IR. Baard (2004) used IR[2] in a cross-disciplinary(management accounting and information systems) doctoral thesis in the small businessenvironment, see also Baard and van den Berg (2004a, b). This dearth of IR is notsignificantly different with regards to other disciplines, for example informationsystems. Baskerville and Wood-Harper (1996) identified rare instances of IR in NorthAmerica, with slightly more activity in the UK, Scandinavia and Australia.

Given this renewed interest in IR, a fragmented notion the nature of IR is prevalentbecause facets of IR are entrenched in AR and branded with AR derivatives such asaction science, clinical research, design science and constructive research, thusproducing reservations concerning its scientific value. Moreover, the absence of an IRtheory or framework, and the related paucity of the extant empirical research furtherexemplified this haziness. This has therefore contributed to the existence of several gapsin the epistemology and practice of IR. First, conceptually, the theoretical foundations ofIR are not explicit, producing reservations concerning its scientific validity, disciplineand therefore legitimacy as a methodology. Given the aim and apparent substantivenature of IR on the one hand, positivist scepticism concerning its scientific validity onthe other, extant empirical accounting and management research seems silent on atheory peculiar to IR. The extant literature identifies tensions between the positivist andinterpretive research perspectives concerning the scientific validity of IR as a qualitativemethodology (Susman and Evered, 1978; Aguinis, 1993; Kasanen et al., 1993; Atkinsonand Shaffir, 1998; Breu and Peppard, 2003; Labro and Tuomela, 2003; Ahrens andChapman, 2006; Davila and Oyon, 2008). Researchers intervening in organisationalsystems have dual objectives: to advance knowledge in their field and help improve thesystem under study. Lindenburg et al. (2001, p. 132) concur and indicate that in manyinstances the goals and methods of science are incongruent with the priorities andrealities of the clients, institutions and/or communities where research is conducted. Incontrast, descriptive or normal science researchers undertaking field work aim toproduce new scientific knowledge through objective study of an empirical object; theobject of study is assumed to be unchanged by the process of studying it (Edmondsonand Moingeon, 1999, p. 170). Positivist research represents rigorous research, it relies onthe importance of results describing knowledge, and the reproduction of these resultscaptured in a different study, it provides external validity, advances knowledge andcontributes to the evolution of theory. AR is confronted by problems including a lack ofimpartiality, discipline or rigor, a lack of differentiation from consulting and is contextbound (Baskerville and Wood-Harper, 1996). Breu and Peppard (2003) assert thatprogress concerning IR output in information systems is retarded with interventionistapproaches being dismissed as unscientific because of the purported absence ofphilosophical foundations compelling it to draw on a “hazy theoretical base” (p. 181).This said, Susman and Evered (1978) propose that AR can be legitimated as a science bylocating its foundation in philosophical viewpoints which differ from those used tolegitimate positivist science; the contribution of this AR legitimisation to a theory of IRhowever remains to be seen.

Second, from a research methodology perspective, are there any distinctiveIR frameworks or methodologies that can be used to stimulate a structured approachto this sort of research, which provides a greater degree of rigour and legitimacy to IR?

QRAM7,1

14

Third, knowledge gathered from disciplines where IR utilisation is largely prevalent(e.g. Nursing, Education, Public Health, Dementia, Social Work, OrganisationalDevelopment and Clinical Psychology), raises the question as to whether consolidatingthis knowledge can contribute to and influence IR practice in accounting andmanagement. In order to address and improve our understanding of theabove-mentioned issues, a critical review of IR in the context of the extant literaturewas required to develop our knowledge and appreciation of these issues. Reviews of IR,for example Jonsson and Lukka (2007) provides a useful introduction to the discourseof IR, through its exploration of the nature of IR, its variations, a suggestedphilosophical basis and the forms of IR output, in a management accounting context;disappointingly it seems to be the only one.

Therefore, the purpose of this paper is to further develop the epistemological base ofIR as a valid accounting and management research methodology, through theidentification of intervention theory and an IR framework derived from social sciences.Moreover, this paper seeks to contribute to empirical knowledge of IR through a criticalreview of limited empirical evidence relating to intervention theory and the extant IRframeworks derived from AR. This paper is organised into five sections. The firstsection specifies how the literature review was conducted, followed by a discussion ofthe nature of IR. The third section presents an overview of intervention theory,followed by the introduction of an intervention framework. The fourth sectiondescribes other key IR issues such as research design and validity and reliability.

Review methodologyTo construct a critical review of IR providing useful outcomes for readers and future IRresearchers, a systematic review of the extant literature was conducted. This reviewcommenced with the analysis of several texts concerning IR, in my personal collection.Additionally, two electronic databases were used to source other relevant publishedstudies ( journal articles), namely ABI/Inform Global Proquest and EBSCO. Thesedatabases were searched using the keywords “interventionist research” “interventionresearch” and “intervention theory” with searches limited to “citations and abstracts”.In the former database 40 references were found; in the latter 1,178; the relevant studieswere then selected from these databases. The studies deemed relevant for review inthis paper fulfilled the following selection criteria:

. research published in scholarly peer reviewed journals (excluded were editorials,letters to the editor, practice notes, exchanges between authors, book reviews andso forth);

. research focusing on intervention research; and

. studies were not restricted to accounting and management.

Any references cited in each of the works that met the selection criteria, was also made.It must be noted that an exhaustive review of publications relating interventionresearch was not the sole object of this paper; the research selected representedreasonably the primary thrusts to support a critical review of IR. A limitation thereforemight be that some works have been overlooked. This does not however invalidate thecontribution of this paper toward the existing body of knowledge of IR.


research

15

The nature of IRAR is considered the origin of all IR in the area of social sciences ( Jonsson and Lukka,2007, p. 376), initiated by Kurt Lewin (1946, 1951) who posited the idea of doing changeexperiments, in the field rather than the laboratory. There literature on AR isconsiderable (Burns, 2007; Stringer, 2007; Reason, 2006; Reason and Bradbury, 2006;Breu and Peppard, 2003; Chandler and Torbert, 2003; Paisey and Paisey, 2003; Bradburyand Reason, 2003; Baskerville and Wood-Harper, 1996; Eden and Huxham, 1996;Edmondson, 1996; Argyris and Schon, 1989; Argyris et al., 1985; Rickards and Bessant,1980; Susman and Evered, 1978; Foster, 1972, Rapoport, 1970; Blum, 1955). Theseauthors offer definitions, perspectives and empirical knowledge that have been useful inenlightening our understanding of IR. However, current IR literature is not forthcomingin attempting to distinguish between it and other forms of applied social research. Thissaid, how do we define intervention and what is IR? “To intervene is to enter into anongoing system of relationships, to come between or among persons, groups or objectsfor the purpose of helping them” (Argyris (1970, p. 15). Intervention is action takenwithin a system of which the researcher is not a part of or constitutes behaviours thatinterject into the ongoing social processes of a system or are novel approaches toaddressing problems involving intrusion into complex health, social, political, culturaland/or technical environments (Cope, 2000; Beckhard, 1979; Lindenburg et al., 2001).Finally, Carkhuff (1983, p. 163) offers an extension on the prior definitions and states:

An intervention is both a response and an initiative. It is a response to a situation that definesa need. It is a response to a deficit or to what is not present. At the same time, it is an initiativeto influence that situation – to fill in what is not present, to transform the deficits into assets.In short, an intervention is an attempt to make a difference.

Mullen (1994, p. 167) determines that the objective of intervention research is thedevelopment of a social technology to ameliorate a social problem; a statement of desiredchange in a social problem or in some condition related to a social problem. IR drawsfrom such areas as evaluation research, behavioural assessment, technologyassessment, technological transfer, simulation and modelling, meta-analysis,knowledge utilisation (KU), practice technology and system engineering; it is theconfiguring of these methodological elements into a phased system of action thatharnesses their potential as general practical intervening innovations (Thomas andRothman, 1994). Thomas and Rothman (1994) and Fawcett et al. (1994) also suggest thatIR has two outputs, namely a knowledge product for both researcher and practitionerand a practice product or intervention developed for problem solution. The aim of thisform of research is therefore to improve community life and well-being (including theorganisational context) through the development of interventions, which are effective invarious real-life contexts, involving a co-ordinated effort of all participants who areactually experiencing the problem, resulting in the generation of knowledge forresearchers and practitioners. This introduction to the nature of IR serves as thefoundation for distinguishing IR its variants and is followed by a discussion onintervention theory and IR frameworks may provide further points of differentiation.

Intervention theoryPreviously, Breu and Peppard (2003) asserted that interventionist approaches draw ona “hazy theoretical base”. Jonsson and Lukka (2007) informed us that the role of theory

QRAM7,1

16

in IR currently is and ought to be debated. This role of theory in IR serves twopurposes. First, theory is used to diagnose problems, to construct an appropriateintervention and to position the findings to contribute towards the production ofpublishable scientific knowledge. Second, there is the use of a theoretical framework toprovide strategies to accomplish IR activities. Numerous studies have been published,containing empirical evidence of interventions that have stimulated significant changein their elected social system (Christopher et al., 2008; McCarty Kilian et al., 2005;Robitaille et al., 2005; Stronks and Mackenbach, 2005; Fishbein et al., 2001; Snyder et al.,2001; Camp, 2001). However, no explicit reference to intervention theory or frameworkoccurs. An intervention theory is therefore necessary and relevant because it serves asguide for intervention researchers, it may provide discipline, exactitude and ultimatelylegitimacy to IR as a research method. Argyris (1970) constructed a theoreticalframework of consulting, referred to as intervention theory; this theory stems fromorganisational theory and the integration of borrowed concepts and empirical researchfrom relevant behavioural sciences (p. 14). Argyris cautions that this theoreticalframework regarding intervention activities is a primitive attempt. However, rareinstances of studies using intervention theory and the paucity of empirical researchusing IR research approaches, is insufficient to epitomise its primitiveness. Rather it isa useful starting point for its contestation to be a worthy opponent for the positivistapproaches.

Intervention theory presents a broad spectrum of intervention activity principlesincluding but not limited to: the primary tasks of intervention, intervention activities,qualities and behaviour of an effective interventionist, difficulties associated withclient and interventionist interaction, effective and ineffective intervention activity,organisational entropy including resistance to change and modifications to establishedconcepts of rigorous methodology. However, given that the primary tasks constituteessential processes supporting intervention activity, they enjoy the primary focus inthis section. The primary tasks of intervention (see Figure 1) are:

. the generation of valid and useful information;

. free, informed choice that maintains effective interventionist behaviour andclient system integrity; and

. internal commitment to choices made for decision implementation.

These primary tasks influence the client system and the interventionist. Beforediscussing intervention theory, clarification of some terminology is required; theseinclude “client system” “competence” “effective” and “intervention system”. Argyris(1970, p. 16) refers to the “client system” as consisting of individuals, groups andorganisations. Susman and Evered (1978, p. 588) refer to a “client system” as the socialsystem in which the members face problems to be solved by AR; it may be one of theface-to-face groups, an organisation, network of organisations, or a community.Therefore, to clarify the term “client system” and avoid association with consultation,“client system” is hereinafter referred to as the participant system; the participantsystem may therefore comprise of individuals or groups or organisations, or a networkof organisations, or a community.

Reference to participant systems as competent and effective and effectiveinterventions is also made. Argyris (1970, p. 36) indicates that the core activities of anysystem are:


research

17

. to achieve its objectives;

. to maintain its internal environment; and

. to adapt to, and maintain control over, the relevant external environment.

Thus, a participant system is competent when it achieves these activities over time andunder different conditions, and is effective when it achieves these activities in anygiven situation. An intervention is effective when it assists the participant systemachieve the core activities, thus augmenting (not reducing) its competence andeffectiveness through autonomous problem-solving, decision-making and decisionimplementation. Hasenfeld and Furman (1994, p. 299) defines “effectiveness” as:

. the intervention design was correct and the associated procedures andtechniques ameliorated the problem;

. the intervention has been tested (even refined perhaps) and works; and

. the organisation has adopted the innovation aiding service deliveryoperationally.

This definition is important when considering validity and reliability concerns of IR.Finally, the intervention system refers to the collaborative relationship between theparticipant system and interventionist (Figure 1).

The generation of valid (relevant and understandable) and useful information forthe participant and the interventionist provides an accurate and collaborative problem

Figure 1.Diagrammaticrepresentation ofintervention theory

Effective andcompetent

Effective

Decisionimplementing

Problemsolution

Participantsystem

Interventionist

Problem-solvingability

Diagnosis

Alternativecourses of action

Decision-making

Generate valid and usefulinformation

Create conditions for freeand informed choice

Internal commitment

Source: Adapted from Argyris (1970)

QRAM7,1

18

diagnosis, describes factoral interrelationships contributing to the problem, and toinitiate the participant learning process required for problem solution. To certifyvalidity of this information, it must be:

. publicly verifiable;

. negligibly attributive to participant system behaviour, negligibly evaluative ofparticipant behaviour to reduce their reticence for defensiveness thusencouraging openness for learning; and

. non-contradictory information that reduces the effectiveness of relationshipsbetween the participant system, the intervention and the interventionist.

The second task concerns the creation of conditions for the participant system to makefree (voluntary and proactive) and informed choices. The interventionist devisesalternative courses of action to remedy the problem (congruency between course ofaction and intended outcome) and ensures that the alternatives are congruent with thecapacity (time, people and material resources) of the participant system. This taskpromotes participant system autonomy, taking responsibility for their future andproviding the ability to explore several significant alternatives, crucial to their needs;this system therefore engages in decision-making to select the option pivotal toproblem solution. Argyris (1970) emphasises that free choice is important for: first, theparticipant system to achieve significant willingness and motivation to workon the problem to facilitate successful change and second, the interventionist to resistthe pressures of being controlled by participant system angst and proposingalternative courses of actions that driven by needs of participant system control andcoercion. Internal commitment means that the participant system should experience ahigh degree of ownership and a feeling of responsibility about the course of actionselected and implemented, and its implications (Argyris, 1970, p. 20). Internalcommitment is related to the execution of decisions in such as way that the problemdoes not recur and that the existing level of competence is not deteriorated (Argyris,1970, p. 37). Given this, the intervention can be deemed effective and effectiveintervention activity assists the participant system to operate more competently.The participant system will be effective through learning how to solve a problem to thedegree that it generates valid and useful information employs free and informed choiceand demonstrates internal commitment.

Argyris (1970) proposes that these primary tasks have implications for interventionactivity. These implications provide useful information for conducting IR andstimulate philosophical discussions thereon. First, there should be congruence betweeneffective intervention activities and participant system activities forming part ofnormal operations. This would reduce the probability of intervention rejection andpromote participant system competency for recurring problem solution, exclusive ofthe interventionist. Second, that change is not a principle interventionist task, nor ischange production a criterion for effective interventions, even though primary taskaccomplishment inevitably leads to change. It should rather be seen as a by-product ofintervention, which preserves the long-range effectiveness of change. Third,primary tasks serve as criteria for selecting participant systems (Argyris, 1970,p. 24); the participant system must want to be helped to increase the probability ofintervention strategy, based the primary tasks, being successful. This implicationalso suggests the involvement of top management is important for effective


research

19

intervention activity stimulation. Fourthly, both participant system and interventionmust retain autonomy through minimisation of manipulation. In this instance,manipulation refers to the interventionist allowing or encouraging participant systemdependence on the researcher, rather than learning to make choices themselves. Fifthly,the primary tasks can be used to determine the conditions, under which theinterventionist will enter, inhabit and depart the participant system as well asdetermine the principles of engagement. Finally, primary task accomplishment shouldresult in the advancement of knowledge in two areas, that of science (academic) andprofessional knowledge (practice). The requirement to obtain valid informationencourages the interventionist to add to the basic knowledge in his field.

Finally, Argyris argues that the primary tasks should be aligned with theinterventionist researchers’ activities. Therefore, rigorous research, traditional orscientific research (“mechanistic research”), should not be used throughout the IRprocess. Using rigorous research results in unintended consequences (e.g. participantsystem dependence on the interventionist, physical and psychological withdrawalfrom intervention process, ineffectiveness of the problem-solving process and hostilitytoward the research being undertaken) that do not encourage achievement of theprimary tasks. Instead an “organic” approach should be used where there iscollaboration between participant system and interventionist researcher concerning allaspects of research design and accomplishment, see Table I.

Empirical evidence demonstrating the use of intervention theory is rare. Foster(1972, p. 536) asserted that the practice of AR, without its guiding theory frequentlymakes little sense. Therefore, in search of a theoretical foundation for AR, Fosterexplored elements of theory and practice in the literature, including interventiontheory. Similarly to Argyris (1970), Foster also uses organisational theory, the theory ofchange, organisational development, field experimentation and consulting in hisconsideration of AR theory. Foster makes no specific reference to intervention theoryas a construct or even acknowledges its existence; specific reference to Argyris’s workis very limited. Rather, it is confined to a discussion on approaches to planned changeand behavioural issues that arise from the application of interventions and theinteraction of the client system with the interventionist. Yet, Fosters’ discourse on ARtheory yields many similarities to intervention theory, including but not limited to:reference to the client system, the criteria for selecting a client system, the collaborativerelationship between research and client termed “a third organisation”[3] (Foster, 1972,p. 538), the importance of interventionist and client autonomy and the behaviouralimprovements sought in the client system. These similarities in AR theory tointervention theory, suggests that Argyris’ work contains merit and may validatethe explanatory concepts contained in intervention theory. Finally, Foster, refers to theresearcher as a “change agent” not interventionist, and focuses largely on the intentionof AR to render change; the matter of change as a primary intention contrasts withArgyris’ view.

Van De Vliert (1977) offers, to my knowledge, the only critical evaluation ofintervention theory, which is useful since it presents an opportunity for theoryrefinement or concept clarification. Theory refinement may support our quest toconduct IR that may be considered valid, scientific and not “sloppy” (Eden andHuxham, 1996). Argyris (1970) argues that intervention theory represents an internallyconsistent framework; it is the issue of “consistency” that Van De Vliert uses as the

QRAM7,1

20

crux of his evaluation. Van De Vliert (1977, p. 557) contends that every theory shouldaim to comply with logical[4] and empirical[5] consistency; intervention theorycontains five logical and empirical inconsistencies regarding the primary tasks ofintervention. Given the scope of this paper, only two inconsistencies are addressed. Thefirst inconsistency relates to Argyris’s postulation that change is not a primary task ofthe interventionist; Van De Vliert and Beckhard (1979) stipulate otherwise and agreethat the interventionist has a need to influence change in the participant system.Previously, Foster (1972) argued that the primary intention of AR is change, thereforereferring to the researcher as a “change agent”. However, this is intervention theoryand perhaps one of the distinguishing characteristics from other forms of IR is the viewthat change is not the only product of the process. Change and adaptability could beconstrued as a side effect of a successful intervention, given the knowledgeand practical product outputs of the IR process. The second inconsistency, relatesto mechanistic versus organic research, especially regarding the generation ofvalid information; seemingly this is a contentious issue subject to further debate.

Mechanistic research Organic research

The interventionist drives goal definition of theprogram

Participant participation in goal definition by,confirming and disconfirming, modifying oradding to those goals defined by theinterventionist

The interventionist assumes that this relationshipof being strictly professional cannot be influencedby the participants. He maintains his power ofexpertise and therefore keeps a professionaldistance from the participantsParticipants have dependent and submissive roleswith low feelings of essentiality in the program

The interventionist acknowledges that they arestrangers (tourists) in the institution. Participantsshould be encouraged to confront and test theirrelationship with him. Interventionist power overparticipants, due to academic competence isequalised through encouragement to question theinterventionist and the entire programParticipants are provided with opportunities forpsychological success, feelings of essentiality,confidence development and trust in others,effective group relations and minimiseddependent and submissive relationships

The amount of participant participation in theentire project is controlled by the interventionist

The amount of participation is influenced by thesubject and the interventionist

The interventionist depends upon theparticipants’ need to be helped or need tocooperate as being the basis for their involvement.He expects participants to be used as informationgivers

The interventionist depends upon the participantsneed to be helped for encouraging them to controland define the program so that they becomeinternally involved and feel that they are asresponsible as the interventionist

If participation is encouraged, it tends to be skin-deep, designed to keep subjects “happy”

Participation is encouraged in terms of instrumentdesign, research methods and change strategy

The costs and rewards of the change program aredefined primarily by the interventionist

The costs and rewards of the change program aredefined by the participants and the interventionist

The feedback to subjects is designed to informthem how much the diagnostician

The feedback to participants is designed tounfreeze them, as well as to help them developmore effective interpersonal relations and groupprocesses

Source: Adapted from Argyris (1970, p. 104)

Table I.Mechanistic and organic

research


research

21

Van De Vliet argues that organic research is not used throughout an interventionproject, specifically during the generation of valid information (diagnostics) stage,thereby infringing free choice and internal commitment achievement. Argyris (1970,p. 104) argues that rigorous research methods are the interventionists’ best chance forobtaining valid data. However, modifications to mechanistic research are required toreduce the unintended consequences of traditional research in IR, previously described;Van De Vliert omits this from his argument. The notion of using organic research isbased on fortifying the relationship between interventionist and participant system to“enhance the effectiveness of the problem-solving process so that the clients couldcontinue to maintain or increase its effectiveness after the interventionist has left”(Argyris, 1970, p. 105). This collaborative relationship will increase the participants’probability of primary task completion. Baard (2004), in this context employed aquasi-experimental field approach in an interventionist study to generate valid datafrom which intervention design could commence, but also used observations andinterviews to infuse an organically-centred approach, to check inconsistencies in theinterview material and to promote intervention system harmony through participantsystem engagement.

Dirks et al. (1978) applied intervention theory, specifically the primary tasks, in asmall informal organisation, concerning organisational change. They found theprimary tasks useful for planning their research, the immersion of the researcher in theresearch process an important stimulant for collaboration, and through internalcommitment encouraging participants to achieve effective long-term change. Argyrisand Kaplan (1994) used intervention theory principles in the organisationalimplementation of a new technical theory, activity-based costing (ABC); an actionscience approach was used. This implementation had intended consequences, that oforganisational change. In response to this change the actors within the organisationapply defensive strategies to obstruct implementation, which are reflective of Model Ibehaviour (Argyris and Schon, 1974). These strategies include unilateral control of therelevant environment and tasks and the suppression of negative feelings, which areimplemented through making evaluations and attribution, discouraging inquiry andtreating ones’ own view as correct. The consequences of Model I strategies includedefensive interpersonal and group relationships, low freedom of choice and reducedproduction of valid information (Argyris et al., 1985). It is at this point that the actionscientist or interventionist seeks to help participant systems and their members learn,especially with a view to change. Model II, therefore hypothesised to enhance learningthrough three governing variables also known as the primary tasks associated withintervention theory, namely:

(1) valid information;

(2) free and informed choice; and

(3) internal commitment; these primary tasks are invoked to change the behaviourof the actors to reduce resistance to change or the implementation of ABC.

Argyris et al. (1985) used intervention theory as a theory of action (theory-in-use) intheir discourse of theories of action in an action science context, as did Argyris andKaplan (1994). This demonstrates the interchangeability between action theory andintervention theory; I would however discourage this practice as it can be confusing tothe novice researcher who may want to engage in IR.

QRAM7,1

22

Intervention research frameworksThomas and Rothman (1994) constructed an IR framework, stemming from informationscience and social science (social work and allied disciplines of Psychology, Psychiatryand Sociology). Atkinson and Shaffir (1998, p. 59) assert that sociology has a long historyand experience with field research from which accounting researchers should learn andprofit. IR is a genre of applied research (Thomas and Rothman, 1994, p. 3), scaffolded onbasic research that explores problems in real-world contexts, with the aim of providingpractical solutions (applicable to organisational practice) and creating knowledge. ThisIR framework consists of varied approaches used to address applied research. Thomasand Rothman (1994, p. 3) describe these approaches as:

. empirical research to extend the knowledge of human behaviour relating tohuman service intervention (intervention KD);

. the means by which the findings from intervention KD may be linked to, andutilised in practical application (intervention KU); and

. research directed toward developing innovative interventions (interventiondesign and development (D&D)).

Figure 2 illustrates an integrative perspective on intervention research through theinterrelationships of KD, KU and D&D: the arrows with solid lines indicate explicitlinkage between aspects; arrows with broken lines represent indirect or developinglinkage.

Table II emphasises the differences between the KD, KU and D&D concerning theirobjectives, methods and outcomes and expands on the nature of their interrelationships(linkages). The linkages between the facets provide flexibility when conducting IR. Forexample, KD can be conducted on its own, without D&D, or it can be followed by KUactivities, or it can be incorporated into D&D. KD, KU and D&D can also be conductedsequentially.

Figure 2.An IR framework

Social and behaviourialscience research

Applied research General knowledgeutilisation

Interventionknowledge

development (KD)

Knowledgeutilisation for

intervention (KU)

Interventiondesign and

development (D&D)

Other uses andusers

Intervention research

Source: Thomas and Rothman (1994, p. 5)


research

23

The extant literature did not reveal an IR framework containing similar facets to thatof Thomas and Rothman (1994). However, Argyris (1970) refers to three interventionactivities prevalent in intervention theory. The first activity (analogous to KU)concerns the utilisation of an existing body of knowledge, experience and techniquesfor solving problems common to different participant systems. The second activity(analogous to KD) is described as the manipulation of an existing body of knowledge,experience and techniques in an innovative manner. Thomas and Rothman include thetheoretical thrust of IR in both KU and KD, which also supports the D&D facet,however Argyris does not appear to include theoretical notions in the first twoactivities. The third activity (analogous to D&D) is described as the development andconstruction of new conceptual models through the combination of the participantsystem and the interventionists’ resources. Thomas and Rothman operationalise KD,KU and D&D in terms of prescribed objectives, methods, outcomes and linkages, butare not specific about participant and interventionist resources. Argyris refers to asliding scale of resources including skills, time, incumbent degrees of effectiveness and

Facets of intervention researchAreas ofdifference KD KU D&D

Objectives To contribute to knowledgeof human behaviourDistinct effort to createfindings applicable to theunderstanding and/orsolutions of practicalproblems

To apply knowledge ofhuman behaviourTo make knowledge fromresearch practical; putknowledge in a usable form

Systematic methodology toevolve new human servicetechnology (e.g. treatmentmethods, programs, servicesystems or policies)

Methods Conventional, social andbehavioural science researchmethodse.g. single-caseexperimentationQuestions posed are moreinstrumental and practicalthan is the case with basicresearch

Transformation andconversion of availableknowledge (theory andempirical research) intoapplication concepts andtheories relevant to giventarget populations, problemsand intervention methods(methods include meta-analysis)

Emerging methods includethe means of problemanalysis, interventiondesign, development,evaluation anddissemination and relatedtechniquesConsists of six main phases

Outcomes Information about humanbehaviour in the form, forexample, of concepts,hypotheses, theories andempirical generalisations

Such applications aschanges in theunderstanding or practicesrelating to populations,problems, or interventions inhuman service

Such technical means ofachieving human serviceobjectives as assessmentand intervention methods;and service programs,systems and policies

Linkages Distinctive activityexplicitly linked to KU andD&DIncorporated into D&D(problem analysis & projectplanning phase)

KU may lead directly into orbe an integral part of theInformation gathering anddesign phases of D&DConversion of knowledge forother users

Distinctive activity OR KUgenerally occurs in theformative stages of D&D

Source: Adapted from Thomas and Rothman (1994, p. 7)

Table II.KD, KU and D&Dfacets of IR

QRAM7,1

24

competences required for each activity, with activity three characterised as extremelydemanding on the resources of intervention system. Finally, Argyris does notexemplify the linkages between the activities or provide operational prescriptions.

D&D has emerged as an explicit paradigm, largely out of frustrations with theinability of conventional research methods to guide the generation of human serviceinterventions; it is the methodology and practice of D&D that provides the uniquenessof intervention research (Thomas and Rothman, 1994) (see Table III).

Therefore, in contrast to the accentuation of interrelationships between variablescharacteristic of traditional research, D&D enjoys a primary focus on the evolution ofintervention technology. Examples of such technologies could include ABC inmanufacturing organisations, a service system, a new or modified policy, a newstrategy, training materials or a motivational system. It can be conceptualised as aproblem-solving process for seeking effective intervention tools to deal with givenhuman and social difficulties; this process is systematic, deliberate and immersed inresearch procedures, techniques and other instrumentalities. The D&D facet, seeTable III, constitutes a model for conducting IR, specifically intervention D&D.It involves sequential and interconnected activities intended to guide researchersand practitioners for effecting change in problem situations, concluding with thecreation of innovative intervention tools. D&D incorporates several differentoverlapping paradigms that seek to construct a systematic methodology forintervention development (Fawcett et al., 1994), similarly to the “borrowed concepts”constituting intervention theory. First, developmental research (Thomas, 1984) usedfor intervention design through the incorporation of applied research methods,empirical practice and other AR strategies. Second, social research and development(Rothman, 1980) which applies a physical sciences engineering model to embody theintervention development process, similar to design science. Third, behaviouralresearch (Fawcett, 1990, 1991) applied to intervention design and implementation,using concepts and methods of behaviour analysis and psychology. Fourth,experimental social innovation (Fairweather, 1967) used for intervention (innovation)evaluation using quasi-experimental designs. Finally, model development research(Paine et al., 1984), focusing on intervention dissemination from innovation to standardpractice.

There are two studies that have used the IR framework. Abell and Wolf (2003)undertook a partial application and adaptation[6] of this framework, to harness acreative approach to social work doctoral education. This particular study applied theD&D and curricular components of one doctoral student over a three-year period; thisfacet was adapted to include a research agenda for a doctoral project encompassingintervention development. Baard (2004) used IR in a doctoral study, specially the D&Dfacet, for intervention design, development and implementation in the small businessenvironment. This study only utilised the first five phases of the D&D process, adaptedto accommodate time and resource constraints and limitations of the environment inwhich innovation implementation occurred. The obvious challenge for adapting themodel to doctoral training begins with recognising that D&D cannot be accommodatedentirely in any single product (Fawcett et al., 1994). Baard found accomplishing D&Dactivities in each phase could occur linearly, but re-cycling back to earlier phasesoccurred during the course of the project. Phases 1 and 2 were performed almostverbatim to the D&D model, whilst Phase 3 included design criteria and intervention


research

25

Operations (Thomas andRothman) Activities (Fawcett et al.)

Phase 1: problem analysis andproject planning

Key problems are identified andanalysedBroad state of the art review isinitiated to provide generalorientation to the problemFeasibility of the D&D project isdetermined

Identifying and involvingparticipantsGaining entry and cooperationfrom settingsForming collaborativerelationshipsIdentifying concerns of thepopulationAnalysing identified problemsSetting goals and objectives

Phase 2: information gatheringand synthesis

Various types of data serve asthe basis for interventiondevelopment:

Basic and applied researchMain contributions of KDresearchOriginal research conductedincluding a needs assessment,survey of practitionersregarding current practiceProfessional/personalexperience

Using existing informationsources (e.g. a literature review)Studying natural examplesIdentifying functional elementsof successful models(interventions)

Phase 3: design Interlaced with researchers andusers in close interactionDesign must include usabilityDesign domain, requirementsand problems must be outlined

Designing an observationalsystemSpecifying procedural elementsof the Interventions

Phase 4: early development andpilot testing

Pilot testing to explorefeasibility of interventionInterventional refinementTrial implementation techniques

Developing a prototype orpreliminary interventionConducting a pilot testApplying design criteria tothe preliminary interventionconcept

Phase 5: evaluation andadvanced development(experimental social research)

Determine effectiveness ofinterventionAdditional development ofintervention

Selecting an experimentaldesignCollecting and analysing dataReplicating the Interventionunder field conditionsRefining the intervention

Phase 6: dissemination (modeldevelopment research)

Process of diffusion andadoption

Packaging/Fabrication ofintervention for userreadinessReinventions if requiredEmploy modes of motivationand persuasion to useintervention

Preparing the product fordissemination (e.g. price,standards for useCreating a demand for theinterventionEncouraging appropriateadaptationProviding technical support foradopters

Sources: Adapted from Thomas and Rothman (1994); Fawcett et al. (1994)

Table III.The lifecycle ofintervention D&D

QRAM7,1

26

prototype development, an intervention pilot test and intervention refinement. Phase 4was used for intervention implementation and Phase 5 concluded with interventionevaluation. Whilst Phase 6 was essentially excluded from the study, dissemination oftwo “products” did occur. Research participants retained ownership of the interventionand were provided with a professional report detailing the results of the study in“laymans” terms. This use of dissemination is consistent with Goldenhar et al. (2001,p. 620) who stipulate that results should be reported to intervention participants ina form that is readily understood. Finally, academic journal publications constitutingthe knowledge product occurred, see Baard and van den Berg (2004a, b). The abovetwo studies indicate the applicability of this framework in IR research. I wouldhowever caution researchers when approaching Phase 6 of the D&D model. It suggestsintervention commercialisation, which may overshadow the true purpose of IR, hinderscientific knowledge production and brand IR as glorified consulting (Kasanen et al.,1993; Vaivio, 2008).

Alternative IR frameworksGiven the very limited application of the D&D facet of IR in accounting andmanagement, a review of similar frameworks to put this facet into some perspective,occurred. The alternative models emanating from different schools of IR include that ofJonsson and Lukka (2007); Goldenhar et al. (2001), Avenier and Nourry (1999), the CRA(Kasanen et al., 1993; Labro and Tuomela, 2003) and design science (van Aken, 2004;Andriesson, 2007).

Jonsson and Lukka (2007) introduced a philosophy of doing IR (see Figure 3) withina management accounting context.

Figure 3.Philosophy ofconducting IR

Field of practice (realm of practical reason)

Academicresearcher

(pure logic ofacademia)(classical

rationality - purereason)

Practioner(practical logic of the

field)

2. What should aperson like me doin a situation like

this?

2. What shouldI do now?

1. Understand the practical situation

3. Deliberation(Generate desire-independent action)

4. Initiate justifiable action together withpractioners

(cause the consequences)

5. Maintain intention-in-action skilfully

Source: Adapted from Jonsson and Lukka (2007)


research

27

We cannot suppose that a philosophy is necessarily a theory or an implied framework.However, having diagrammatically arranged the concepts comprising this philosophy,I would propose its further development either into a competing intervention theory, oran IR framework, or it can make a significant contribution to intervention researchtheory and framework refinement. Moreover, this philosophy is not necessarilyconfined to a management accounting context. The philosophy is “practical reason”centric, whilst evolving from multiple authors, is based primarily on classicalrationality (Searle, 2001), ethnomethodology (Garfinkel, 1967; Coulon, 1965, Lynch,2001), decision-making (March, 1994), practice theory (Schatzski, 2001) andanthropological theory (Hastrup, 1997). Thus, the philosophy too, utilises “borrowedtheories and concepts”. It stresses “the shift between emic[7] and etic[8] perspectives inIR[9]” insists on researcher and practitioner collaboration and prescribes theoreticalknowledge and practical result (the intervention) outcomes, similarly to Argyris (1970),Thomas and Rothman (1994), Avenier and Nourry (1999), Kasanen et al. (1993) andLabro and Tuomela (2003).

Goldenhar et al. (2001) present an IR framework, sans any AR “branding” consistingof sequentially conducted developmental, implementation and effectiveness researchphases in the context of occupational safety and health interventions (see Figure 4).

The aim of this model is to develop a research agenda designed to carry anintervention through all three phases, thereby establishing a cycle of continuousimprovement. The developmental phase addresses:

. the best way to incorporate theoretical principles/constructs into the interventionitself; and

. the kinds of measurement tools should be developed to assess any change inthose constructs (p. 618).

Figure 4.

Gather backgroundinformation (conduct

needs assessment)

1

2

34

5 Report anddisseminate

Developmentalresearch

Implementationresearch

Effectivenessresearch

Completedevelopment,

implementation,or evaluation

Choosemethods or

designs

Developpartnerships

Source: Goldenhar et al. (2001, p. 617)

QRAM7,1

28

The implementation phase should provide feedback for intervention refinement, tointerpret effectiveness study findings and to replicate an intervention effective in onecontext to another, whilst avoiding the pitfalls. The effectiveness phase attempts todetermine whether an intervention did or did not work in a real life setting. Theauthors contend that research design option such as quasi-experimental and datacollection methods (qualitative case studies) can be used for intervention evaluation(p. 619). The IR tasks appear to be consistent with that of D&D, with specific emphasison collaboration and theory. Finally, the goals of applying research designs shouldinclude that the intervention made a difference (internally validity), addressingresources (time, funding and expertise), feasibility and limitations (p. 620).

Avenier and Nourry (1999) present an IR process (Figure 5), consisting of threephases. This IR process is termed “intervention research” given that the intervention isthe focal point of the research project.

Phase 1, “Negotiation” consisting of two initially different projects, the organisationalproject (rooted in the logic of action) and the research project (rooted in the logicof knowledge). Through “initial negotiations” identification of congruency betweenthese projects, results in the organisation and the researcher engaging in anintervention project, where researchers position themselves as participants in the field,in contrast to positions of neutrality and indifference. Phase 2 therefore represents theintervention project embracing various interactions between researchers andpractitioners; this phase is consistent with the “intervention system” referred to byArgyris (1970). It is here that the authors tell us that the intervention is based onan “ongoing construction” principle, which is consistent with a CRA. In fact whencomparing Figure 3 and Figure 4, the similarities seem to suggest that this framework isa CRA variant. However, discussion concerning the role of the theoretical framework inthe intervention project is muted, which is something that Goldenhar et al. (2001)were reasonably emphatic about. It is unclear whether this theoretical framework:first, is the theory underpinning the research project component that will secure

Figure 5.Schematic representation

of an intervention researchprocess

Culture and values ofthe organization

Project of theorganization

Researchproject

Theoretical framework

Knowledge-projects

Intervention

Context of organization

Practitioners

Researches

Response to theorganization's projects

Meta-knowledge aboutthe organization

“Publishable”knowledge

Knowledgeproduced

Culture and valuesof the research team

Source: Avenier and Nourry (1999, p. 57)


research

29

scientific knowledge production, publishable in an academic journal, or second, pertainsexclusively to intervention design, (McCoy et al. (2002) refers to using traditionallearning and communications theories as theoretical strategies for intervention D&D), orfinally, is derived in part from the research project and in part to the intervention project.Moreover, there is no reference to intervention design or the processes that support it;this is a critical component of IR. Phase 3 relates to three streams of knowledgeproduction which is generated over the course of an IR project.

CRA, also a form of applied research, is an approach for producing models,diagrams and plans that solve managerial problems in running businessorganisations, through explicit linkage between the problem, the solution andtheoretical knowledge (Kasanen et al., 1993), see Figure 6.

Labro and Tuomela (2003) expand on the original model of CRA by Kasanen et al.(1993), see Figure 7, therefore perspectives from both studies will be discussed intandem. These authors describe CRA as a “sequential process” that addressesmethodological aspects (validity and theoretical connections), whilst providing apractical focal point. Kasanen et al., 1993 deliberately emphasises the practicalcontribution of the CRA, to exemplify and identify the scientific merits of researchersproducing practical managerial problem-solving constructions.

Figure 6.

Practicalrelevance

Theoryconnection

Practicalfunctioning

Construction,problem-solving

(Innovate)Theoreticalcontribution

Source: Adapted from Kasanen et al. (1993)

Figure 7.

Preparatory phase

Source: Labro and Tuomela (2003)

Fieldwork phase

Theorizing phase

1. Finding a practically relevant and theoretically interesting problem

2. Examining the potential for long-term co-operation

4. Creating a novel construct

3. Obtaining a profound understanding of the topic

5. Implementing and testingthe construct

6. Examining the scope of applicability of the construct

7. Showing the theoretical contributions

QRAM7,1

30

These authors use arrows to demonstrate typical time lapses in each phase. Thearrows (with dotted lines) used for steps 3 and 4 indicate that a theoretical KD of thetopic starts in the preparatory phase of CRA, with theoretical linkages andcontributions receiving consideration for the duration of the project. Incorporated inthe three phases are steps guide researchers through the CRA approach, illustrated inTable IV; Labro and Tuomela made some adaptations to the original model. The CRArefers to an “intervention” as a “construction” or “innovation”.

This approach is similar to the Thomas and Rothman approach although thetheoretical components are KD and KU, which are then embedded into D&D, whilsttheory is explicit within the Labro and Tuomela (2003) adaptation. The Labro andTuomela framework is superior to that of Avenier and Nourry (1999) because of theexplicit theoretical references. Theory is an important source of legitimisation for IR.

Both Kasanen et al. and Labro and Tuomela provide transparency concerningvalidity issues; discussion concerning validity and reliability occurs later on. However,an interesting point relates to the external validity of the innovation, specifically itstransferability to other contexts. This issue should be considered at the early stages ofthe intervention design phase, however Kasanen et al. (1993) provides some useful“validity tests” or “market based validations” that can be considered in interventionresearch irrespective of the IR model that is selected. A weak market test is passed if theparticipant is agreeable to implementing the intervention. A semi-strong market test ispassed if there is intervention adoption by a wide range of organisations. A strongmarket test is passed if organisations employing the intervention systematicallyproduce better results than those who are not using it. This “market based validation”may align with Jonsson and Lukka’s (2007) reference to modest and strong interventionforms. Finally, there are also two products stemming from CRA, namely a novel“construct” and theory refinement, development, testing and abandonment.

Kasanen et al. Labro and Tuomelo

Find a practically relevant problem which alsohas research potential

Find a practically relevant problem which alsohas research potential. Theoretical significancemust be identifiedTo examine the potential for long-term researchco-operation with the target audience, includingavailable resources

Obtain a general and comprehensiveunderstanding of the topic

Obtain a general and comprehensiveunderstanding of the topic. Knowledge fromprevious literature required

Innovate (construct a solution idea such asmodels, diagrams, plans)

To innovate and construct a theoreticallygrounded solution idea

Demonstrate that the solution works To implement the solution and test whether itworks in practice

Show the theoretical connections and researchcontribution of the solution concept

To examine the scope of the solution’sapplicability (including issues of validity)

Examine the scope of applicability of the solution Show the theoretical connections and the researchcontribution of the solution

Sources: Adapted from Kasanen et al. (1993, p. 246); Labro and Tuomela (2003, p. 415) Table IV.


research

31

Andriessen (2007) developed an organisational development intervention using ARand a DBR methodology, also known as design science, design research, designstudies, see Figure 8. Design science is a form of IR introduced by van Aken (2004)derived from engineering sciences, medical science and psychotherapy; this is similarto the social research and development (Rothman, 1980) paradigm embedded in D&D.The mission of a design science is to develop knowledge for the design and realisationof artefacts, i.e. to solve construction problems, or to be used in the improvement of theperformance of existing entities, i.e. to solve improvement problems (van Aken, 2004,p. 224).This model consists of three phases, namely design, intervention testing anddesign KD, with some underlying tasks. This framework also provides theoretical andpractical contributions, manufactured by the interconnectivity of the knowledge (DBR)and practice (AR) streams.

The conceptual framework embedded theoretical considerations, which was used todefine a research problem that resulted in an initial solution concept design; theoryseems confined to this task only. Seemingly the solution concept was tested in practicewithout researcher and participant organisation collaboration. Moreover, it isanomalous that the solution concept is designed and refined before the practiceproblem is diagnosed. Post practice problem diagnosis is a match between the solutionconcept and practice problem sought. Is this suggesting that “one size fits all”? I wouldconsider these as major pitfalls in this approach in the context of IR. Another point ofdifferentiation is the employment of consultants with respect to design andimplementation tasks, suggestive of “expert” driven approach rather than “organic” asexemplified by Argyris (1970), with negligible researcher visibility in the process.Finally, some of the participant organisations had to pay a fee for the implementationproject, which is also consistent with consultation practice. Kaplan (1998) states that toensure commitment of the case company a charge for the researchers’ services wouldnot be untoward. The greater emphasis on AR and consultant led activity found inAndriessens’ DBR approach, results in greater divergence from the Thomas andRothmans’ framework and the other frameworks examined. It is therefore not an IR

Figure 8.Research methodology ofa design-based researchstudy using actionresearch

Knowledge stream (design-based research using the reflective cycle)

1. Theorizing

Conceptualframework

Case nCase 2

Case 1

Client agenda

Specific solutionPractice problemRecord of the

evolving processConsequences of

actionsFindings

4. Diagnosing

Practice stream (action research using the problem solving cycle)

5. Action planning 6. Action taking 7. Evaluating8. Specifying

learning

Research problemSolution concept

and plausible rivalexplanations

Successes andimprovements Design knowledge

Mat

ch?

2. Agenda setting 3. (RE) Designing 9. Reflecting 10. Developingknowledge

Source: Andriesson (2007)

QRAM7,1

32

“friendly” approach and may prove to be a confusing methodology for the noviceresearcher or the researcher who is seeking a “recipe” for engaging in IR.

Additional key issues for doing IRThis section reviews the utilisation of consulting theory to supplement IR practice,addresses IR validity and reliability issues, followed by a discussion on researchdesign.

IR not consultingThe activity of intervening is usually considered by researchers as consulting(Arygyris, 1970, p. 12). Evidently, academics conducting IR, whatever its form, areaware of the similarities between IR and consulting and have sought to differentiatethe two, in order to preserve the integrity of IR and minimise discouraging academicsfrom IR engagement (Argyris, 1970; Pasmore and Friedlander, 1982; Kasanen et al.,1993, Kurpius, 1993; Eden and Huxhum, 1996; Kaplan, 1998, Labro and Tuomelo,2003). The purpose of this section is not to debate the merits of research versusconsulting, rather to identify some practical considerations, emanating from consultingtheory, for interventionist researchers wanting to conduct IR, including types ofinterventions, levels of interventions and practical guidelines for intervention selection.A researcher may ask: What type of intervention do I want to make? Am I going tofocus this intervention on an organisation or community as whole, perhaps only oncertain groups or individuals within an organisation or community? IR literature doesnot offer much support for these questions. Reddy (1994), see Table V and Beer(1980)[10], see Table VI, identifies different types of interventions.

The “focus” of the intervention may be an entire organisation or community,specific groups, two or more individuals interfacing at an interpersonal level, or aspecific individual. Reddy (1994) also indicates that the intensity of a selected type ofintervention should be considered. As the intensity of a type of intervention increasesso does the risk associated with a higher probability of intervention resistance orrejection. Intensity of an intervention increases as the focus moves from organisation toan individual. Harrison (1970) and Reddy (1994) suggest that interventionists shouldconsider how deep they want to intervene with any given project (see Table VII); this isimportant from an interventionist resource and skill requirement perspective.

Cognitive interventions Interventions that are abstract, intellectual or idea orientedSkill and activityinterventions

Interventions suggesting training or skill learning (e.g. problem-solving) activities that are structured and defined

Behaviour descriptioninterventions

Intervention that describes what behaviour the consultant observes inthe group (group processes)

Emotional or reflectiveinterventions

Intervention is the emotional or feeling component observed in thegroup surrounding an event. These interventions require a great dealof skill, particularly when the focus is interpersonal or personal

Interpretive interventions This intervention is most appropriate following a behaviourdescription of emotional/reflective interventionIntervention is a hypothesis (conjecture or speculation) or insight ofwhat is occurring at a dynamic level

Source: Adapted from Reddy (1994)Table V.

Types of interventions


research

33

The diagnosed problem, the intervention type, the focus of the intervention, the natureof the intervention tasks, the alternative courses of actions and many other facetsidentified in intervention theory and the IR frameworks will influence the depth ofinterventions. Baard (2004) incorporated the type and depth of intervention fromtwo perspectives, namely as an interventionist and the intervention itself. As theinterventionist, a diagnostic intervention (see Table VI, questionnaire, interviews andother informal communication) was employed for information gathering to assist in theanalysis of identified problems within the small business environment, identifyingconcerns within specific small businesses and specific elements required forintervention design (Phases 1, 2 and partly Phase 3 of IR framework). Second, anindividual intervention (see Table VI) was employed since the intervention was using

Diagnosticintervention

Interventions used for learning about individuals, groups or systems; gatheringdata about the total system or its parts and creating a set for feedback anddiagnosis; assessment of organisation’s problems, determination of factors andopportunities for change (e.g. survey feedback)

Individualintervention

Interventions focus on helping humans develop to a higher level of functioning,using techniques for selecting, training and developing individuals to improvematching between people and social systems. Takes the form of workshops,seminars and other educational approaches (e.g. counselling and coaching)

Process intervention Interventions that assist people experiencing the problem that have an intrinsicability to solve the problem but need help to examine and understand theproblem more clearly (e.g. group and intergroup development)

Structuralintervention

Interventions to change human behaviour, beliefs and feelings or to change theorganisational structure and so improve organisational effectiveness (e.g. jobdesign/enrichment, organisational culture building, strategic planningapproaches)

Source: Adapted from Beer (1980)Table VI.Types of interventions

Levels Name Description

1 Content Work, task, project or service accomplishment2 Overt group issues Basic and obvious observable member interactions

Task behaviours, problem-solving, decision-making, conflictresolutionAll behaviours are observable

3 Covert and core group issues Inferred from level 1 and 2Inclusion, belonging, control and power, independence,competence, autonomy

4 Individual values, beliefs &assumptions

Least manageable characteristics of the individual, e.g. defencemechanisms, history and personalityNot targeted in organisational settingsNot recommended for IR studies

5 Unconsciousness Reservoir of individuals’ basic instincts, motivations, impulsesand so forthNot targeted in organisational settingsNot recommended for IR studies

Source: Adapted from Reddy (1994)Table VII.Depth of Interventions

QRAM7,1

34

an educational approach to assist small firm owner-managers in aspects of strategicplanning to promote business sustainability, to appropriately harness informationtechnology and systems as a strategic asset to provide critical information for controland improved performance. Given the intervention type, the organisation andindividual focus of the intervention, the depths of intervention was level 1, 2 and to avery small degree level 3 (competence and autonomy), see Table VII. Finally, theintervention integrated a cognitive and a skills and activity intervention type(Table VII) in its design.

Validity and reliability of IRPositivist and interpretive research differ in many aspects, such as the quantitativephenomena (objective facts) versus structured and unstructured facts, measurement ofvariables, generalisations, intermingling of validity and reliability issues in qualitativeresearch, the subjectivity inherent in interpretative research, and the qualitative insightrequired in societal realms in which these studies occurs (Becker, 1970; Kasanen et al.,1993; Needleman and Needleman, 1996; Avenier and Nourry, 1999; Labro and Tuomela,2003; Ahrens and Chapman, 2006). These differences also apply to IR, wheresubjectivity is a larger point of differentiation given the collaboration betweenresearcher and participant system. Therefore, it seems that a different perspective ofvalidity and reliability within the IR context is warranted. Second, IR is referred to as aform of case study or a kind of “field experimentation” ( Jonsson and Lukka, 2007),where these underlying methods are subject to criticism regarding their ability toachieve scientific research criteria of validity and reliability. There has been significantresponse in the literature to this problem (McKinnon, 1988; Lukka and Kasanen, 1995;Atkinson and Shaffir, 1998; Davila and Oyon, 2008). McKinnon (1988) identifies threatsto validity and reliability in field research and presents strategies to overcome thesethreats. Ahrens and Chapman (2006) argue that notions of validity were developed toevaluate the objective reality of positivistic research which is incongruous with fieldstudies assuming a social reality subjectively created through human interaction. Giventhis, the literature is not forthcoming with a notion of validity from an IR context. Third,apart from the generation of theoretical knowledge, the D&D of an intervention is thefocal point of IR. The intervention has to correspond with the needs of the participantsystem; it has to solve an inherent problem thus the validity and reliability of theintervention also requires consideration. Therefore, I propose that we incorporate caseand field study validity and reliability issues in IR, as previous authors have done, butwe also examine these issues in the IR framework and the intervention context. Our goalis helping society through concepts or frameworks to make action effective in improvingsociety Davila and Oyon (2008, p. 891).

IR conducted in the social realm generally cannot avail itself of the controlleddesigns possible in evaluation settings; thus the challenge of IR is to conduct crediblestudies where truly rigorous experimental research methods are not possible (Schulteet al., 1996, p. 286). A means to achieve this “credibility” or “plausibility” (Covaleskiet al., 1998) is through a scientific view of the IR framework. Niiniluoto (1984 cited inKasanen et al., 1993) suggests criteria under which research is scientific, namely,objectivity, criticalness, autonomy, progressiveness; these criteria were applied in aconstructive research context. Table IV and Figure 7 presented a framework forconducting constructive research, which constitutes a series of phases and steps.


research

35

This facilitates a critical assessment and verification of each phase and steprelating to the construct development and indeed the construct itself. The developmentof the construct is “an independent self-supporting activity” and is objectivebecause the constructions either work or they do not. Furthermore, constructions areprogressive because of their capacity to solve problems and highlight emergingproblems scaffolding new research questions. Therefore, the application of scientificcriteria to an IR framework provides validity and reliability to IR. Labro and Tuomela(2003) indicate that steps 3, 4 and 5 (Table III) are related to ensuring internal validity;during these steps the authors engaged McKinnon’s (1998) assistance in identifying thevalidity threats and implement the combatant strategies. This demonstrates that we donot need to discount the merits of validity strategies, but that the IR framework can usethese strategies and the framework itself to promote validity. Given these examples,the IR framework of Thomas and Rothman (1994), specifically the D&D facet, can alsobecome a tool for achieving validity within an IR project, through the application of thescientific research criteria. Therefore, in the case of D&D, Phases 2, 3 and 4 could beused to address internal validity.

There are also external validity issues in IR. External validity refers to the extent towhich we can generalise the results of a research study to people, settings, times,measures and characteristics other than those used in that study (Gravetter andForzano, 2009, p. 159). Broadly speaking (Needleman and Needleman, 1996, p. 330)argue that IR is not to uncover distributions and typical characteristics for the purposeof generalisations, but rather to gain depth insight into the complexities of humaninteraction and socials meanings evident in the particular case under study. Morespecifically within the IR context, generalisability relates to whether interventionswould also work in settings similar to those which they were originally constructed for(Kasanen et al., 1993; Lindenburg et al., 2001). Replicating interventions under variousfield conditions assists in the assessment of the generality of the effects of theintervention (Fawcett et al., 1994). Davila and Oyon (2008, p. 890) indicate thatinterpretive research rely to a larger extent on case studies where external validityhappens through an audit of the research process and theoretical generalisations(a common approach in case studies also under a “positivist” perspective) rather thanfrom reproducing the study. Therefore, the IR framework can also be used to addressexternal validity issues. Labro and Tuomela (2003) indicate that step 6 addressesexternal validity issues, which relates to the solutions applicability. This is congruentwith the market-based validation of managerial constructs (Kasanen et al., 1993). In theD&D context this would be Phase 5 (see Table IV) examining intervention evaluationand advanced development. However, when using an IR framework for validity issues,common sense should prevail and these issues should be considered in the first phase.Finally, generalisations also include arriving at statistical generalisations, which giventhe subjective data, behavioural factors and the small samples found in IR raisesanother challenge. Argyris (1970) and Thomas (1984) suggest that we should considerstatistical significance and behavioural significance when doing IR. Statistical resultsmay not always be useful or meaningful to the participant system because theyprovide insufficient information concerning the effects for the participants involved;these results may judge an implemented intervention as ineffective when in fact theopposite stands true. Behavioural significance occurs when an event (implementationof an intervention) results in a difference in the behaviour and values of the participant

QRAM7,1

36

system and constitutes a significant departure from the steady state prevalent prior tointervention implementation. Baard (2004) tested this hypothesis, when evaluating theintervention after implementation. The results indicated statistically insignificantresults concerning the effectiveness of the intervention and the changes produced inthe participant system. However, there was strong evidence of behavioural changes,reflected in the documented observations, interviews and other informalparticipant-researcher interaction results, within the small businesses; learning wasoccurring and owner-managers were engaging with managerial activities differentlyas compared to pre-intervention implementation.

A primary condition for intervention validity is that they work or are truly effective,incorporate objective capability, contain ethical suitability, are feasible given thetargeted environment, the degree to which inference is made that the interventionrather than uncontrolled extraneous factors are responsible for the observed effects,they solve the problems in question, they are relevant, simple and easy to use (Ramp,1984; Paine et al., 1984; Thomas, 1984; Kasanen et al., 1993; Lindenburg et al., 2001;Baard, 2004). Therefore, the validity of the intervention is dependent upon its design;research undertaken by Baard (2004) and Baard and van den Berg (2004a) illustratesthis point. The intervention requirement specifications were determined through arigorous literature review and analysis of the results of a semi-structured survey andinformal interviews conducted in small service[11] businesses. Intervention design wasuser-centric and incorporated design features including simplicity, relevance andcompatibility with small firm owner-manager characteristics to support intervention“user-friendliness”. Small firm environmental considerations including the nature ofthe small business environment in general terms, managerial practices, resourceavailability, existing operational routines were also incorporated given thatimplementation would occur in small businesses. Implementation strategiesconsidered the capability (skills, cognitive knowledge, behavioural components) ofthe people conducting the implementation; to influence the success in interventionimplementation and retention. Design occurred within the IR framework (Thomas andRothman, 1994) and procedural elements (including detailed records) were carefullydocumented. Given the breadth of the service industry, the intervention wasstandardised to enable its implementation into other service environments. Manyinterventions have been unsuccessful or discarded because they were overly ambitiousin their development and design and with implementation in mind, the complexity ofthe intervention was carefully monitored throughout design.

Whereas the reliability in field research can be replicated with the same results(Atkinson and Shaffir, 1998, p. 62), the reliability of an intervention is dependent uponits implementation (Cook and Campbell, 1979, p. 43). Implementation serves twopurposes:

(1) it means that the research process has been successful; and

(2) that the construct (intervention) is technically feasible (Labro and Tuomela,2003, p. 428).

There is dearth of intervention implementation literature. Rosen et al. (1999, p. 12)indicates that only three percent of all published articles (social work literature)could inform a practitioner of how to implement reliably the intervention thatwas studied. Baard (2004) also found the literature wanting in this regard and


research

37

therefore project management, which offers a structured approach, was used forintervention implementation. It is important to note that rigorous pilot-testing ofthe intervention and the selected implementation methodology to reduce risk ofintervention rejection during implementation are also factors that should be consideredto preserve validity and reliability of the intervention; see Phase 4 of D&D, Table III.

Research designThe discourse thus far has suggested that the research design used in the various formsof IR is a case or field study. Thomas and Rothman (1994) indicated that there is no oneparticular research technique that is employed in D&D; both quantitative andqualitative research modalities are used in relation to the particular type of interventionthat is being produced. These authors, as with Argyris (1970), do not discourageresearchers from harnessing traditional research methods for information gathering topromote the validity of the information on which intervention design is dependent.Kasanen et al. (1993) typify case study method utilisation in constructive research, butthey also indicated that CRA may either be quantitative or qualitative or both. The keyissues in research design selection are the type of intervention produced, the context forwhich production occurred and the necessary reliance in IR on non-experimentalmethods. Social and health science literature advocates the use of quasi-experimentalresearch within the IR context, specifically in field settings; social program andinnovation evaluation (Fairweather, 1967; Fairweather and Tornatzky, 1977), fallprevention (Robitaille et al., 2005), performance appraisal systems (Mayer and Davis,1999) nursing and health care interventions (Pruitt and Privette, 2001), feedback systems(Smither et al., 1995) and management research (Punnett, 1988). Zwerling et al. (1997) andSchulte et al. (1996) acknowledge that rigorous experimental research methods are notpossible in field research, yet refer extensively to the work of Cook and Campbell (1979)who use quasi-experimental designs in field settings. Quasi-experiments can be used innatural settings (e.g. organisations) when random assignment is not practical, andprovides an opportunity to determine what would happen in a participant systemwithout the intervention. Cook and Campbell also explicate sources of invalidity thatmay occur in field settings and provide strategies for their management, not thatdissimilar from McKinnon (1988). Baard (2004) used a quasi-experiment design in a fieldsetting (small service businesses) with intervention and “control groups” to provide anappropriate test of the effectiveness of the intervention when implemented in areal-world setting. The selection of the research design occurred taking into account theprimary tasks of intervention theory and the D&D intervention framework andaddressed the validity issues based on Cook and Campbell’s work.Quasi-experimentation in field settings may not be appropriate in all IR research dueto the key issues mentioned previously. Jonsson and Lukka (2007) referred to IR as a“kind of field experimentation where the researcher, not having completed control overthe design of the experiment, seeks to determine the experimental situation throughobservation”. Therefore, I propose that a quasi-experimental design in field settings maybe an alternative design for utilisation in IR.

Concluding commentsThis paper, through a systematic and critical review of the literature and IR,provided insights into the conceptualisation and research methodological bases of IR.

QRAM7,1

38

This review defined IR and identified an intervention theory that provided explanatoryconcepts to assist researchers in conducting IR. An intervention research framework,derived from social sciences was explicated, and compared to variant forms of IRnamely, AR, constructive research and DBR. A philosophical approach to IR wasidentified which provides future research opportunities which may include itsdevelopment into a competing intervention theory or framework or may make valuablecontributions to theory and framework refinements. A discussion on some additionalIR practical considerations, emanating from consulting theory ensued. Finally,I examined validity and reliability issues from an IR perspective, ending with aproposed IR design alternative to the case and field study.

Avenier and Nourry (1999, p. 55) state:

[. . .] whether called clinical research, AR methods, intervention research, or engineeringresearch, these “transformative” fieldwork methods all seem to have the same intention whichconsists in succeeding in an intentionally change-inducing project and, in doing so advancefundamental knowledge in human sciences.

However, these alternate forms of IR lead to misconceptions concerning the processesand outcome from this methodology. My argument in this paper is for IR to be a“stand-alone” methodology embraced by its own theory and frameworks. Indeed it isthis lack of cohesion as a methodology that undermines it scientific value and thereforeits perpetuation in future research. Argyris (1970) constructs an intervention theory tounderpin IR, and provide guidance for conducting this form of research. He admits thathis theory may have loopholes, which empirical research may not identify since theresearch is guided by the theory; this danger is inherent in any new theory. It doeshowever provide opportunities for further theory testing and refinement, Van De Vliertoffered us a solitary critical review in almost three decades. Empirical researchutilising intervention theory is rare; this also provides opportunities for future IR toutilise this theory.

Thomas and Rothman (1994) present a comprehensive IR framework thatis grounded in “overlapping research paradigms” providing the interventionistresearcher with a comprehensive guide to conducting this type of research, especiallygiven the literature paucity in this regard. The CRA of Kasanen et al. (1993) andLabro and Tuomela (2003) is a worthy contender. This may therefore provide anopportunity to refine, if required, the Thomas and Rothman approach to provide anauthentic IR framework. Both approaches however emphasise the use of theoryscaffolding IR processes and the intervention itself. Theory is an important source oflegitimacy for IR, “the very essence of an applied science lies in preparing theoreticallygrounded solutions for practical purposes” (Mattessich, 1995). For IR to enjoylegitimacy as research methodology and more respect from the positivists, theory asindicated is important, but so is attending to the validity and reliability aspects of IR.I proposed an alterative view of validity issues in the IR context, with somefortification from the scientific criteria proposed by Niiniluoto and camaraderie of CRAadvocates, that uses the IR framework as tool to attempt validity management; thisnotion too may need some further exploration. Clearly there is much work to be done toexemplify the untapped potential of IR, there are opportunities for doctoral students toengage with IR in their research and there are significant contributions to be made tothe body of academic knowledge and practitioners and organisations.


research

39

Notes

1. This study used an AR and DBR approach.

2. Baard (2004) used an IR framework originally developed for social work research.

3. Argyris (1970) uses the term intervention system.

4. Logical consistency refers to concepts, relations and assertions in theory withoutcontradiction.

5. Empirical consistency refers to concepts and assertion in theory not in conflict withempirical facts.

6. Ideally the entire D&D process if undertaken can range from three to ten years.

7. Emic perspective is the studying of human behaviour within a system.

8. Etic perspective is the studying of human behaviour outside of a system.

9. Atkinson and Shaffir (1998, p. 43) assert that qualitative methodology recognises thathuman behaviour cannot be adequately understood by observing it from the outside, ratherunderstanding a social world from the perspective of the actors, experiencing reality asothers experience it.

10. Beer (1980) in an organisational change and development, rather than consulting contextoffers a complementary approach to Reddy (1994). Scapens (2008, p. 917) also refers tovarious types interventions.

11. There is a complex service typology, see Dicken 1998, only business providing “BusinessServices” were included in this study.

References

Abell, N. and Wolf, D. (2003), “Implementing intervention research in doctoral education:maximising opportunities in training for outcome evaluation”, Journal of Teaching inSocial Work, Vol. 23 Nos 1/2, pp. 3-19.

Aguinis, H. (1993), “Action research and scientific method: presumed discrepancies and actualsimilarities”, Journal of Applied Behavioural Science, Vol. 29, pp. 416-31.

Ahrens, T. and Chapman, C. (2006), “Doing qualitative field research in management accounting:day-to-day contests of accountability in a UK restaurant chain”, Accounting,Organizations and Society, Vol. 31 No. 8, pp. 819-41.

Andriessen, D. (2007), “Designing and testing an OD intervention: reporting intellectual capital todevelop organisations”, Journal of Applied Behavioural Science, Vol. 43 No. 1, pp. 89-107.

Argyris, C. (1970), Intervention Theory and Method: A Behavioural Science View,Addison-Wesley, Reading, MA.

Argyris, C. and Kaplan, R.S. (1994), “Implementing new knowledge: the case of activity-basedcosting”, Accounting Horizons, Vol. 8 No. 3, pp. 83-105.

Argyris, C. and Schon, D. (1974), Theory in Practice, Addison-Wesley, Reading, MA.

Argyris, C. and Schon, D.A. (1989), “Participatory action research and action science compared”,American Behavioural Scientist, Vol. 32 No. 5, pp. 612-23.

Argyris, C., Putnam, R. and McLain Smith, D. (1985),Action Science, Jossey-Bass, San Francisco, CA.

Arnaboldi, M. and Azzone, G. (2004), “Benchmarking university activities: an Italian case study”,Financial Accountability & Management, Vol. 20 No. 2, pp. 205-21.

Atkinson, A.A. and Shaffir, W. (1998), “Standards for field research in management accounting”,Journal of Management Accounting Research, Vol. 10, pp. 41-68.

QRAM7,1

40

Avenier, M.J. and Nourry, L. (1999), “Sciences of the artificial and knowledge production:the crucial role of intervention research in management sciences”, Design Issues, Vol. 20No. 2, pp. 205-21.

Baard, V.C. (2004), “The design and implementation of an interactive IT consulting systems toimprove performance in South African small businesses”, Doctoral thesis, CentralUniversity of Technology, Bloemfontein.

Baard, V.C. and van den Berg, A. (2004a), “Interactive information consulting system forSouth African small businesses, part 1”, South African Journal of InformationManagement, Vol. 6 No. 2.

Baard, V.C. and van den Berg, A. (2004b), “Interactive information consulting system forSouth African small businesses, part 2”, South African Journal of InformationManagement, Vol. 6 No. 4.

Baskerville, R.L. and Wood-Harper, A.T. (1996), “A critical perspective on action research as amethod for information systems research”, Journal of Information Technology, Vol. 11,pp. 235-46.

Becker, H.S. (1970), Sociological Work: Method and Substance, Aldine, Chicago, IL.

Beckhard, R. (1979), “Organization changing through consultation and training”, in Sinha, D.P. (Ed.),Consultants and Consulting Styles, Vision Books, New Delhi, pp. 17-47.

Beer, M. (1980), Organization Change and Development: A System View, Scott Foresman,Glenview, IL.

Blum, F.H. (1955), “Action research: a scientific approach?”, Philosophy of Science, Vol. 22, pp. 1-7.

Bradbury, H. and Reason, P. (2003), “Action research: an opportunity for revitalising researchpurpose and practices”, Qualitative Social Work, Vol. 2 No. 2, pp. 155-75.

Breu, K. and Peppard, J. (2003), “Useful knowledge for information systems practice:the contribution of the participatory paradigm”, Journal of Information Technology,Vol. 18, pp. 177-93.

Burns, D. (2007), Systematic Action Research: A Strategy for Whole System Change, Policy Press,Bristol.

Camp, C.J. (2001), “From efficacy to effectiveness to diffusion: making the transitions in dementiaintervention research”, Neuropsychological Rehabilitation, Vol. 11 Nos 3/4, pp. 495-517.

Carkhuff, R.R. (1983), Sources of Human Productivity, Human Resource Development Press,Amherst, MA.

Chandler, D. and Torbert, B. (2003), “Transforming inquiry and action: interweaving 27 flavoursof action research”, Action Research, Vol. 1 No. 2, pp. 133-52.

Christopher, S., Watts, V., McCormick, A. and Young, S. (2008), “Building and maintaining trustin a community-based participatory research partnership”, American Journal of PublicHealth, Vol. 98 No. 8, pp. 1398-406.

Cook, T.D. and Campbell, D.T. (1979), Quasi-Experimentation: Design and Analysis Issues forField Settings, Houghton Mifflin, Boston, MA.

Cope, M. (2000), Seven C’s of Consulting, Pearson Education, Harlow.

Coulon, A. (1965), Ethnomethodology, Sage, Thousand Oaks, CA.

Covaleski, M.A., Dirsmith, M.W., Heian, J. and Samuel, S. (1998), “The calculated and the avowed:techniques of discipline and struggles over identify in big six public accounting firms”,Administrative Science Quarterly, Vol. 43 No. 2, pp. 293-327.


research

41

Davila, T. and Oyon, D. (2008), “Cross-paradigm collaboration and the advancement ofmanagement accounting knowledge”, Critical Perspectives on Accounting, Vol. 19,pp. 887-93.

Dirks, M.J., Rottinghaus, M.K. and Lansky, L.M. (1978), “Argyris’ intervention theory: a smallgroup application”, Group & Organization Studies, Vol. 3 No. 3, pp. 317-30.

Eden, C. and Huxham, C. (1996), “Action research for management research”, British Journal ofManagement, Vol. 7, pp. 75-86.

Edmondson, A. and Moingeon, B. (1999), “Learning, trust and organisational change”,in Easterby-Smith, M., Burgoyne, J. and Araujo, L. (Eds), Organisational Learning and theLearning Organisation: Developments in Theory and Practice, Sage, London, pp. 157-75.

Edmondson, A.C. (1996), “Three faces of Eden: the persistence of competing theories andmultiple diagnoses in organisational intervention research”, Human Relations, Vol. 49No. 5, pp. 571-95.

Fairweather, G.W. (1967), Methods for Experimental Social Innovation, Wiley, New York, NY.

Fairweather, G.W. and Tornatzky, L.G. (1977), Experimental Methods for Social Policy Research,Pergamon Press, New York, NY.

Fawcett, S.B. (1990), “Some emerging standards for community research and action”, in Tolan, P.,Keys, D., Chertok, F. and Jason, L.E. (Eds), Researching Community Psychology:Integrating Theories and Methodologies, American Psychological Association,Washington, DC, pp. 64-75.

Fawcett, S.B. (1991), “Some values guiding community research and action”, Journal of AppliedBehaviour Analysis, Vol. 24, pp. 621-36.

Fawcett, S.B., Suarez-Balcazar, Y., Balcazar, F.E., White, G.W., Paine, A.L., Blanchard, K.A. andEmbree, M.G. (1994), “Conducting intervention research – the design and developmentprocess”, in Rothman, J. and Thomas, E.J. (Eds), Intervention Research: Design andDevelopment for Human Service, Haworth Press, Binghamton, NY, pp. 3-23.

Fishbein, M., Hennessy, M., Kamb, M., Bolan, G.A., Hoxworth, T., Iatesta, M., Rhodes, F.and Zenilman, J.M. (2001), “Using intervention theory to model factors influencingbehaviour change: project respect”, Evaluation and the Health Professions, Vol. 24 No. 4,pp. 363-85.

Foster, M. (1972), “An introduction to the theory and practice of action research in workorganisations”, Human Relations, Vol. 25 No. 6, pp. 529-56.

Garfinkel, H. (1967), Studies in Ethnomethodology, Prentice-Hall, Englewood Cliffs, NJ.

Goldenhar, L.M., LaMontagne, A.D., Katz, T., Heaney, C. and Landsbergis, P. (2001),“The intervention research process in occupational safety and health: an overview fromthe national occupational research agenda intervention effectiveness research team”,Journal of Occupational Environmental Medicine, Vol. 43 No. 7, pp. 243-64.

Gravetter, F.J. and Forzano, L.B. (2009), Research Methods for the Behavioural Sciences, 3rd ed.,Wadsworth Cengage Learning, Belmont, CA.

Harrison, R. (1970), “Choosing the depth of organisational intervention”, Journal of AppliedBehavioural Science, Vol. 5, pp. 243-64.

Hasenfeld, Y. and Furman, W. (1994), “Intervention research as an interorganizationalexchange”, in Rothman, J. and Thomas, E. (Eds), Intervention research: Design anddevelopment for human service, Haworth Press, Binghamton, NY, pp. 297-311.

Hastrup, K. (1997), “The dynamics of anthropological theory”, Cultural Dynamics, Vol. 9 No. 3,pp. 351-71.

QRAM7,1

42

Jonsson, S. and Lukka, K. (2007), “There and back again: doing interventionist research inmanagement accounting”, in Chapman, C.S., Hopwood, A.G. and Shields, M.S. (Eds),Handbook of Management Accounting Research, Vol. 1, Elsevier, Oxford, pp. 373-97.

Kaplan, R.S. (1998), “Innovation action research: creating new management theory and practice”,Journal of Management Accounting Research, Vol. 10, pp. 89-118.

Kasanen, E., Lukka, K. and Siitonen, A. (1993), “The constructive approach in managementaccounting research”, Journal of Management Accounting Research, Vol. 5, pp. 243-64.

Kurpius, D.J. (1993), “Consultation interventions: successes, failures and proposals”,The Counseling Psychologist, Vol. 13 No. 3, pp. 368-89.

Labro, E. and Tuomela, T. (2003), “On bringing more action into management accountingresearch: process considerations based on two constructive case studies”, EuropeanAccounting Review, Vol. 12 No. 3, pp. 409-22.

Labro, E., Degraeve, Z. and Roodhooft, F. (2005), “Constructing a total cost of ownership supplierselection methodology based on activity based costing and mathematical programming”,Accounting & Business Research, Vol. 35 No. 1, pp. 409-22.

Lewin, K. (1946), “Action research and minority problems”, Journal of Social Issues, Vol. 2,pp. 34-46.

Lewin, K. (1951), Field Theory in Social Science, Harper & Row, New York, NY.

Lindenburg, C.S., Solorzano, R.M., Vilaro, F.M. and Westbrook, L.O. (2001), “Challenges andstrategies for conducting intervention research with culturally diverse populations”,Journal of Transcultural Nursing, Vol. 12 No. 2, pp. 132-9.

Lukka, K. (2006), “Interventionist research”, Journal of Social Issues, November, p. 36.

Lukka, K. (2007), “Management accounting change and stability: loosely coupled rules androutines in action”, Management Accounting Research, Vol. 18, pp. 76-101.

Lukka, K. and Kasanen, E. (1995), “The problem of generalizability: anecdotes and evidence inaccounting research”, Accounting, Auditing & Accountability Journal, Vol. 8 No. 5,pp. 71-90.

Lynch, M. (2001), “Ethnomethodology and the logic of practice”, in Schatzki, K., Knorr Cetina, K.and von Savigny, E. (Eds), The Practice Turn in Contemporary Theory, Routledge,London.

McCarty Kilian, C., Hukai, D. and McCarty, C.E. (2005), “Building diversity in the pipeline tocorporate leadership”, The Journal of Management Development, Vol. 24 No. 2, pp. 155-68.

McCoy, H.V., Messian, S.E. and Zhao, W. (2002), “Improving access to primary health care forchronic drug users: an innovative systemic intervention for providers”, Accounting,Auditing & Accountability Journal, Vol. 1 No. 1, pp. 34-54.

McKinnon, J. (1988), “Reliability and validity in field research: some strategies and tactics”,Accounting, Auditing & Accountability Journal, Vol. 1 No. 1, pp. 34-54.

Malmi, T., Jarvinen, P. and Lillrank, P. (2004), “A collaborative approach for managing projectcost of poor quality”, European Accounting Review, Vol. 13 No. 2, pp. 293-317.

March, J.G. (1994), A Primer on Decision Making, The Free Press, New York, NY.

Mattessich, R. (1995), “Conditional-normative accounting methodology: incorporating valuejudgements and means-ends relations of an applied science”, Accounting, Organizationsand Society, Vol. 20 No. 4, pp. 259-84.

Mayer, R.C. and David, J.H. (1999), “The effect of the performance appraisal system on trust formanagement: a field quasi-experiment”, Journal of Applied Psychology, Vol. 84 No. 1,pp. 123-37.


research

43

Mullen, E.J. (1994), “Design of social intervention”, in Rothman, J. and Thomas, E.J. (Eds),Intervention Research: Design and Development for Human Service, Haworth Press,Binghamton, NY, pp. 3-23.

Needleman, C. and Needleman, M.L. (1996), “Qualitative methods for intervention research”,American Journal of Industrial Medicine, Vol. 29, pp. 329-37.

Paine, S.C., Bellamy, G.T. and Wilcox, B. (1984), Human Services that Work: From Innovation toStandard Practice, Paul H. Brooks, Baltimore, MD.

Paisey, C. and Paisey, N.J. (2003), “Developing research awareness in students: an action researchproject explored”, Accounting Education, Vol. 12 No. 3, pp. 283-302.

Pasmore, W. and Friedlander, F. (1982), “An action-research program for increasing employeeinvolvement in problem solving”, Administrative Science Quarterly, Vol. 27, pp. 343-62.

Pruitt, R.H. and Privette, A.B. (2001), “Planning strategies for the avoidance of pitfalls inintervention research”, Journal of Advanced Nursing, Vol. 35 No. 4, pp. 514-20.

Punnett, B.J. (1988), “Designing field experiments for management research outsideNorth America”, International Studies of Management & Organization, Vol. 18 No. 3,pp. 44-55.

Ramp, K.K. (1984), “Effective quality control for social service programs: one piece of the puzzle”,in Paine, S.C., Bellamy, G.T. and Wilcox, B. (Eds), Human Services that Work: FromInnovation to Standard Practice, Paul H. Brooks, Baltimore, MD, pp. 261-8.

Rapoport, R.N. (1970), “Three dilemmas of action research”, Human Relations, Vol. 23 No. 6,pp. 499-513.

Reason, P. (2006), “Choice and quality in action research practice”, Journal of ManagementInquiry, Vol. 15 No. 2, pp. 187-203.

Reason, P. and Bradbury, H. (2006), Handbook of Action Research, Sage, London.

Reddy, W.B. (1994), Intervention Skills: Process Consultation for Small Groups and Teams,Pfeiffer and Company, San Diego, CA.

Rickards, T. and Bessant, J. (1980), “The creativity audit: introduction of a new research measureduring programmes for facilitating organisational change”, R & D Management, Vol. 10No. 2, pp. 67-75.

Robitaille, Y., Laforest, S., Fourniet, M. and Gauvin, L. (2005), “Moving forward in fall prevention:an intervention to improve balance among older adults in real-world settings”, AmericanJournal of Public Health, Vol. 95 No. 11, pp. 2049-56.

Rosen, A., Proctor, E.K. and Staudt, M.M. (1999), “Social work research and the quest for effectivepractice”, Social Work Research, Vol. 23 No. 1, pp. 4-14.

Rothman, J. (1980), Social R&D: Research and Development in the Human Services, Prentice-Hall,Upper Saddle River, NJ.

Scapens, R.W. (2008), “Seeking the relevance of interpretive research: a contribution to thepolyphonic debate”, Critical Perspectives on Accounting, Vol. 19, pp. 915-9.

Schatzki, T.R. (2001), “Introduction-practice theory”, in Schatzki, K., Knorr Cetina, K. andvon Savigny, E. (Eds), The Practice Turn in Contemporary Theory, Routledge, London.

Schulte, P.A., Goldenhar, L.M. and Connally, L.B. (1996), “Intervention research: science, skillsand strategies”, American Journal of Industrial Medicine, Vol. 29, pp. 285-8.

Seal, W., Cullen, J., Dunlop, A., Berry, T. and Ahmed, M. (1999), “Enacting a European supplychain: case study on the role of management accounting”, Management AccountingResearch, Vol. 10, pp. 303-22.

Searle, J.R. (2001), Rationality in Action, MIT Press, Cambridge, MA.

QRAM7,1

44

Smither, J.W., Wohlers, A.J. and London, M. (1995), “A field study of reactions to normativeversus individualised upward feedback”, Group & Organization Studies, Vol. 20 No. 1,pp. 61-80.

Snyder, M., Tseng, Y., Brandt, C., Croghan, C., Hanson, S., Constantine, R. and Kirby, L. (2001),“Challenges of implementing intervention research in persons with dementia: example of aglider swing intervention”, American Journal of Alzheimer’s Disease and Other Dementias,Vol. 16 No. 1, pp. 51-6.

Stringer, E.T. (2007), Action Research, 3rd ed., Sage, Thousand Oaks, CA.

Stronks, K. and Mackenbach, J.P. (2005), “Evaluating the effect of policies and interventions toaddress inequalities in health: lessons from a Dutch programme”, European Journal ofPublic Health, Vol. 16 No. 4, pp. 346-53.

Susman, G.I. and Evered, R.D. (1978), “An assessment of the scientific merits of action research”,Administrative Science Quarterly, Vol. 23 No. 4, pp. 582-603.

Thomas, E.J. (1984), Designing Interventions for the Helping Professions, Sage, Beverley Hills, CA.

Thomas, E.J. and Rothman, J. (1994), “An integrative perspective on intervention research”,in Rothman, J. and Thomas, E.J. (Eds), Intervention Research: Design and Development forHuman Service, Haworth Press, Binghamton, NY, pp. 3-23.

Vaivio, J. (2008), “Qualitative management accounting research: rationale, pitfalls and potential”,Qualitative Research in Accounting & Management, Vol. 5 No. 1, pp. 64-86.

van Aken, J. (2004), “Management research based on the paradigm of the design sciences:the quest for field-tested and grounded technological rules”, Journal of ManagementStudies, Vol. 41 No. 1, pp. 221-46.

Van De Vliert, E. (1977), “Inconsistencies in the argyris intervention theory”, The Journal ofApplied Behavioural Science, Vol. 13 No. 4, pp. 557-64.

Zwerling, C., Daltroy, L.H., Fine, L.J., Johnston, J.J., Melius, J. and Silverstein, B.A. (1997), “Designand conduct of occupational injury intervention studies: a review of evaluation strategies”,American Journal of Industrial Medicine, Vol. 32, pp. 164-79.

Corresponding authorVicki Baard can be contacted at: [email protected]

To purchase reprints of this article please e-mail: [email protected] visit our web site for further details: www.emeraldinsight.com/reprints


research

45

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.

IR 1

Documents

interventionist research

conducting of ir

ir proposals

dearth of ir

empirical knowledge

withempirical research

management accounting

optimising ir utilisation