Ethnographic and Action

Research Methods:

A Comparison

Joyce Tanjuakio

Introduction

Ethnographic and action research methods are both used in the field of information systems research. Each research method consists of their own goals and target audience. This paper examines the nature of each research method, sample case studies in which each research method is used and a comparison of both research methods.

Ethnographic Research Method

The Nature of the Method

Ethnographic research is a research method that stems from social and cultural anthropology. Ethnographic research involves the ethnographer to be immersed into the lifestyle of the people or organization that they are studying. Ethnographic research aims to place the phenomena or hypothesis into the social and cultural context studied. (Myers, 1999) Because of its immersive nature and its ability to study phenomena in context, ethnography has become an important tool to study social and organizational contexts of information systems. Different perspectives gained from ethnographic research can also be incorporated in subsequent system designs.

There are different schools of ethnography that represent different approaches to ethnographic research. The different schools of ethnography are:

Holistic

Semiotic

Behaviouristic

The holistic approach advocates empathizing with the social group or organization being observed. The holistic approach requires the ethnographer to “go native” or become blank slate or a sponge in order to soak up all the information and fully understand social and cultural practices (Myers, 1999).

The semiotic approach is broken down further into two categories: thick description and ethnoscience. Thick description is the opposite of ethnographic where empathy and identification with the local and social grouping is unnecessary. The goal of thick description is to analyze symbols (words, images, institutions, behaviours etc) in comparison to each other and the organization as a whole. (Myers, 1999) Thick description aim to understand the ‘web of significance’ or group of symbols that make up the cultural context. The data sources are mainly ‘thick descriptions’ of the situation and the context (Myers, 1999).

These schools of thought of ethnography range from two extremes: the positivist researchers and post-modernist ethnographers. The positivist researchers aim to describe the real world and whose approach include preparation for experiments on the field, ethical issues and writing up the results (Myers, 1999). However, post-modern ethnographers’ equate writing up results to writing a novel and approaches ethnography more as a genre rather than only a method. In between these two extremes are other approaches that combine viewpoints of method and genre (Myers, 1999).

Critical ethnography, on the other hand, sees research as a process that’s upcoming. Critical ethnography involves an exchange between the ethnographer and the social grouping (Myers, 1999). Unlike the other aforementioned approaches, critical ethnography has the propensity to scrutinize hidden agendas and assumptions about what were perceived as commonsense (Myers, 1999).

The Rigour of the Method

Ethnographic research is the most rigourous method in information systems research. Ethnographic research demands the ethnographer to stay with the organization that is studied for extended periods of time, for instance, eight months. The amount and nature of the data collected is evidence of the rigour of ethnographic research. Some of the data sources of ethnographic research include: interview, annual reports and meeting minutes. In addition to these sources, the data is supplemented by data collected through participant observation through information social contacts with the organization. (Myers, 1999) The emphasis on detailed and meticulously documented observational evidence shows the depth and rigour of the method. The result of the rigour of ethnographic research yields in depth understanding of the social grouping and the organization within its context. Because the researcher is immersed in the context, the researcher is able to gain familiarity with the environment including dilemmas, relationships, risks, routines and frustrations, making this method the most intensive method.

The Knowledge the Method Provides

Ethnographic research results are used in many ways including: the study of information systems in organizations, development of information systems, the management of information systems and overall impacts of information systems.

In order to evaluate whether the knowledge ethnographic research provides is a valuable contribution, Myers suggests that there are factors that the research must be weight against. Such factors include evaluating the extent to which the ethnographer offers new insights (Myers, 1999). The ethnographer must convince their audience of the value of the research and that it offers something to help the progress of ethnographic research in information systems. Myers also suggests to identify whether the manuscript contradicts any established knowledge and whether the research utilizes quality and significant data (Myers, 1999). There must be evidence that the ethnographer was involved in the research and that the ethnographer addresses hidden agendas, disagreements or personal and organizational issues. Finally, the research must provide significant information on the method the researcher used (Myers, 1999).

Ethnographic research, if done well, can contribute substantial knowledge since it puts the researchers ‘where the action is’ (Myers, 1999). However, the topics that ethnographic research covers are limited to what the organization does and what the information research community demands (Myers, 1999).

Ethnographic Research Paper: Brown

The ethnographic research paper by Barry Brown entitled, “The Artful Use of Groupware: An Ethnographic Study of How Lotus Notes are Used in Practice”, Brown uses results from ethnographic study of groupware system, Lotus Notes, to argue against widely accepted views that are taken for granted with regards to groupware systems (Brown, 2000). This study argues that groupware is useful with supporting daily existing organizational process instead of an agent of radical change (Brown, 2000). Brown discusses how an British oil company, Narajo uses Lotus Notes to support unradical everyday applications such as tracking repairs and encouraging knowledge sharing (Brown, 2000). However, as an agent of radical change, Brown argues that Lotus Notes is unable to change system processes.

In this study, the method that Brown uses in data collection is participant observation (Brown, 2000). Participant observation includes study of social groupings and context in depth by interacting in the activity studied (Brown, 2000). The time span of this research study was for three months, where researchers observed and participated in the information systems studied (Brown, 2000). In order to analyze this data, the researchers used re-reading, re-arranging and recollection of the data (Brown, 2000).

It was found that in Narajo, Lotus Notes supports accountability. Finally, this research study determines that the use of groupware systems is beneficial if it is inventively integrated into the work processes by the organization or social grouping that use them (Brown, 2000). The utility or the success of the groupware system is not found in the system itself, but instead, in the participants’ usage of the system.

ACTION RESEARCH

The Nature of the Method

Action research method stems from use in social and medical sciences. Action research possesses strong positivist assumptions that consist of two factors; ideographic and interpretive research ideas. Action research takes on different forms that respond to different domains of problems and issues, however, the most common is the ‘participatory method’, which is based on a five-step model (Baskerville, 1999).

There are two key assumptions in action research, according to Baskerville:

social settings cannot be reduced for study

action brings understanding

(Baskerville, 1999)

These key assumptions and philosophy behind action research consists of beliefs that complex human organizations can only be studied as a whole entity and cannot be dissected into components or variables for meaningful studies (Baskerville, 1999) . It is deemed that if the complex social system is factored into parts, then this would not yield useful knowledge about the entire organization.

The action research process is initiated by the establishment of a research framework within the organization and subsequently, five phases are iterated:

Diagnosing

Action Planning

Action Taking

Evaluating

Specifying Learning

(Baskerville, 1999)

However, the essence of the action research process can be distilled to the diagnostic stage and therapeutic stage. The diagnostic stage involves collaborative analysis of the social grouping and situation by the researcher as well the practitioner who are the subjects of the research (Baskerville, 1999). Based on the the domain of the research, theories can then be formulated. The therapeutic stage involves collaborative change experiments where changes are introduced into a process or an environment and the effects of these changes are studied.

The IS action research forms and characteristics are listed in the following table.

Forms of Action Research	Characteristics of Action Research
Canonical Information Systems Prototyping Soft Systems Methodology ETHICS Multiview Action Science Participant Observation Action Learning Clinical Field Work Process Consultation	Process Model Iterative , Reflective , Linear Structure Rigorous , Fluid Typical Involvement Facilitative, Expert, Collaborative Primary Goals Organizational Development, Systems Design, Scientific Knowledge, Training

(source: Baskerville, 1999)

Although there are various forms of action research as illustrated by the above table, there are common characteristics across all forms: an action and change orientation, problem focus, organic process involving systematic and iterative stages and collaboration among participants (Baskerville, 1999). Action research aims to understand immediate social situations with a focus on complex social settings as whole. Action research also participates in practical problem solving while contributing to scientific knowledge. Some data sources are produced by highly interpretive assumptions on participant observations while the researcher intervenes through collaborative work with the subjects (Baskerville, 1999).

The Rigour of the Method

The rigour of the action research method is widely questioned by the IS research community because of the nature of the method. Action research is difficult to validate as it takes the form of social enquiry (reductionism, repeatability and refutations) rather than a social science which are not not deemed to be as valid knowledge (Baskerville, 1999). The collaborative nature of action research also poses questions in its rigour as researchers both observe and participate in the phenomena under study (Baskerville, 1999). There is a possibility the researcher will develop bias. In addition, action research studies are commonly funded by the client which can further affect the validity of the study.

The Knowledge the Method Provides

The data sources that action research provides are not only the observations by the researcher but it also consists of inevitable effects of the action research method. These inevitable effects are caused by researcher intervention, ideographic viewpoint of research enquiry and qualitative data and analysis. According to Myers, these unavoidable effects include:

Researcher intervention - the researcher becomes a part of the study and as a result, the interpretive statements incorporate the observer’s values and prior knowledge. On the negative side, the attempts of the researcher to understand the focus, the researchers’ personal understanding may invade and observation and the resulting conclusions.

Ideographic viewpoint - each social setting in action research requires a unique set of subjects and frames of references and the social values underlying of the subjects must be considered. In this effect, the subjects are considered as powerful collaborators and contributors to the study.

Soft data - as a result of the action research method, soft data, which is the ideographic description of the utility of the actions adhering to the mental schema and cognitive structure of the subjects, causes researchers to use the terminology of the social grouping. Soft data can sometimes be analyzed in its original state by using techniques such as hermeneutics, deconstruction and theoretical sampling.

(Baskerville, 1999)

The idea domain that action research addresses influences the knowledge the method provides. According to Baskerville, the ideal setting or domain for action research are:

Active involvement of the researcher with the expected benefit for both the researcher and the organization.

The results and knowledge yielded by the research can be readily applied. This is due to the researcher as active participant willing to use new knowledge based on a clear conceptual framework.

The research is in the form of usually, a cyclical iterative process that links theory and practice.

(Baskerville, 1999)

As a result of conducting research based on these aforementioned domains, the action research method produces highly relevant knowledge. Action research, since it is based on collaboration between researchers and practitioners, is inherently grounded in practical action. However, action research does not only solve immediate problems, it also informs theory (Baskerville, 1999) . These results are readily usable by the practitioners that the researcher is studying. In addition, the results that are yielded are in the form of theoretical framework that is developed through the experimentation of untested theory rather than experienced solutions (Baskerville, 1999).

However, the merit of the knowledge yielded from action research are commonly questioned because of action research’s collaborative nature with practitioners. Because of this, there could be ethical and bias issues on the knowledge generated. In addition, because of its financial support from the subject studied, the knowledge yielded may be akin to consulting (Baskerville, 1999) In action research, the client members are empowered by the collaborative process that may pose problems to the researcher’s ability to control the process and outcomes of the research.

Action Research Paper: Kock & McQueen

In 1998, a study by Ned Kock and Robert J. McQueen titled “Groupware support as a moderator of interdepartmental knowledge communication in process improvement groups: an action research study” reviews email conferencing (EC) support software using action research. This research paper evaluated the effects of process improvement (PI) and PI groups. Kock and McQueen use action research cycle to analyze the impact of EC support on obstacles faced by PI groups such as physical barriers, functional departmentalization and established social norms. The study explores the following questions:

1. Does EC support reduce the obstacles to desirable interdepartmental heterogeneity in PI groups

2. Does EC support increase individual process-related and social learning in PI groups?

(Kock and Mcqueen, 1998)

The nature of Kock and McQueen’s empirical analysis examines asynchronous collaborative software or groupware used by PI groups in the form of either quality circles or business process re-engineering groups. The benefits of knowledge sharing and the learning that results between interdepartmental and geographically dispersed PI groups with the use of asynchronous groupware support (EC) are also examined. Kock and McQueen define EC as “…email tools used to support group communication, beyond one-to-one message exchanges.” (Kock and McQueen, 1998) EC functions like an Internet listserv where messages are sent to all subscribers to that particular emailing list. EC is used in this case as a productivity tool along with a group process methodology called MetaProi. MetaProi “provides a group process, guidelines and graphical tools to be used by PI groups” (Kock and McQueen, 1998). The MetaProi methodology consists of three main processes – process definition, process analysis and process redesign.

The data from this study was collected between a span of eight months through the process of participant observation, unstructured interviews, structured open-ended interviews, questionnaires with open-ended questions and automatic computer generated transcripts of electronic group discussions (Kock and McQueen, 1998). The data collected were then tabulated, plotted and correlations were drawn from these information.

Two previous research support Kock and McQueen’s assumption of PI groups being effective mechanisms for organizational learning and knowledge communication. One study’s conclusion suggested that PI groups foster learning in non-managerial staff. The second study suggests that PI group-initiated process instances included more than double the proportion of knowledge shared found in the core and support businesses. These two previous research shows relevance in the study and that shows the likelihood that the decrease in PI group obstacles will lead to increased communication and organization knowledge dissemination and learning.

The conclusion reached by Kock and McQueen’s study suggested that EC’s effects on PI groups are that (a) there is a reduction in the influence of geographic or other physical obstacles in the participation of PI group members (b) there is a reduction in interdepartmental conflict barriers to the formation of PI groups and (c) an increase or reduction in individual learning in PI groups depending on the complexity and clarity of the issues and messages contributed by members (Kock and McQueen, 1998).

ETHNOGRAPHIC RESEARCH VS. ACTION RESEARCH

There are several key differences that differentiate ethnographic and action research from each other:

Rigour

Ethnographic research requires meticulous, detailed documentation of observations. In ethnographic research, the ethnographer immerses him or herself into the social grouping and documents every observation. However, in action research, there is less emphasis on documentation and quantity of data collected as there is in the quality of interpretation from participant observation. Because of the shorter time period that the researcher spends with the social grouping in action research, there is less depth in the research data obtained.

Method Approach

The approach of ethnographic research mainly focuses on the researcher’s observation of the study. However, in action research, there is an emphasis on collaborative work between the practitioners, who are the subjects, and the researcher. This participation and intervention of the researcher is a key difference between ethnographic research and action research since it also reflects the applicability of the study and the knowledge yielded.

Knowledge Yielded

The knowledge produced in ethnographic research are with respect to the study of information systems in organizations, development of information systems, the management of information systems and overall impacts of information systems. However, in action research, the knowledge obtained are highly relevant to practitioners while also informing theory. These results are readily usable by the practitioners in action research while the knowledge produced in ethnographic research are mainly targeted to the IS research community and is not immediately implemented in the practitioner community.

Time Span

Ethnographic research is a more rigourous and intensive approach than action research. The researcher, in ethnographic research is required to immerse themselves in the local culture or organization for an extended period of time whereas action research does not involve a long time span.

Data Collected

The knowledge yielded from ethnographic research is primarily highly descriptive documented materials such as meeting minutes, annual reports and interviews. On the other hand, in action research, the source of data are interpretations of observations. In action research, soft data is the resulting knowledge and data yielded, which is the ideographic description of the meaningfulness of actions and using the social grouping’s terminology.

Iterative vs. Linear

The nature of ethnographic research is a linear function whereas in action research, the method is cyclical or iterative. In ethnographic research, the method occurs in a single phase. On the other hand, in action research, the research method is executed in phases and these phases are iterated until the research goals are met.

CONCLUSION

Which method is more appropriate for IS research? As evidenced by the two different groupware studies done by Kock and McQueen and Brown, the results of each research method, although addressing similar topics, yields different conclusions. The difference in the conclusion is mainly on the perspective. In the action research study by Kock and McQueen, the conclusion focused on the effects on the participants, showing the interpretive quality of action research knowledge. In the ethnographic research study by Brown, the conclusion is focused on the information systems itself rather than the affects on the participants which shows how the knowledge yielded comprises of observations and little interpretations of these observations. These two different conclusions demonstrate the varying focuses of the ethnographic and action research studies. Each method has its own merits and limitations and both are valid research methods in terms of information systems research. The choice between which research method to use will depend on the topics and questions addressed and what knowledge is wished to be obtained.