Ambiguity in Research: Not Necessarily a Bad Thing
Michael S. Cole
Auburn University
The significant discoveries, the best science, requires us to be more
venturesome and heretic in research design, and to explore fundamental questions without
knowing the answer in advance. The worth of the research outcome is measured by surprise.
The greater the surprise, the more interesting the result, and the greater the new
knowledge about organizations (Daft and Lewin, 1990)
In a recent issue of TIP, Holland, Hogan, and Shelton (1999)
provide a harsh, yet honest outlook as to why psychologists are often negatively perceived
by the public. In their introspective review, the authors suggest the use of psychological
jargon and lack of clearly defined constructs contributes to our unpopular image. In
addition, they proffer the disregard of measurable data as detrimental to our reputation
with the public (i.e., business, government). To overcome these issues, Holland et al.
(1999) suggest spending more time operationally defining the constructs and making
measurement decisions based on data. Although I agree with their suggestions (for the most
part), I am troubled with Holland et al.'s contention suggesting "… without
common definitions, it is impossible to establish convincing generalizations and
meaningful relationships between variables (p. 35)." As a psychologist, I agree with
the methodological desirability of precision when measuring organizational constructs.
However, as an I-O practitioner, I understand that organizations are comprised of robust,
multidimensional interactions occurring at the individual, group, and organizational
levels. In fact, the actual equivocality of meaning experienced by real people is part of
the very dynamic of organizational life. Therefore, the complexity of the organization and
purpose of the research should decide the level of precision and/or ambiguity present in
the exploration.
The purpose of this response is to discuss these issues and consider
the utility and appropriateness of ambiguity in a research design. First, I will discuss
precision and the inherent flaws that accompany a rigidly defined construct. Then I will
introduce ambiguity as a research tool and present an example from the literature when
ambiguity was found to be helpful. Finally, I will conclude with an emphatic assertion
that we should refrain from discounting the use of ambiguity and balance the two designs
within research.
Precision Vision
Psychologists often argue that to further scientific knowledge we must
precisely define the constructs in question. For example, Holland et al. (1999) suggest
that communication will be unsuccessful unless essential concepts are clearly defined. St.
Clair and Quinn (1997) report precision can be useful, but also has the capability to
close down dialogue. When too much attention is directed toward definitional precision,
researchers lose sight of the initial purpose of the investigation. Moreover, researchers'
designed constructs may impose a reality on the Ss population that doesn't exist (which is
part of the argument for more qualitative research approaches). For example, researchers
often spend more effort building an argument for their interpretation of the construct
than time spent designing the study. Weick (1979) notes that researchers often forget
construct measurement is only a means to understanding. Once the organization is forced
into a countable form, it's stripped of what made it worth counting in the first place
(Leach, 1967). Such a hazard is especially dangerous when precision overtakes a nascent
research domain. Kuhn (1962) notes normal science establishes a paradigm for a way of
thinking about a phenomenon. Once institutionalized, the shared paradigm becomes a basis
of scholarly conviction. In the worst case scenario, a research domain may be adversely
affected by positivist techniques that trivialize research with concern for minor
definitional problems unconnected with relevant issues (e.g., Mills, 1959).
Ambiguity as a Research Tool
Before we can agree on a precise, clear operationalization of a
construct in question, we must fully understand what we are researching. Such an
understanding commonly requires concerted attention to the ambiguity inherent in the
exploration (St. Clair & Quinn, 1997). Assuming the question is even worth asking,
Daft and Lewin (1990) suggest why ask the question if the researcher understands the
phenomenon well enough to operationally define the constructs, state hypotheses, and
control confounds. Similarly, Kirk and Miller (1986, p.17) note "… in order to
test a hypothesis, the investigator must already know what he or she is going to
discover." Thus, constructs that allow some degree of interpretation may not only
benefit from researchers' willingness to scientifically test the phenomenon, but also
extend our knowledge and understanding in new and creative ways. It is when the research
"focus becomes so tightly constrained that there is no disagreement about our
constructs that we run the risk of failing to obtain creative insights into
phenomena" (St. Clair & Quinn, 1997, p. 102).
Consider for example implementing an organizational change process
without some degree of ambiguity. Quinn and colleagues (St. Clair & Quinn, 1997;
Spreitzer & Quinn, 1996) discuss the Ford-University of Michigan LEAD Program as an
example when ambiguity was appropriate. The goal of the program was to empower middle
management to make important organizational changes and become better leaders. The
managers received a week of leadership and organizational change education. At no time
were they provided a definition of empowerment. Over a 6-month period, managers were urged
to empower themselves and make organizational changes. Of 191 LEAD managers involved in
the study, 17% made changes in their interpersonal skills, 21% made changes that improved
their business unit, 16% made changes at the organizational level while another 36%
performed more "global" organizational changes. The authors concluded if they
would have provided an empowerment definition, managers might have added an evaluative
component that would have restricted their range of thought. Rather, leaving empowerment
equivocally defined, managers created their own understanding of empowerment by
re-evaluating themselves, managerial roles, and the organization.
The above discussion is not intended to suggest that precision is the
death knell of I-O psychology; however, when we continually extol and reward (i.e.,
publish) scholars for contributions that are irrelevant to real world problems, we are in
fact contributing to our positivist image. In the eyes of the practitioner, this
"focused" scholarly research is of little use in a manager's attempt to empower
or motivate the workforce. We have to do a better job of relating our work to the
practicing manager.
A Question of Balance
It is my argument that both precision and ambiguity can be utilized
beneficially in organizational research. Without a doubt, precision has allowed the field
to empirically examine otherwise amorphous psychological constructs. In addition, without
some degree of precision scholars would be divided into camps unable to communicate with
one another. Finally, precision allows multi-faceted constructs (e.g., conscientiousness)
to be segmented into empirically discrete phenomenon. On the other hand, ambiguity is
essential when exploring underdeveloped domains to gain a broad understanding of the
phenomenon in question. In addition, ambiguity is a useful tool when implementing
organizational change processes, and is a pathway to innovation. As professionals in a
field whose reputation and image depends on our furthering scientific knowledge while also
increasing managers' "bottomline," we must be aware of the impact of our
explorations. Precision is helpful in certain situations; however, ambiguity is not
necessarily a bad thing!
References
Daft, R. L. & Lewin, A. Y. (1990). Can organization studies begin to break out of
the normal science straitjacket? An editorial essay. Organization Science, 1(1),
1_9.
Holland, B., Hogan, R. T., and Shelton, D. (1999). From phrenology to fraud: The
breakdown of science in the practice of I-O psychology, The I-O Psychologist, 36(3),
35_36.
Kirk, J. & Miller, M. L. (1986). Reliability and validity in qualitative
research. Beverly Hills, CA: Sage Publications.
Kuhn, T. S. (1962). The principles of scientific management. New York: Harper
& Row.
Leach, E. R. (1967). An anthropologist's reflections on a social survey. In D. G.
Jongmans and P. C. Gutkins (Eds.), Anthropologists in the field, 75-88. Atlantic
Highlands, NJ: Humanities Press.
Mills, C. W. (1959). The sociological imagination. New York: Oxford University
Press.
Spreitzer, G. M. & Quinn, R. E. (1996). Empowering middle managers to be
transformational leaders. Journal of Applied Behavioral Science, 32(3),
237_261.
St. Clair, L. & Quinn, R. E. (1997) Progress without precision: The value of
ambiguity as a tool for learning about organizational phenomena. In R. W. Woodman and W.
A. Pasmore (Eds.), Research in organizational change and development: 105_129.
Greenwhich, CT: JAI Press.
Weick, K. E. (1979). The social psychology of organizing. Reading, MA:
Addison-Wesley Publishing.