Int. J. Entrepreneurship and Small Business, Vol. 2, No. 1, 2005 7
9
Copyright © 2005 Inderscience Enterprises Ltd.
Expanding the scope of methodologies used in
entrepreneurship research
Leo Paul Dana
University of Canterbury, Commerce Building,
Private Bag 4800, Christchurch, New Zealand
E-mail: leo.dana@canterbury.ac.nz
Teresa E. Dana*
Christchurch College of Education, School of Business,
Dovedale Avenue, P.O. Box 31 065, Christchurch, 8030, New Zealand
E-mail: teresa.dana@cce.ac.nz
*Corresponding author
Abstract: If we accept that entrepreneurs are influenced by culture, then the
simplest possible unit of research for understanding entrepreneurship is not the
entrepreneur. Nor is it a laboratory study of individuals nor even a field study
comparing entrepreneurs, but rather it is useful to have a case study in which
the important aspects of environment are analysed and understood. It is
doubtful, then, that a mail survey or even interviews could be sufficient for a
researcher to truly gain a holistic understanding of the entrepreneurial process.
Keywords: research methodology; qualitative methods; entrepreneurship.
Reference to this paper should be made as follows: Dana, L.P. and Dana, T.E.
(2005) ‘Expanding the scope of methodologies used in entrepreneurship
research’, Int. J. Entrepreneurship and Small Business, Vol. 2, No. 1,
pp.79–88.
Biographical notes: Leo Paul Dana is senior advisor to the World Association
for Small and Medium Enterprises. He is currently based at the University of
Canterbury, in New Zealand. He formerly served as Deputy Director of the
Nanyang Business School International Business MBA Program, and Visiting
Professor of Entrepreneurship, at INSEAD. He is founder of the Journal of
International Entrepreneurship and author of over 100 papers. His reference
volume, Entrepreneurship in Pacific Asia: Past, Present and Future, has been
on the best-seller list for over two years. His biography appears annually in the
Canadian Who’s Who as well as in Who’s Who in the World.
Teresa Dana is senior lecturer at the Christchurch College of Education in
New Zealand. She received her MBA from McGill University in Montreal.
Her research interests are in the areas of marketing and self-employment.
She has published papers in the Journal of Small Business Management and
elsewhere and has also written several case studies.
80 L.P. Dana and T.E. Dana
This paper discusses an alternative to the traditional hypothetico-deductive approach.
It recommends the application of a more inductive approach, making use of more
non-quantitative data. Having traced the successful development and implementation of
qualitative methods in other social sciences, including anthropology and sociology,
we demonstrate that the use of inductive and non-quantitative research is a useful
strategy, applicable to research in small business and entrepreneurship as well.
Advantages of qualitative (as opposed to quantitative) research, include the ability to
learn directly from the research subject, thereby reducing measurement errors common in
survey studies which often need to make assumptions. The result is a deeper holistic
understanding. Such an approach, because of the interaction between researcher and
subject, also reduces Type III error (asking the wrong question) and Type IV error
(solving the wrong problem). Effective in yielding internal validity, a qualitative
approach can also be used to verify validity of quantitative research.
1 The entrepreneur and the environment
According to Cantillon (1756), the entrepreneur is the one who takes the risk of being
self-employed. Knight (1921) described the entrepreneur as a taker of non-quantifiable
risks, and profits as a reward that owner-managers receive for bearing risk. Weber (1930)
described entrepreneurship as the expression of cultural values. Schumpeter (1931)
emphasised the innovative nature of the entrepreneur, an individual experimenting with
new combinations. McClelland (1961) interpreted entrepreneurship as the expression of
high need for achievement. Hagen (1962) saw entrepreneurship not as a goal in itself, but
as a means to obtain recognition in compensation for social marginality. Barth (1967)
described the entrepreneur as a social agent for change.
Considerable research has been done on the personality of the entrepreneur.
Examples include Begley and Boyd (1987), Brockhaus (1982), Brockhaus and
Horwitz (1986), Gasse (1977), Kets de Vries (1977), Sexton and Bowman (1985) and
Sexton and Upton (1990). Whereas much of this behavioural literature suggests the
entrepreneur usually has certain characteristics which are innate, to others including
Gibb (1986) and Knight (1987), entrepreneurial behaviour can be learned.
Summarising the resolutions of an international summit of leading researchers on
entrepreneurship, Leighton (1988) concluded that “studying entrepreneurs as individuals
is a dead end” and “environment, culture, etc., the context of entrepreneurial behaviour is
important” (p.74). Shapero and Sokol (1982), Shapero (1984) and Peterson (1988) had
given credibility to the concept that the environment may encourage or discourage
entrepreneurial activity. Dana (1990), Gasse and Neff (1990), Soon and Huat (1990),
Giamartino (1991), Dana (2002), and Bjørklund (2004) are examples of entrepreneurship
research conducted in contexts where sensitivity to the cultural environment is important.
A methodological difficulty in conducting such research is that North American-style
questionnaires may be ambiguous in foreign environments, and another problem with
questionnaires is that their low response may yield data which is non-representative of
some environments. Telephone surveys also have limitations as in some areas of the
world very few people have phones. Furthermore, the conceptual tools with which an
investigator approaches research, influences what is observed. To help avoid selective
observation, a non-reactive mode of inquiry is necessary. The objective of this paper is to
promote naturalistic inquiry; we believe that qualitative/non-quantitative empirical data
Expanding the scope of methodologies used in entrepreneurship research 81
resulting from naturalistic inquiry may help researchers and policy makers better
understand entrepreneurship in the context of its environment.
2 Research and its purpose
Considerable research is taking place to learn about entrepreneurs and their activities.
Having recognised the social and economic value of entrepreneurial activity, various
governments around the world have been trying to foster entrepreneurship. Australia
(Meredith, 1984), Canada (Dana, 1988) and the UK (Gibb, 1986–1987) are among the
classic examples.
Loucks (1988) pointed out, however, that entrepreneurship is culture-bound and that
policy on entrepreneurship is therefore culture-specific. Replicating a US-developed
conceptual framework is ethnocentric, as this will not function effectively in all
environments. Applying a universal framework across varied cultures would be naive.
If policy-makers are to formulate policies which will actively create entrepreneurs and
increase the wealth of nations, then research is necessary to understand the values and
aspirations of cultures and their people, before imposing a policy on them.
The research which needs to be done cannot be conducted using mail questionnaires,
surveys or brief interviews. Whereas traditional research in small business tried to
discover why entrepreneurs do what they do, future research should focus on how they do
it, and how others can be encouraged to succeed as well. How is business conducted in
various environments? How do individuals from different cultures perceive opportunity?
How can entrepreneurship be fostered in different environments? Wortman and
Roberts (1982) suggest that whereas quantitative research designs typically answer why
questions, “in employing the qualitative approach, the focus is on how (…) rather than
why” (pp.2, 3).
3 Ontology and epistemology
According to the positivists, reality can be observed objectively. Yet, we can argue that
the researcher influences the research by virtue of selecting hypotheses, thereby
restricting the outcome of the research to either supporting or disproving each hypothesis.
Findings which are not directly linked to the predetermined hypotheses are usually
ignored. Although this approach is ideal for some research in some fields
(e.g., engineering), in others (e.g., social science research of cross-cultural nature) a
positivist approach may lead to applying incorrect theories and methods and
consequently a Type III error. The alternative perspective, that of the idealists, is that
reality is subjective, and relative; reality is a social interpretation, a function of
environment and other factors.
The structuralists see social systems as stable, and change as a crisis which threatens
status quo. In their view, structure controls the behaviour of individuals. The
entrepreneur, however, is described by Schumpeter (1931), Barth (1967) and others as an
agent of change. To the functionalists, each individual in society adapts and functions
according to the needs of the social system.
82 L.P. Dana and T.E. Dana
A dominant methodological approach in sociology is termed symbolic interactionism.
It considers the individual as conditioned by the environment as well as causing change
within it. The focus is on interaction of elements.
4 Non-quantitative research designs
Pasquero suggested that quantitativists often let themselves be carried away by strict but
limited methods and by unrealistic assumptions, so they “miss a true understanding of
real-world behaviours in alien cultures” (1988, p.184). Indeed, the dominant paradigm in
the administrative sciences is positivist, with hypothetico-deductive methodology,
imitating that of the pure sciences. It involves pre-selected constructs in an attempt to
obtain meaningful quantitative data which is easily analysed by means of sophisticated
statistical software. Survey results are analysed relative to an average. However,
Kuhn (1962), Von Bertalanffy (1968), Tinker et al. (1982), Morgan (1983) and
Berry (1986), among others, emphasise that hypotheses are value-laden, and this limits
research findings. Furthermore, surveys and short interviews run the risk of obtaining
socially desirable responding, with respondents presenting themselves favourably
regarding current social norms (Adair, 1984; Arnold et al., 1985; Crowne and
Marlowe, 1960; Golembiewski and Munzenrider, 1975; Lopez, 1982; Rahim, 1983;
Zerbe and Paulhus, 1987). Numerous studies have raised concerns about the
contamination of research findings, by such socially desirable responding (Golembiewski
and Munzenrider, 1975; Thomas and Kilmann, 1975; Stone et al., 1979; Arnold and
Feldman, 1981; Rosenkrantz et al., 1983).
An alternative to traditional, positivist, hypothetico-deductive quantitative research, is
the holistic-inductive qualitative option which is already the dominant paradigm in
anthropological research. It relies on naturalistic inquiry, i.e., the research setting is
naturally occurring with no manipulation by the researcher (Willens and Rausch, 1969).
Qualitative methodology is based on personal observation of situations, events,
individuals, interactions and transactions, as well as document analysis (including
quantitative records) and open-ended interviews yielding in-depth and oral testimonies.
Qualitative data thus includes thick description (Geertz, 1973), and direct quotations from
people about their attitudes, beliefs, thoughts, intentions, actions and experiences.
While a hypothetico-deductive methodology imposes the researcher’s values
(e.g., predetermining which variables are worth measuring, i.e., what to ask on a survey),
in contrast, a holistic-inductive design allows the researcher to be open to whatever
emerges from the data (Patton, 1982). Whereas qualitative data may be reduced to
quantitative codes for statistical analysis, the conversion of quantitative findings into
detailed qualitative descriptions would be more difficult.
4.1 Type of design
Hypothetico-deductive methodology involves a structured and predetermined formal
design. The survey questionnaire, for example, has pre-set questions. In contrast, the
holistic-inductive approach requires a flexible design which is in constant evolution.
Since the researcher does not impose a priori categories or hypotheses, but rather
attempts to understand phenomena based on field research, new questions must
constantly be formulated. The researcher is inspired by observations, to seek more
Expanding the scope of methodologies used in entrepreneurship research 83
answers, which in turn inspire new questions. This allows the researcher to acquire an
understanding of the environment for entrepreneurship, as well as the entrepreneur as an
individual.
Hypothetico-deductive research is designed to test pre-formulated hypotheses.
Variables have been defined. Such a design is a cost-effective efficient means of
obtaining quantitative data from a large sample size for the purpose of statistical analysis.
The alternative flexible design of qualitative research is more appropriate for yielding
new theories. While a survey may be designed to test hypotheses about an entrepreneur,
qualitative research may be better suited to understanding the entrepreneur’s interaction
with the environment.
4.2 Research strategy
The strategy of using case studies in research involves the thorough study, in depth and
detail, of a limited number of objects, individuals or environments. Ideally, data
collection in such research should include observation and interviews, as is done in
ethnography. In contrast to ethnography, however, most entrepreneurship case studies
have not researched intensively environmental explanatory variables such as culture.
Small business and entrepreneurship research often focuses on the firm or the
entrepreneur. A quantitative strategy often limits the researcher’s ability to study context
and environment. Adopting an ethnographic approach in non-quantitative research would
enhance our knowledge and understanding of such pertinent and critical factors. As noted
by Bherer et al. (1989) among others, exploratory research in developing fields requires
more specialised instrumentation and different strategy than classical research of
phenomena in familiar domains. Crozier and Friedberg (1977), for example, suggest a
strategy involving an inductive approach with qualitative interpretation which leads the
way to an understanding of culture and society.
A fundamental advantage of a non-quantitative strategy in research is its flexibility.
A research plan serves as a basis for the researcher, but it is constantly modified, adapting
to the constraints and opportunities which the researchers encounter in their operating
environment. In quantitative research, a survey must be finalised and printed before
researching the respondents. With a non-quantitative ethnographic strategy, the
researcher may formulate each question based on each individual respondent’s
personal experiences and willingness to answer. Non-quantitative research strategy is
interactive, as is the relationship between an entrepreneur and the environment of the
firm. The non-quantitative researcher may consider the following strategy while keeping
in mind the possibility at any time, to return to previous steps and modify them:
• Development of a ‘loose evolving framework’ (Patton, 1987), according to research
objectives (this might involve formulating some possible research questions).
• Pre-testing on the field to test viability of project.
• Data collection including observation of interaction and in-depth interviewing.
• Preparation of ‘site summaries’, resumes of preliminary analysis, as described in
detail by Miles (1979).