Background for research approaches

The research philosophy forms a basis for the research strategy, data collection methods and data analysis of the research. Philosophical aspects and questions exist behind every research method and methodological approach and thus define the ways it is possible to provide new knowledge through research (Eriksson and Kovalainen, 2008; Hirsjärvi et al., 2009). The research tradition in the field of industrial management is partly contradictory and is seen to locate somewhere in the middle ground of two opposite research philosophies: positivism and hermeneutic science (Olkkonen, 1993).

Positivism, inherited from the theories of Auguste Comte (1798-1857) and originated from realism, refers to the assumption that the only legitimate knowledge can be found from experience. Positivists believe in empiricism – the knowledge is seen to be restricted in the observation and measurement of existing facts. The aim of research should be in finding causal explanations and regularities (Eriksson and Kovalainen, 2008; Olkkonen, 1993).

At the opposite end is hermeneutics, also called ‘interpretivism’, a philosophy which originates from idealism and was launched by Friedrich Schleiermacher (1768-1834). Hermeneutics refers to interpretation and understanding as part of the research process. It is seen that human intentions mould and change the reality crucially, and that is why understanding human intentions is needed (Eriksson and Kovalainen, 2008; Olkkonen, 1993).

Based on the basic philosophical traditions, certain prevailing ways of thinking on the used data, methodologies and the gathered results exist in every field of science. This has given birth to certain bunches of value choices and working methods which are called paradigms or research approaches (e.g. Modell, 2009; Olkkonen, 1993). I use the term “research approach” which I consider to include decisions e.g. about the following aspects:

Research logic: deduction, abduction or induction Research data: large vs. concise

Research design: quantitative, mixed or qualitative

Research results: descriptive, normative or explorative

(compiled from Creswell, 2003; Eriksson and Kovalainen, 2008: Kasanen et al., 1993:

Olkkonen, 1993).

Deduction rests on the idea that theory is the first source of knowledge, and on the basis of what is known about a phenomenon theoretically, a researcher can deduce one or more hypotheses.

Induction, on the other hand, follows the logic of proceeding from empirical research to theoretical results. These two research logics or traditions are not, however, clear-cut alternatives, but many researchers use both induction and deduction in different phases of their study. This is often called abduction (Eriksson and Kovalainen, 2008, Modell, 2009). In practice, it may be difficult to dissect abduction from the iterative work taking place in all empirical research. That is why some researchers talk about a hermeneutic circle in much the same meaning (Eriksson and Kovalainen, 2008). However, in this thesis abduction is seen as a research logic between the two opposites, which offers an opportunity to modify the original framework successively, partly as a result of unanticipated empirical findings, but also of theoretical insights gained during the process (Dubois and Gadde, 2002).

Most research methods are based on either qualitative or quantitative methodologies. Qualitative methods involve collecting data that is mainly in the form of words, and quantitative methods involve data which is either in the form, or can be expressed as, numbers. However, it is sometimes difficult to distinguish quantitative and qualitative research from each other accurately, and they should thus be seen as complementary approaches (Hirsjärvi et al., 2009).

This has raised interest in the mixed methods approach, which involves both quantitative and qualitative information (Creswell, 2003; Silverman, 2005).

There are many ways to categorize the research results. In this thesis, I use the categorization of e.g. Olkkonen (1993) into descriptive, normative and explorative research results. Descriptive research aims at describing the characteristics of the studied phenomenon and quantitative dependencies. Explorative research aims at increasing the understanding on the phenomenon, and normative research at recommending a way of acting.

Positivism is typically related to quantitative research, facts and large research data. In positivist research deductions from large congruent data sets are made, and descriptive results are often produced. Hermeneutics is involved in research on new fields without a vast material or opportunities for statistical analyses, in which it is difficult to form a single structured research problem. Its interpretive nature accepts certain subjectivity, and the results gathered are often explorative or qualitatively descriptive or normative (Hirsjärvi et al., 2008; Kasanen et al., 1991;

Olkkonen 1993).

For example, natural sciences lean strongly on positivism, whereas humanities use a lot of hermeneutics-based research methodologies. The reality is not, however, so black and white, as positivism and hermeneutics rarely totally rule out each other in research. For example in business economics, these approaches alternate even within a same study (Olkkonen, 1993). The abductive reasoning and mixed method approaches in management research have gained attention among management scholars recently because of their potential to provide more credible results through interpretive logic, and at the same time an opportunity for more multifaceted theoretical reasoning leaning on the positivist logic (e.g. Creswell, 2003; Modell, 2010). A suggestive representation on the research philosophies and the research approaches related to them is presented in Figure 5.

Figure 5: Research approaches in the continuum of research traditions (adapted from Hirsjärvi et al., 2008; Kasanen et al., 1991; Olkkonen, 1993; Modell, 2010)

Business and management research is applied research, which aims at both a theoretical and a managerial contribution (Vafidis, 2002). There are some research approaches which have been typically used in business and management research. Neilimo and Näsi (1980) have developed a framework for business research categorization with two dimensions: theoretical-empirical and descriptive-normative. The original framework included four categories of research: conceptual, nomothetical, decision-orientated and action-orientated research. Kasanen et al. (1993) have added a fifth category to the framework: constructive research. The extended framework is presented in Figure 6.

Figure 6: Framework for business research categorization (Neilimo and Näsi, 1980; Kasanen et al., 1993)

In the framework, theoretical research means reasoning. Theoretical knowledge is prior knowledge that is observable without experiment. Empirical research means that the data is collected in the field or in a laboratory. As already discussed above, in descriptive research the emphasis is on describing, explaining and forecasting “what is” and “how is”. Normative research, on the other hand, is target-oriented, aiming at recommending a way of acting in practical situations (Lukka, 1991; Olkkonen, 1993).

1) The conceptual approach is mainly based on existing concepts and their analysis, and analysis and synthesis are used to create new concepts and frameworks. The research findings may be mere statements or recommendations by nature. There are no clear rules on how the research must be performed (Neilimo and Näsi, 1980; Vafidis, 2002).

2) The nomothetical approach can be characterized as a causal and analytical approach in generating ideas and assuming relationships. Usually, a hypothesis is put forward and tested with empirical evidence, using e.g. different scales and methods for measuring.

The explanation of the findings is central but understanding is not a major concern. The approach highlights the objectivity of the researcher (Neilimo and Näsi, 1980; Vafidis, 2002).

3) In decision-oriented research, the emphasis is on building a model that solves a certain type of problem under certain conditions. Mathematics and logic are used for making

selections in the model. Observations from the environment are ignored, and empirical evidence is not important (Neilimo and Näsi, 1980; Vafidis, 2002).

4) The action-oriented approach leans strongly on hermeneutics. The role of empirical data is central, and empirical evidence is examined by informal but versatile methods. The approach assumes the human to be intentional, and thus accepts subjectivity. The results are often multi-level conceptual frameworks that are used to analyze or plan the surrounding world (Neilimo and Näsi, 1980; Vafidis, 2002).

5) The fifth approach, constructive research aims at building a solution construct, based on theoretical knowledge and the research process itself. The researcher and his research project affect the setting and the development of the research material – thus, results from the previous stages as well as the researcher himself become part of the data of the research. The role of the researcher is thus highly subjective (Kasanen et al., 1993;

Vafidis, 2002). To be accurate, the solution to the problem should be linked with earlier knowledge, its novelty and functionality should be demonstrated, and moreover, the construction should be validated in a real-life context, i.e. the marketplace, by using weak and strong market tests (Kasanen et al., 1991).

As stated above, combining quantitative and qualitative methods is typical in the business and management research field. Both quantitative and qualitative research have some typically used research methods, which are listed in Table 1.

Table 1: Examples of quantitative and qualitative research (adapted from Eriksson and Kovalainen, 2008; Myers, 2013)

Quantitative research: focus on

numbers Qualitative research: focus on text

Surveys Action research

Laboratory experiments Case study research

Simulation Ethnography

Mathematical modelling Grounded theory Statistical analysis Discourse analysis

Econometrics Narrative

In addition to combining quantitative and qualitative methods, a researcher can mix several other elements in a study, also within quantitative and qualitative researches. This basic idea that more than one thing can be done in a single study, is called triangulation (e.g. Denzin, 1978; Myers,

2013). It may feature throughout the research or only at the analysis phase (e.g Jonsen and Jehn, 2009).

The initial objective of triangulation was the confirmation of findings in order to increase validity and objectivity. More recently, triangulation has been used to increase the completeness of the data and comprehensiveness of the study (Jonsen and Jehn, 2009). According to Myers (2013), it offers a way to look at the same topic from different angles and provides a fuller understanding of the phenomenon.

Denzin (1978) has categorized triangulation into four types:

Data triangulation, in which the same problem is tried to be solved by analyzing various data sources

Theoretical triangulation, which involves the use of more than one theoretical perspective in interpreting the data

Investigator triangulation, which means that the data is collected and analyzed by several researchers

Method triangulation, which involves the use of several different research methods.

Method triangulation and mixed method research are sometimes considered as synonyms in the literature. However, this thesis follows the definitions of e.g. Creswell (2003) and Creswell and Plano Clark (2007) for mixed methods, and considers it more widely as a research approach which includes different strategies for collecting and analyzing quantitative and qualitative data, interpreting the combined data and reporting the research results as a single unity or as a multi-method research.