There has been heated discussion over the use of positivism and social constructivism in management research over the past few decades. And one can find that most of the theorists take a neutral view with regard to this situation. After the 1980s there has been a trend towards social constructivism but even then no one is sticking completely to the principles of one single philosophy.
The main idea behind positivism is that society is an external agent and hence its properties or characteristics must be measured objectively rather than by any subjective means like intuition, sensation etc. This view was first proposed by French theorist Auguste Comte (1853) according to whom “All good intellects have repeated, since Bacon’s time, that there can be no real knowledge but that which is based on observable facts”. Hence his view has two main assumptions i.e. reality is external and objective and knowledge is based upon observations of this reality. Thus there are certain postulates of this philosophy:
Independence: The researcher must be independent in his thoughts and is not influenced by any factors while observing the reality.
Value-freedom: This involves deciding upon the area to be studied and how to study it. Positivism suggests doing this on the basis of objective research and not on human beliefs.
Causality: The main objective of social and behavioral sciences is to evaluate the causes and fundamental laws behind the fluctuations in the behavioral pattern of people.
Hypothesis and deduction: every research will be based on certain hypothesis and the observations will be made to either prove or disprove this hypothesis statement.
Operationalization: The research concepts must be formulated in such a way that the facts obtained from observations can be measured empirically.
Reductionism: This concept believes that a bigger problem will be easy to solve if it is divided into smaller elements which can be separately solved.
Generalization: In order to make general conclusions on fluctuations in human behavior, it is essential that the research must be conducted among samples of appropriate size. On the basis of results obtained from the sample size, the researcher can make wider judgments on the whole population.
Cross-sectional analysis: This kind of regularities can be easily identified by comparing the differences in the samples.
All the above given implications are the collective view points of different theorists of positivism. But there were many theorists who were having contrasting opinions on these postulates too. Even Comte was not supporting the reductionism concept. Wittgenstein proposed in his past studies that all factual intentions can be broken down into smaller elementary propositions that were entirely independent from each other. He himself later challenged this theory arguing that elementary propositions can be logically related to each other (Pears, 1971). Thus this philosophy is under an unending debate among the theorists. Aiken (1956) came up with the view that positivism is the best way to find out the behavioral patterns in human beings which may be originated as a result of metaphysical speculation.
The concept of social constructivism on the other hand was developed based on the works of Berger and Luckman (1966), Watzlawick (1984) and Shotter (1993). This philosophy is based upon idea that people use language as a medium for sharing their experiences and it is through this way, they make sense of the world. Hence this theory suggests that people form, create and make themselves and their worlds by means of the verbal skills. It is more concerned about the process and product peoples’ way of interacting with each other (Leeds-Hurwitz, 1995). It is considered as social because it focuses on the interaction rather than the individual.
Hence even though there exist a clear dichotomy between the two philosophies, there exists a high difference in opinion among the researchers about the selection of methods. Therefore a number of compromising has to be done between these two philosophies in order to conduct a research.
Figure: Matrix of research designs
The following are pure researches mentioned in the figure above:
Action research: Action research cannot completely belong to pure research as it contains certain elements of applied research too. In this research, the researcher tries to be as much close to the research area or subject. This is because the researcher intends to make a direct influence for which he may make certain alterations even in the research process. The ground concept behind this research was that if one wants to know a subject well he must try to change it. This theory was widely accepted in Organizational Development where the trainers use it for enhancing the team effectiveness (French & Bell, 1978; Homan, 1979).
Survey Research: This is a kind of research process which requires direct interaction with the samples. If done properly, this research can give accurate information which can be highly useful in solving the research problem. Survey research gives much importance to primary data and involves systematic observation (Sapsford, 2007). Therefore this research helps to take better business decisions by which way the business activities can be handles much more effectively.
Experimental design: In this research, the experimenter assigns subjects at random to a test or control group. The test group conditions are then manipulated by the researcher so as to evaluate their effect with regard to the control group who are in a normal setting. But this kind of study is quite difficult to conduct in real organizations since its not easy to obtain a captive population for the research.
Quasi-experimental designs: This was developed when the researchers found it difficult to create pure experimental designs under the positivist concept. This was clearly explained by Campell and Stanley (1963), who assessed a number of designs which made use of multiple scales so as to minimize the impact caused by incomplete matching of control and test groups.
Ethnography: This is a theory based on social constructivist principle where the experimenter will try to be a part of the research so as to understand the mindset of the experimental group and the meanings and significances that the samples bring in their behavior of themselves and others.
Case Method: There are several research designs at present and some of them are similar to fieldwork methods while some others fall in the intermediate position. Case study method is the kind of research design which comes under the latter category. The most important deliberations to this topic were made by Robert Yin (1993). The other popular contributors to this method are Stake (1995), Eisenhardt (1989), and Hamel et.al (1993).
Grounded theory: This theory is based on constructivist principles and helps the researcher to understand how he will proceed with the research. This concept got worldwide attention from the studies of Glaser and Strauss (1967). In this method, the researcher mainly has to develop a theory by means of comparative method, i.e. viewing the same event in different situations and analyzing the difference.
Co-operative enquiry: This is a type of action research where all the participants of the research contribute equally to the development of enquire and show active participation in the research process (Heron, 1996; Reason, 2003).
Apart from the above research designs, there exist certain other designs which can be used for management researches. The most important ones among them are critical inquiry, participant observation and narrative methods. All these methods have certain similarities with the above mentioned designs as well as certain unique features of their own. Hence they are also widely accepted in business research and used according to the requirements.
2. The different areas of research that holds high significance are:
a) Research topic: Finalizing the research topic is the first and most important part of a research since it is based on the depth of title that the research proceeds (Kumar, 2005). In the broad sense, research title can be any question or problem that you need to answer or any assumption that needs to be challenged. As per the words of Powers, Meenaghan & Twoomey (1985), “Potential research questions may occur to us on a regular basis, but the process of formulating them in a meaningful way is not at all an easy task”. In order to frame a research title, it is essential to have a thorough knowledge of research methodology as well as on the theoretical aspects of the topic. The research title can be formulated based upon four sources namely daily life, practical issues, past research and theory (Johnson & Christensen, 2010). They form the initial base for the structuring of research title. Thus the research topic serves as a foundation to the research and hence if it is framed properly, the study will also move in the right direction. The research problem can vary in complexity and hence the way it is formulated determines the research approach, study design, sampling strategy, research instrument as well as many other factors. Therefore there must be certain considerations while selecting a research topic. The main considerations are interest, magnitude, measurement concepts, expertise level, relevance, ethical issues etc. (Kumar, 2005). If the research topic fulfills most of these considerations, then it can be regarded as a suitable one.
b) Research question: A research question refers to the problem that the experimenter seeks to answer through the research. The research question is an essential element of every research despite of the fact that it is qualitative and quantitative. But they will vary according to the nature of research. If the research question is qualitative, then it will be intended for finding out the relationship between two variables. A variable is something which has more than one value (Vaus, 2002). Hence research questions for these kinds of researches are not much specific. While on the hand, a quantitative research question tends to more precise as it will be based on mathematical results. Business research usually applies qualitative research questions since they ask more general questions which can be used for exploring the sample’s views or responses on a particular phenomenon. It is usually an interrogative sentence which aims to relate two or more variables. It can be of three types namely: descriptive, predictive and causal. The researcher must have thorough understanding of the variables that are related in the research. These questions are open-ended, general, and over-reaching regarding an issue, problem or phenomenon.
c) Research hypothesis (es): They can be referred to as tentative answers to the research questions. This is because the solution to them can be found out after statistical analysis only and hence it forms a critical step in the evidence-gathering process of a research (Guerrero & Nachmias, 2010). There will be two hypotheses while conducting this process. The first hypotheses will be referred to as research hypotheses (H1) and the other one will be termed as null hypothesis (H0). The research hypotheses are formulated based upon the parameters of population but its form is different from research to research. Null hypothesis is the one set as a counter the research hypothesis. It is because every research hypothesis is set in order to prove a relationship. Hence it is necessary to set an alternate hypothesis that could disprove the relationship. For hypothesis testing, the researcher aims to disprove the null hypothesis in order to provide support to the research hypothesis. Therefore research hypotheses are oriented towards a particular direction. This means it states whether a population mean is greater than or lesser than a particular value.
d) Research objectives: A researcher intends to summarize what must be achieved by means of the research. The objectives of the study must be in alignment with the problem statement. There are mainly two types of objectives set for a research- general objective and specific objective. General objective states what the researcher intends to achieve in general terms through conducting this study. This general objective when broken down to smaller parts for better solving becomes specific objectives. Research objectives are better than research questions since they lead to higher specificity in the research (Saunders et.al. 2003).
3. A research design process that could be adopted in business researches is as follows:
Web 01: An overview of the research design process
This research design can be used in most management studies with little variations. This process starts with identification of research problem. Every research must have a problem which needs to be solved through the research process. Business researches usually intend to find answers to some social issues, business problem or a phenomenon. After identifying the research problem, the next step is determining the research objectives and questions. A research question is the problem or issue that the research intends to solve while research objectives tend to summarize the area that is to be achieved by the study. Hence both these factors lie in close relation to each other. The next task in the research design process is conducting the secondary research. This is done by means of an extensive literature study. This way the researcher gathers details on the theoretical aspects of the research. Hence the researcher will be searching in books, journals, articles, online databases, websites etc for collecting the data. An appropriate methodology for conducting the research must be developed by this time. Research methodology helps in determining the research philosophy, approach, method, study design etc. All these factors are highly important for the successful completion of research. After collecting the secondary data, the researcher will look out for methods for collecting the primary data. There exist several research methods for data collection. The research instrument will be decided based upon the literature study (Taylor et.al. 2006). The most common research instruments include surveys, interviews, questionnaires, etc. In some cases, the researcher conducts a pilot study in order to obtain responses from a small portion of the population. These research instruments help to gain valid data from the samples. After collecting the primary and secondary information, the researcher decides upon the methods for their analysis. Literature review analysis is quite critical for researches as they would help in gaining insights on the core issues and provide suitable recommendations. The analysis of primary data is done by means of certain analysis techniques. The most common methods include percentage analysis, chi-square test, ANOVA, z-test etc. They help in providing empirical findings for the research which are most accurate. After obtaining the findings, they are interpreted to reach at final judgments and conclusion. On the basis of these interpretations, the researcher could provide suitable suggestions or recommendations for solving the research problem.
a) A survey is used for describing a population and it usually involves systematic observation or interviewing. Surveys involve what the researcher wants to find out and the answers also will be defined under a specific range (Sapsford, 2007). The most common survey instruments used in management and social researches are questionnaires and interviews. For constructing a questionnaire, the responses must be provided in a range. Hence it must have certain measurement scales. There are mainly four types of measurement scales used in surveys. They are:
Nominal Scale: This is a kind of measurement scale where the objects are assigned to categories without any numerical properties. These scales have definite identity of their own but have no other characteristics (Jackson, 2007). Those variables which are measured using these scales are named as categorical variables since they try to classify the information collected. But these variables have no empirical value. Examples of variables measured through nominal scales are ethnicity, gender etc. These variables cannot be hence used in mathematical analysis. Hence these scales are considered as the lowest level of measurement since the variables are different by means of quality rather than quantity in this (Bordens & Abbott, 2006).
Ordinal Scale: They form the next level of measurement used in surveys. They constitute numbers which are in a predetermined order so that there will be certain relationships which can be inferred from them (Sapsford, 2007). Hence the objectives are classified into certain categories which form a rank order through a range. The data in this scale do have certain identity and properties but they don’t have equal unit size and absolute zero.
Interval Scale: In this scale, the measurement units are all of equal size. Therefore it fulfills the three main criteria of an ideal scale i.e. identity, magnitude and equal unit size. It forms the third level of measurement used in surveys. Hence this scale contains the characteristics of both nominal and ordinal scale but is much more effective than the other two (Malhotra & Dash, 2008). Fahrenheit scale is the most prominent interval scale used.
Ratio scale: This forms the highest measurement scale and has all the characteristics of nominal, ordinal and interval scale. It is the most complicated scale since it allows the researcher to identify the absolute differences between scale points as well as helps him in comparing the responses obtained (Hair et.al. 2007). This scale possesses an absolute zero which indicates the absence of variables in measurement.
b. Questionnaire is the most commonly used research instrument. The main characteristics of a good questionnaire are as follows:
It must be short and precise. So while framing a questionnaire, ensure that the research objectives will be met from the analysis of responses of the questionnaire.
Usage of simple language: The questionnaire must be framed in such a way that the respondents find it easy to understand.
Start with interesting questions: The questionnaire will include questions that will generate interest in the samples to answer and encourage them to fill out the complete form.
Absence of leading questions: If there are leading questions in the form, it will encourage the respondents to answer in a particular way. But this is not considered as a good practice since the sample must not be influenced by any means in writing an answer.
The number of choices won’t be too long because the respondents will find it difficult to assess all of them. Hence the choice list must be small.
Include simple concepts which the samples could understand. Including complicated concepts will make it a tough task for the respondents to answer the questions properly.
The elements of a bad questionnaire are included below.
It will have more open-ended questions than close-ended ones. This will make it difficult for the respondents to answer the questions and also their answers won’t be similar. The analysis of these responses is not possible through any statistical method.
The questions will be vague. The respondents will not be able to understand the real meaning of this question and this will influence their answer.
The research questions are not communicated through this questionnaire. This way the questionnaire fails to serve its purpose.