Research Process in Social Sciences- Part 2

Short introduction to Research Process-Part- 2

Salman Ata

i)               RESEARCH QUESTIONS

Research questions are specific questions about an event, phenomenon or behavior and are mode of conveying researchers’ interest in a research problem.  Examples include: what are the factors which affect the learning ability of students of B.Sc (Hons.) Agriculture at University of Agriculture Faisalabad? How creative thinking can be improved in Pakistani students? A good research question is one which can be generalized to a broader population, address real, complex, and un-answered problems. In qualitative research, research questions are often used instead of objectives or hypotheses. Narrowly focused research questions are not desirable because often they are not able to capture complex social phenomena and lead to un-publishable research findings. Such type of questions should be avoided in which researcher ask questions which are interesting to him/her but not for the scientific community and it does not has the worth to add significantly in scientific knowledge. As research process consume much time, resources and energy of researcher, the researcher should be sure that research questions address the real problems, not the imaginary ones and has not been previously addressed properly (Anol, 2012, pp-21).

The research questions often take two forms: central question and sub-questions. The central question is more general or broader question which is designed to explore the core phenomenon or concept under investigation. The associated questions help to answer the central question. For example, in the field of agriculture, a central question can be: “what major factors are responsible for low yield of cotton in Punjab, Pakistan?”. The associated sub-questions can be as follows: 1. What are the associated problems (and to what extent) of farmers regarding production of cotton? what are associated causes? 2. What are the problems (and to what extent) regarding plant protection measures? 3. What type of problems cotton growers faced regarding marketing? and so on.

In social sciences, there are two types of research questions:

        1.     Descriptive (what is going on?)

        2.     Explanatory (why is it going on)

1.     Descriptive Research Questions

              Descriptive research performs a key role in research by immeasurably adding knowledge about the nature and shape of the society we live. Government funded researches such as population census (what is the population in the country?), economic survey (what is the status of countrys’ economy?), crime statistics (what is the nature and rate of crimes in the country) are mainly encompasses descriptive research. Good description provides direction for explanatory research questions (why). For example if we find that yield of major crops in Pakistan is declining over the last 5 years than we can ask “why it is happening?”

2.     Explanatory Research Questions

               Explanatory research centers on “why?” questions. Why yield of cotton in Pakistan is low as compared to India? Why food security status in South Punjab is not satisfactory? and Why terrorist attack in Pakistan increased over last 10 years? are examples of explanatory research questions. As already mentioned, descriptive research provides a way to ask explanatory research question.

     (1)   Searching relevant literature

The next step in the research process is searching of relevant literature. A literature review is an evaluative report of studies found in the literature related to the topic under investigation. According to Boote and Beile (2005) “a researcher cannot perform significant research without first understanding the literature in the field”. A literature review is not just a search for information but comprises of recognition and articulation of relationships between the literature and topic under study. Anol (2012) describes three main purposes of review of literature: 1) assessment of the present status of knowledge in the field of research 2) identification of main researchers, theories, articles, and finding in the research area, and 3) identification of research/knowledge gaps in that area. In reality perfect review does not exist because all reviews are written from a specific perspective or view point of the reviewer. Therefore, reviewer should be aware of his/her own value judgment and try to give scholarly respect to the ideas of others.

      (2)   Selecting a Theory

             “A theory is a set of systematically interrelated constructs and propositions intended to explain and predict a phenomenon or behavior of interest, within certain boundary conditions and assumptions” (Anol, 2012; pp-14). It is a systematic collection of connected theoretical propositions (tentative and hypothetical relationship between constructs). In general, propositions relate two or three construct (an abstract concept chosen or created to explain a given phenomenon), theories constitute a system of multiple constructs and propositions. Therefore theories have a wider scope and are more complex and abstract than the proposition and/or hypothesis. These are the explanations of a social or natural phenomenon, event or behavior. More specifically, “a scientific theory is a system of constructs (concepts) and propositions (relationships between those constructs) that collectively presents a logical, systematic, and coherent explanation of a phenomenon of interest within some assumptions and boundary conditions” (Bacharach 1989, cited by Anol, 2012). A good theory not only describes and predicts things, but also explain why things happen. It is important to mention here that some events or phenomenon can be predicted without explaining the causes. In simple words, we can say that any idea or hypothesis validated by testing is called a theory. While, the review of literature may explore a broad range of constructs relevant to phenomenon, a theory may identify which constructs are rationally related to the research subject. Agency Theory, Innovation Diffusion Theory, and Theory of Planned Behavior are some examples of theories in social sciences.

     (3)   Deriving Hypotheses

                In social phenomena, one or more relationships could be under investigation in order to learn more about their functions. However, it might be possible that some relationships under observation might be due to chance and not due to relationships between the variables. In this case there a method is needed to evaluate the probability that relationships exist by chance. The formulation and testing of hypothesis is such a method. “Hypotheses are single tentative guesses, good hunches –assumed for use in devising theory or planning experiments intended to be given a direct experimental test when possible” (Rogers, 1966). According to Creswell (1994) “Hypothesis is a formal statement that presents the expected relationship between an independent and dependent variable”. Whereas Anol (2012) defined hypothesis as “the empirical formulation of propositions, stated as relationships between variables”. Yield of cotton crop increases with the adoption of Bt cotton in Punjab, education significantly increase women empowerment index in district Faisalabad, and ICT has potential to improve agricultural growth in Pakistan, and poverty significantly increases crime rates in Sindh, are some examples of Hypotheses. Scientific hypothesis is one which is testable. It means that through test or experiment we can show that variables might be related. The outcomes of the experiment will decide whether to accept or reject hypothesis. Scientific hypothesis which do not indicate directionality (whether relationship is positive or negative) and/or causality (which variable is “cause”) are considered as weak hypothesis.  For example: “child education is associated with long term household food security” is a weak hypothesis because it is not clear whether child education reduces or increases household food security. A strong hypothesis should be “child education increases household food security in long term” because it defines both directionality and causality (child education improves household food security and not the reverse). Hypothesis are specified in empirical plane, unlike the proposition which are intended for theoretical plane. Therefore, hypotheses are observable and testable empirically and if not supported empirically, they are rejected. Statistical procedures are usually uses to test the hypothesis. Hypotheses take two forms: null and alternate hypotheses. A null hypothesis makes a prediction that there is no significant difference or relationship between groups on a variable in the general population. Hypothesis can be derived from any source such as a social sciences theory, experience or previous research. The alternative hypothesis is a statement of what a hypothesis test is set up to establish: 1. Opposite of a null hypothesis; 2. Only reached if H0 is rejected. In statistical testing, it is not possible to test alternative hypothesis directly. Instead, it is indirectly tested by rejecting the null hypothesis with a specific level of probability. Unidirectional hypotheses, also called “one-tailed” states which way one variable will affect the other. Bidirectional hypotheses also called “two-tailed” states a relationship between two variables but does not indicate which way one variable will affect the other. 

      A)   DATA AND METHODS

      (1)   Collecting data

                The next step in the research process is data collection. It is the process of collecting and measuring data on concerned variables in a systematic way that enables the researcher to answer research questions, test hypotheses and interpret the findings. The data collection part of research process is common in both social and natural sciences although there is a difference in methods. The significance of accurate and honest data collection is also remains same in all fields of studies. In social sciences, different types of research instruments are used to obtain different types of data. Focus group discussions, key informant interviews and personal observation are the qualitative data collection tools whereas structured interview schedules, questionnaires, and telephonic interviews are common quantitative data collection instruments. Recently combination of both quantitative and qualitative methods has gained much popularity (e.g., Greene and Caracelli, 1997; Tashakkori and Teddlie, 1998).

              Researchers are increasingly using triangulation (mix method techniques) for in-depth research and to widen the scope of study.  According to the proponents of mix-method research, complex research designs are required to capture complex nature of human phenomenon. But this combination is not always desirable and should be used when research questions and context of study need such combination.