BBA RESEARCH METHODOLOGY TYPES OF RESEARCH DESIGN
Types of Research Design:
(i) Exploratory or Formulative Design:
The main purpose of exploratory study is to gather information which will help in future for formulation of a precise research problem. On the basis of the collected facts the researcher may be able to formulate sound hypotheses for further research. It may also enable the researcher to get himself acquainted with the phenomena which he expects to investigate at a later stage. The aim of an exploratory or formulative study may be clarification of concepts, establishing priorities for future research and collection of data about the actual conditions which affect an intended research.
Requirement of Exploratory Design:
The essentials for exploratory or formulative design are:
(a) Review of pertinent literature
(b) Experience Survey
(c) Analysis of Insight Stimulating cases.
(a) Review of pertinent literature:
While proceeding in the path of research the researcher has to take help from the work already done by his predecessors. By doing so, he will not only save himself from the problem of trial and error but also minimize the expenditure of his energy. Apart from reviewing available literature pertaining to the problem under investigation, the researcher may also take into account the literature pertinent to analogous problems.
(b) Experience survey:
Because of the complicated nature of social problems, the researcher is not in a position to collect all the required materials about a particular problem from one place. At times the researcher has to contact the persons who have earned enough of experience to understand and analyze the social reactions. The researcher should take advantage of their experience in a very intelligent manner.
Taking good advantage of the experience of the persons involves the following steps:
(i) Selection of respondents:
Formulation of a correct exploratory design requires that the investigator should make proper selection of the respondents. For this purpose he should select only those respondents who are dependable and who have actual knowledge regarding the problem under investigation.
The selection of the respondents may be made either directly or indirectly. In direct selection the investigator chooses those persons who are well known for their knowledge in the problem area. In case of indirect selection the investigator chooses those persons who are indirectly concerned with the problem. Hence, the selection of the respondents should not be confined to a particular group; rather it should be many sided.
(ii) Questioning of the respondents:
Proper questioning of the respondents ensures relevant information. Therefore while framing the questions, due attention should be given on clarity of concepts. For this purpose, the investigator should consult the books and the relevant portions of the bibliographical schemes adequately.
(c) Analysis of insight stimulating cases:
Analysis of insight stimulating cases includes all those events, incidents and phenomena that stimulate the researcher. Such cases invoke in the investigator the thinking regarding the formulation of the hypotheses. In this regard, the attitude of the investigator, intensity of the case study and integrative power of investigators appear to be very important.
As regards the attitude of the investigator, receptivity and sensitivity are needed. These qualities enable the investigator to take stock of various developments occurring in his field of study and make steady progress.
Intensive case study involves studying the subject matter in all its dimensions and verifications, in the background of history.
In this regard, the groups, the community and groups of individuals may be treated as the units of study.
Integrative power of the investigator is considered important because on that basis he is able to collect even the minutest possible information regarding the subject matter. What appears significant, in this regard, is his attention on new observations rather than on experimentation.
(ii) Descriptive Research Design:
The purpose of descriptive type of design is to describe some event, situation, people, group or community or some phenomena. Fundamentally, it is a fact finding exercise which focuses on relatively far dimensions of a well defined entity, aiming at precise and systematic measurement of some dimensions of a phenomenon.
Usually a descriptive design involves detailed numerical descriptions, such as distribution of the population of a community by age, sex, caste or education. The researcher may also take recourse to descriptive design for estimating the proportion of people in a particular geographical locality in respect of their specific views or attitudes.
However, the procedure followed in descriptive design is broadly analogous, notwithstanding the differences evinced in their field, formulation of hypotheses, objectives, for treatment of the problem and in matters of field expansion.
(iii) Diagnostic Research Design:
Being concerned with the express characteristics and existing social problems, the diagnostic research design endeavors to find out relationship between express causes and also suggests ways and means for the solution. Thus, the diagnostic studies are concerned with discovering and testing whether certain variables are associated. Such studies may also aim at determining the frequency with which something occurs or the ways in which a phenomenon is associated with some other factors.
Diagnostic studies are mostly motivated by hypotheses. A primary description of a problem serves the basis so as to relate the hypotheses with the source of the problem and only those data which form and corroborate the hypotheses are collected. As regards the objectives of diagnostic research design, it is based on such knowledge which can also be motivated or put into practice in the solution of the problem. Therefore, it is obvious that the diagnostic design is concerned with both the case as well as the treatment.
Diagnostic studies seek immediate to timely solution of the causal elements. The researcher, before going through other references, endeavors to remove and solve the factors and the causes responsible for giving rise to the problem.
The research design of diagnostic studies demands strict adherence to objectivity for elimination of any chances of personal bias or prejudice. Utmost care is taken while taking decisions regarding the variables, nature of observation to be made in the field, the type of evidence to be collected and tools of data collection. Simultaneously the research economy should not be lost sight of. Any faulty decision in these regard will result in wastage of time, energy and money.
Usually the first step in such designing is accurate formulation of research problem wherein research objectives are precisely stated and principal areas of investigation are properly linked. Otherwise the investigator will find it difficult to ensure the collection of required data in a systematic manner. Simultaneously, the clarification of concepts and the operational definition of the terms should also be ensured so as to make them emendable to measurement.
At the next stage certain decisions regarding collection of data are taken. In this regard, the researcher should always bear in mind the advantages and disadvantages of the method to be employed and at the same time the nature of research problem, type of data needed, degree of desired accuracy etc. should be considered. That apart, while collecting data, effort must be made to maintain objectivity to the maximum possible extent.
In order to surmount the financial constraints, paucity of time, a representative sample of the research universe should be drawn so as to gather relevant information. A wide range of sampling techniques is prevalent which must be made use of, appropriately by the researchers.
At the stage of analysis of data, the researcher must take proper care in placing each item in the appropriate category, tabulating of data, applying statistical computations and so on.
Sufficient care must be taken to avoid potential errors due to faculty procedures of analysis of data. Advance decisions regarding the mode of tabulation, whether manual or by machine, accuracy of tabulating procedures, statistical application etc. will be of immense help in this regard.
(iv) Experimental Design:
The concept of experimental design in sociological research refers to systematic study of human relations by making the observations under conditions of control. In the words of Jahoda and Cook, ‘an experiment maybe considered as a way of organizing the collection of evidence so as to permit one to make inference about the tenability of a hypothesis. According to Chapin, “experiment is simply observation under controlled conditions. When observation alone fails to disclose the factors that operate in a given problem, it is necessary for the scientist to resort to experiment.”
In real terms, experimentation is resorted to when it is not possible to solve the problem through observation and general knowledge. The core of the experimental method lies in drawing inferences by observation of human relations under controlled conditions. Since a number of factors are in operation in every complex social situation, the social scientist, while seeking to describe the single causal relation of factor A to factor B, must attempt to create an artificial situation wherein all other factors, such as C, D, E etc., are controlled.
Such a state is achieved by selecting two groups which are equal in all significant receipts and choosing either of the groups as experimental group, and the other as the ‘control group’, and thereafter exposing the ‘experimental group’ to the assumed causal variable, while keeping the ‘control’ group under control. After a specific time period, the two groups are compared in terms of the ‘assumed effect’.
The assumed causal variable and the assumed effect are otherwise called the independent variable and dependent variable respectively. Required evidence for testing causal relations among variables, already stated in the form of a hypothesis, is generated by the above method of experiment.
Demonstration of causal relationship among variables in experimental design involves three clear-cut operations; such as demonstrating co-variation, eliminating spurious relationships and establishing the time order of occurrence.
Here we will discuss the third operation which is concerned with establishing the time order of occurrence. This necessitates that the researcher should demonstrate that one phenomenon occurs first or gets transformed before the other phenomenon with the premise that the phenomenon which is yet to occur cannot be the determinant of the present or past phenomena.
Experimental design enables the researcher to draw causal inferences. It also smoothens, the observation of independent variable causing assumed effect.
The three components of experimental design are: comparison, manipulation, and control.
Through comparison, the correlation between variables is known. It also enables us to demonstrate the association between two or variables.
Through manipulation the researcher establishes the time order of events. The major evidence which become essential to determine the sequence of events is that a change occurs only after the activation of the independent variable. In other words the independent variable precedes the dependent variable.
Types of Experimental Design:
There are numerous ways in which experiments can be done in the field of social sciences. In their work “Experimental and Quasi-Experimental Designs of Research on Teaching”, Donald T. Cambell and Julian C. Stanley have mentioned more than a hundred ways of conducting experiments which may be designated as experimental design.
But from the analytical point of view seven broad categories may be mentioned:
(i) After only Design:
Among all categories of experimental designs, after only design appears to be the simplest. This consists in measuring the dependent variable only after the experimental subjects have been exposed to the experimental variable. This design is considered more appropriate as an exploratory study than a real experiment.
(ii) Before-After Design:
As the name suggests, in this design measurement of the dependent variable is taken before as well as after exposure of the subject to the experimental variable, and the difference between the two measurements is taken to be the effect of the experimental variable. For example if the measured value of the dependent variable before exposure of the subject to the experimental variable is noted as ‘A’ and its measured value after exposure of the subject to experimental variable is noted as ‘B’ then the effect of the experimental variable is taken to be (B—A).
(iii) Before-After with Control Group Design:
In this design the research has a control group against which the results of the experimental groups are compared. The control group and experimental groups are selected in such a way that both the groups are similar and interchangeable. The control group is measured before as well as after without being exposed to the experimental variable.
Hence, there may hardly be any difference between before and after measurements. But if there is any difference between before and after measurement, it represents the result of uncontrolled variables.
On the other hand, the experimental variable is introduced in the experimental group. The difference between before and after measurements in respect of the experimental group is construed as the result of experimental variable as well as the uncontrolled variables. To know the exact effect of the experimental variable, the researcher deducts the difference between the two measurements of the controlled group from the difference of the two measurements of the experimental group.
(iv) Four Group-Six Study Design:
In this type of design two experimental groups and two control groups are taken. Measurements are made in six cases, i.e. before- measurement, and after-measurement in respect of experimental group-I, after-measurement in experimental group-II, before and after measurements in respect of control group-I; and only after measurement in control group-II.
Before measurements in all the four identical groups will be almost the same. If the before-measurements have no effect on the variable being studied, the two experimental groups should provide the same after-measurements and, similarly, the two control groups should also give the same after measurements. However, the results of in the two experimental groups are most likely to be different from the results of the two control groups, if the experimental variable exerts any influence.
(v) After Only with Control Group Design:
This is also known as two group-two study design, which is a modification of the four group-six study design. Here, the researcher does not study the experimental variable under different conditions. Hence, the effect of experimental variable is determined simply by finding out the differences between the after-measurements in respect of experimental and control groups. It so happens because if before-measurements of the experimental group-II and control group-II are taken, those are likely to be the same due to the identical characteristics of the groups. On this presumption, the researcher may very well ignore them.
(vi) Ex-Post Facto Design:
In Ex-post facto design the experimental and control groups are selected after the introduction of the experimental variable. Thus, it can be called as a variation of the after-only design. The main advantage of this design is that the test subjects are not influenced towards the subject by their knowledge of being tested. It also enables the researcher to introduce the experimental variable according to his own will and to control his observations.
(vii) Factorial Design:
All categories of experimental designs discussed above are designed to test experimental variable at one level only. But, on the other hand, the factorial designs enable the experimenter the testing of two or more variables simultaneously.