Research Design is the approach with which an empirical research is planned to be done. It discusses about the data collection process, instrument development process and the sampling process.
Research designs are broadly classified into positivist research designs for theory testing and interpretive research designs for theory building. Positivist designs use deductive approach while interpretive research use inductive approach.
The type of data used in positivist designs is mainly quantitative but can be qualitative also. Similarly the qualitative data is predominantly used in interpretive designs but quantitative data are also used. In certain phenomenon, both quantitative and qualitative data are used and such designs are referred to as ‘mixed-mode designs‘.
The quality of research designs is specified in terms of few attributes like internal validity, external validity, construct validity and statistical conclusion validity (Bhattacherjee, A. 2012). Internal validity is about the causal relation between independent variable and dependent variable. The causality requirements are covariation of cause and effect; temporal precedence of cause before effect; and no other possible explanations. External validity is the extent to which the observed relationships of the sample can be generalized to the population. Extent of internal validity and external validity achievable varies with research designs and the choice of the research design is based on the expected internal validity and external validity. Construct validity is the extent to which a measurement scale measures the theoretical construct. This is assessed in positivist research using correlational or factor analysis through pilot test data. Statistical conclusion validity is the extent to which the statistical test used is valid in terms of the right test in terms of sample size, distributions etc.
Dr. Selvaraj Vadivelu 