Designing A Research: Dr. Sasmita Mishra

Designing A Research Dr. Sasmita Mishra

Research Design:its meaning A research design is the arrangement of conditions for collection and analysis of data in a manner that aims to combine relevance to the research purpose with economy in procedure. It is the conceptual structure within which the research is conducted; it constitutes the blueprint for the collection,

Research Designs and Methods A Research Design provides a framework for the collection and analysis of data. Choice of research design reflects decisions about priorities given to the dimensions of the research process. • A Research Method is simply a procedure

for collecting data. Choice of research method reflects decisions about the type of instruments or techniques to be used.

Design Decision Answer the following questions before deciding to select a research design. What is the study about? Why is the study being made? Where will the study be carried out? (Study setting) What type of data is required?(type of investigation) Where can the required data be found? What periods of time will the study include? (time horizon) What will be the sample design? (sampling design)

Segments of Research Design Research design constitutes: Sampling design:selecting items to be observed from the given study. Observational design:conditions under which the observations are to be made. Statistical design:how much information to be collected and how the data to be analyzed. Operational design: the techniques by which the procedure specified in the sampling, statistical and observational design can be carried out.

Purpose of Research Design To provide a conclusive answer to the problem posed by the researcher. To control the potentially relevant variables, that is those which may also influence the research outcome, but in which the researcher is not interested at the moment.

Characteristics of A Good Research Design Reliability – are measures consistent? Replication/replicability – is study repeatable? Validity – are conclusions well-founded?

Types of Validity Measurement (or construct) validity – do measures reflect concepts? Internal validity – are causal relations between variables real? External validity – can results be generalized beyond the research setting? see Box 2.7 Ecological validity – are findings

Important concept Relating to Research Design

Independent and Dependent Variable Extraneous Variable Control Confounded Relationship Experimental and non-experimental Hypothesis-testing research Experimental and control group Experiment

What is a variable? A variable that can take on differing or varying values. The value can differ at various times for the same object or person, or at the same time for different objects or persons. Ex. Production unit, absenteeism, motivation etc.

Types of Research Design Exploratory Descriptive Hypothesis-testing research Observational Studies & Survey

Exploratory Research Design Exploratory research is of three types: Survey of concerning literature Experience survey Analysis of insight stimulating examples

Descriptive Design The goal of descriptive study is to offer to the researcher a profile or to describe relevant aspects of the phenomena of interest from an individual, organisational, industryoriented or other perspective.

Ex: descriptive study A marketing manager might want to develop a pricing, sales, distribution, and advertising strategy for her product and for this she may want to know: – The percentage of companies who have price higher and lower than the industry norm, a profile of the terms of scale etc. – Percentage of competitors hiring inhouse staff to handle sales and those who use independent agents – The types of distribution channel used and the percentage of customers using each.

What is experimental design Experimental designs are those where the researcher tests the hypotheses of causal relationship between variables. To establish that variable X causes variable Y, all three of the following conditions should be met: Both X & Y should covary X should precede Y No other factor should cause the change in the Y.

Types of Experimental Design Laboratory Experiment: Experiments done in an artificial or contrived environment. Field Experiment: Done in natural environment in which activities regularly take place.

Lab Experiment The success of lab experiment depends upon proper control of extraneous variable and and manipulation of the independent variable.

Classical Experimental Design T1 Obs1 experimental

8 months

T2

experimental treatment

Obs2 group

Random Assignment Obs3

no

Obs4

experimental group

control

Threats to Internal Validity History Testing Instrumentation Mortality Maturation Selection Ambiguity about the direction of causal influence

Threats to External Validity 1. Interaction of selection and treatment 2. Interaction of setting and treatment 3. Interaction of history and treatment 4. Interaction effects of pretesting 5. Reactive effects of experimental arrangements