Business Research for Decision Making Sixth Edition

Preview:

Citation preview

Business Research for Decision MakingSixth Edition

by

Duane Davis

Chapter 6

Fundamentals of Research Design

PowerPoint Slides for the Instructor’s Resource Manual for

Research Design

Is the structure of the research project to solve a particular problem?

Its purpose is to guide researchers in answering problems.

Is a series of tradeoffs and compromises.

Potential Sources of Error in the Design Process: 1. Planning – faulty designs 2. Collection – actual collection of data 3. Analytical – the way the data is analyzed 4. Reporting- errors in interpretation

MaxiMinCon Principle

• Researchers are guided by the MaxiMinCon Principle

Maximize systematic variance Minimize error variance Control extraneous variance

Design is the researcher’s attempt at variance control.

MaxiMinCon Principle, continued

• Four Primary Means to Control Variance 1. Build the variable into the study 2. Exclude the variable from the study 3. Statistically control the variable(s) through covariance

analysis 4. Randomization

Potential Sources of Error in the Research Design Process

Table 6.1

Managerial Strategies for Dealing with Error

• Planning Error - well thought out proposals that are specific in design aspects– evaluated impartially

• Collection Error – careful execution of planned design – periodic managerial evaluations

• Analytical Error – justification of analytical methods – outside evaluation

• Reporting Error – unbiased and knowledgeable reviewers – commonplace in rigorous research environments

Schematic Diagrams Illustrating Moderation & Intervening Variables

Figure 6.1

Source: Adapted from Uma Sekaran, Research Methods for Managers (New York: Wiley, 1984), pp. 50–58.

Internal Validity

Are the results of the study true?Is what has taken place due to the variables the researcher claims to be

operating or could something else have taken place?

Internal Validity• History: events outside the study affect the

results. E.g. high unemployment• Maturation: changes in respondents over time• Testing: halo effect, Hawthorne effect – are the

respondents tipped about what your are studying?

• Instrumentation: changes in the instrument over time.

• Selection: differences among groups; respondents vs. non-respondents

• Mortality: people drop out of a study over time

External Validity

What is causality and proof?

Can anything be proven?

Causality

• Theory – a reason why one variable should cause an effect in another

• Correlation• Time order: cause MUST proceed effect• Rule out other explanations for the effect

or spurious variables

Comparison of the Major Types of Research Designs

Table 6.2

©2005 by DuxburyA division of Thomson Learning

Sources of Invalidity for Basic Research Designs*Table 6.3

Source: Donald T. Campbell and Julian C. Stanley: Experimental and Quasi-Experimental Designs for Research, copyright © 1963, Houghton Mifflin Company, used by permission.

R - Randomly A - Assign to conditions of I - Independent variable, then observe on D - Dependent variable, forE- ExperimentalR - Research 

major research methods tend to

have strengths, as well as

weaknesses. 

Ability to Infer Causality

(INTERNAL VALIDITY)

Low High

Ability to Represent Real-life Situations(EXTERNAL VALIDITY)

Low EXPERIMENTS

High SURVEYS

This would be the ideal – strong in both areas

Managerial Concerns

• No single correct design• Design to answer the research

problem• All research design represents a

compromise• A design is not a framework to be

followed blindly

Key Managerial Questions Pertaining to Research Design

Recommended