Research Design Notes

The Research Process: Elements of Research Design

Topics Discussed

• Research Design Topics:
  • Purpose of the study: exploratory, descriptive, hypothesis testing, case study analysis
  • Type of investigation: causal versus correlational
  • Extent of researcher interference with the study
  • Study setting: contrived versus noncontrived
  • Unit of analysis: individuals, dyads, groups, organizations, cultures
  • Time horizon of study: cross-sectional versus longitudinal
  • Managerial implications

Chapter Objectives

Upon completion of this chapter, you need to:

1. Understand different aspects relevant to designing a research study.
2. Identify the scope and end use of results for any given study.
3. Decide on the type of investigation needed, the study setting, the extent of researcher interference, the unit of analysis, and the time horizon of the study for a given situation.
4. Identify if a causal or correlational study is more appropriate in a specific situation.

The Research Design

• After identifying variables and developing the theoretical framework, the next step is to design the research to gather and analyze data for a solution.
• Key issues in research design include:
  • Purpose of the study (exploratory, descriptive, hypothesis testing).
  • Study setting.
  • Type of investigation.
  • The extent of researcher interference.
  • Time horizon.
  • Unit of analysis.
• Decisions also involve:
  • Sampling design.
  • Data collection methods.
  • Measurement.
  • Data analysis to test hypotheses.
• Each component of the research design provides critical choice points.
• Scientific rigor depends on carefully selecting appropriate design alternatives based on the study's purpose.
• For critical financial decisions (e.g., investing millions of dollars), ensure precision and acceptable confidence levels through careful attention to sampling, measurement, and data collection.
• In contrast, generating a profile of managers for a newsletter requires less elaborate research design decisions.
• More sophisticated and rigorous research designs require greater time, costs, and resources.
• At each decision point, consider whether the benefits of a more sophisticated design (accuracy, confidence, generalizability) justify the larger investment of resources.

Basic Aspects of Research Design

The six basic aspects of research design are:

1. Purpose of the study.
2. Types of investigation.
3. Extent of researcher interference.
4. Study setting.
5. Unit of analysis.
6. Time horizon of the study.

Measurement, data collection methods, sampling design, and data analysis are elaborated in later chapters.

Purpose of the Study

Studies can be:

• Exploratory.
• Descriptive.
• Hypothesis testing (analytical and predictive).
• Case study analysis.

Case Study: Examines studies in similar organizational situations to solve problems or understand phenomena.

• The nature of the study depends on the stage of knowledge about the research topic.
• Design decisions become more rigorous from exploratory to descriptive to hypothesis testing stages.

Exploratory Study

• An exploratory study is undertaken when little is known about the situation, or no information is available on how similar problems or research issues have been solved in the past.
• Extensive preliminary work is needed to:
  • Gain familiarity with the phenomena.
  • Understand what is occurring.
  • Develop a model.
  • Set up a rigorous design for comprehensive investigation.
• Exploratory studies help to comprehend the nature of the problem.
• Extensive interviews may be necessary to understand the situation.
• Qualitative studies (e.g., data collected through observation or interviews) are exploratory in nature.
• When data reveal patterns, theories are developed and hypotheses formulated for subsequent testing.
• Example:
  • Henry Mintzberg interviewed managers to explore the nature of managerial work.
  • He formulated theories of managerial roles, the nature and types of managerial activities, and so on.
  • These have been tested in different settings through both interviews and questionnaire surveys.
• Exploratory studies are also necessary when some facts are known, but more information is needed for developing a viable theoretical framework.
• Example:
  • If we want to get at the important factors that influence the advancement of women in organizations, previous studies might indicate that women are increasingly taking on qualities such as assertiveness, competitiveness, and independence.
  • There is also a perception that a judicious blend of masculine and feminine traits – such as being strong but not tough, kind but not soft – is conducive to women's organizational advancement.
  • Interviewing women managers who have made it to the top to explore relevant variables will help build a robust theory.
• Exploratory studies are crucial for obtaining a good grasp of the phenomenon and advancing knowledge through theory building and hypothesis testing.
  Example: A manager wants to know if the work ethic values of employees in a subsidiary in Pennathur City are different from those of Americans. Since there is little information about Pennathur, an exploratory study is needed, interviewing the employees in organizations in Pennathur.
• Religion, political, economic, and social conditions, upbringing, cultural values, and so on play a major role in how people view their work in different parts of the world.
• Many topics of interest and concern to management in the management and organizational behavior areas have been studied, and information is available in the library on these subject areas.
• Although few exploratory studies are currently undertaken in the management area, researchers do explore new grounds from time to time with the changing dynamics that occur in the workplace.
• Exploratory studies about organizationally relevant differences in race, ethnic, and country origins are being undertaken so that sound theories about managing a diverse work group can be evolved for the future.
• Exploratory studies can be undertaken by interviewing individuals and through focus groups.
  *Focus groups are discussed further in Chapter 8.

Descriptive Study

• A descriptive study aims to ascertain and describe the characteristics of the variables of interest.
• For example, describing a class composition by year, sex, age, semesters left, and business courses taken.
• Descriptive studies are undertaken in organizations to learn about and describe the characteristics of a group of employees, for example, the age, educational level, job status, and length of service of Hispanics or Asians working in the system.
• Descriptive studies are also undertaken to understand the characteristics of organizations that follow certain common practices.
• For example, one might want to know and be able to describe the characteristics of the organizations that implement flexible manufacturing systems (FMS), or that have a certain debt-to-equity ratio.
• The goal is to offer a profile or describe relevant aspects of the phenomenon of interest from an individual, organizational, industry-oriented, or other perspective.
• Such information may be vital before even considering certain corrective steps; for example, should the organization consider changing its practices?

Example: A brief description of the study of advances in the textile industry might look something like this: US Textiles has used high-tech to make huge advances in productivity and innovation. An army of reels surrounds a complex circular knitting machine at Malden Hills, feeding yarn to thousands of needles producing polyester fabric. The 3-dimensional loom has added value utility in the competitive market. Jacquard looms, which are computer controlled at Burlington, weave miles of intricately patterned materials. Even faster looms use jets of compressed air to move the weft. The industry is being revolutionized with a 3-dimensional loom that weaves the fibers in the shape of the end product, skipping the laborious process of lamination. Denims, which are difficult to make, are now made by computer controls and electric sensors to a so-called dye range, which ensures a shade perfectly matching customers' samples. The fastest looms, automated spinning processes, and Sanfmi machines programmed from a desktop computer and capable of producing any type of garment are other innovations in the textile industry.

Descriptive studies help to:

1. Understand the characteristics of a group in a given situation.
2. Think systematically about aspects in a given situation.
3. Offer ideas for further probe and research.
4. Help make certain simple decisions (such as how many and what kinds of individuals should be transferred from one department to another).

Examples:

• A bank manager wants to have a profile of the individuals who have loan payments outstanding for six months and more. The profile will include details of their average age, earnings, nature of occupation, full-time/part-time employment status, and the like. This might help him to elicit further information or decide right away on the types of individuals who should be made ineligible for loans in the future.
• A CEO may be interested in having a description of organizations in her industry that follow the LIFO system. In this case, the report might include the age of the organizations, their locations, their production levels, assets, sales, inventory levels, suppliers, and profits. Such information might allow comparison later of the performance levels of specific types of companies.
• A marketing manager might want to develop a pricing, sales, distribution, and advertising strategy for her product. With this in mind, she might ask for information regarding the competitors.

Descriptive studies are essential in many situations.

• Qualitative data from interviews may help understand phenomena at the exploratory stages.
• Quantitative data (frequencies, mean, standard deviations) are necessary for descriptive studies.

Example: a descriptive study of organizational member reaction to an on-site child care facility.

• 30% of employees were in favor.
• 40% felt it was unnecessary.
• 20% indicated it would only benefit those with preschool children.
• 10% suggested a cafeteria-style benefits system.
• More women than men were favorably inclined (2:1).
• Parents with two or more preschool children desired it; others opposed.
• Employees over 50 and below 25 did not favor it.
• Women between 25 and 45 desired it most.
• The mean on the preference scale was low (1.5 on a five-point scale) with high dispersion (standard deviation = $1.98$).
• The average preference indicated by women between the ages of 30 and 45 with children is the highest ($4.75$ on a five-point scale) with very little dispersion (the standard deviation for this group of 42 women is $0.38$).
• Introductory descriptive narratives in some research reports are drawn from government statistical publications such as the Bureau of Labor Statistics, census, and the like, from which data are culled for presentation, as and when appropriate.

Hypothesis Testing

• These studies explain relationships, establish differences among groups, or test the independence of factors.
• Hypothesis testing explains variance in dependent variables or predicts organizational outcomes.

Examples:

• A marketing manager wants to know if the sales of the company will increase if he doubles the advertising dollars. Here, the manager would like to know the nature of the relationship that may be established between advertising and sales by testing the hypothesis: If advertising is increased, then sales will also go up.
• Given people's tensions on the subject of purchase of guns in these days of crime in cities big and small, a marketing researcher might be interested in predicting the factors that significantly account for the variance in people's decision to purchase guns. Here, the researcher might theorize the factors that influence people's decision to possess guns (through literature search and interviews) and then test the hypothesis that four specific variables significantly account for the variance in people's intention to buy a gun.
• More men than women are whistleblowers (establishes a difference between groups).
• Working the night shift (as opposed to the day shift) is related to whether or not one is married: In this case, a $\{χ^2$ test of independence$$ will easily provide the answer to this question.
• As may be seen, in hypothesis testing the researcher goes beyond mere description of the variables in a situation to an understanding of the relationships among factors of interest.

Case Study Analysis

• Involves in-depth, contextual analyses of similar situations in other organizations.
• Case studies, as a problem-solving technique, are not frequently resorted to in organizations because finding the same type of problem in another comparable setting is difficult due to the reluctance of companies to reveal their problems.
• Case studies that are qualitative in nature are, however, useful in applying solutions to current problems based on past problem-solving experiences.
• They are also useful in understanding certain phenomena, and generating further theories for empirical testing.

Review of the Purpose of the Study

• In exploratory studies, the primary interest lies in exploring situational factors to understand the characteristics of the phenomena.
• Descriptive studies are carried out when the characteristics or phenomena are known to exist, and the aim is to describe them by offering a profile of the factors.
• Hypothesis testing enhances the understanding of relationships among variables and may establish cause-and-effect relationships.
• Case studies are generally qualitative and used in managerial decision-making.
• Methodological rigor and research costs increase progressively from exploratory to hypothesis-testing studies.
• Increases in sample size, multiple data collection methods, sophisticated measuring instruments, and so on, add to research costs but contribute more to testability, accuracy, precision, and generalizability.

Type of Investigation: Causal Versus Correlational

• A manager should determine whether a causal or a correlational study is needed to find an answer to the issue at hand.
• A causal study establishes a definitive cause-and-effect relationship.
• A correlational study identifies important factors "associated with" the problem.
• In a causal study, the researcher aims to state that variable X causes variable Y. When variable X is removed or altered, problem Y is solved.
• Due to multiple factors influencing each other, a researcher might identify crucial factors associated with the problem rather than establish a cause-and-effect relationship.
• Whether a study is causal or correlational depends on the type of research questions asked and how the problem is defined.

Examples:

• Causal Study Question: Does smoking cause cancer?
• Correlational Study Question: Are smoking and cancer related? OR Are smoking, drinking, and chewing tobacco associated with cancer? If so, which of these contributes most to the variance in the dependent variable?
• The answer to the causal question establishes whether not smoking will avoid cancer.
• The answer to the correlational question determines if smoking and cancer are correlated.
• The intention is not to establish a causal connection but to see if a relationship exists among the variables investigated.

Examples:

• Causal: Fears of an earthquake predicted recently in the New Madrid fault zone were instrumental (i.e., causal) in an unprecedented number of house owners in the Midwest region taking out an earthquake insurance policy.
• Correlational: Increases in interest rates and property taxes, the recession, and the predicted earthquake considerably slowed down the business of real estate agents in the Midwest.

Extent of Researcher Interference with the Study

• The degree of researcher interference affects whether the study is causal or correlational.

• Correlational Study: Conducted in the natural environment with minimal interference.
  Example: Studying factors influencing training effectiveness requires developing a theoretical framework, collecting data, and analyzing findings.

• Causal Study: The researcher manipulates certain variables to study the effects on the dependent variable.
  Example: Studying the influence of lighting on worker performance involves manipulating lighting intensities.

• The researcher might create an artificial setting (e.g., a laboratory) to study cause-and-effect relationships by manipulating certain variables and tightly controlling others.

• Thus, there could be varying degrees of interference by the researcher. From minimal to extreme.