BUsiness research methods

BUS519 – Business Research Methods Study Notes BUS519 – Business Research Methods Student Study Notes Copyright 2010, 2011 The Taft University System, Inc All rights reserved No part of this publication may be reproduced or transmitted in any form or by any means, electronic or mechanical, including photocopy, recording, or any information storage and retrieval system, without permission in writing from the copyright holder BUS519 – Business Research Methods Study Notes Required Materials Business Research Methods (Tenth Edition, 2008), by Donald R Cooper and Pamela S Schindler ISBN 978-0-07-340175-1 Optional Readings and Internet sites: Journals such as the Journal of Small Business Management and Internet sites such as www.merlot.org provide current topical and supplemental business research coverage for those areas of student interest beyond required coursework BUS519 – Business Research Methods Study Notes Lesson Business research is a systematic inquiry that provides information More specifically, it is a process of planning, acquiring, analyzing, and disseminating relevant date, information, and insights to decision makers in ways that mobilize the organization to act in ways that maximize business performance Managers use this information to guide business decisions and reduce risk Multiple types of projects can be labeled “business research” Decision scenarios and decision makers can be found in every type of organization, whether for profit, non-profit, or public Decision makers rely on information to make more efficient and effective use of their budgets At no other time in history has so much attention been placed on measuring and enhancing return on investment (ROI) At a basic level, measurement of the ROI means calculating the financial return for all expenditures Over the past dozen years, technology has improved our measuring and tracking capabilities, while managers simultaneously realized their need for a better understanding of employee, stockholder, and customer behavior in order to meet goals Although business research helps managers choose better strategies, the cost of such research is being scrutinized for its contribution to ROI A management dilemma is a problem or opportunity that requires a management decision There are many factors that should stimulate your interest in studying research methods: Information overload, Technological connectivity, shifting global centers of economic activity and competition, increasingly critical scrutiny of big business, more government intervention, Battle for analytical talent, Greater computing power and speed (includes lower-cost data collection better visualization tools, powerful computations, more integration of data, real-time access to knowledge and new perspectives on established research methodologies) Understanding the relationship between business research and information generated by other information sources is critical for understanding how information drives decisions relating to organizational mission, goals, strategies, and tactics Even very different types of businesses have similar types of goals, which are related to such things as: Sales (membership), Market share, Return on investment, Profitability, Customer acquisition, Customer satisfaction, Employee productivity, Machine efficiency and Maximization of stock price or owner’s equity The need to complete exchanges with prospective customers drives every organization An exchange can be a purchase, a vote, attendance at a function, or a donation to a cause Each exchange, along with the activities required to complete it, generates data If organized for retrieval, these data constitute a decision support system (DSS) During the last 25 years, advances in computer technology made it possible to share this collected data among an organization’s decision makers, over an intranet or an extranet Sophisticated managers have developed DSSs where data can be accessed in real time These managers have a distinct advantage in strategic and tactical planning A business intelligence system (BIS) provides managers with ongoing information about events and trends in the business environment In our restaurant scenario, it might be collecting customer comments In the Mind Writer example, it might be data about laptops needing repair It costs less to retain a customer than to capture a new one, so businesses place a high value on keeping customers buying That is why customer satisfaction, customer loyalty, and customer assessment studies represent a significant portion of research studies Microsoft recently decided BUS519 – Business Research Methods Study Notes to tie its 600 managers’ compensation to levels of customer satisfaction rather than to sales and profits Strategy is defined as the general approach an organization will follow to achieve its goals A firm usually implements more than one strategy at a time The discovery of opportunities and problems, and the resulting strategies, often result from a combination of business research and BIS Business research contributes significantly to the design and selection of tactics Tactics are specific, timed activities that execute a strategy The purposes of business research include: Identifying and defining opportunities and problems; Defining, monitoring, and refining strategies; Defining, monitoring, and refining tactics; and Improving our understanding of the various fields of management Not all organizations use business research to help make planning decisions Increasingly, however, the successful ones Exhibit 1-2 shows an emerging hierarchy of organizations in terms of their use of business research In the top tier, organizations see research as the first step in any venture They use creative combinations of research techniques to gain insights that will help them make better decisions They may partner with outside research suppliers Every decision is guided by business research There is generally enterprise-wide access to research data and findings In the middle tier, decision makers periodically rely on research information Decision makers turn to business research when they perceive the risk of a particular strategy or tactic to be too great to proceed without it They rely heavily on tried-and-true methodologies, such as surveys and focus groups They often choose the methodology before fully assessing its appropriateness to the problem at hand There is limited access to research data and findings In the base tier, managers primarily use instinct, experience, and intuition to facilitate their decisions Decisions are supported with secondary data searches They often rely on informal group discussion, a small number of individual interviews, or feedback from the sales force Large firms that occupy this tier are often influenced by culture, smaller companies because they think formalized research is too expensive to employ Managers who not prepare to advance up the hierarchy will be at a severe competitive disadvantage The research process begins with understanding the manager’s problem the management dilemma In other situations, a controversy arises, a major commitment of resources is called for, or conditions in the environment signal the need for a decision In every chapter, we refer to this model as we discuss each step in the process Exhibit 1-4 is an important organizing tool because it provides a framework for introducing how each process module is designed, connected to other modules, and then executed Researchers often are asked to respond to “problems” that managers needed to solve Applied research has a practical problem-solving emphasis The problem-solving nature of applied research means it is conducted to reveal answers to specific questions related to action, performance, or policy needs Pure research or basic research is also problem-solving based It aims to solve perplexing questions or obtain new knowledge of an experimental or theoretical nature that has little direct or immediate impact on action, performance, or policy decisions Basic research in the business arena might involve a researcher who is studying the results of the use of coupons versus rebates as demand stimulation tactics, but not in a specific instance or in relation to a specific client’s product Both applied and pure research is problem-solving based, but applied research is directed much more to making immediate managerial decisions Is BUS519 – Business Research Methods Study Notes research always problem-solving based? The answer is yes Whether the typology is applied or pure, simple or complex, all research should provide an answer to some question Good research generates dependable data that are derived by professionally conducted practices, and that can be used reliably for decision making It follows the standards of the scientific method: systematic, empirically based procedures Exhibit 1-5 shows actions that guarantee good business research Characteristics of the scientific method are as follows: Purpose clearly defined, Research process detailed, Research design thoroughly planned, High ethical standards applied, Limitations frankly revealed, Analysis adequate for decision maker’s needs, Findings presented unambiguously, Conclusions justified, and Researcher’s experience reflected Good business research has an inherent value only to the extent that it helps management make better decisions that help achieve organizational goals The value of information is limited if the information cannot be applied to a critical decision Business research finds its justification in the contribution it makes to the decision maker’s task and to the bottom line Ethics are norms or standards of behavior that guide moral choices about our behavior and our relationships with others The goal of ethics in research is to ensure that no one is harmed or suffers adverse consequences from research activities Unethical activities are pervasive and include such things as: Violating nondisclosure agreements, breaking respondent confidentiality, misrepresenting results, deceiving people, invoicing irregularities and avoiding legal liability A recent study showed that: 80 percent of the responding organizations had adopted an ethical code There was limited success for codes of conduct There is no single approach to ethics Advocating strict adherence to a set of laws is difficult because of the constraint put on researchers Because of their war history, Germany’s government forbids many types of medical research Sometimes, an individual’s personal sense of morality is relied upon This can be problematic because each value system claims superior moral correctness Clearly a middle ground is necessary The foundation for a middle ground is an emerging consensus on ethical standards for researchers Codes and regulations guide both researchers and sponsors Review boards and peer groups examine research proposals for ethical dilemmas Many design-based ethical problems can be eliminated by careful planning and constant vigilance Responsible research anticipates ethical dilemmas and adjusts the design, procedures, and protocols during the planning process Ethical research requires personal integrity from the researcher, the project manager, and the research sponsor Exhibit 2-1 relates each ethical issue under discussion to the research process In general, research must be designed so that a respondent does not suffer physical harm, discomfort, pain, embarrassment, or loss of privacy To safeguard against these, the researcher should follow three guidelines: Explain study benefits Explain participant rights and protections Obtain informed consent Whenever direct contact is made with a participant, the researcher should discuss the study’s benefits, without over- or understating the benefits An interviewer should begin an introduction with: His or her name, the name of the research organization, A brief description of the purpose and benefit of the research, knowing why one is being asked questions improves cooperation, inducements to participate, financial or otherwise, should not be disproportionate to the task or presented in a fashion that results in coercion Sometimes, the purpose and benefits of the study or experiment must be concealed from respondents in order to avoid introducing bias The need for concealing objectives leads directly to the problem of deception BUS519 – Business Research Methods Study Notes Deception occurs when the participants are told only part of the truth, or when the truth is fully compromised There are two reasons for deception: To prevent biasing the participants and to protect the confidentiality of a third party Deception should not be used to improve response rates When possible, an experiment or interview should be redesigned to reduce reliance on deception Participants’ rights and well-being must be adequately protected Where deception in an experiment could produce anxiety, a subject’s medical condition should be checked to ensure that no adverse physical harm follows The American Psychological Association’s ethics code states that the use of deception is inappropriate unless deceptive techniques are justified by the study’s expected value and equally effective alternatives that not use deception are not feasible Participants must have given their informed consent before participating in the research Securing informed consent from respondents is a matter of fully disclosing the procedures of the proposed study or other research design before requesting permission to proceed It is always wise to get a signed consent form when: Dealing with children, Doing research with medical or psychological ramifications, There is a chance the data could harm the participant, If the researchers offer only limited protection of confidentiality For most business research, oral consent is sufficient Exhibit 2-2 presents an example of how informed-consent procedures are implemented In situations where respondents are intentionally or accidentally deceived, they should be debriefed once the research is complete Debriefing involves several activities following the collection of data: Explanation of any deception, Description of the hypothesis, goal, or purpose of the study, Post-study sharing of results, and Post-study follow-up medical or psychological attention It explains the reasons for using deception in the context of the study’s goals Where severe reactions occur, follow-up attention should be provided to ensure that the participants remain unharmed Even when research does not deceive the participants, it is good practice to offer them follow-up information This retains the goodwill of the participant and provides an incentive to participate in future projects Follow-up information can be provided in a number of ways: with a brief report of the findings and with descriptive charts or data tables For experiments, all participants should be debriefed in order to put the experiment into context Debriefing usually includes a description of the hypothesis being tested and the purpose of the study Debriefing allows participants to understand why the experiment was created Researchers also gain insight into what the participants thought about during and after the experiment, which can lead to research design modifications The majority of participants not resent temporary deception, and debriefed participants may have more positive feelings about the value of the research than those who didn’t participate in the study Nevertheless, deception is an ethically thorny issue and should be addressed with sensitivity and concern for research participants Privacy laws in the United States are taken seriously All individuals have a right to privacy, and researchers must respect that right Desire for privacy can affect research results Example: Employees at MonsterVideo did not guarantee privacy, so most respondents would not answer research questions about their pornographic movie viewing habits truthfully, if at all The privacy guarantee is important not only to retain validity of the research but also to protect respondents Once the guarantee of confidentiality is given, protecting that confidentiality is essential Obtain signed nondisclosure documents Restrict access to participant identification Reveal participant information only with written consent Restrict access to data instruments where the participant is identified Do not disclose data subsets Researchers should restrict access to BUS519 – Business Research Methods Study Notes information that reveals names, telephone numbers, addresses, or other identifying features Only researchers who have signed nondisclosure, confidentiality forms should be allowed access to the data Links between the data or database and the identifying information file should be weakened Interview response sheets should be accessible only to the editors and data entry personnel Occasionally, data collection instruments should be destroyed once the data are in a data file Data files that make it easy to reconstruct the profiles or identification of individual participants should be carefully controlled For very small groups, data should not be made available because it is often easy to pinpoint a person within the group This is especially important in human resources research Privacy is more than confidentially A right to privacy means one has the right to refuse to be interviewed or to refuse to answer any question in an interview Potential participants have a right to privacy in their own homes, including not admitting researchers and not answering telephones They have the right to engage in private behavior in private places, without fear of observation To address these rights, ethical researchers: Inform participants of their right to refuse to answer any questions or participate in the study Obtain permission to interview participants Schedules field and phone interviews Limit the time required for participation Restrict observation to public behavior only Some ethicists argue that the very conduct that results in resistance from participants— interference, invasiveness in their lives, denial of privacy rights—has encouraged researchers to investigate topics online The growth of cyberstudies causes us to question how we gather data online, deal with participants, and present results Issues relating to cyberspace in research also relate to data mining The information collection devices available today were once the tools of spies, the science fiction protagonist, or the superhero For instance: Smart cards, Biometrics, Electronic monitoring, Global surveillance and Genetic identification (DNA) All these things are used to track and understand employees, customers, and suppliers The primary ethical data-mining issues in cyberspace are privacy and consent (See Exhibit 2-3) Smart cards that contain embedded personal information can be matched to purchase, employment, or other behavior data Use of such cards offer the researcher implied consent to participant surveillance Smart cards are commonly used by grocers, retailers, wholesalers, medical and legal service providers, schools, government agencies, and so on In most cases, participants provide the personal information requested by enrollment procedures In others, enrollment is mandatory, such as when smart cards are used to track convicted criminals in correctional facilities or those attending certain schools In some cases, mandatory sharing of information is for personal welfare and safety, such as when you admit yourself for a medical procedure In other cases, enrollment is for monetary benefits The bottom line is that the organization collecting the information gains a major benefit: the potential for better understanding and competitive advantage General privacy laws may not be sufficient to protect the unsuspecting in the cyberspace realm of data collection 15 European Union (EU) countries started the new century by passing the European Commission Data Protection Directive Under this directive, commissioners can prosecute companies and block websites that fail to live up to its strict privacy standards The directive prohibits the transmission of names, addresses, ethnicity, and other personal information to any country that fails to provide adequate data protection This includes direct mail lists, hotel and travel reservations, medical and work records, orders for products, and so on U.S industry and government agencies have resisted regulation of data flow, but the EU insists that it is the right of all citizens to find out what information is in BUS519 – Business Research Methods Study Notes a database and correct any mistakes Few U.S companies would willingly offer such access due to the high cost Whether undertaking product, market, personnel, financial, or other research, a sponsor has the right to receive ethically conducted research With regards to confidentiality; some sponsors wish to undertake research without revealing themselves Types of confidentiality include: Sponsor nondisclosure, Purpose nondisclosure, and Findings nondisclosure Companies have a right to dissociate themselves from the sponsorship of a research project This is called sponsor nondisclosure This is often done when a company: Is testing a new product idea, to avoid having the company’s current image or industry standing influence potential consumers Is contemplating entering a new market, to keep from tipping off competitors and in such cases, it is the responsibility of the researcher to devise a plan that safeguards the identity of the sponsor Purpose nondisclosure involves protecting the purpose of the study or its details Even if a sponsor feels no need to hide its identity or the study’s purpose, most sponsors want the research data and findings to be confidential, at least until the management decision is made Thus, sponsors usually demand and receive findings nondisclosure between themselves or their researchers and any interested but unapproved parties With regards to the Sponsor-Researcher Relationship, the obligations of managers include: Specify their problems as decision choices Provide adequate background information Provide access to company information gatekeepers The obligations of researchers include: Develop a creative research design that will provide answers to manager’s questions Provide data analyzed in terms of problems/decision choices specified Point out limitations of research that affect results Make choices between what manager wants and what research thinks should be provided Manager-Researcher conflict arises due to: Knowledge gap between researchers and the manager; job status and internal political coalitions to preserve status; unneeded or inappropriate research; the right to quality research Managers have limited exposure to research and often have limited formal training in research methodology Explosive growth in research technology has led to a widening of this gap in expertise Researchers challenge a manager’s intuitive decision making skill Managers feel requesting research is equivalent to indicating their decision making skills are lacking One research function—to challenge old ideas—as well as to challenge new ideas threatens insecure managers by inviting a critical evaluation of a manager’s ideas by others who may be seen as rivals Research has inherent value only to the extent that it helps management make better decisions Not all decisions require research Decisions requiring research are those that have potential for helping management select more efficient, less risky, or more profitable alternatives than would otherwise be chosen without research An important ethical consideration for the researcher and the sponsor is the sponsor’s right to quality research This right entails: Providing a research design appropriate for the research question Maximizing the sponsor’s value for the resources expended Providing data-handling and –reporting techniques appropriate for the data collected BUS519 – Business Research Methods Study Notes From the proposal to final reporting, the researcher guides the sponsor on the proper techniques and interpretations The researcher should propose the design most suitable for the problem A researcher should not propose activities designed to maximize researcher revenue or minimize researcher effort at the sponsor’s expense We’ve all heard “You can lie with statistics.” It is the researcher’s responsibility to prevent that from occurring The ethical researcher reports findings in ways that minimize the drawing of false conclusions The ethical researcher also uses charts, graphs, and tables to show data objectively, despite the sponsor’s preferred outcomes Occasionally, research specialists may be asked by sponsors to participate in unethical behavior Compliance by the researcher would be a breach of ethical standards Examples of things to avoid: Violating participant confidentiality, changing data or creating false data to meet a desired objective, changing data presentations or interpretations, interpreting data from a biased perspective, omitting sections of data analysis and conclusions, and making recommendations beyond the scope of the data collected Behaving ethically often requires confronting the sponsor’s demand and educating the sponsor to the purpose of research, explaining the researcher’s role in fact finding versus decision making, explaining how distorting the truth or breaking faith with participants leads to future problems, failing moral suasion, terminate the relationship with the sponsor, researchers and team members Researchers are responsible for their team’s safety, as well as their own Responsibility for ethical behavior rests with the researcher who, along with assistants, is charged with protecting the anonymity of both the sponsor and the participant Researchers must design a project so that the safety of all interviewers, surveyors, experimenters, or observers is protected Factors that may be important when ensuring a researcher’s right to safety: Some urban and undeveloped rural areas may be unsafe for researchers If persons must be interviewed in a high-crime district, it may be necessary to provide a second team member to protect the researcher It is unethical to require staff members to enter an environment where they feel physically threatened Researchers who are insensitive to these concerns face both research and legal risks Researchers should require ethical compliance from team members Assistants are expected to: Carry out the sampling plan Interview or observe respondents without bias Accurately record all necessary data The behavior of the assistance is under the direct control of the responsible researcher or field supervisor If an assistant behaves improperly in an interview, or shares a respondent’s interview sheet with an unauthorized person, it is the researcher’s responsibility Consequently, all assistants should be well trained and supervised Each researcher handling data should be required to sign a confidentiality and nondisclosure statement Many corporations, professional associations, and universities have a code of ethics The impetus for these policies and standards can be traced to two documents: The Belmont Report of 1979 and The Federal Register of 1991 Society or association guidelines include ethical standards for the conduct of research One source contains 51 official codes of ethics issued by 45 associations in business, health, and law Without enforcement, standards are ineffectual BUS519 – Business Research Methods Study Notes Effective codes: Are regulative Protect the public interest and interests of the profession served by the code Are behavior-specific and Are enforceable A study that assessed the effects of personal and professional values on ethical consulting behavior concluded that “… unless ethical codes and policies are consistently reinforced with a significant reward and punishment structure and truly integrated into the business culture, these mechanisms would be of limited value in actually regulating unethical conduct.” The U.S government implemented the Institutional Review Boards (IRBs) in 1966 The Dept of Health and Human Services (HHS) translated the federal regulations into policy Most other federal and state agencies follow the HHS-developed guidelines Each institution receiving funding from HHS, or doing research for HHS, is required to have its own IRB to review research proposals Exhibit 2-4 describes some characteristics of the Institutional Review Board process IRBs concentrate on two areas: The guarantee of obtaining complete, informed consent from participants and the risk assessment and benefit analysis review The need to obtain informed consent can be traced to the first 10 points in the Nuremberg Code Complete informed consent has four characteristics: The participant must be competent to give consent Consent must be voluntary Participants must be adequately informed to make a decision Participants should know the possible risks or outcomes associated with the research In the risk assessment and benefit analysis review: Risks are considered when they add to the normal risk of daily life The only benefit considered is the immediate importance of the knowledge to be gained Possible long-term benefits are not considered Right to Privacy laws that influence the ways in which research is carried out: Public Law 95-38 (Privacy Act of 1974): the first law guaranteeing Americans the right to privacy Public Law 96440 (Privacy Act of 1980): carries the right to privacy further These two laws are the basis for protecting the privacy and confidentiality of the respondents and the data There are many resources for Ethical Awareness According to the Center for Business Ethics at Bentley College over a third of Fortune 500 companies have ethics officers and almost 90 percent of business schools have ethics programs Exhibit 2-5 provides a list of recommended resources for business students, researchers, and managers The Center for Ethics and Business at Loyola Marymount University provides an online environment for discussing issues related to the necessity, difficulty, costs, and rewards of conducting business ethically Its website offers a comprehensive list of business and research ethics links When we research, we seek to know what is, in order to understand, explain, and predict phenomena This requires asking questions These questions require the use of concepts, constructs, and definitions A concept is a generally accepted collection of meanings or characteristics associated with certain events, objects, conditions, situations, and behaviors When you think of a spreadsheet or a warranty card, what comes to mind is not a single example, but your collected memories of all spreadsheets and warranty cards From this, you extract a set of specific and definable characteristics Concepts that are in frequent and general use have been developed over time, through shared language usage These concepts are acquired through personal experience That is why it’s often difficult to deal with an uncommon concept or a newly advanced idea One way to handle this problem is to borrow from other languages (gestalt) or to borrow from other fields (impressionism) Sometimes, we must adopt new meanings for words or develop new labels for concepts When we adopt new meanings or develop new labels, we begin to develop a BUS519 – Business Research Methods Study Notes Pearson’s Product Moment Coefficient r The Pearson correlation coefficient is an estimate of strength of linear association and its direction between interval or ratio variables The coefficient ρ represents the population correlation Correlation coefficients reveal the magnitude and direction of relationships The magnitude is the degree to which variables move in unison or opposition Direction tells us whether large values on one variable are associated with large values on the other (and small values with small values) The absence of a relationship is expressed by a coefficient of approximately zero Scatterplots for Exploring Relationships Scatterplots is a visual technique that depicts both the direction and the shape of a relationship between two variables Both the direction and the shape of a relationship are conveyed in a plot The shape of linear relationships is characterized by a straight line, whereas nonlinear relationships have curvilinear, parabolic, and compound curves representing shapes Pearson’s r measures relationships in variables that are linearly related Careful analysts make scatterplots an integral part of the inspection and exploration of their data The Assumptions of r The first requirement for r is linearity (e.g., the assumption that data can be described by a straight line passing through the data array The second assumption for correlation is a bivariate normal distribution (e.g., data are from a random sample where two variables are normally distributed in a joint manner) If these assumptions cannot be met, the analyst should select nonlinear or nonparametric measures of association Computation and Testing of r During computation, covariance is the amount of deviation that the X and Y distribution have in common Common Variance as an Explanation Coefficient of determination (r2) amount of common variance in two variables in regression On a visualization, the area of overlap (X and Y) represents the percentage of the total relationship accounted for by one variable of the other Testing the Significance of r The observed significance level for a one-tailed test is half of the printed two-tailed version in most programs Interpretation of Correlations A correlation coefficient of any magnitude, whatever its statistical significance, does not imply causation Take care to avoid so-called artifact correlations (e.g., where distinct subgroups in the data combine to give the impression of one) Another issue affecting interpretation of 104 BUS519 – Business Research Methods Study Notes correlation coefficients concerns practical significance Even when a coefficient is statistically significant, it must be practically meaningful With large samples, even exceedingly low coefficients can be statistically significant A coefficient is not remarkable simply because it is statistically significant SIMPLE LINEAR REGRESSION Relationships, among other things, may serve as a basis for estimation and prediction Simple prediction—when we take the observed values of X to estimate or predict corresponding Y values Regression analysis uses simple and multiple predictors to predict Y from X values With respect to similarities and differences of correlation and regression (see Exhibit 18-9), their relatedness would suggest that beneath many correlation problems is a regression analysis that could provide further insight about the relationship of Y with X The Basic Model A straight line is fundamentally the best way to model the relationship between two continuous variables Regression coefficients are the intercept and slope coefficients Slope (β1) is the change in Y for a 1-unit change in X This is the ratio of change (∆) in the rise of the line relative to the run or travel along the X axis Intercept (β0)—one of two regression coefficients, is the value for the linear function when it crosses the Y axis or the estimate of Y when X is zero Concept Application Unfortunately, one rarely comes across a data set composed of four paired values, a perfect correlation, and an easily drawn line A model based on such data is deterministic in that for any value of X, there is only one possible corresponding value of Y A probabilistic model also uses a linear function Error term is the deviations of values of Y from the regression line of Y for a particular value of X Method of Least Squares The method of least squares is a procedure for finding a regression line that keeps errors of estimate to a minimum When we predict the values for Y for each Xi the difference between the actual Yi and the predicted Y is the error This error is then squared and then summed Residuals A residual is the difference between the regression line value of Y and the real Y value When standardized, residuals are comparable to Z scores with a mean of and a standard deviation of It is important to apply other diagnostics to verify that the regression assumptions (normality, linearity, equality of variance and independence of error) are met Predictions Prediction and confidence bands are bow-tie shaped confidence interval around a predictor Confidence intervals can be expanded or narrowed 105 BUS519 – Business Research Methods Study Notes Testing for Goodness of Fit Goodness of fit is a measure of how well the regression model is able to predict Y The most important test in bivariate linear regression is whether the slope,β1, is equal to zero Zero slopes result from various conditions: Y is completely unrelated to X, and no systematic pattern is evident There are constant values of Y for every value of X The data are related but represented by a nonlinear function The t-Test To test whether β1 = 0, we use a two-tailed test The F Test The F test has an overall role for the model in multiple regression Coefficient of Determination In predicting the values of Y without any knowledge of X, our best estimate be Y mean Each predicted value that does not fall on Y contributes to an error estimate Based on the formula (see chapter), the coefficient of determination is the ratio of the line of best fit’s error that incurred by using Y One purpose of testing is to discover whether the regression equation is a re effective predictive device than the mean of the dependent variable The coefficient of determination is symbolized by r squared It has several purposes: As an index of fit, it is interpreted as the total proportion of variance in Y explained by X As a measure of linear relationship, it tells us how well the regression line fits the data It is also an important indicator of the predictive accuracy of the equation Typically, we would like to have an r squared that explains 80 percent or more of the variation NONPARAMETRIC MEASURES OF ASSOCIATION Measures for Nominal Data Nominal measures are used to assess the strength of relationships in cross- classification tables There is no fully satisfactory all-purpose measure for categorical data Technically, we would like to find two characteristics with nominal measures: When there is no relationship at all, the coefficient should be and When there is a complete dependency, the coefficient should display unity, or Chi-Square-Based Measures The first chi-square-based measure (e.g., tests to detect the strength of the relationship between the variables tested with a chi-square test) is called phi (φ) (e.g., used with chi-square, a measure of association for nominal, nonparametric variables) Cramer’s V is used with chisquare, a measure of association for nominal, nonparametric variables with larger than x tables The contingency coefficient C is used with chi-square, a measure of association for 106 BUS519 – Business Research Methods Study Notes nominal, nonparametric variables This measure is not comparable to other measures and has a different upper limit for various table sizes Proportional Reduction in Error Proportional reduction in error (PRE) are measures of association used with contingency tables to predict frequencies The lambda (λ) coefficient is a measure of how well the frequencies of one nominal variable predict the frequencies of another variable Goodman and Kruskal’s tau (τ) is a measure of association that uses table marginals to reduce prediction errors Measures for Ordinal Data When data require ordinal measures (e.g., measures of association between variables generating ordinal data), there are several statistical alternatives Illustrations will include: Gamma, Kendall’s tau b and tau c, Somers’s d, and Spearman’s rho All but Spearman’s rankorder correlation are based on the concept of concordant (e.g., when a participant that ranks higher on one ordinal variable also ranks higher on another variable) and discordant (e.g., nature of the association when a subject that ranks higher on one ordinal variable ranks lower on another variable) pairs of individual observations may be calculated Goodman and Kruskal’s gamma (γ) uses a preponderance of evidence of concordant pairs versus discordant pairs to predict association Kendall’s tau b (τb) is a refinement of gamma for ordinal data that considers “tied” pairs, not only discordant or concordant pairs, for square tables Kendall’s tau c (τc) is a refinement of gamma for ordinal data that considers “tied” pairs, not only discordant or concordant pairs, for any-size table Somers’s d is a measure of association for ordinal data that compensates for “tied” ranks and adjusts for direction of the independent variable Spearman’s rho (ρ) correlates ranks between two ordinal variables Rho’s strengths outweigh its weaknesses 107 BUS519 – Business Research Methods Study Notes Lesson In recent years, multivariate statistical tools have been applied with increasing frequency to research problems Multivariate analysis—statistical techniques that focus upon and bring out in bold relief the structure of simultaneous relationships among three or more phenomena SELECTING A MULTIVARIATE TECHNIQUE Classifications of multivariate techniques may be classified as dependency and interdependency techniques Dependency techniques are techniques where criterion or dependent variables and predictor or independent variables are present Examples include multiple regression, multivariate analysis of variance MANOVA, and discriminant analysis Interdependency techniques are techniques where criterion or dependent variables and predictor or independent variables are not present Examples include factor analysis, cluster analysis, and multidimensional scaling Measures to be checked: Metric measures—statistical techniques using interval and ratio measures Nonmetric measures—statistical techniques using ordinal and nominal measures DEPENDENCY TECHNIQUES Multiple Regression Multiple regression—statistical tool used to develop a self-weighting estimating equation that predicts values for a dependent variable from the values of independent variables Multiple regression is used as a descriptive tool in three types of situations: It is often used to develop a self-weighting estimating equation by which to predict values for a criterion variable (DV) from the values for several predictor variables (IVs) A description application of multiple regression calls for controlling for confounding variables to better evaluate the contribution of other variables Multiple regression can be also used to test and explain causal theories This approach is referred to as path analysis (e.g., describes, through regression, an entire structure of linkages advanced by a causal theory) Multiple regression is also used as an inference tool to test hypotheses and to estimate population values Multiple regression is an extension of the bivariate linear regression discussed in Chapter 18 Dummy variables—nominal variables converted for use in multivariate statistics Regression coefficients are stated either in raw score units (the actual X values) or standardized coefficients (regression coefficients in standardized form [mean = 0] used to determine the comparative impact of variables that come from different scales When regression coefficients are standardized, they are called beta weights (β) (standardized regression coefficients where the size of the number reflects the level of influence X exerts on Y), and their values indicate the relative importance of the associated X values, particularly when the predictors are unrelated Most statistical packages provide various methods for selecting variables for the equation Forward selection—sequentially adds the variable to a regression model that results in the largest R2 increase Backward elimination—sequentially removes the variable from a regression model that changes R2 the least Stepwise selection—a method for sequentially adding or removing variables from a regression model to optimize R2 Collinearity—when two independent variables are highly correlated Multicollinearity—when more than two 108 BUS519 – Business Research Methods Study Notes independent variables are highly correlated Both of the above can have damaging effects on multiple regression Another difficulty with regression occurs when researchers fail to evaluate the equation with data beyond those used originally to calculate it A solution to the above problem can be the holdout sample (the portion of the sample is excluded for later validity testing when the estimating equation is first computed) Discriminant Analysis Discriminant analysis is frequently used in market segmentation research Discriminant analysis is a technique using two or more independent interval or ratio variables to classify the observations in the categories of a nominal dependent variable Once the discriminant equation is found, it can be used to predict the classification of a new observation The most common use for discriminant analysis is to classify persons or objects into various groups; it can also be used to analyze known groups to determine the relative influence of specific factors for deciding into which group various cases fall MANOVA Multivariate analysis of variance (MANOVA) assesses the relationship between two or more dependent variables and classificatory variables or factors MANOVA: Is similar to univariate ANOVA, with the added ability to handle several dependent variables, Uses special matrices to test for differences among groups, and Examines similarities and differences among the multivariate mean scores of several populations Centroids—term used for the multivariate mean score in MANOVA Before using MANOVA to test for significant differences, you must first determine that the assumptions for its use are met Structural Equation Modeling Since the 1980s, marketing researchers have relied increasingly on structural equation modeling to test hypotheses about the dimensionality of, and relationships among, latent and observed variables Structural equation modeling (SEM) uses analysis of covariance structures to explain causality among constructs It is most commonly called LISREL (linear structural relations) models The major advantages of SEM are: Multiple and interrelated dependence relationships can be estimated simultaneously It can represent unobserved concepts, or latent variables, in these relationships and account for measurement error in the estimation process Researchers using SEM must follow five basic steps: Model specification: A formal statement of the model’s parameters Specification error—an overestimation of the importance of the variables included in a structural model Estimation: Often uses an iterative method such as maximum likelihood estimation (MLE) Evaluation of fit: Goodness-of-fit tests are used to determine if the model should be accepted or rejected Respecification of the model and the last basic step is Interpretation and communication: SEM hypotheses and results are most commonly presented in the form of path diagrams (presents predictive and associative relationships among constructs and indicators in a structural model) Conjoint Analysis The most common applications for conjoint analysis are market research and product development 109 BUS519 – Business Research Methods Study Notes Conjoint analysis measures complex decision making that requires multiattribute judgments The objective of conjoint analysis is to secure utility scores (e.g., a score in conjoint analysis used to represent each aspect of a product or service in a participant’s overall preference ratings—also called partworths), that represent the importance of each aspect of a product or service in the subject’s overall preference ratings The first step in a conjoint study is to select the attributes most pertinent to the purchase decision Possible values for an attribute are called factor levels After selecting the factors and their levels, a computer program determines the number of product descriptions necessary to estimate the utilities Conjoint analysis is an effective tool used by researchers to match preferences to known characteristics of market segments and design or target a product accordingly INTERDEPENDENCY TECHNIQUES Factor Analysis Factor analysis is a technique for discovering patterns among the variables to determine if an underlying combination of the original variables (a factor) can summarize the original test Factor analysis begins with the construction of a new set of variables based on the relationships in the correlation matrix The most frequently used approach is the principle components analysis (one of the methods of factor analysis that transforms a set of variables into a new set of composite variables) These linear combinations of variables, called factors (the result of transforming a set of variables into a new set of composite variables through factor analysis), account for the variance in the data as a whole The best combination makes up the first principal component and is the first factor (and so on) The process continues until all the variance is accounted for Loadings: the correlation coefficients that estimate the strength of the variables composing the factor Eigenvalues: the proportion of total variance in all the variables that is accounted for by a factor Communalities: the estimate of the variance in each variable that is explained by the factors being studied Rotation: a technique used to provide a more simple and interpretable picture of the relationship between factors and variables If factor analysis’s results are examined with care, it can be a powerful tool Cluster Analysis Cluster analysis identifies homogeneous subgroups of study objects or participants and then studies the data by these subgroups Often used in the fields of medicine, biology, and other sciences Cluster analysis offers a means for segmentation research and other marketing problems where the goal is to classify similar groups Cluster analysis starts with an undifferentiated group of people, events, or objects and attempts to reorganize them into homogeneous subgroups Five steps are basic to the application of most cluster studies: 1) Selection of the sample to be clustered 2) Definition of the variables on which to measure the objects, events, or people 3) Computation of similarities among the entities through correlation, Euclidean distances, and other techniques 4) Selection of mutually exclusive clusters or hierarchy arranged clusters 5) Cluster comparison and validation 110 BUS519 – Business Research Methods Study Notes The average linkage method (evaluates the distance between two clusters by first finding the geometric center of each cluster and then computing distances between the two centers) is demonstrated The resulting diagram is called a dendogram Multidimensional Scaling Multidimensional scaling (MDS) is a scaling technique to simultaneously measure more than one attribute of the participants or objects A perceptual map is created With MDS, items that are perceived to be similar will fall close together on the perceptual map, and items that are perceived to be dissimilar will be farther apart We may think of three types of attribute space, each representing a multidimensional map: Objective space—in which an object can be positioned in terms of its measurable attributes (i.e., its flavor or weight) Subjective space— where perceptions of the object’s flavor, weight, et cetera value may be positioned The third map could describe respondents’ preferences using the object’s attributes Cluster analysis and MDS can be combined to map market segments and then examine products designed for those segments Presenting Insights and Findings: Written and Oral Reports Exhibit 20-1 Sponsor Presentation and the Research Process provides a picture of the report development process As part of the research proposal, the sponsor and the marketing researcher agree on what types of reporting will occur both during and at the end of the research project Depending on the budget for the project, a formal oral presentation may not be part of the reporting A research sponsor, however, is sure to require a written report THE WRITTEN RESEARCH REPORT A final report or presentation can destroy a study is not handled correctly Most readers are influenced by the quality of the reporting This fact should prompt researchers to make special efforts to communicate clearly and fully Research reports contain findings, analyses of findings, interpretations, conclusions, and sometimes recommendations Research reports must be objective in their nature Research reports may be defined by their degree of formality and design Short Reports Short reports are appropriate when the problem is well defined, is of limited scope, and has a simple and straightforward methodology Most informational, progress, and interim reports are of this kind Short reports are about five pages in length Format—a brief statement at the beginning describing the authorization for the study, the problem examined, and its breadth and depth Next are the conclusions and recommendations, followed by the findings that support them Section headings should be used The letter of transmittal is a vehicle used to convey short reports Short reports are produced for clients with small, relatively inexpensive research projects A letter report is often written in a personal style 111 BUS519 – Business Research Methods Study Notes Long Reports Long reports are of two types: A technical report is a report written for an audience of researchers and A management report is a report written for the non-technically-oriented manager or client The Technical Report This report should include full documentation and detail With respect to completeness, a good rule to follow is to provide sufficient procedural information so that others (if they chose to) could replicate the study A technical report should include a full presentation and analysis of significant data In a short technical report the emphasis is placed on the findings and conclusions The Management Report The management report is for the non-technical client The reader needs prompt exposure to the most critical findings Thus, this report is presented in inverse order with the findings presented first The order allows the reader to grasp the conclusions and recommendations quickly without much reading The management report should make liberal use of visual displays Headlines and underlining for emphasis help with comprehension It helps to have a theme running through the report that the reader can follow RESEARCH REPORT COMPONENTS Research reports, long and short, have a set of identifiable components (see Exhibit 20-2) Prefatory Items Prefatory materials not have a direct bearing on the research itself They assist the reader in using the research report Letter of Transmittal A letter of transmittal is the element of the final report that provides the purpose of, scope of, authorization for, and limitations of the study This is appropriate when a report is for a specific client and when is generated for an outside organization Internal projects not require this letter Title Page The title page should include four items: 1) Title of the report 2) Authorization Letter 3) Executive Summary (An executive summary is a concise summary of the major findings, conclusions and recommendations It can serve as a miniature (Topline report) report Two pages are generally sufficient.) 4) Table of Contents: If the report is totals more than to 10 pages, it should have a table of contents 112 BUS519 – Business Research Methods Study Notes Introduction The introduction prepares the reader for the report by describing the parts of the project: the problem statement, research objectives, and background material Problem Statement The problem statement contains the need for the research project Research Objectives The research question addresses the purpose for the project The objectives may be research questions and associated investigative questions Background It may be preliminary results of exploration form an experience survey, focus group, or another source It could also be secondary data from the literature review Previous research, theory, or situations that led to the management question are discussed in this section Methodology In short reports and management reports, the methodology should not have a separate section; it should be mentioned in the introduction, and details should be placed in an appendix For the technical report, the methodology is an important section, and contains at least five parts: Sampling Design The researcher explicitly defines the target population being studied and the sampling methods used Research Design The coverage of the design must be adapted to the purpose Strengths and weaknesses should be identified Data Collection This part describes the specifics of gathering the data Contents of this section depend on the selected design Relevancy of secondary data would be discussed here Any instructions should be placed in an appendix Data Analysis This section summarizes the methods used to analyze the data A rationale for choices should be provided A brief commentary on assumptions and appropriateness of use should be presented 113 BUS519 – Business Research Methods Study Notes Limitations The section should be a thoughtful presentation of the significant methodology or implementation problems if any exist All studies have their limitations Honesty and professionalism are the watchwords Findings This is generally the longest section of the report Exhibit 20-3 provides a sample findings page The objective is to explain the data rather than draw interpretations or conclusions Quantitative data should be presented with charts, graphs, and tables The data need not include everything you have collected Make this portion of the report convenient for the reader Conclusions Summary and Conclusions The summary is a brief statement of the essential findings In simple descriptive research, a summary may complete the report because conclusions and recommendations may not be required Findings state facts; conclusions represent inferences drawn from the findings Conclusions may be presented in tabular form for easy reading and reference Recommendations In applied research the recommendations will usually be for managerial action, with the researcher suggesting one or several alternatives that are supported by the findings Appendices The appendices are the place for complex tables, statistical tests, supporting documents, copies of forms and questionnaires, detailed descriptions of the methodology, instructions to field workers, and other evidence important for later support Bibliography The used of secondary data requires a bibliography A bibliography documents the sources used by the writer WRITING THE REPORT Judging a report as competently written is often the key first step to a manager’s decision to use the findings in decision making and also to consider implementation of the researcher’s recommendations 114 BUS519 – Business Research Methods Study Notes Prewriting Concerns Before writing, one should ask again, “What is the purpose of this report?” Another prewriting question is, “Who will read the report?” The technical background—the gap in subject knowledge between the reader and the writer—should be considered Next, ask, “What are the circumstances and limitations under which I am writing?” Lastly, “How will the report be used?” is a useful piece of information for the writer to possess The Outline A topic outline is a report planning format using key words or phrases A sentence outline is a report planning format using complete sentences The Bibliography Style manuals provide guidelines on form, section and alphabetical arrangement, and annotation Bibliographic retrieval software allows researchers to locate and save references from online services and translate them into database records Writing the Draft Once the outline is complete, decisions can be made on the placement of graphics, tables, and charts Each writer uses different mechanisms for getting thoughts into written form Readability Sensitive writers consider the reading ability of their audience to achieve high readership A readability index measures the difficulty level of written material (see Exhibit 20-4) Using a readability index allows the writer to revise the draft if necessary to match the audience of the report Advocates of readability measurement argue for written reports that are appropriate for the audience Comprehensibility Good writing varies with the writing objective Words and sentences should be carefully organized and edited Don’t confuse readers by mixing subordinate with major ideas Pace which is the rate at which the printed page presents information to the reader Writers use a variety of methods to adjust the pace of their writing (see chapter section for illustrations and details) Review the report to ensure the tone is appropriate Final Proof It is helpful to put the draft away and return to it the next day with a fresh objective eye and review it one more time before transmittal and presentation 115 BUS519 – Business Research Methods Study Notes Presentation Considerations The final consideration in the report writing process is production Overcrowded text may be avoided in the following ways: Use shorter paragraphs Indent parts of text that represent listings, long quotations, or examples Use headings and subheadings to divide the report and its major sections into homogeneous topical parts Use vertical listings of points Label correctly to avoid problems PRESENTATION OF STATISTICS There are four basic ways to present statistical data: 1) A text paragraph 2) A semitabular form 3) Tables 4) Graphics Text Presentation This is probably the most common method of presentation when there are only a few statistics The drawback is that the statistics are submerged in the text and require the reader to scan the entire paragraph to extract the meaning Tabular Presentation Tables are generally superior to text for presenting statistics, although they should be accompanied by comments directing the reader’s attention to important features Tables may be either general (tend to be large, complex, and detailed) or summary in nature (only a few key pieces of data closely related to specific findings) Any table should contain enough information for the reader to understand its contents Graphic Presentation Compared with tables, graphs show less information and often only approximate values However, they are more often read and remembered than tables See Exhibit 20-6 for a summary of the most common forms of graphic presentation formats Line Graphs Line graphs are a statistical presentation technique used for time series and frequency distributions over time Area (Stratum or Surface) Charts An area chart (consisting of a line that has been divided into component parts) may be used for a time series (see Exhibit 20-9) Pie Charts Pie charts are a graphical presentation using sections of a circle to represent 100 percent of a frequency distribution (see Exhibit 20-9) They are often used with business data Research shows readers’ perceptions of the percentages represented by pie slices are consistently inaccurate See text section for ideas on how to improve comprehension and perception 116 BUS519 – Business Research Methods Study Notes Bar Charts Bar charts are graphical presentation technique that represents frequency data as horizontal or vertical bars A computer charting program (e.g., Excel or the newest version of SPSS) easily generates charts Bar charts come in a variety of patterns (see associated exhibits; note designated exhibits from previous chapters that might also be helpful with explanations) Pictographs and Geographs These graphics are used in popular magazines and newspapers because they are eye-catching and imaginative A pictograph is a bar chart using pictorial symbols rather than bars to represent frequency data (see the PicProfile on the Ohio Lottery) Geographic charts use a map to show regional variations in data Stacked data sets produce variables of interest that can be aligned on a common geographic referent See Chapter 16 for an example of mapped data 3-D Graphics Virtually all charts can now be made three-dimensional 3-D graphic is a presentation technique that permits a graphical comparison of three or more variables Surface charts and 3-D scatter charts are helpful for displaying complex data patterns if the underlying distributions are multivariate ORAL PRESENTATIONS Researchers often present their findings orally A briefing is a short presentation to a small group where statistics constitute much of the content An oral presentation normally lasts between 20 minutes and one hour The presentation is normally followed by questions and discussion Preparation A successful briefing typically requires condensing a lengthy and complex body of information Speaking rates should not exceed 100 to 150 words per minute In preparing the presentation answer the following questions: How long should you talk? What is the purpose of the briefing? Major parts of the presentation include: Opening—a brief statement, probably not more than 10 percent of the allotted time, sets the stage for the body of the report Findings and conclusions— the conclusions may be stated immediately after the opening remarks, with each conclusion followed by the findings that support it Recommendations—where appropriate, these are stated in the third stage Each recommendation may be followed by references to the conclusions leading to it At this stage you must decisions about the use or non-use of audiovisuals Be sure to practice in advance with any AV equipment Type of presentation forms include: Memorization is risky and time-consuming Reading is not advisable and usually boring Extemporaneous presentation—a conversational-style oral presentation made from minimal notes A plus is that it is audience-centered This is the best choice for most presentations Some speakers use note cards to assist in this presentation 117 BUS519 – Business Research Methods Study Notes Delivery The delivery is very important in a presentation A polished delivery adds to the receptiveness of the audience Sometimes the delivery can overpower the message Speed of speech, clarify of enunciation, pauses, and gestures all play their part Voice pitch, tone quality, and inflections are proper subjects for concern Speaker Problems Inexperienced speakers may have difficulties in making presentations Areas to watch include: Vocal characteristics and Physical characteristics Audiovisuals The choice of visual aids is determined by your intended purpose, the size of the audience, meeting room conditions, time and budget constraints, and available equipment Visual aids help the speaker to clarify major points The continuity and memorization ability of the speaker’s message is improved with the use of visual aids There are two major groupings of audiovisual aids: low tech and high tech Low Tech Usually used for small audiences, less formal situations Types: Chalkboards and whiteboards, Handout materials and Flip charts, Overhead transparencies and Slides High Tech Usually used for large audiences, more formal settings, more complex information, and when a large group of presenters is involved Types: Computer-drawn visuals and Computer animation 118 ...BUS519 – Business Research Methods Study Notes Required Materials Business Research Methods (Tenth Edition, 2008), by Donald R Cooper and Pamela... applied and pure research is problem-solving based, but applied research is directed much more to making immediate managerial decisions Is BUS519 – Business Research Methods Study Notes research always... of business research and BIS Business research contributes significantly to the design and selection of tactics Tactics are specific, timed activities that execute a strategy The purposes of business