The book‘s editor, Pamela Schindler, argue that the business environment has changed faster than the pedagogy. It’s never been more important for students of business to learn and master the tools and processes of business research, she writes.
What has shifted is the fundamental nature of the managerial dilemma. Companies no longer face predictable challenges that can be addressed with standard templates. They face what former Intel CEO Andrew Grove famously called strategic inflection points – moments when the fundamentals of a business are about to change. These disruptions can emerge from unexpected directions. Grocery chains find themselves competing with Amazon. CBS and NBC find themselves competing with Netflix. Managers increasingly find themselves blindsided when having to read and interpret the signals of change in an unfamiliar environment.
Research, in this new environment, cannot also be same as before.
The Data Paradox
The book explains that businesses are collecting more data than ever before. It lists a range of sources: transaction data from online and in-store purchases, observational data from web visits and biometric measures, conversational data from call centers and social media posts, and internet analytics from keyword searches and click analysis. Yet despite this abundance, managers have historically struggled to use data effectively.
Schindler identifies three key obstacles. The first is the state of digital transition. Until recently, most data was collected manually and stored on paper, making it cumbersome, tedious, time-consuming, and costly to reorganize or aggregate. Even today, many small businesses have not fully migrated to digital collection, putting them at a competitive disadvantage.
The second obstacle is what researchers call data silos. Businesses often collect and store data at the business-unit level, sharing it only on a need-to-know basis. Forrester Research estimates that between 60 and 73 percent of a company’s data is inaccessible for analysis. The result is duplication, inconsistency, and an inability to see the bigger picture.
The third obstacle is perhaps the most profound: a severe shortage of people with the skills to interpret the expanding ocean of data. The competition for individuals who can analyze data and deliver insights is fierce. KPMG’s most recent CEO Outlook Survey found that a majority of CEOs plan to upskill between 40 and 60 percent of their employees in the next decade, recognizing that a “skills revolution” is needed.
The Language of Research
Schindler devotes a chapter to what she calls the language of research. Considering the distinction between data, information, and insights. Data, she writes, are raw, unprocessed facts in the form of numbers, text, pictures, audio or video.
Information is what we get when we process data by accumulating and organizing it; applying rules of measurement, statistical summary, and testing; and developing data presentations that make its true meaning clear.
Insights are what we generate when we analyze information in the context of a management dilemma and the business’s mission, goals, resources, and culture.
The difference is not just semantic. It is operational. It determines whether research leads to action or simply to more reports.
Considering the concept of customer lifetime value, which the book explores through the work of Wharton professor Peter Fader. Fader argues that businesses are tracking the wrong metrics. They still focus on the average customer, giving that imaginary customer everything he or she wants. But there is no such thing as an average customer. Some customers are far more valuable than others. Businesses should invest in research that identifies those customers who generate the greatest lifetime value and focus all strategic and tactical decisions to attract and keep them.
— Pamela Schindler
Data are raw, unprocessed facts in the form of numbers, text, pictures, audio, or video collected by either quantitative or qualitative means.
Information is processed data. We process data by accumulating and organizing it; applying rules of measurement, statistical summary, and testing; and developing data presentations that make its true meaning clear.
Insights are generated by analyzing information in the context of a management dilemma and the business’s mission, goals, resources, and culture. Insights determine whether the hypotheses and theories are supported or refuted.
Key insight: Data alone is insufficient. The value of research lies in its ability to generate actionable insights—conclusions that lead to recommendations for specific decisions and are aligned with key business goals.
Management dilemma: The problem or opportunity that requires a decision—a symptom of an actual problem, such as rising costs, declining sales, or increasing employee turnover.
Management question: A restatement of the manager’s dilemma(s) in question form. Example: “What can we do to increase profits?”
Research question(s): Questions that explore the various options and form one or more hypotheses that best state the objective of the research. Example: “Should the bank position itself as a modern, progressive institution or maintain its image as the oldest, most reliable institution in town?”
Investigative questions: Questions the researcher must answer to satisfactorily answer the research question—what the manager feels he or she needs to know. Example: “What is the public’s position regarding financial services and their use?”
Measurement questions: The actual questions that researchers use to collect data in a study.
Evaluation of solutions: “How can we achieve the objectives we have set?”
Choice of purpose: “What do we want to achieve?”
Troubleshooting: “Why is our program not meeting its goals?”
Control: “How well is our program meeting its goals?”
Exploration is loosely structured research with a defined purpose. Through exploration, researchers:
- Understand the management dilemma and research question
- Discover how others have addressed similar problems
- Establish priorities for dilemmas to be addressed
- Identify action options and develop hypotheses
- Gather background information to refine the research question
- Develop operational definitions for concepts and variables
Research design is a time-based, procedural plan for every research activity. It is always focused on the research question. It guides the selection of sources of information (cases) and provides the framework for specifying the relationships among the study’s variables.
The tasks covered by research design include: sampling design, data collection design, and the measurement instrument(s).
Objective of the study: Reporting, descriptive, causal-explanatory, causal-predictive.
Researcher’s ability to manipulate: Experimental vs. ex post facto.
Topical scope: Statistical study vs. case study.
Measurement emphasis: Qualitative vs. quantitative.
Method of data collection: Monitoring vs. communication study.
Research environment: Field vs. laboratory.
Time dimension: Cross-sectional vs. longitudinal.
Probability sampling is based on random selection—a controlled procedure that assures that each case is given a known nonzero chance of selection. Only probability samples provide estimates of precision.
Nonprobability sampling is arbitrary and subjective. Each member of the target population does not have a known nonzero chance of being included. It is less expensive and often used in exploratory research.
Key consideration: When a researcher is making a decision that will influence the expenditure of thousands, if not millions, of dollars, an estimate of precision is critical.
Measurement in research consists of assigning numbers to empirical events, objects or properties, or activities in compliance with a set of rules. This definition implies that measurement has three tasks:
- Select variables to measure.
- Develop a set of mapping rules: a scheme for assigning numbers or symbols to represent aspects of the variable being measured.
- Apply the mapping rule(s) to each observation of that event.
Nominal scale: Classification but no order, distance, or natural origin. Example: Gender (male, female).
Ordinal scale: Classification and order but no equal distance or natural origin. Example: Doneness of meat (well, medium well, medium rare, rare).
Interval scale: Classification, order, and equal distance but no natural origin. Example: Temperature in degrees.
Ratio scale: Classification, order, equal distance, and natural origin. Example: Age in years.
Validity is the extent to which a chosen or developed scale actually measures what we wish to measure. Three forms:
- Content validity: Adequate coverage of the investigative questions.
- Criterion-related validity: Success of measures used for prediction or estimation.
- Construct validity: The degree to which the instrument conforms to predicted correlations.
Reliability is concerned with the degree to which a measurement is free of random or unstable error. A measure is reliable to the degree that it supplies consistent results.
Data analysis involves reducing accumulated data by developing summaries (descriptive statistics of the variables), looking for patterns by looking for relationships among variables, and applying statistical techniques.
When the researcher interprets the findings in light of the manager’s research questions, he or she develops insights. Insights determine whether the hypotheses and theories are supported or refuted.
Used when data are derived from interval and ratio measurements.
Assumptions: Observations must be independent; observations should be drawn from normally distributed populations; populations should have equal variances; measurement scales should be interval or ratio.
Examples: t-test, Z test, ANOVA, F test.
Used to test hypotheses with nominal and ordinal data. Parametric techniques are the tests of choice if their assumptions are met, as they are more powerful than nonparametric tests.
Examples: Chi-square (χ²) test, McNemar test, Kolmogorov-Smirnov test, Mann-Whitney U test, Kruskal-Wallis test.
Advantage: Nonparametric tests have fewer and less stringent assumptions.
Data-centric planning: A focus on the data or information and sharing as much data/information as was discovered. The report is more factual and statistical.
Audience-centric planning: A focus on gaining the audience’s embrace of data insights and recommendations. The report is more persuasive and tells a story where statistics is a tool, not the focus.
Introduction: Title, researcher profile, executive summary, table of contents.
Background: Problem statement, research objectives.
Methodology: Sampling design, research design, data collection, data analysis, limitations.
Findings, Insights, and Recommendations: The largest and most important section.
Appendices: Complex tables, statistical tests, supporting documents, measurement instruments, bibliography.
Ethos: The audience believes the researcher is qualified to conduct the research. Established through the researcher’s profile and credibility.
Pathos: An appeal to an audience’s sense of identity, self-interest, and emotions. Requires understanding the audience’s predispositions and the effects of recommendations.
Logos: The logical argument. Describes facts and findings that support the researcher’s insights. The core of most research presentations.
Key insight: “Without appropriate support materials, a researcher’s insights are perceived as nothing more than a series of unsupported claims. Essentially, support materials are the leaves on the branches of your organizational framework.”
The Art of Asking the Right Questions
The book is built around a single, integrated research process model. It divides the overall process into five stages: clarify the research question, design the research, collect and prepare the data, analyze and interpret the data, and report insights and recommendations. Every chapter is explicitly linked to a specific stage of this process.
But the model is not a straight line. It is a recursive, iterative process. And the most important stage is the first one: clarifying the research question.
One task of a researcher is to assist the manager in formulating a research question that fits the management dilemma,” Schindler argues. Incorrectly defining the research question is the fundamental weakness in the business research process. Time and money can be wasted studying an option that won’t help the manager rectify the original dilemma.
The Researcher as Storyteller
The book’s approach in the final stage of the research process is reporting. Schindler argues that research reporting has experienced three major shifts in the last decade: the emphasis on reporting insights rather than sharing data or information, the use of improved web-based features incorporated into oral and electronic reports, and the dominance of audience-centric planning rather than traditional data-centric planning.
The distinction between data-centric and audience-centric planning is crucial. In the old model, the report was a factual summary of statistical findings. In the new model, the report is a persuasive narrative that tells a story grounded in the data. With data-centric planning, the report is more factual and statistical, Schindler writes. With audience-centric planning, the report is more persuasive and tells a story where statistics is a tool, not the focus.
The Ethical Dimension
The book is also notable for its sustained attention to ethical issues. Ethics is not treated as an afterthought but as an integral part of the research process. At every stage, from clarifying the research question to reporting the results, ethical considerations are embedded.
The primary issues, Schindler argues, are deception, privacy, quality, notice, choice, access, security, respect, and bias. For the participant, the primary ethical responsibilities are truthfulness and completion of research tasks. For managers, they are record accuracy, purpose transparency and truthfulness, discouraging falsification of research results, and fulfilling contractual obligations. For researchers, they are obtaining informed consent, accepting the subject’s choice on whether to participate, following appropriate safety procedures, keeping participant identity confidential, protecting manager confidentiality, following acceptable industry practices and procedures, and protecting data, findings, insights, and recommendations.