case report

CASE ANALYSIS PROCESS

These cases are designed to create a problem-solving environment that closely matches the environment experienced in an actual business. Case solutions require significant statistical analysis and interpretation of results. The focus is on creating a concise professional report that responds to the original business or public policy question. In contrast, textbook problems often require only a set of statistical analyses and computations. To simulate an actual business setting, you should use software to perform the statistical computations and a text editor to prepare a professional-looking report. These cases are perfect for team work. Please try to keep everyone involved in discussions. Working in teams permits a broader examination of the case questions and fosters useful discussion. In addition, most real-life business problems are solved by teams. Thus students gain practical experience by working with others on a common problem.

At the beginning of the semester, I will give a series of questions designed to help guide your analysis. The ultimate goal of your analysis, though, is a formal written report that helps solve the original problem. After you get comfortable with the cases, I will stop giving specific questions and let you work on the business problem.

Statistical studies follow a general structure, with a number of side diversions resulting from discoveries made along the way. The process usually consists of the following steps:

1. Identifying the business analysis questions

2. Converting business questions into statistical questions

3. Performing a descriptive analysis of the data

4. Applying formal analysis procedures

5. Developing conclusions and recommendations

6. Identifying unusual outcomes and future analysis

Each of these steps should be accompanied by written descriptions in the final report. After all of the analysis and blocks of writing are completed, you may wish to move blocks of text and statistical output around to produce a more coherent report. The final step is then to prepare the Executive Summary, which is the very first section of your report.

Each step entails a wide range of options.

1. IDENTIFYING THE ANALYSIS QUESTIONS

Your first task is to read the entire case carefully and make notes concerning key details. From this reading, you can begin to identify the important business questions in the case. In a real business application, this process consists of talking with key managers and other knowledgeable persons. That process usually leads to an analysis proposal, which is then "signed off' by the executive responsible for underwriting the study.

2. CONVERTING ANALYSIS QUESTIONS INTO STATISTICAL QUESTIONS

Statistical analysis can answer specific well-structured questions. You have seen many examples of these questions in textbook problems and examples. An important task in any applied study is to convert the case analysis questions into statistical questions. This process requires thorough understanding of the problem requirements and of the available data. In many cases the available data, including the data form and source, limit the kinds of analysis that are possible.

3. PERFORMING A DESCRIPTIVE ANALYSIS OF THE DATA

Performing a descriptive analysis provides the analyst with a general picture of the data. Means, variances, ranges, and simple correlations are computed for appropriate variables. The analyst may also prepare histograms, box-and-whisker plots, and other graphical displays. If the data have been collected over time, creating time plots will help identify patterns that change over time. From this analysis, you can define the range of applicability for your analysis conclusions.

4. APPLYING FORMAL ANALYSIS PROCEDURES

By this point in the study, you will have developed a good understanding of the problem issues and the basic structure of the data. In addition, if you have been writing intermediate results, you will probably have additional questions. These can be compared against the assumptions required for the various standard statistical procedures. Do you have everything needed for a formal hypothesis test? Do the data satisfy the assumptions required for regression analysis? Which of the available analyses will answer the questions you have formulated?

The analysis should be driven by the questions and not by a desire to demonstrate how clever you can be with various statistical tools. You should be willing to try a variety of procedures to test the stability of the conclusions. For example, means, standard deviations, and other descriptive statistics should show the same results as bar charts and graphs; and regression models of different forms should provide similar conclusions. Analysis of residuals and analysis of outliers in general evince a careful analyst. Study of the distribution patterns for important variables indicates the validity of your analysis. Many strange patterns can occur in data, and you should try the analysis in enough different ways to ensure reliability of the results. Incidentally, the ability to support different analyses is a major advantage of the computer analysis environment.

This is also the time to perform some basic reality checks on your conclusions. Do the results make sense, given what you know about the area being studied? You can also ask other persons with experience in problems of this type to react to your initial conclusions.

Computer-based problem solving can create undeserved support for certain answers, if there were initial errors in the analysis. Older, more tedious non-computer-based analysis procedures were closer to the data and the detailed steps. Thus, errors in computation or analysis approach were more likely to be detected. In our modem computer-based statistical analysis environment, results must be examined for reasonableness and reality.

5. DEVELOPING CONCLUSIONS AND RECOMMENDATIONS

At this point you should review the initial objectives and questions to confirm that all of the required analyses have been completed. If you have been writing after completing each analysis step, you already have the basis for discussion of your major conclusions. Key ideas should be extracted and combined for a smooth-flowing discussion. Some of the more basic details can be moved to an appendix. This is also the time to write the Executive Summary.

Ideally this step involves bringing together your results and then writing a clear discussion that links them. However, real studies do not always proceed linearly (such that the previous step is completed before the next step is begun). Thus you may identify additional questions when you are preparing your conclusions. In that case you must cycle back to perform additional analysis.

6. IDENTIFYING UNUSUAL OUTCOMES AND FUTURE ANALYSIS

Research, analysis, and problem solving are an ongoing process. Answers at one point identify additional questions. Analysis often indicates additional paths that should be followed. However, studies in a business or public policy environment must reach a close. Decision makers require timely answers, and limited resources are allocated to each study. Personnel and financial resources are allocated to a study, and results must be produced within resource limits. As a student, you have only so many hours to devote to each course and/or to special projects. Thus completion of a study is important. Closure typically includes a final written report and/or an oral presentation.

Given the need for closure, you will often be left with unexplored questions in your study. These should be documented and noted as questions for future work. In some cases, the questions might be important enough to justify another study. Your task is to document the additional questions objectively and to suggest how they might be answered. Information users must then decide whether additional work should be undertaken.

PREPARING THE WRITTEN REPORT

Since producing a written report is your ultimate objective, you would be wise to begin writing your report after you have identified the analysis questions but simultaneously with performing the statistical calculations. For example, start the case report by preparing the introduction and the project scope. Initial statistical analysis can then be transferred to your emerging report. Next, prepare comments on the initial statistical work, and move back to perform more statistical computations. This strategy has the advantage of encouraging you to keep your results well organized as you proceed. More importantly, the process of writing forces you to define specific questions that you can then answer by using statistical analysis. And good writing almost always requires several revisions.

For these cases, your report will typically be limited to two (or three pages only if necessary) with appendices that present specific analyses and data displays, again, if necessary. Integrating the charts and tables to your report always makes it an easier read for the audience. The report should begin with an Executive Summary of one to three paragraphs. This summary-which is the last item written-should identify the problem, indicate your approach to solving it, and concisely state your conclusion.

The body of your report should indicate how you developed your conclusions and recommendations. Begin with a concise presentation of the question from the business perspective and explain how you conducted the analysis. Define the data set and specify the statistical procedures you used. Include specific statements of statistical models and hypothesis tests, and outline the results. Discuss the statistical results, indicating how they provide a solution to the case problem. Include additional observations and extensions of your results, as appropriate, given the case objective.

Business writing tends to be short and concise. Managers and executives seek results and conclusions. They need to understand your analysis and conclusions. In addition, any limitations and alternatives need to be clearly expressed. Managers do not have time for extensive reading of analysis details. In many instances, the Executive Summary is the only part of a report that is read. Thus you should prepare the Executive Summary with great care. The results of all of your hard work may be revealed or hidden depending on the quality of the Executive Summary.

Persons who seek more detailed information will read the entire report. The report should describe your analysis and present conclusions together with supporting evidence. Well-designed graphs and figures greatly enhance communication. Mathematical equations and data tables provide strong reinforcement for readers who desire rigorous and complete understanding. However, you need to avoid both intimidation and boredom. Analysis details, including detailed statistical computer outputs and supporting graphs, should be in an appendix. These reports are a series of layers that supply further depth as more accessible layers are peeled away. The first layers, however, must provide the most important results and key conclusions. Think about this design for your report as you prepare case solutions.

CONCLUSION

Work on the cases that you will be assigned can provide a valuable capstone to your study of applied statistics. By developing solutions to applied problems you will think carefully about the assumptions behind various statistical procedures.