Abstract
This chapter takes many of the research project design steps and analysis at this point and looks at things that can (and may) go wrong and what can be done about this. The author discusses how the issue may have been mitigated with effective research design decisions ahead of time and how to recover from an issue when it is identified later. As such, this chapter provides an overview of design decisions and why they are essential. It provides a range of remedies and actions that can be taken if issues and problems are identified part-way through a project. Several examples are provided of how different scholars have resolved the issues in their published articles, showing that resolutions are possible and practical.
TopProblems And Troubleshooting
It is often difficult to isolate and separate out specific problems you may have. Consequently, it may be necessary to read through the chapter quickly and identify whether or not any of the issues are apparent with your project. In other cases, the troubleshooting problem has been reported to me in similar language terminology by some of my postgraduate students.
Key Terms in this Chapter
Heteroscedasticity: The error variance that is different or varies with some level of dispersion.
Multicollinearity: A case where several of the variables in the model are highly correlated. This may include both the control variables as well as the variables of interest and used to test the hypotheses.
Sample Size: The number of observations in the sample. In an event study, one firm may have different events impact them over time, and so while there may be many observations in the sample, there may be fewer firms.
Abnormal Return: The divergence from the expected, ‘normal’ return that we would expect to see, indicating that the event has influenced the stock returns or that there is evidence of a stock market reaction.
Sampling: the process of creating a sample for the analysis. Errors in sampling create consequential problems in the analysis.
Event List: The list of observations of an event, including the date and the firm affected.