While measurement and data collection are typically focused on the constructs, variables, factors, and the context – the analysis is focused on the relationships between the variables, factors, contextual factors. The type and level of data that is collected along with the questions and purpose(s) will determine the data analysis options that are available. In addition, for quantitative data the level (nominal, ordinal, interval, and ratio) will also determine the specific statistical tests that are available.
We measure variables and we analyze relationships. If we have quantitative data from the data collection phase then we can use statistical analysis methods to analyze relationships between the variables. The advantage of using mathematics is that the formulas, when executed the same way each time, produce the same result (assuming no math error). This is not necessarily the case for qualitative analysis where the researcher’s brain is ultimately the analysis instrument and doesn’t follow the exact path each time it analyzes the data.
While quantitative analysis is more objective, it does not always provide a rich understanding of the details behind the numbers. For example, the correlation between employee turnover and employee satisfaction as measured by a survey might be significant at the .05 level. What does that mean? What satisfaction factors were the most important to determining whether an employee would leave or not? These are the types of questions qualitative methods are best suited to answer. Then quantitative methods can often be used to test the new insights.
Given the limitations of each method, quantitative and qualitative, the use of mixed methods has grown in popularity. Most problems or topics in organization research involve both easily measurable variables (e.g., time, money, quality) and constructs that are not so easily measurable such as culture and individual behavior.
Developing an Analysis Strategy
Developing an analysis strategy, like the rest of the research design is an iterative process. If you are doing a fixed design then a detailed analysis strategy, including specific statistical tests, can be developed prior to the research and included in the proposed research plan. If, on the other hand, you are using a flexible design it might not be possible to know in advance all the analysis techniques that might provide insight into your questions. In the case of flexible studies, the challenge is to pre-think as much as you can the analysis options then describe that in the proposal. If you are using qualitative analysis software to assist in the process then that will impact the types of analysis methods that you choose. However, flexible does mean flexible and the actual analysis methods used might be quite different than those that you predict at the writing of the proposal.
- Based on the research questions, the overall approach, and the data collected choose the analysis methods (be specific).
- Align the analysis methods with the individual research questions.
- Identify the bias, validity, and reliability issues and methods to address those issues.
- Support your discussion with solid peer-reviewed references and research methods texts.
Align and Integrate
- The data analysis methods should provide the findings necessary to answer the research questions or test the hypothesis and draw conclusions.
- The data analysis methods should be derived from and consistent with the type and level of data that is collected in the previous step.
- As with all the components of the research methodology, the data analysis methods should be appropriate for the variables, relationships, context, and so forth identified in the conceptual framework.