Question

I am doing a statistics research paper and it is asking Data Analysis. Explain the form of analysis that you used to manipulate your data. Consider the following questions in your explanation. From the point that you had the raw data, what test did you run? Why did you select that test? Why did you choose to manipulate the data? How did you manipulate the data? What results did you desire? My variables are highest degree, and quality of life

Answer 1

First of all, almost all quantitative research needs statistical data analysis, and does not manipulate your data. Actually statistical analysis supports your subjective judgement or validate your conjecture for possible acceptance or rejection based on the available data. Now the following discussion is based on your study variables highest degree and quality of life.

Here both the variables are categorical variables. Highest degree is an ordinal variable and quality of life is mostly an ordinal variable but may be represented by a nominal variable also. Here the research question should be to study the dependence or association between these two variables. They might be dependent or independent, or a subjective knowledge might tell you that their would be positive association. Now the data available on this occasion needs to be analyzed to check the validity of the subjective knowledge or to study the nature of association.

To analyze the data, first we need to concise the data from raw data. Suppose, highest degree has P categories and quality of life has q categories. Then all data points can be arranged in a pxq table. This is known as contingency table.

Here to study the association, or validate the conjecture based on subjective knowledge, we need to use test of Independence using Pearsonian chi- square test. This test is appropriate as the variables are categorical and the entire data can be represented in a contingency table.

Based on the p-value of this test we will decide whether those variables are associated or independent. It can be said that the desirable result should be associated, but that may not be found in the present data. And we use statistical analysis to validate that only which is supported by the available data.