Identification and Description of Variables Used
In our selected data set (Hawaii Dataset ss05phi), we find that many variables are used to explain the housing dynamics in Hawaii. Among them are age, sex, citizenship, class of worker, educational level, annual wages, poverty status, age of respondents and age groups (Hawaii, 2016). Although these variables are instrumental in understanding the Hawaiian housing market, in this paper, we only focus on using three variables – income (annual wages and salary), work class and educational level to answer our research question, which is centered on understanding whether there is an association between income, working class and educational levels. The research question is instrumental in exploring our research topic, which focuses on understanding the relationship between income, working class and educational levels.
Types of Variables Used
According to Westfall and Henning (2013), understanding the types of variables applicable in research studies is an important first step to answering quantitative research questions. There are two main types of variables applicable in quantitative research. They are the independent and dependent variables. Independent variables are those that are subject to manipulation because they help to determine whether the dependent variable would change, depending on the variations created (Westfall & Henning, 2013).
Based on this definition alone, we find that the dependent variable is that which changes (or fails to change) depending on how researchers manipulate the independent variable. In the context of our analysis, income is the dependent variable, while class and educational levels are independent variables. Thus, we will be manipulating the independent variable (working class and educational level) to ascertain whether it has any effect on the dependent variable – income.
Levels of Measurement for Each of the Variables
There are different types of measures for different variables in quantitative research. The most common ones are categorical, ordinal, ratio (scale) and interval (Westfall & Henning, 2013). A categorical variable is that which allows users to assign values, but prevents them from ranking the variables in order of the highest to the lowest, or best to worst (Choosing a statistical test, 2012). An ordinal variable shares close similarities with the categorical measure, except that it allows for an ordering or ranking system. An interval is similar to an ordinal variable, except that it only allows for equal spacing between variables (Westfall & Henning, 2013).
Lastly, the ratio measuring technique is the most sophisticated one because it contains attributes of all the three measurement metrics described above. Each of the variables mentioned in this study would have a unique measurement system. Education level would have an ordinal measurement system because education is often ranked from the highest to the lowest levels. The class of worker would also be measured using an ordinal measuring system because we can rank class from the highest to the lowest. We would measure income, as a variable, using ratios because the data provided for this metric is complex and spread across a wide income pool.
Indeed, since the income data provided in the assigned data set are not uniformly spaced, we cannot use interval as a measurement technique. Similarly, we cannot use ordinal measurement to analyze this variable because the data is not ranked according to a set of measurable values. Thus, we will use the ratio method because its features accommodate those of the selected variable. For example, one criterion for using ratio is the presence of a zero value. Income (our dependent variable) could have a zero value. Therefore, our dependent variable (income) qualifies for this data measurement technique.
Statistical Tests to be used
The statistical tests to be used in this quantitative data analysis largely depend on the type of data measurement tools used. Two types of data measurement tools are described above – ordinal measurement and ratio (scale) measurement. Ordinal measurement is used in two variables – educational levels and worker class. According to Westfall and Henning (2013), ordinal measures are usually analyzed using four methods – the Mann-Whitney test, Wilcoxon test, Kruskall Wallis test, and Friedman test. We cannot use Kruskall Wallis and the Friedman tests for our data analysis because they measure three or more treatments (populations) (Choosing a statistical test, 2012).
We only have two variables to quantify using the ordinal technique. Therefore, we will measure educational levels and worker class using the other two remaining tests – Mann-Whitney test and Wilcoxon test. These tests would evaluate the treatment between two sets of variables – educational and class variables, which are our independent variables.
Regression is also another data analysis tool we could use to understand the relationship between the independent and dependent variables in our analysis. This data analysis method is appropriate for our study because it could establish the strength of relationships between the independent and dependent variables. Researchers have also used it to investigate whether a variable could be a predictor of another variable (Shi & Johnson, 2014).
This feature aligns with the nature of our study because we strive to investigate whether educational levels and worker class could predict income levels. Using the regression analysis, we could find out whether this relationship is weak or strong. Comprehensively, the regression method, Mann-Whitney test and Wilcoxon test are our selected data analysis methods.
Reference List
Choosing a statistical test. (2012). Web.
Hawaii. (2016). Web.
Shi, L., & Johnson, J. A. (Eds.). (2014). Novick & Morrow’s public health administration: Principles for population-based management (3rd ed.). Burlington, MA: Jones & Bartlett.
Westfall, P., & Henning, K. (2013). Understanding advanced statistical methods. New York, NY: CRC Press.