Using the General Social Survey dataset, last week I discovered using dummy variables that union members hold higher degrees and more income compared to nonunion members. To expand on that research, I am using a Chi Squares test to examine the relationship between the two variables; respondents’ highest degree (dependent variable) and Does respondent or spouse belong to a union (independent variable). Since both variables are measured as categorical, the Chi-Square test will provide additional insights into the relationship (Frankfurt-Nachmias & Leon-Guerrero, 2018).
The results show a p-value of .007, significantly less than the threshold of .05, indicating the rejection of the null hypothesis that union members do not hold higher degrees than nonunion members. Since the Chi-Square test does not reveal the significance of the statistically significant relationship, a Cramer’s V test is used, which results in a value of .074. The Cramer V relationship is between 0 and 1.0, with 1.0 being a strong relationship; thus, in this example, the value of .074 shows a weak relationship between education levels and unionized labor. However, even though the relationship is weak, the dataset does show a statistically significant relationship and therefore, we can assume unionized members earn higher degrees than nonunion members.
Frankfort-Nachmias, C., & Leon-Guerrero, A. (2018). Social statistics for a diverse society (8th ed.). Thousand Oaks, CA: Sage Publications.
Laureate Education (Producer). (2016). Bivariate categorical tests [Video file]. Baltimore, MD: Author. span>