response to classmates paper on models using dummy varaibles
Respond to at least one of your colleagues’ posts and provide a constructive comment on their assessment of diagnostics.
- Were all assumptions tested for?
- Are there some violations that the model might be robust against? Why or why not?
- Explain and provide any additional resources (i.e., web links, articles, etc.) to provide your colleague with addressing diagnostic issues.
Numerical discussion and analysis
Jason Evans RE: Discussion – Week 10 Collapse Top of Form Total views: 13 (Your views: 2) Using the data from the SPSS General Study Set (Wagner, 2016), I decided to formulate my research question from the socioeconomic index and sexual orientation variables. From these variables, I formed this research question: Can one’s sexual orientation predicate an impact on one’s socioeconomic status? Of course, my null hypothesis is that sexual orientation has no impact whatsoever on one’s socioeconomic status. As stated, the variables chosen were socioeconomic index, which is measured on a scale and is our dependent variable. Our independent variable is sexual orientation which is measured nominally and has been broken down into 3 dummy variables (gay, bisexual, and unsure). For the exercise, we are utilizing a multiple regression to analyze our data and determine our results. Looking at the model summary, we can see that r = .045 shows that there is a small, positive linear relationship (Frankfort-Nachmias & Leon-Guerrero, 2015) between sexual orientation and socioeconomic status, but given our significant value of .222 is greater than the standard p = .05 and which tells us that we cannot reject our null hypothesis and that our data shows there is a significant relationship between sexual orientation and socioeconomic status. Looking at our coefficients, we can see determine the statistical significance, compared with our reference category, for the difference in means for our dummy variables of 2.283 (Gay), -5.576 (Bisexual), and -4.690 (Unsure) (Laureate Education, 2016). Using the collinearity statistic and VIF, we can see that while there is quite a bit of variance in the means of each variable, the fact that they are all below a value of 10.0 would suggest a higher level of correlation between each variable (Laureate Education, 2016m). After analyzing the data and utilizing the Durbin-Watson and Cook’s distance statistics, we can see that the 1.733 Durbin-Watson statistic falls within the 1.0-3.0 range allowing us to assume that our data is free of independent error and using Cook’s distance, all the data follows the rule of being below 1 and we can assume that there is no undue influence in our model. We can conclude that our model has met all assumptions and we do not have a significant deviation from normality (Laureate Education, 2016). In lay terms, we can see that there is a small, but significant relationship between socioeconomic status and sexual orientation. Since this data was pulled from 2010, we should not assume that it is still relevant, however we can use this data as a baseline to reconduct the study and analyze how this has changed and whether it has been better or worse. References Frankfort-Nachmias, C., & Leon-Guerrero, A. (2015). Social statistics for a diverse society (7th ed.). Thousand Oaks, CA: Sage Publications. Laureate Education (Producer). (2016). Dummy variables [Video file]. Baltimore, MD: Author. Laureate Education (Producer). (2016m). Regression diagnostics and model evaluation [Video file]. Baltimore, MD: Author. Wagner, W. E. (2016). Using IBM® SPSS® statistics for research methods and social science statistics (6th ed.). Thousand Oaks, CA: Sage Publications. SPSS Statistical Analysis.docx (15.664 KB)