8210 Week 10 Dis

2

Respond to at least one of your colleagues’ posts in 100 words and provide a constructive comment on their assessment of diagnostics.

1. Were all assumptions tested for?

2. Are there some violations that the model might be robust against? Why or why not?

3. Explain and provide any additional resources (i.e., web links, articles, etc.) to provide your colleague with addressing diagnostic issues.

Terell Johnson

Using the General Social Survey (2014) dataset, two independent variables (highest year of school completed – respondent income in constant dollars) and a dependent variable (socioeconomic index) was chosen to analyze a multiple regression. In addition, a dummy categorical variable (male) was chosen as for recoding practice and interpretation.

RQ: To what extent does highest year of school completed, respondent income in constant dollars, and male (sex) predict socioeconomic index?

H0: Highest year of school completed, respondent income in constant dollars, and male (sex) do not predict socioeconomic index?

Ha: Highest year of school completed, respondent income in constant dollars, and male (sex) do predict socioeconomic index?

As displayed in the model summary (Figure 1), Pearson R is moderately strong at .643 with 41% SES explained. According to Walden University (2016m), Durbin-Watson statistic values range from 0 to 4.0 with values below 1.0 and above 3.0 being considered dangerous and indicates the model has serious serial correlation. For this analysis, the Durbin-Watson statistic has a value of 1.9 and indicates no correlation between the residuals. The ANOVA (Figure 2) test for overall significance of the regression model is .001 and below the threshold of .05 or p < .05, indicating the model has statistical significance and the R Square can be interpreted. The coefficients summary (Figure 3) displays the variance inflation faction (VIF) between 1.0 and 1.2 for the predictor variables, which are well below the general rule of 10 and meets assumption that multicollinearity is not an issue. The Cook’s Distance (Figure 4) values range from a minimum of .000 to a maximum of .085, which is below the genal rule of 1.0 and indicates no undue influence. The histogram (Figure 4) displays normal distribution of standardized residuals, and the scatterplot (Figure 5) indicates no pattern (funnel or cone shape) reflecting homoscedasticity or a linear relationship. Based on this analysis, the null hypothesis is rejected.

Reference

Walden University, LLC. (Producer). (2016m). Regression diagnostics and model evaluation [Video file]. Baltimore, MD: Author.

Check attachment for Module pictures