e-Articles (Chapter 16) Here are the titles of some full-length research articles that illustrate concepts discussed in this chapter. To view any article on-line, simply click on its title. The Impact of Influenza Epidemics on Hospitalizations Illustrates the application of bivariate regression. Here, the independent and dependent variables were "flu-related mortality" and "flu-related hospitalizations," respectively, with data points corresponding to the 26 years of the study. Figure 2 shows the scatter diagram for the data points and the line-of-best-fit. Correlations Among Depression Rating Scales and a Self-RatingAnxiety Scale in Depressive Outpatients Illustrates the use of a stepwise multiple regression involving 6 independent variables. The final regression model contained 2 independent variables, with beta weights presented for each one. Illustrates the use of a stepwise multiple regression . The researchers present beta weights (with each tested for significance), an F-test on the full regression equation, and an R-square index to show how much variability in the dependent variable is "explained by" the combination of independent variables. (See the final two paragraphs of the "Methodology" section" and the entirety of the "Results" section.) Illustrates the use of a logistic regression, with focus directed to the full model (evaluated via chi-square tests) and to the individual independent variables (evaluated via Wald tests). See the subsections of "Results" entitled "Model Evaluation", "Assessing the Fit of the Overall Model," and Assessing the independent Variables." In this study, a stepwise logistic regression was used, with attention directed toward the crude and adjusted odds ratios. Nicely illustrates how covariates can be included in the regression model in an effort to "control" for certain variables such as sex and age. Illustrates the use of control variables in a logistic regression. Results focus on odds ratios (ORs) with 95% CIs. Experience of Non-Fatal Overdose Among Heroine Users in Adelaide, Australia Logistic regression was used in this study, with a focus on odds ratios (and accompanying CIs) and the "fit" of the regression equation. See the last section of the "Results" section (called "Predictors of Overdose") as well as Table 4. |

Copyright © 2012 Schuyler W. Huck |
| Book Info | Author Info | Site URL: www.readingstats.com |