Directions:
Run the appropriate regression (linear regression or logistic regression) using your main explanatory and response variable and additional covariates of interest. Describe your findings.
Optional (encouraged): Construct a linear regression or logistic regression plot to accompany your model. The code for this can be found in graphing resources.
- Example of how to write results for Multiple Linear Regression: Remember use this model when your response variable is quantitative. Suppose that our research question is: Does major depression significantly relate to number of nicotine dependence symptoms after controlling for age, socioeconomic status, and biological sex?
- Major depression (Beta=1.34, CI 1.29-1.39, p=.0001) was significantly and positively associated with number of nicotine dependence symptoms after controlling for age, socioeconomic status, and biological sex. On average, someone with major depression is expected to have 1.34 additional symptoms more than someone without major depression holding all other variables fixed.
- You can also discuss any findings from interesting covariates. It may become burdensome to discuss them all. Perhaps:
- Age (Beta=0.12, CI: 0.08, 0.16, p<0.001) is significantly and positively associated with number of nicotine dependence symptoms in the model stated above. In particular, the older a participant is the higher number of expected symptoms are expected holding all other variables fixed.
- Example of how to write results for Multiple Logistic Regression: Remember use this model when your response variable is binary categorical. Suppose that our research question is: Does major depression significantly relate to whether someone has nicotine dependence after controlling for age, socioeconomic status, and biological sex? [Notice that in this version – the response variable is binary categorical – the participant either has nicotine dependence (1) or does not (0)].
- Major depression (O.R. 3.2, CI 1.9, 5.6) is significantly associated with the likelihood of meeting criteria for nicotine dependence after controlling for age, socioeconomic status, and biological sex. Those with major depression have an expected odds of nicotine dependence that is 3.2 times higher than those without major depression holding all other variables fixed.
- You can also discuss any findings from interesting covariates. It may become burdensome to discuss them all. Perhaps:
- Age (O.R. 1.1, CI: 1.05, 1.2, p<0.001) is significantly and positively associated with odds of nicotine dependence in the model stated above. In particular, the odds of nicotine dependence is expected to increase by a factor of 1.1 for each additional year older a participant is holding all other variables fixed.