A total of 3141 variables were measured during the NSDUH 2013 and are included in the dataset, so my recommendation could be a lot more comprehensive, but I would like to focus specifically on the student population. That is why my recommendation is to use student status (full- and part-time) and ethnicity as independent variables and explore their relationship with alcohol abuse.
- Tobacco
- Alcohol
- Marijuana
- Crack
- Cocaine
- Heroin
- Hallucinogens
- Inhalants
- Pain relievers
- Tranquilizers
- Stimulants
- Sedatives
Several measurements were taken for each of these variables, including: ever used, age of first use, time since last used, use in past 12 months, and use in past 30 days. For some variables like alcohol or cigarettes, the survey also measured estimated number of days used within the past 30 days and number of units used per day in the past 30 days.
Numerous independent variables were measured during the survey, including demographics (state, sex, age, ethnicity, income, household composition, military status, marital status, etc.), participant background (pregnancy, criminal behavior, mental health, mental health treatment, youth experiences, active military duty, military combat experiences, etc.), social environment, religious beliefs, insurance, overall health, employment, workplace, and various others.
Some additional options for the analysis include the following:
Comparing alcohol use in students to the rest of the population. (e.g., participants with 4+ college education were more likely to report drinking in the past 30 days and ever drinking alcohol than participants with 1-3 years of college or less)
Comparing alcohol use among states. (Students make up only 4-5 thousand participants, so it is possible that some states are over- or under-represented)
Comparing alcohol use between employed and unemployed students.
Comparing alcohol use among students based on their household composition and household income levels.
Comparing substance use among different age groups to determine which substances those groups prefer.
However, given the comprehensive nature of this survey, I would recommend focusing on the student population to make the proposal specific and avoid diluting the analysis.
Thanks again for your continuous support!
