Add Health data is one of the most widely utilized datasets in public health. Add Health is the largest, most comprehensive, nationally-representative longitudinal survey of adolescents ever undertaken. Beginning with an in-school questionnaire administered to a nationally representative sample of students in grades 7-12, the study followed up with a series of in-home interviews conducted in 1995, 1996, 2001-02, 2008, and 2016-18. As a result, there are 5 waves of publicly available data that are can be downloaded and analyzed.

To access the data, click here
Download the Wave 1 dataset by clicking on 'Download Data.' This link will take you to the ICPSR website. From there, you will need to enter some information. You can create an account or log in with your google account. You also need to provide a brief description of the project and the name of the Dean, as provided below.

Make sure you download the SPSS datasets.
When you unzip the file you get a folder with this structure and you wonder where in the world the data are located

The first thing to do is take a look at the documentation data guide datasets/AddHealth/ICPSR_21600-V25/ICPSR_21600/21600-Documentation-dataguide.pdf
Locate the folders containing ==Wave I: In-Home Questionnaire, Public Use Sample <== and Wave III: In-Home Questionnaire, Public Use Sample <
The in-home questionnaire contains the survey administered to the respondents

You can see that from the dataguide that we need to use the data located in DS0001 and DS0008
We are going to learn how to clean, merge and analyze data using multiple waves of the Add Health survey.

We don't have time to go into details about how to formulate a research question. However, you should be able to come up with many ideas by looking across waves.
From yesterday, I have come to understand that there is an interest in exploring the relationship between status as veteran and criminal justice system involvement. Let's see if we can use the Add Health data to inform this question.

• I have no idea what variables are in the survey, so I go through the codebook. Usually this is a tedious task but the codebook for Add Health is searchable online. I searched the word "military" and came up with the following
  Codebook Explorer – Add Health (unc.edu)
  Here are some potential questions we can use
  

• Let's use the variable "Have you ever served in the military?" Click on the link to get the variable name
  

NOTE: you must understand the data. Wave 1 and II were asked of children (~15/16) so we would not anticipate finding a question about military involvement in Waves I or II. We would expect to find this question in a later wave, as shown above.
Let's see if there is any criminal justice system involvement in Wave III we can use.

I searched the word "arrest") and here are the results.

• Let's use the variable "Have you ever been arrested or taken into custody by the police?"

• Click on the link to see the variable name
  

• We will now do the following

  • filter the data to make it more manageable
    - From wave I take AID and BIO_SEX
    - From Wave II take AID, H3CJ3, H3LM38, H3MA1, H3MA2, H3MA3 and H3MA4 (the child abuse and neglect variables)
    

  • merge the data (note, you should expect some data loss because not everyone who took the survey in Wave I also took the survey in Wave III)

    • review the types of merges
  • clean the data

  • recode the data

    • recode the child maltreatment variables so that 0 = never happened and 1 = happened one or more times -
  • create a composite variable using the data

    • using the original variables, create an index representing severity of child maltreatment
  • make a table comparing veterans and non-veterans on the arrest variable
    

  • compare the mean of child abuse severity by arrest status