Population Assessment of Tobacco and Health

alcohol and drug use problems questions from GAIN-SS Tips for Using the PATH Study Data User Forum

Return to the Population Assessment of Tobacco and Health (PATH) Study Series page.

4 posts / 0 new
Last post
bjevans
alcohol and drug use problems questions from GAIN-SS

Hi,

I am having trouble with understanding a skip pattern in wave 1 adult data (though it appears there is the same issue at other waves too) for which I'm hoping someone can provide some insight. There are 7 variables from the GAIN-SS that assess alcohol or other drug problems in the dataset:

R01_AX0170, R01_AX0171, R01_AX0193, R01_AX0172, R01_AX0173, R01_AX0174, R01_AX0194

For those who report lifetime use of 2 or more substances and report an alcohol/drug problem they are asked which substance(s) caused these problems (R01_AX0176_01- R01_AX0176_07). However in the codebook and in looking at the data, it looks like the item R01_AX0171 was not factored into the skip logic such that those who reported ONLY that alcohol or drug problem (R01_AX0171) are NOT asked which substance caused the problem. When I do crosstabs it looks like this is confirmed that those who report only the problem of R01_AX0171 are coded as "Inapplicable" [-1]. Is this correct? and is there an explanation for why these individuals were not asked this question? This is an important variable in my analysis so I want to make sure I code it correctly.

As a side note I also see that R01_AX0170 is not included in this skip logic but individuals are asked in a separate question what substance they are using weekly (R01_AX0175_01-R01_AX0175_07). 

Thanks in advance for any insight or suggestions!

-Rebecca

 

Joy Jang
 

 

Hi Rebecca,

Among the 7 variables that assess alcohol or other drug problems in the dataset (R01_AX0170, R01_AX0171, R01_AX0193, R01_AX0172, R01_AX0173, R01_AX0174, R01_AX0194), I understand "R01_AX0171" only asked about the time that a respondent spent a lot for "getting" alcohol or other drugs, not about time spent for "using" alcohol or drugs like other 6 variables. (* R01_AX0171 was asked to respondent who report ever using alcohol or ever using drugs --> we don't know if the respondents ever had any problems with the use from R01_AX0171). 

 

I understand that if respondents only reported a problem with "getting" alcohol or drugs (R01_AX0171), they would not have been asked about what substances caused alcohol/drug problems (R01_AX0176_01- R01_AX0176_07). Thus, I believe your code is correct -- those who report only the problem of R01_AX0171 should be coded as "Inapplicable" to R01_AX0176_01- R01_AX0176_07.

 

Hope this helps! Thank you! 

- NAHDAP

 

 

bjevans
Thanks for that point and

Thanks for that point and that is an interesting distinction. It is strange though because that in papers like this one https://doi-org.proxy.lib.umich.edu/10.1016/j.drugalcdep.2017.03.032 it seems to be treated as a problem equal to the other 6 items. The substance use problem scale is created by summing all seven items.

But I appreciate the clarity and confirmation that individuals who reported only this item were not asked this follow up question about what substance these problems were associated with.

Joy Jang
 

 

That is an interesting point! Thank you for the link to the paper. I hope other users chime in. Would like to hear from others too.

In the meantime, you may want to contact PATHDataUserQuestions@Westat.com for further clarifications. Although the email address is not for questions about statistical analysis or analytic guidance, they might have some insights on skip patterns. 

 

Thank you again!

- NAHDAP 

Log in or register to post comments