topic Re: Dataset within a variable? in SAS Programming

Dataset within a variable?

SAS93 — Sat, 14 Dec 2019 11:09:51 GMT

I’m getting ready to start helping with a report for my (brand) new job. The data we’re working with is the ACEs module from BRFSS, but the data is apparently stored within a single variable called SEQNO (“sequence number” from the BRFSS codebook) as a string of letters. This string has to be teased apart into its individual questions (I think there are 11 total).

I’m trying to get an initial understanding of what this means exactly & how to go about tackling it. I’m decent with SAS, no expert, but I’m capable of learning.

Re: Dataset within a variable?

PaigeMiller — Sat, 14 Dec 2019 11:59:16 GMT

Give us a few examples of this string.

Re: Dataset within a variable?

VDD — Sat, 14 Dec 2019 12:02:35 GMT

@SAS93 congratulations on the new job. First thing I would ask my supervisor is where is the cookbook of rules dealing with ACEs.

does the record contain a pic code that defines how the string used for each record?.

Re: Dataset within a variable?

ed_sas_member — Sat, 14 Dec 2019 16:02:30 GMT

Hi @SAS93

The SEQNO seems to be only an identification number.

In the BRFSS documentation, Adverse Childhood Experience is referenced as module 22. Responses to the 11 questions are respectively stored in variables ACEDEPRS, ACEDRINK, ACEDRUGS, ACEPRISN, ACEDIVRC, ACEPUNCH, ACEHURT1, ACESWEAR, ACETOUCH, ACETTHEM and ACEHVSEX (https://www.cdc.gov/brfss/questionnaires/pdf-ques/2019-BRFSS-Questionnaire-508.pdf)

However, all modules are not used every single year and in every single state (eg. https://www.cdc.gov/brfss/questionnaires/modules/category2018.htm)

What dataset do you have?

Re: Dataset within a variable?

ballardw — Mon, 16 Dec 2019 21:20:04 GMT

I've never used the ACES module but in working with BRFSS data for over 20 years now I have never seen any question that had a before such as you are supposing. The fun part is tracking down the variable names for the questions of interest.

SEQNO is just that, a sequence number to identify records.

I suggest looking at the Labels in your data set or find the documentation used to create your local data set from the CDC supplied version to identify the variables holding the responses to your questions of interest.

The variables whose names start with an underscore are generally calculated variables such as _BMI5 which uses questions about height and weight to calculate Body Mass Index values and then create the categorical variable _BMI5CAT to assign the Underweight/normal/overweight/obese categories.