My Research Blog

Posts to this blog present my work in Applied Data Analysis (QAC201).

[ilanaeladis]

Logistic Regression

Proc logistic; 

Class S2BQ1A27 SEX; 

model S2CQ1= S2BQ1a27 SEX; 

After adjusting for potential confounding factors (SEX, having been arrested because of alcohol abuse), seeking help for drinking was positively and significantly associated with sex/having been arrested.

[ilanaeladis]

Moderation Assignment

PROC SORT; by SEX; Proc freq; tables S2CQ1*S2BQ1A27/Chisq;

by SEX;

I saw the moderation assignment as an opportunity to explore some new variables which I have not previously incorporated into my research question. Using sex as a moderating variable, I wanted to see whether males or females were more likely to seek treatment after having an alcohol related arrest. 

The freq procedure shows that more men are likely to seek help for drinking after having been arrested for drinking (68 as opposed to only 27 women). However, this is most likely because men are more likely to be arrested than women because of drinking related activities in the first place- this is demonstrated by another chi square test for independence that measured the influence on sex (explanatory variable) on whether or not an individual was arrested for drinking (response variable). 

I probably wouldn't have decided to explore the sex variable without the moderation assignment, and the results of this exploration gave me valuable insight into the population I've chosen to study. 

[ilanaeladis]

Chi Square Assignment (X^2)

                                    Explanatory Variable: (x-axis) 

                                                          Response Variable: (y-axis) 

                        Chi Square Syntax: 

PROC SORT; BY IDNUM; 

PROC freq; tables S2CQ1*S2BQ1A27/Chisq; 

When examining the association between getting arrested for alcohol related reasons and seeking help for drinking, a chi square test of independence revealed a very strong correlation between the two variables. Those who were arrested for drinking sought help 35.32% of the time, as opposed to those who had never been arrested (1.83%). X2=710.53, DF=1, p<.001. 

No post hoc tests for the chi-square were necessary as my response variable only has two levels.

[ilanaeladis]

ANOVA Procedure

For the Anova Procedure, I decided to test whether there is a relationship between attending a drug/alcohol treatment program and the age at full remission of alcohol abuse. 

Explanatory (Categorical)--->  Ever went to drug/alcohol treatment program 

Response (Quantitative)----> Age at full remission of alcohol abuse 

First, I subsetted my data with the following code to make sure I only get results from 18-24 year olds, and only from individuals who have had at least one alcoholic drink in their lives: 

If AGE in (18:24); If S2AQ1=1;

This is the code I used to run my ANOVA: 

PROC ANOVA; CLASS S3DQ2a6; MODEL S2BQ3F=S3DQ2a6; MEANS S3DQ2a6;

When examining the association between age of full alcohol remission (quantitative response) and whether or not an individual went to treatment (categorical explanatory), an Analysis of Variance (ANOVA) revealed that among young adults, those who attended treatment achieved full remission of alcohol abuse (Mean=19.26, s.d. ±2.40) around the same age as those young adults who did not attend treatment (Mean=19.15, s.d. ±2.25).

By looking at the mean table, I can see that the age of alcohol remission for both those who did go to treatment (1) and those who did not go to treatment (2) is about 19 years old. Because my p value is greater than 0.05, I'm going to accept the null hypothesis, which states that there is no relationship between attending treatment and age of alcohol remission. In this case, I do not post ad hoc results because my variables are only "Yes" and "no." 

[ilanaeladis]

The Association between Motivation, Coercion, and Alcohol Abuse Treatment in Young Adults

The Association between Motivation, Coercion, and Alcohol Abuse Treatment in Young Adults

1. Background

            Alcohol and substance abuse/addiction is an increasing problem among adolescents and young adults in the United States, and the onset of problematic drinking usually occurs between the ages of 18 and 24 (Bukstein, Cornelius, Trunzo, Kelly, & Wood, 2005). Alcohol abuse and binge drinking among young adults in college and non-college populations can lead to sexual assault, injury, and in some cases, death. (Stone, Becker, Huber, & Catalano, 2012) According to a cross-sectional survey of college students from a sample of 119 public and private colleges, binge drinkers are over eight times more likely to have gotten hurt or gotten in trouble with the campus police or local police. (Stone et al., 2012) While there is no scientific cure for substance abuse/addiction, effective treatment can help an individual to obtain sobriety.

            Three fundamental clinical issues are associated with treatment outcomes for populations with drug and alcohol addiction: coercion, motivation, and therapeutic alliance (Wolfe, Kay-Lambkin, Bowman, & Childs, 2013). The best predictors of who enters treatment and how successful they are in treatment are environmental factors such as family factors, motivation, and access to treatment (Bukstein et al., 2005). Motivated individuals are more willing to engage in therapy and to take an active role in the treatment process (Wolfe et al., 2013).

            Motivation is categorized into two primary types: internal and external motivation. Internal motivation stems from an intrinsic desire to change upon recognition of one's substance use as problematic. (Goodman, Peterson-Badali, & Henderson, 2011) An example of internal motivation might be the desire to change upon noticing unpleasant results (e.g. getting fired from one's job due to inappropriate behavior caused by excessive drinking). Contrarily, external motivation is a result of pressure from family and friends (Wolfe et al., 2013). 

            The present study will examine young adults from the National Epidemiologic Survey of Alcohol and Related Conditions (NESARC). The goals of the analysis will include 1) establishing the association between alcohol abuse and treatment seeking patterns in young adults, and 2) determining whether motivation (internal/external) or coercion has a higher success rate in terms of treatment outcome (e.g. will the individual continue to abuse alcohol after leaving treatment, or will they stay sober?)

2. Method

            2.1 Sample

            The sample from the first wave of the National Epidemiologic Survey on Alcohol and Related Conditions (NESARC) represents the civilian, non-institutionalized adult population of the United States, and includes persons living in households, military personnel living off base, and persons residing in the following group quarters: boarding or rooming houses, non-transient hotels and motels, shelters, facilities for housing workers, college quarters, and group homes. The NESARC included over sampling of Blacks, Hispanics and young adults aged 18 to 24 years. The sample included 43,093 participants. The study was done by computer-assisted interviews. 

            2.2 Measures

            Survey data from 18-24 year old respondents will be included, as the study will focus exclusively on young adult behavior. The study measures three categorical variables: internal motivation, external motivation, and coercion. Internal motivation was assessed through problems that directly affected the drinker ("Ever continue to drink even though causing health problem," "More than once want to stop or cut down on drinking" "Ever had job or school troubles because of drinking"). External motivation was assessed based on the influence of people in the drinker's social circle ("Ever continue to drink despite causing trouble with family/friends", "Did not seek help because afraid of what boss, friends family, or others might think.") [WCM3] Coercion was assessed by a history of arrest or legal problems tied to drug use/drinking ("Ever get arrested/have legal problems because of alcohol/drug use".) These variables are compared against the primary response variable ("Ever Went to Drug/Alcohol Rehabilitation Program").

3. Implications

            The effects of alcohol abuse are disastrous, especially on college campuses. Hardly a week goes by without a newspaper headline announcing an alcohol-related death or accident. It is important for colleges to become better equipped to deal with alcohol abuse. For example, the study may find that young adults are less inclined to go to treatment and stay sober if they are worried about what their friends think. If that is the case, colleges should take that information and find ways of providing confidential treatment on campus to enable students to get help without worry about external judgment.

References

Bukstein, O. G., Cornelius, J., Trunzo, A. C., Kelly, T. M., & Wood, D. S. (2005). Clinical predictors of treatment in a population of adolescents with alcohol use disorders. Addictive Behaviors, 30(9), 1663-1673. doi: 10.1016/j.addbeh.2005.07.013

Goodman, I., Peterson-Badali, M., & Henderson, J. (2011). Understanding motivation for substance use treatment: The role of social pressure during the transition to adulthood. Addictive Behaviors, 36(6), 660-668. doi: 10.1016/j.addbeh.2011.01.011

Stone, A. L., Becker, L. G., Huber, A. M., & Catalano, R. F. (2012). Review of risk and protective factors of substance use and problem use in emerging adulthood. Addictive Behaviors, 37(7), 747-775. doi: http://dx.doi.org/10.1016/j.addbeh.2012.02.014

Wolfe, S., Kay-Lambkin, F., Bowman, J., & Childs, S. (2013). To enforce or engage: The relationship between coercion, treatment motivation and therapeutic alliance within community-based drug and alcohol clients. Addictive Behaviors, 38(5), 2187-2195. doi: 10.1016/j.addbeh.2013.01.017

[ilanaeladis]

I am testing several associations and for this assignment I have chosen to focus on the association between third party perspective and whether or not an individual attended treatment. The first graph shows the relationship between age of onset of alcohol dependence and if an individual did not seek help because they were afraid of what others (family/friends) might think. 1= yes and 2= no, so based on this graph there does not appear to be an initial causal relationship between the two (most respondents said "no"). 

In my second graph, a third variable is added: S3DQ2A6 which is "Ever Attended Treatment for Drugs/Alcohol." Once again, yes=1 and no= 2. This graph shows that some people went to drug/alcohol treatment even if they were worried to seek help because of what others might think. I plan to test more associations and see if people are more likely to attend treatment after having been arrested/ had legal troubles for an alcohol related crime, or if having trouble at school or work correlates with treatment seeking. 

[ilanaeladis]

I have further managed my data to ONLY include information from respondents who answered "yes" to the survey question, "Have you ever been to drug/alcohol treatment."

I used the following code to obtain these results: 

IF S3DQ2A6=1;

The information in this graph supports what I found in my literature review: that most people begin to abuse alcohol from the ages of 16- 26. This graph shows the ages of when people began to abuse alcohol ONLY if they have been to drug/alcohol treatment. 

[ilanaeladis]

I have begun to create univariate graphs with the main variables in my codebook. I'm comparing how things like getting arrested and doing poorly at work/school might translate into someone seeking out treatment for substance abuse disorder. Interestingly enough, 18 years old seems to be the prime age that respondents said they began to abuse alcohol. 

[ilanaeladis]

Data Management

Here is my data management code. I have added the following quantitative variable: "How often used cannabis in the last twelve months." 

options ps=50 ls=95;

libname out "P:\QAC\qac201\STUDENTS\section2\iladis"; libname mydata "/courses/d1406ae5ba27fe300/c_3054" access=readonly;

data mynewdata; set mydata.nesarc_pds;

keep S3BD5Q2C S3DQ2A6 S3CQ12A9 S3DQ4D9;

If S3BD5Q2C=99 then S3BD5Q2C=.; If S3DQ2A6=9 then S3DQ2A6=.; If S3CQ12A9=9 then S3CQ12A9=.; If S3DQ4D9=9 then S3DQ4D9=.;

*label new variables; LABEL S3BD5Q2C=HOW OFTEN USED CANNABIS IN THE LAST TWELVE MONTHS; LABEL S3DQ2A6=EVER WENT TO DRUG/ALCOHOL REHABILITATION PROGRAM; LABEL S3CQ12A9=EVER GET ARRESTED/HAVE OTHER LEGAL PROBLEM BECAUSE OF DRUG USE; LABEL S3DQ4D9=DID NOT SEEK HELP BECAUSE AFRAID OF WHAT BOSS, FRIENDS, FAMILY, OR OTHERS WOULD THINK;

If S3BD5Q2C NE .;

proq freq; tables S3BD5Q2C S3DQ2A6 S3CQ12A9 S3DQ4D9;

run;

[ilanaeladis]

Updated Program

I have been playing around with my code a bit and wanted to post an updated version of my code/frequency tables in anticipation of the assignment due Friday, September 26th. It is the same as my last assignment essentially but I have added a few variables from my codebook and labeled the titles of the frequency tables. 

options ps=50 ls=95;

libname out "P:\QAC\qac201\STUDENTS\section2\iladis"; libname mydata "/courses/d1406ae5ba27fe300/c_3054" access=readonly;

data mynewdata; set mydata.nesarc_pds;

keep S3DQ2A6 S3DQ2B6 S3DQ4D9 S3DQ4D16;

If S3DQ2A6=9 then S3DQ2A6=.; If S3DQ2B6=9 then S3DQ2B6=.; If S3DQ4D9=9 then S3DQ4D9=.; If S3DQ4D16=9 then S3DQ4D16=.;

*label new variables; LABEL S3DQ2A6=EVER WENT TO DRUG/ALCOHOL REHABILITATION PROGRAM; LABEL S3DQ2B6=WHEN WENT TO DRUG/ALCOHOL REHABILITATION PROGRAM; LABEL S3DQ4D9=DID NOT SEEK HELP BECAUSE AFRAID OF WHAT BOSS, FRIENDS, FAMILY, OR OTHERS WOULD THINK; LABEL S3DQ4D16=DID NOT SEEK HELP BECAUSE FAMILY THOUGHT SHOULD GO BUT I THOUGHT IT WAS UNNECESSARY;

proq freq; tables S3DQ2A6 S3DQ2B6 S3DQ4D9 S3DQ4D16;

run;

[ilanaeladis]

Creating Your Own Data Set

This week, I was introduced to SAS for the first time. I have never worked with data analysis software before, so it was a very new experience for me. I was initially very frustrated when my program would not run, but after numerous trials I was finally successful in calling in my data set. 

As mentioned in my previous posts, I am working with the NESARC data set. My research question investigates what factors motivate young adults to seek treatment for drug addiction/alcoholism. The variables I explored in my frequency tables are as follows: 

1. "Ever Sought Help Because of Drinking" (S2CQ1)

2. "Ever Went to Drug/Alcohol Rehabilitation Program" (S3DQ2A6)

3. "Did Not Seek Help Because Afraid of What Boss, Friends, Family, or Others Would Think" (S2CQ4D9) 

I'm looking forward to further refining my research question and preparing my research outline due in a few weeks from now. I'm also in the process of narrowing my code book in order to find specific correlations I wish to investigate. 

[ilanaeladis]

Citation Assignment

[ilanaeladis]

This is a spreadsheet of data I complied for the Medical Records assignment. It contains four variables: one unique identifier, two quantitative variables, and one categorial variable.