finne2-blog

[finne2-blog]

DATA TOOLS - Assignment 1

Research Question:  Are children raised by foster parents more likely to be associated with earlier  drug experimentation than children raised by adoptive parents?

Code:

Output:

The output above shows that the variable has two levels, 1=adoptive parents and 2=foster. Four hundred twenty-nine cases were reviewed and  cases were included. 

A total of 21 cases with adoptive parents first started experimentation with opioids at a mean age of 20.5 and 5 cases first started experimentation with opioids at a mean age of 18.80.

The F result is not significant at p > 0.05.   The null hypothesis cannot be that there is no difference in the age of first experimentation with opioids between those cases with adoptive parents and those cases with foster parents can not be rejected.   Since there was no significance and only two categories in the explanatory variable, post hoc tests were not necessary. 

[finne2-blog]

Assignment 4

Research Question:  Are children raised by foster parents more likely to be associated with earlier  drug experimentation than children raised by adoptive parents?

1) The first variable created was a discrete or categorical variable of “Family Type” that will be used as the explanatory variable.  Familytype=1 for children raised by adoptive parents and Familytype=2 for child raised by foster parents.  The graph below shows that 85.8% were raised by adoptive parents and only 10 percent were raised by foster parents. 

The second variable is a quantitative variable response variable.  This could have been done for all classes of drugs, but for the sake of this assignment, I will modify the research question so it only looks for association between familytype and opioid drug experimentation.  The research question becomes 

Are children raised by foster parents more likely to be associated with earlier opioid experimentation than children raised by adoptive parents.  The univariate tables and graphs were examined to determine if I needed to break the response variable into categories or leave it as a quantitative variable.  

There were a total of 26 cases that were either raised by an adoptive parent or foster parent that had entered an age for first time opioid usage on the survey. The mean age at the time of first opioid use was 20.15 years with a range of 11 to 35. The mode or most common age was 18 as shown on the frequency graph below.  The standard deviation was 5.56.  Looking at the distribution below with the relatively small range of years and number of bins, I decided to keep the response variable quantitative as it was simply viewed and understood in that state. 

When graphed for association below, it appears that the foster status is associated with a slightly younger age of opioid experimentation than those with adoptive status as hypothesized (approximately 18.75 years versus 20.5 years).  However, no statistical tests of significance were done and the number of cases with foster status was very low.  Follow-up statistical testing would be required. 

[finne2-blog]

Assignment 3

Code:

My selected data set needed very little data management due to the narrow topic and low frequencies of observations.  Collapsing responses or aggregating variables would have caused too low observations for statistical significance to be determined.  The only elements of data management in the code were to 

1) Delete cases in which observations were selected for having both foster and adoptive parents because I wanted to compare observations in which they had one or the other

2) Aggregate Independent Variable as new Family Type Variable so have one independent categorical variable for statistical analysis

2) Recode unknown age at first drug use for each of the drugs to missing

The frequency table below is an example is for the new variable of FamilyType.  This variable =1 if adopted and 2 if foster care. There were 368 Respondents who had adoptive parents and 61 who had foster care.  Ten cases had been removed because they had both types of parents.  These would not provide information needed for my research questions. 

The frequency table for age at the first use of opiods did demonstrate a change in the number of missing values when the code reassigned “unknown age at first use” to blank.  This was because none of my subset of interest had answered as unknown.  403 answered as they had never used and 26 gave an age at first use, leaving none in the missing category.  This frequency table with missing data is shown below:

The second frequency table shows that prior to recoding there is one case of unknown age at first cannabis use coded as 99.  Since this provides no information for my research question, this was recoded and missing and the frequency table below it shows that it is gone with the frequency of missing data changed from 305 to 306 after it was recoded to missing. 

[finne2-blog]

Assignment 2

Code:

The code below selects a subset of the data:  observations for children with adoptive or foster parents and recorded ages of onset of drug use for each of the drug types listed in the code.  Since 10 observations had both adoptive and foster parents, a line of code was written to exclude those cases because I am interested in comparing those children with adoptive parents only with those with foster parents only - not that had both types of parents.  

Frequency Tables for Adoptive Parents:

Three of the frequency tables from the run are attached below for discussion.  There were additional tables for each drug in the code that have not been attached, but are similar to the third table, the opioid table.

The first frequency tables shows that there were 429 for the selected subset, those with adoptive parents and those with foster parents.  As shown in the frequency table below, 368 said yes to the adoption question and 61 said no.  These same 61 said yes to the foster question in the second table.  Therefore, a total of 368 had adoptive parents and 61 had foster parents.  As explained in the code explanation, there were 10 observations which had both types of parents that were removed from the subset. 

The third table below shows the distribution of ages when cases started using opioids in my selected subset.  A total of 26 observations gave an age at which they started using (the column on the left). The most common age of start of use was 16, with a frequency of 6 cases, but the range in ages was from 11 years to 35  years. The data set was missing 403 responses which means that 403 out of 429 of the subset did not answer this question or it was not applicable because they did not ever use opioids.  

Frequency Table for Foster Parents:

Frequency for age at start of  opioid use in children with adoptive or foster parents:

[finne2-blog]

Assignment 1

Chosen Data Set: U.S. National Epidemiological Survey on Alcohol and Related Conditions (NESARC)

Research Question:  Are children raised by foster parents more likely to be associated with earlier  drug experimentation than children raised by adoptive parents?

Hypothesis:  Children raised by foster parents are more likely to be associated with earlier drug experimentation.

Literature Review:

Search terms:  “Foster Care and Substance Abuse” and “Adoption and Substance Abuse”  and “Adoption and Drug Use”

    A child’s status as being raised by an adoptive parent or a foster care parent versus those children who have not has been shown to be associated with and increased prevalence of lifetime substance abuse (Yoon et, al., 2012; Blome et al., 2009, National Survey on Drug Use and Health, 2005). Overall, adoptees have been determine to have a 1.87-fold increase (adjusted odds ratio [AOR] 1.87, 95% CI 1.51–2.31) in the odds of any lifetime SUD compared to nonadoptees (Yoon et. al., 2012).  Interviews with 320 adolescents in foster care found approximately 40% used alcohol, 35% reported marijuana use, and 25% described using both in the prior six months (Thompson & Auslander, 2007). It is though that children living with adoptive parents versus foster care have more stability.  Children in foster care are typically there for a short time and usually return to their families.  However, longer stays and multiple placements are also possible and these have been found to be related to negative developmental outcomes (Shin, 2004).  In comparison, adoption is associated with many positive aspects.  Placement with and adoptive family most likely results in better childhood experiences, healthcare, family stability, and family relationships

 Blome W, Shields J, Verdieck M. The Association Between Foster Care and Substance Abuse Risk Factors and Treatment Outcomes: An Exploratory Secondary Analysis. Journal Of Child & Adolescent Substance Abuse [serial online]. July 2009;18(3):257-273. Available from: Education Research Complete, Ipswich, MA. Accessed April 1, 2018.

Dube, S., Felitti, V., Dong, M., Chapman, D., Giles, W., & Anda, R. (2003) Childhood abuse, neglect and household dysfunction and the risk of illicit drug use: The adverse childhood experiences study. Pediatrics, 111(3), 564-572.

Hawkins, J. David,Catalano, Richard F.,Miller, Janet Y. Risk and protective factors for alcohol and other drug problems in adolescence and early adulthood: Implications for substance abuse prevention. Psychological Bulletin, Vol 112(1), Jul 1992, 64-105

 Langbehn DR, Cadoret RJ, Caspers K, Troughton EP, Yucuis R (2003) Genetic and environmental risk factors for the onset of drug use and problems in adoptees. Drug Alcohol Depend 69: 151–167.

National Survey on Drug Use and Healt. (2005). Substance use and need for treatment among youths who have been in foster care- February 18, 2005. Washington, DC: SAMHSA.

Thompson, R., & Auslander, W.  (2007). Risk factors for alcohol and marijuana use among adolescents in foster care. Journal of Substance Abust Treatment, 32, 61-69.

 Westermeyer J, Bennett L, Thuras P, Yoon G (2007) Substance use disorder among adoptees: a clinical comparative study. Am J Drug Alcohol Abuse 33: 455–466

Yoon G, Westermeyer J, Warwick M, Kuskowski MA (2012) Substance Use Disorders and Adoption: Findings from a National Sample. PLoS ONE 7(11): e49655. https://doi.org/10.1371/journal.pone.0049655