In an advanced attempt to use the dataset I had selected before to analyze some original data, I again, turn to the data from the General Social Survey collected in 2014. Previously, (visit http://rpubs.com/lsmolar/65376 for reference) I tried to analyze the surveyed population’s opinions on abortion in relation to their demographic information. I speculated that if we examine the relationships of these variables, females may be more likely to be pro-abortion in all cases, while males may be less likely in all cases.
library(Zelig)
library(foreign)
library(DescTools)
d <- read.dta("/Users/laurenberkowitz/Downloads/GSS2014.DTA", convert.factors = FALSE)
names(d)library(dplyr)
Abortion <- select(d, age, sex, marital, educ, abdefect, abnomore, abhlth, abpoor, abrape, absingle, abany)
names(Abortion)Findings included a small positive relationship (0.002) between age and the likelihood that someone will support abortion which are significant with a 99% confidence rate. T he relationship between sex of the person surveyed does not significantly impact their likelihood to be in support of abortion when pregnancy is as a result of rape. There seems to be a small but significant negative relationship (-0.042) between the education level of the person surveyed and their likelihood to be in support of abortion when pregnancy is as a result of rape. This means that the higher a person’s education, the less likely they are to be in support of abortion when a pregnancy is as a result of rape. We see from the results that there is some room for further exploration. The age and education variables alone already present significant results just from independent relationships between the variables and support for an abortion if the pregnancy is as a result of rape. Further exploration of combined variables which may increase the effect and also exploration of these variables on questions of abortion in relation to other pregnancy scenarios can be examined in the future. #Corrections It is worth noting that it appears my results from the previous regression analysis were actually related to abortions for any reason, not specifically only in cases of pregnancy as a result of rape. This means that findings were misleading. Here I have rerun the regressions to produce updated results:
demog1 <- lm(abany ~ age, data=Abortion)The first regression analysis is showing the relationship of age to people’s support for any type of abortion.
demog2 <- lm(abany ~ sex, data=Abortion)The second regression analysis is showing the relationship of sex (gender) to people’s support for any type of abortion.
demog3 <- lm(abany ~ educ, data=Abortion)The third regression analysis is showing the relationship of education to people’s support for any type of abortion.
library(stargazer)##
## Please cite as:
##
## Hlavac, Marek (2014). stargazer: LaTeX code and ASCII text for well-formatted regression and summary statistics tables.
## R package version 5.1. http://CRAN.R-project.org/package=stargazerstargazer(demog1, demog2, demog3, type="html")| Dependent variable: | |||
| abany | |||
| (1) | (2) | (3) | |
| age | 0.002*** | ||
| (0.001) | |||
| sex | 0.023 | ||
| (0.025) | |||
| educ | -0.042*** | ||
| (0.004) | |||
| Constant | 1.460*** | 1.513*** | 2.124*** |
| (0.036) | (0.040) | (0.053) | |
| Observations | 1,646 | 1,653 | 1,652 |
| R2 | 0.004 | 0.001 | 0.070 |
| Adjusted R2 | 0.003 | -0.0001 | 0.069 |
| Residual Std. Error | 0.497 (df = 1644) | 0.498 (df = 1651) | 0.480 (df = 1650) |
| F Statistic | 6.705*** (df = 1; 1644) | 0.858 (df = 1; 1651) | 124.001*** (df = 1; 1650) |
| Note: | p<0.1; p<0.05; p<0.01 | ||
We see here that there is a small positive relationship (0.002) between age and the likelihood that someone will support abortion and are significant with a 99% confidence rate. Surprisingly it appears that the relationship between sex of the person surveyed does not significantly impact their likelihood to be in support of abortion generally. There seems to be a small but significant negative relationship (-0.042) between the education level of the person surveyed and their likelihood to be in support of abortion genderally. This means that the higher a person’s education, the less likely they are to be in support of abortion in any case. #Review As a reminder, here are the variable I have selected to use in the “Abortion” dataset:
AGE Respondent’s Age
SEX Respondent’s Sex
RACE Respondent’s Race
MARITAL Marital Status
EDUC Highest year of school completed
ABDEFECT Abortion if there is a strong chance of a serious defect for the baby
ABNOMORE Abortion if the woman is married but wants no more children
ABHLTH Abortion if the woman’s health is seriously endangered
ABPOOR Abortion if the family cannot afford more children
ABRAPE Abortion if the woman became pregnant as a result of rape?
ABSINGLE Abortion if the woman does not want to marry the father of the child
ABANY Abortion if the woman wants one for any reason
Based on our previous results, it is logical to consider the next steps. We saw that when abortion is related to a pregnancy generally, education and age show signifcant influences on decisions. For this assignment, I would like to explore those two variables in relation to other situations that may lead to abortion.
I would like to now analyze some other specific instances of abortion to see where these factors play a role. For this I am making a new dataset:
Abortion2 <- select(d, age, premarsx, educ, res16, teensex, hapmar, abrape)
names(Abortion2)## [1] "age" "premarsx" "educ" "res16" "teensex" "hapmar"
## [7] "abrape"Here are the meanings of the variables:
AGE Respondent’s Age
EDUC Highest year of school completed
RES16 Respondent’s Residence at age 16 from smaller population to larger population
TEENSEX Agree with teens having sex
HAPMAR Happy with marriage
ABRAPE Support of abortion from a pregnancy as a result of rape
For this one I am going to show a Generalized Linear Model for demographics affecting support of abortion if the pregnancy is as a result of rape. Here I am comparing age and education as factors
library(tidyr)
library(pander)
library(car)##
## Attaching package: 'car'
##
## The following object is masked from 'package:DescTools':
##
## Recode
##
## The following object is masked from 'package:boot':
##
## logitabrape2 <- as.numeric(Abortion$abrape == "yes")abr1 <- glm(abrape2 ~ age + educ + teensex, family = binomial, data = Abortion2)## Warning: glm.fit: algorithm did not convergeabr2 <- glm(abrape2 ~ age + educ + teensex + res16, family = binomial, data = Abortion2)## Warning: glm.fit: algorithm did not convergeabr3 <- glm(abrape2 ~ age + educ + teensex + res16 + hapmar, family = binomial, data = Abortion2) ## Warning: glm.fit: algorithm did not convergestargazer(abr1, abr2, abr3, type = "html")| Dependent variable: | |||
| abrape2 | |||
| (1) | (2) | (3) | |
| age | 0.000 | 0.000 | 0.000 |
| (725.938) | (730.108) | (1,246.682) | |
| educ | -0.000 | -0.000 | 0.000 |
| (4,074.019) | (4,156.640) | (6,174.678) | |
| teensex | 0.000 | 0.000 | -0.000 |
| (14,403.580) | (14,479.840) | (23,033.070) | |
| res16 | -0.000 | -0.000 | |
| (8,305.547) | (12,334.480) | ||
| hapmar | -0.000 | ||
| (34,437.760) | |||
| Constant | -26.566 | -26.566 | -26.566 |
| (70,321.240) | (74,971.920) | (132,572.100) | |
| Observations | 804 | 802 | 356 |
| Log Likelihood | -0.000 | -0.000 | -0.000 |
| Akaike Inf. Crit. | 8.000 | 10.000 | 12.000 |
| Note: | p<0.1; p<0.05; p<0.01 | ||
We see from here there is really no measured relationship with any of these combined variables in relation to abortion. I also tested out these variables with the question of being in favor of abortion when the pregnancy was as a result of rape as well as testing out variables related to premarital sex, drinking excessively, and gender. None of these show any relationship. This assignment was productive in helping me figure out that I need to collect a new dataset to begin with for the next time, or I need to find a new topic to analyze related to the General Social Survey’s 2014 Respondents. Evidentally there are not a lot of logical variables that can predict any kind of effect on answers given being pro or anti abortion.