The overturning of Roe v. Wade marked an important turning point in history as abortion has been gaining increasing attention throughout the years. However, the complete effect of Roe v. Wade has not been fully understood, specifically has the overturning of Roe v. Wade changed the American perspective on abortion? By looking at American's opinions on abortion throughout time, it can be understood what key points in time have affected American's views on abortion.
The Roe v. Wade decision protected women's right to abortion and ensured privacy for them as well. Planned Parenthood has created an infographic detailing how the ruling indicated a change in American's views on abortion as there was a nationwide effort trying to end criminal abortions.
With its overturning, there are numerous groups of women left at a greater disadvantaged, especially women of color. Further information regarding the implications have been explained and statistically-backed by the Kaiser Family Foundation. The perspectives of Americans on abortion is therefore of greater importance as it is not only about sex/gender but about race, culture, and religion as well. With the great range of demographics touched by abortion, it is necessary to understand how the ban can affect people's perspectives. This is important for future political topics as well.
# load dataset
import pandas as pd
# first dataset asks 'are you pro-life or pro-choice?' before asking 'should abortion be legal?'
before_data = pd.read_csv('Q2 Are You Pro-Choice Or Pro-Life (NOT AFTER).csv')
true = []
for row in before_data.values:
true.append(1)
before_data['Before legality'] = true
# second dataset asks 'are you pro-life or pro-choice?' after asking 'should abortion be legal?'
after_data = pd.read_csv('Q2 Are You Pro-Choice Or Pro-Life (AFTER).csv')
false = []
for row in after_data.values:
false.append(0)
after_data['Before legality'] = false
# by having both datasets we can hopefully combat any bias that might have been created from the
# previous question of 'should abortion be legal?'
data = pd.concat([before_data, after_data])
data.head()
| Date | Pro-choice | Pro-life | Mixed/Neither (vol.) | Don't know what terms mean (vol.) | No opinion | Before legality | |
|---|---|---|---|---|---|---|---|
| 0 | 2019 Jun 3-16 | 52 | 43 | 4 | 1 | * | 1 |
| 1 | 2018 Jun 1-13 | 49 | 45 | 3 | 1 | 2 | 1 |
| 2 | 2015 Oct 7-11 | 49 | 44 | 2 | 2 | 2 | 1 |
| 3 | 2012 Sep 24-27 | 47 | 46 | 3 | 2 | 2 | 1 |
| 4 | 2010 Mar 26-28 | 45 | 46 | 4 | 2 | 3 | 1 |
Responses after being asked 'Are you Pro-choice or Pro-life?'
| Date | Date of the corresponding Gallop survey. |
|---|---|
| Pro-choice | Percentage of Americans surveyed who identified themselves as "Pro-choice". |
| Pro-life | Percentage of Americans surveyed who identified themselves as "Pro-life". |
| Mixed/Neither (vol.) | Percentage of Americans surveyed who identified themselves as either mixed or neither on their abortion stance. |
| Don't know what terms mean (vol.) | Percentage of Americans surveyed who did not know what "Pro-choice" or "Pro-life" meant. |
| No opinion | Percentage of Americans surveyed who had no opinion on abortion stance. |
| Before legality | If pro-life/pro-choice was asked before questions on legality of abortion (0 is False, 1 is True). |
# create a graph that shows pro-choice, pro-life, and no opinion percentages
# x-axis is time
# y-axis is percentages
# illustrates how perspectives change over times - special attention toward
# 2022 (when roe v. wade is overturned)
We will use classification to identify whether an American is pro-choice, pro-life, or no opinion based on a date. Then we will conduct a confusion matrix analysis by graphing the data to see how many are correct or incorrect based on the date to see how effective the date is in regards to an American's perspective.
Note: A dataset including specifics such as race, gender, ethnicity, etc. would be more helpful - especially in the machine learning aspect. However, I have been unable to find a dataset with this information. I wil be continuing to look for one, but for now this is what I have been able to find and I still believe there will be useful anlayses that can be done with machine learning as explained briefly above. As we develop a greater understanding of ML in class, this can be adjusted.