SAT Project - The first project I had done at General Assembly, performed EDA, data cleaning and created data visualizations

Overview

Project 1: Standardized Test Analysis by Adam Klesc

Overview

This project covers:

  • Basic statistics and probability
  • Many Python programming concepts
  • Programmatically interacting with files and directories
  • Visualizations
  • EDA
  • Working with Jupyter notebooks for development and reporting

The SAT is a standardized test that many colleges and universities in the United States require for their admissions process. This score is used along with other materials such as grade point average (GPA) and essay responses to determine whether or not a potential student will be accepted to the university.

The SAT has two sections of the test: Evidence-Based Reading and Writing and Math (source).

The SAT changed their format in 2016 to address some problems that students and schools had with their style of questions and questionable grading. The ACT surpassed the SAT in popularity due to these problems and a perception that the SAT was "class-biased" due to their questions. The ACT also did this through a lot of states make it free state-wide and a requirement to pass high school which the SAT did not catch up on.


Problem Statement

When crafting my problem statement, I really was stuck with what to really analyze and what problem to solve. But it really simplified the process for me when looking at the yearly datasets because looking at the years and the progression of both the SAT and ACT participation rates, it really clicked.

My problem statement in a short sentence is:

Has collegeboard solved the problems they were seeking to fix with the 2016 format change?

I really had to go through three factors when making my problem statement. Who is my audience, what data am I using, and what value am I bringing to my audience.

  • The first factor was fairly easy to answer, I think the people who would be most interested in the answer to my problem statement would be collegeboard and the people involved in the change to see whether their change has actually worked.

  • The second factor was what data I would use and primarily I would be using the SAT 2017-2019 data along with the ACT 2017-2019 data but over the course of the project, I added some outside sources as well.

  • The final factor in creating that problem statement was bringing clarity to the changes and whether they have been actually productive in solving what they were mean't to solve.


Datasets

Provided Data

These are the datasets I used in the process that were included in the data folder provided, I cleaned each of them but will still be sharing both the raw and cleaned datasets

Additional Data

These are the datasets I required from outside research I performed.


Data Dictionary

A dictionary compiling the primary columns from the two main datasets used throughout the analysis.

Feature Type Dataset Description
participation17 float SATTotal Participation numbers for the 2017 SAT(in decimal percentage)
participation17 float ACTTotal Participation numbers for the 2017 ACT(in decimal percentage)
participation18 float ACTTotal Participation numbers for the 2018 ACT(in decimal percentage)
participation18 float SATTotal Participation numbers for the 2018 SAT(in decimal percentage)
participation19 float SATTotal Participation numbers for the 2019 SAT(in decimal percentage)
participation19 *float( ACTTotal Participation numbers for the 2019 ACT(in decimal percentage)
total17 int SATTotal Average total score for the 2017 SAT(median score)
total18 int SATTotal Average total score for the 2018 SAT(median score)
total19 int SATTotal Average total score for the 2019 SAT(median score)
composite17 float ACTTotal Average total score for the 2017 ACT(mean score)
composite18 float ACTTotal Average total score for the 2018 ACT(mean score)
composite19 float ACTTotal Average total score for the 2019 ACT(mean score)
sat_free boolean ACTTotal/SATTotal States that have the SAT free statewide
act_free boolean ACTTotal/SATTotal States that have the ACT free statewide
population_rank float ACTTotal/SATTotal Population rank for each state
gdp_rank_19 float/int ACTTotal/SATTotal GDP rank for each state
state object SATTotal The states where the data is located
state object ACTTotal The states where the data is located

Deliverables

Notebook that contains my cleaning of data and creation of new CSV's

cleaning_and_creating.ipynb

Notebook that contains my analysis and visualization

visualization_and_analysis.ipynb

Starter Code

starter-code.ipynb

ReadMe

README-ADAMK.md

Presentation

project_1_presentation.pdf


Technical Report Starter Code

Edited structure of starter code notebook by creating individual notesbooks for cleaning and analysis + visualization. Starter code trials are complete either in those two notebooks or on the starter code itself

Notebook that contains my cleaning of data and creation of new CSV's

cleaning_and_creating.ipynb

Notebook that contains my analysis and visualization

visualization_and_analysis.ipynb

Starter Code

starter-code.ipynb


Analysis

The analysis started with understanding the problem statement, in order to do that, I had to find out the reasons why the SAT changed and look at those problems in isolation each when looking at the data to see if collegeboard has actually succeeded in solving the problems they tried to address with their change.

  • Increase participation to overtake the ACT in market-share

  • Eliminate the stigma that the SAT is class-biased and is unfair towards those of lower incomes

I narrowed it down to these two problems because of these two sources

greentestprep

cnn.com

These sources really helped clarify what the SAT was actually looking to achieve with their change and helped me in my data collection process

  • From looking at the mean and median, it was evident that the SAT was steadily gaining popularity while the ACT was losing popularity but what makes the change a great success in terms of the first problem the change was mean't to address was how the SAT was significantly more popular than the ACT in states with a higher population. The SAT has also notably been made free in 3 states after 2017 (the first full year of the SAT change), those states being Colorado, West Virginia, and Illinois. New Hampshire, Michigan, and Connecticut also made a switch to the SAT, the year of the change. It's clear that the SAT change has done wonders in improving participation and making states believers in their test.

  • This lines up with outside sources claiming that the SAT surpassed the ACT in 2019 with overall market share with 2.2 million students to the ACT's 1.8.

  • From looking at the correlation between the GDP per state + SAT Participation, it was clear to see that the change had not had the effect that was desired from collegeboard execs. The coefficient stayed the same despite the added help in a new Khan Academy course that was free to all + fee waivers (for college applications) for those who met a certain income threshold. source

  • While a lot of studies tend to disagree on the fact that the SAT is a class-biased test, it is clear that a lot of universities are taking note of the potentially inherent class disadvantage in standardized testing as a whole and have outright banned it not only for just times of a pandemic, but also for the foreseeable future. The most prominent of these universities being the University of California


In Conclusion

From my analysis, I was able to conclude that the SAT has successfully managed to overtake the ACT once again as the market leader in standardized testing due to primarily to the SAT changes made to the format and structure. It not only made the test more appealing to students but also states themselves as number of them made it free during the time between 2019 from the date of the change. This is evidenced by the fact that the SAT once again, became the market-leader in standardized testing after the ACT's regression post-SAT change. While solving participation was a massive success in both participation numbers and time, the second problem is where the SAT still needs work.

The SAT change partially occurred due to the reputation that the SAT was a class-biased test, the strange scoring of the 2400 scale and detachment from the actual school work that students were learning in school lead to the ACT being more popular and while they did fix the problem with participation, the lingering problem of class-bias is still there. With more colleges pulling away from standardized testing and geographical location being so heavily tied to participation and success, there need to be more upcoming changes to rectify this issue and make the SAT a more equitable test.

Owner
Adam Muhammad Klesc
Hopeful data scientist. Currently in General Assembly and taking their data science immersive course!
Adam Muhammad Klesc
Examples of using f2py to get high-speed Fortran integrated with Python easily

f2py Examples Simple examples of using f2py to get high-speed Fortran integrated with Python easily. These examples are also useful to troubleshoot pr

Michael 35 Aug 21, 2022
CVPR 2021 Official Pytorch Code for UC2: Universal Cross-lingual Cross-modal Vision-and-Language Pre-training

UC2 UC2: Universal Cross-lingual Cross-modal Vision-and-Language Pre-training Mingyang Zhou, Luowei Zhou, Shuohang Wang, Yu Cheng, Linjie Li, Zhou Yu,

Mingyang Zhou 28 Dec 30, 2022
A python package for generating, analyzing and visualizing building shadows

pybdshadow Introduction pybdshadow is a python package for generating, analyzing and visualizing building shadows from large scale building geographic

Qing Yu 13 Nov 30, 2022
OpenGAN: Open-Set Recognition via Open Data Generation

OpenGAN: Open-Set Recognition via Open Data Generation ICCV 2021 (oral) Real-world machine learning systems need to analyze novel testing data that di

Shu Kong 90 Jan 06, 2023
Keras implementation of Normalizer-Free Networks and SGD - Adaptive Gradient Clipping

Keras implementation of Normalizer-Free Networks and SGD - Adaptive Gradient Clipping

Yam Peleg 63 Sep 21, 2022
Official implementation of SynthTIGER (Synthetic Text Image GEneratoR) ICDAR 2021

🐯 SynthTIGER: Synthetic Text Image GEneratoR Official implementation of SynthTIGER | Paper | Datasets Moonbin Yim1, Yoonsik Kim1, Han-cheol Cho1, Sun

Clova AI Research 256 Jan 05, 2023
Repository for the COLING 2020 paper "Explainable Automated Fact-Checking: A Survey."

Explainable Fact Checking: A Survey This repository and the accompanying webpage contain resources for the paper "Explainable Fact Checking: A Survey"

Neema Kotonya 42 Nov 17, 2022
Reinforcement learning models in ViZDoom environment

DoomNet DoomNet is a ViZDoom agent trained by reinforcement learning. The agent is a neural network that outputs a probability of actions given only p

Andrey Kolishchak 126 Dec 09, 2022
Winning solution of the Indoor Location & Navigation Kaggle competition

This repository contains the code to generate the winning solution of the Kaggle competition on indoor location and navigation organized by Microsoft

Tom Van de Wiele 62 Dec 28, 2022
Machine Learning From Scratch. Bare bones NumPy implementations of machine learning models and algorithms with a focus on accessibility. Aims to cover everything from linear regression to deep learning.

Machine Learning From Scratch About Python implementations of some of the fundamental Machine Learning models and algorithms from scratch. The purpose

Erik Linder-Norén 21.8k Jan 09, 2023
Delving into Localization Errors for Monocular 3D Object Detection, CVPR'2021

Delving into Localization Errors for Monocular 3D Detection By Xinzhu Ma, Yinmin Zhang, Dan Xu, Dongzhan Zhou, Shuai Yi, Haojie Li, Wanli Ouyang. Intr

XINZHU.MA 124 Jan 04, 2023
ESTDepth: Multi-view Depth Estimation using Epipolar Spatio-Temporal Networks (CVPR 2021)

ESTDepth: Multi-view Depth Estimation using Epipolar Spatio-Temporal Networks (CVPR 2021) Project Page | Video | Paper | Data We present a novel metho

65 Nov 28, 2022
Related resources for our EMNLP 2021 paper

Plan-then-Generate: Controlled Data-to-Text Generation via Planning Authors: Yixuan Su, David Vandyke, Sihui Wang, Yimai Fang, and Nigel Collier Code

Yixuan Su 61 Jan 03, 2023
This is the official pytorch implementation for the paper: Instance Similarity Learning for Unsupervised Feature Representation.

ISL This is the official pytorch implementation for the paper: Instance Similarity Learning for Unsupervised Feature Representation, which is accepted

19 May 04, 2022
[ArXiv 2021] One-Shot Generative Domain Adaptation

GenDA - One-Shot Generative Domain Adaptation One-Shot Generative Domain Adaptation Ceyuan Yang*, Yujun Shen*, Zhiyi Zhang, Yinghao Xu, Jiapeng Zhu, Z

GenForce: May Generative Force Be with You 46 Dec 19, 2022
Implementation detail for paper "Multi-level colonoscopy malignant tissue detection with adversarial CAC-UNet"

Multi-level-colonoscopy-malignant-tissue-detection-with-adversarial-CAC-UNet Implementation detail for our paper "Multi-level colonoscopy malignant ti

CVSM Group - email: <a href=[email protected]"> 84 Nov 22, 2022
Code for the prototype tool in our paper "CoProtector: Protect Open-Source Code against Unauthorized Training Usage with Data Poisoning".

CoProtector Code for the prototype tool in our paper "CoProtector: Protect Open-Source Code against Unauthorized Training Usage with Data Poisoning".

Zhensu Sun 1 Oct 26, 2021
TSP: Temporally-Sensitive Pretraining of Video Encoders for Localization Tasks

TSP: Temporally-Sensitive Pretraining of Video Encoders for Localization Tasks [Paper] [Project Website] This repository holds the source code, pretra

Humam Alwassel 83 Dec 21, 2022
PyTorch code for SENTRY: Selective Entropy Optimization via Committee Consistency for Unsupervised DA

PyTorch Code for SENTRY: Selective Entropy Optimization via Committee Consistency for Unsupervised Domain Adaptation Viraj Prabhu, Shivam Khare, Deeks

Viraj Prabhu 46 Dec 24, 2022
Band-Adaptive Spectral-Spatial Feature Learning Neural Network for Hyperspectral Image Classification

Band-Adaptive Spectral-Spatial Feature Learning Neural Network for Hyperspectral Image Classification

258 Dec 29, 2022