CSY-014. An exploratory data analysis of public school demographics.

PROJECT COMPLETED: January 2021

Public School Demographics by State | Exploratory Data Analysis (EDA) by David White

Exploratory Data Analysis

U.S. Department of Education Database | Public School Data by State


What is Exploratory Data Analysis?

Exploratory data analysis (EDA) is a technique used by data scientists to inspect, characterize and briefly summarize the contents of a dataset. EDA is often the first step when encountering a new or unfamiliar dataset. EDA helps the data scientist become acquainted with a dataset and test some basic assumptions about the data. By the end of the EDA process, some initial insights can be drawn from the dataset and a framework for further analysis or modeling is established.

Work Process:

Here’s the process I used to complete this project—

Featured
Case Study 014. Exploratory Data Analysis: Pre-Work
CSY-014, MSM Design
CSY-014, MSM Design

Pre-Work

In this exploratory data analysis I explored a dataset of information on public schools in the United States. The underlying data was made freely available to the public and I obtained it from the U.S. Department of Education website.

CSY-014, MSM Design
Case Study 014. Exploratory Data Analysis: Step 1
CSY-014, MSM Design
CSY-014, MSM Design

Step 1 - Prepare the Workspace

Since I was working with a fairly large dataset, I decided that Python was the best tool to use to analyze it. I used a Jupyter Notebook to run my code.

CSY-014, MSM Design
Case Study 014. Exploratory Data Analysis: Step 2
CSY-014, MSM Design
CSY-014, MSM Design

Step 2 - Describe the Characteristics of the Dataset

First, I wanted to wrap my head around the dataset by finding out exactly what size it was and specifically what kind of information (variables) it contained.

CSY-014, MSM Design
Case Study 014. Exploratory Data Analysis: Step 3
CSY-014, MSM Design
CSY-014, MSM Design

Step 3 - Summarize the Dataset

Next, I did some basic calculations to produce a set of summary statistics.

CSY-014, MSM Design
Case Study 014. Exploratory Data Analysis: Step 4
CSY-014, MSM Design
CSY-014, MSM Design

Step 4 - Visualize the Dataset

Then, I made a number of visualizations that helped me to better understand what the data was saying.

CSY-014, MSM Design
Case Study 014. Exploratory Data Analysis: Step 5
CSY-014, MSM Design
CSY-014, MSM Design

Step 5 - Identify Insights

Finally, I could draw some insights from the dataset. Here are the main takeaways:

  • The dataset consists of 51 rows and 42 columns

  • The dataset consists of:

    • student enrollment

    • school staffing

    • student demographic information

  • There are 51 rows and 42 columns in the dataset. None of the rows are blank.

  • The dataset contains totals per state of the number of students in (2) gender categories and (7) race/ethnicity categories.

  • 2018-19 US public school total enrollments by demographic group are as follows:

    • 25.8 million male students

    • 24.4 million female students

    • 473K American Indian/Alaska Native students

    • 2.6 million Asian or Asian/Pacific Islander students

    • 13.7 million Hispanic students

    • 7.6 million Black students

    • 23.7 million White students

    • 176K Hawaiian Nat./Pacific Isl. students

    • 2 million multiracial students

  • The states with the highest number of Black public school students are: Florida, Georgia and Texas

  • The states with the highest number of Hispanic public school students are: California and Texas

  • The state with the highest number of Asian or Asian/Pacific Islander public school students is California. New York and Texas are a distant second and third.
  • The states with the highest number of American Indian/Alaska Native public school students by far is Oklahoma
  • The state with the highest number of Hawaiian/Pacific Islander public school students by far is Hawaii

  • The states with the highest number of White public school students are: California and Texas

CSY-014, MSM Design

Final Results:

 
case studies, data visualizationDavid Whiteexploratory data analysis, Python, report

3. Work Samples - Content | Work samples of data visualizations, data dashboards, websites, infographics, publication-quality reports, slide decks and visual presentations I've created (detailed)