Austin Animal Center Data Analysis

Background on this project and the company

I’m approaching this project as part of my Google Data Analyis course. I chose to look at pet adoption data because my wife and I have two pets that we “rescued” and they hold a special place in our hearts. So, what better way to learn more about data analysis than by looking at data about something meaningful.

The publicly available data from the Austin Animal Center shows that Austin, Texas is the largest “No Kill” city in the country. The data provided gives details of animal intakes and the outcomes as they leave the Animal Center. This data is something shared through the official city of Austin’s open data portal. I found references to the data through Kaggle as I was researching possible data analysis projects that I could work through to broaden my data analysis skills. I intentionally chose not to review others’ work that is on Kaggle so as not to skew my own work, but it’s very likely that analysis I do in exploriing this data has already been performed by others that are doing similar exploratory data analysis work. Additionally, the city of Austin has some basic charts of the data that was helpful to look at some of the basics. These initial looks at the data helped me to define the questions I wanted to ask of the data.

I initially wanted to look at data closer to home, and I still plan to do so, so I will use the skills learned here looking at Austin’s data to look closer at data for animal shelters in my own state of Idaho, separately in another analysis project. I’m hopeful that this project will allow me to dig deeper into the R programming language, RStudio, and showcase some of the additional analysis and programming skills I’ve learned, along with skills specific to using plotting within R to create meaningful charts and graphs.

Ask

• I’d like to answer three questions with this project:
  1. How does the length of stay in the animal center differ depending on animal type?
  2. How does the outcome differ depending on animal type? Does one animal type survive more than another?
  3. How does the calendar or cycle of adoptions impact the animal center? Is there a time of year that’s busier?
• The Business Task: Through analysis of these guiding questions, I’d like to identify ways that the animal center can advertise, change processes, or have increased awareness to be able to serve the pets better.

Prepare

• Because I’m the dedicated resource for this project, and fully in control of the plan for the analysis, the timeline should be only a few days.
• As mentioned above, the data for this work is available publicly and can be found here and is also referenced on Kaggle here.
• I’ll begin by downloading the available data for intakes and outcomes, which includes data back to 2013 through August 2024 when I downloaded the data.
• The initial analysis includes:
  1. How is the data organized?
  2. Sort, filter, and categorize the data to understand what’s included
  3. Review the credibility of the data
  4. Confirm the completeness of the data to understand capability to answer the questions being asked
• Data Sources Used: All data sources used for this were obtained from the City of Austin’s open data portal as linked above.

Process

• Now that I’ve downloaded the data I can do my initial preparatory analysis. Because the data was provided in a .csv file format, there’s some basic analysis that can be done within Excel quickly. For example, I can see that there are more records in outcomes than there are in intakes, so I will need to review consistency of the data to make sure I understand the differences. I can also see for example the earliest dates of the records and that not all records have the name of the pet.
• My entire cleaning, filtering, transforming, and bias checking of the data will be done within R so that I can maintain a record of the work, document any cleaning of the data, as well as to make the process repeatable.
• Data Cleaning and Manipulation:
  1. Join the data tables together, and normalize the column names, so that the intake details and outcome information can be compared seamlessly.
  2. Remove records that are irrelevant to analysis including animals with multiple intakes.
  3. Calculate the date parts for the difference between intake time and outcome time to understand length of stay.
  4. Calculate the date parts for the intake date and outcome date to understand the cycles through the year to know if there are peaks or valleys.

Analyze

• We’ve asked questions of the data, and also prepared and processed the data, now we need to analyze the data to be able to make recommendations.
• Summary of Analysis:
• Dogs stay about a week less than cats do, seemingly the favorite to adopt in the Austin area.
• Birds have a higher rate of euthanasia than other pets, contributing to the perceived average shorter stay during initial analysis.
• Animals other than pets are euthanized more often than anything else.
• Austin may be indicated as a “no kill” shelter, but there are animals that are euthanized. However, some pets stayed several years without being euthanized.
• As one might expect, in spring, there’s a spike in cat intakes at the animal shelter.

Share

• Let’s take a look at a few of the key findings

Act

• Recommendations
  1. On average, cats take longer to be adopted than dogs. Is there a possibility of public advertising about cats to increase adoption rates?
  2. Birds are euthanized more than other pets. Are there opportunities to find organizations to partner with to take in birds?
  3. There are many “other” animals that are euthanized, such as bats for example. Is there a more appropriate handling for these animals?
  4. In Spring there’s a spike in intakes for cats, presumably because of kittens. How can the animal center partner with other agencies to increase the rate spay/neuter of pets?
  5. Year over year, there’s a spike in the Spring for cats. Do we need to increase staff during these times to account for the increased pets?

Reflect

• With the project complete, we’re now at the phase of reflection.
• What have I learned? What can I do better next time?
• This case study was a bit more “Self-Guided” than the last study I worked on, so it allowed me to do research on different programming in R to learn more and showcase some more things like using Facets.
• This case study also didn’t have clearly defined steps like the prior study did, so I needed to look at data more closely to reconcile what I was seeing and what needed to be answered.
• I also realized in this case study that the template I used/created in the prior case study was helpful to kickstart the process for this case study.