Service, Learning and Feeding the Hungry

Statistics majors Edy Reynolds (left) and Lauren Taylor volunteered for People's Kitchen in San Luis Obispo and helped assess the nonprofit's data analytics and homeless meal serving plans. Courtesy photo

REWARDING REAL-WORLD EXPERIENCES FOR STASTISTICS STUDENTS TRACKING MEAL DATA FOR LOCAL NONPROFIT

OCTOBER 2024
by NICK WILSON
Infographics by Eileen odanaka Vavra

Each year, the nonprofit People’s Kitchen serves several thousand meals to help meet the extensive needs of the unhoused in San Luis Obispo County. 

With churches, schools and social clubs all sharing service duties, the organization required an effective system to better maintain and study its meal count numbers and identify patterns in fluctuating attendance throughout the year. A Cal Poly statistics team stepped up with the expertise needed to take the nonprofit’s data tracking to the next level and make a meaningful, real-world impact.

Starting in spring 2023, statistics Professor Emily Robinson and her student team went to work creating a variety of user-friendly analytics tools to track daily meal counts served at 40 Prado Homeless Services Center in San Luis Obispo. While People’s Kitchen had been logging these counts, the organization recognized a need to improve its data management.

“My background is in data visualization and user experience information, and I was one of the department consultants at that time,” Robinson said. “When Department Chair Andrew Schaffner contacted me, he asked, ‘Is this interesting to you?’ Immediately,I said ‘Yes, this sounds so cool.’” 

Robinson met with Robyn Kontra Tanner, a People’s Kitchen board member and Cal Poly University Communications and Marketing staff member, to learn about the nonprofit’s needs. Robinson then recruited the student research team of statistics majors Edy Reynolds and Lauren Taylor, as well as recent English graduate Payton Swanson.

People’s Kitchen draws upon the goodwill of more than 40 local organizations whose volunteers serve hot noon meals 364 days per year. The organization provided 35,273 meals in 2023, its largest annual total to date, after serving 27,306 in 2022; 23,529 in 2021; and 26,924 in 2020.

“There are different groups serving every day,” said Reynolds. “So, it’s good for them to know how much food to prepare, so there’s enough but not too much left over. What we’re hoping to explore in the future is some actual predictive modeling.”

But the first step was “digitizing everything,” Reynolds said. 

Student researchers started from scratch, pulling from existing People’s Kitchen computer files as well as handwritten tallies of meal counts kept daily in recent years, and by the month in previous years.

“They wanted everything housed in their Google Drive to access it privately, so we entered the records into a spreadsheet and then formatted the variables to come up with a process about how they could input data themselves,” Reynolds said.

The students created a Google Form that links to the spreadsheet, which links to dashboards in the program Looker Studio, formerly Google Data Studio. Automating the process for customizing reports has created an easy-to-use portal.

“Designing that data workflow was very interesting because there was a lot of code that makes the connection from the Google Form to the spreadsheet smooth and easy to use,” Taylor said. 

The informative dashboard tools the research team created now offer People’s Kitchen the ability to pull data quickly and conveniently. Coordinators can spot trends and observe fluctuations to identify what factors cause shifts in the number of meals served, such as weather, major holidays, time of month and serving group.

Swanson’s role involved writing procedural and instructional documents for People’s Kitchen coordinators and volunteers, many of them unfamiliar with the data tools. Swanson’s manuals can be used by future organizational leaders as well. 

Robinson said the more accessible data records may also help People’s Kitchen apply for grants and other types of funding to support the program.

The students gained real-world experience collaborating with clients on data analytics and developed expertise in software they normally don’t work with, including Looker Studio, which converts data into customizable, informative reports and dashboards; Google Sheets (AI-powered spreadsheets) and Google Apps Script that lets users automate tasks across Google products.

From left, Taylor, statistics Professor Emily Robinson and Reynolds. Courtesy photo

“We don’t typically use anything hosted in Google for stats,” Taylor said. “Typically, we’d do something like this in R Shiny (an open source web application package), which helps create dashboards really easily. But it was important for the organization to have full control and a deep understanding of the data workflow. We could not provide this with R Shiny, which is why we opted to host in Google.” 

The research team presented their work at the May Bailey College 2024 Student Research Conference and the June 2024 Symposium in Statistics and Data Science Conference in Richmond, Virginia.

Student researchers also visited the 40 Prado Services Center to serve meals alongside volunteers from various local organizations, making special connections with those who serve meals to people in need on a regular basis.

“The people really are what makes that organization so special,” Taylor said. “It’s just a great community.”