Customer exit interviews provide rich dataset


Posted: May 4, 2015

As the EFSNE project enters its fifth year, the Consumption Team is marking a major milestone: the completion of its shopper intercept survey effort. Over the course of three years, team members surveyed some 2,700 shoppers, paving the way for researchers to answer several questions about the food-shopping experiences of those surveyed.

"Given the urban/rural variation in our locations, size of the sample, and longitudinal nature, we anticipate our findings to contribute to a better understanding of shoppers' behavior in lower income neighborhoods," said Anne Palmer, Consumption Team leader and Johns Hopkins University researcher. 

Collecting so many surveys involved the effort of the project's numerous community liaisons who assist with implementing site-based activities. For some, it provided a valuable opportunity to engage with their neighbors.

"For me, as one who is a community leader and advocate, the intercept survey was a tool that I could use to engage people to talk about food," said Joyce Smith, who is active in a Southwest Baltimore community organization dedicated to educating people about diet, exercise, and the health disparities that affect Baltimore's low-income neighborhoods. "It helped spark an awareness that food has become an issue," she said. Smith also serves on the EFSNE advisory board, and added that these interactions also kindled interest in the project. "A lot of the people who agreed [to answer the survey questions] wanted to know why we were doing this."

Smith and others across the region collected the surveys during three distinct rounds at each of the project's 17 food stores. The team's goal was to administer 100 surveys per round in their urban sites and 50 surveys per round in their rural sites, which turned out to be challenging for some survey collectors, given the low-traffic nature of some stores or reluctance on the part of potential survey takers. But the team's bi-weekly group conference calls allowed survey collectors and site leaders to swap ideas for how to overcome such challenges.

For example, after one team member found that offering an incentive to shoppers resulted in a much better response rate, the team began providing gift cards to survey takers at all the sites in subsequent rounds. "Having the incentive of the gift card made all the difference," said Rachel Samuels, a Penn State Extension staff member and one of the Consumption Team's newest members who helped conduct surveys this spring at the project’s store in Pittsburgh. Samuels, who is proficient in Spanish, found that her language skills also helped her response rate, since she could administer the survey to several shoppers who preferred to be interviewed in Spanish.

Carol Giesecke, a team member based at Delaware State University, found people to be very willing to participate, and felt that having students involved in the survey collection provided some extra interest. "They're young and vivacious, and I think people were more willing to stop for them," she said. "I was very proud of my students. They did a great job and were very professional."

When the last of the surveys are collected, Consumption Team researchers will proceed with analyzing the data and writing manuscripts. Preliminary data analysis has already taken place at different points during the project timeline for various audiences. Most analyses thus far have focused on producing statistics that describe the samples of consumers interviewed, for use in reports given to the owners of stores where the surveys took place, or in working-paper drafts presented to both academic and non-academic audiences.

The thousands of data points collected by team members will allow them to answer a series of interesting questions regarding food-shopping habits of the respondents. For example, one line of research will pursue an investigation into the main barriers perceived by respondents regarding their purchase of "healthy foods." Other research questions are investigating whether there are differences in shopping patterns between respondents who participate in public assistance programs and those who don't, or differences between the days SNAP recipients receive their benefits versus non-benefit days.

More information about the Consumption Team's activities is available here.

-- by Kristen Devlin