Comparing shopper characteristics by online grocery ordering use among households in low-income communities in Maine

Abstract

Objective: Online grocery shopping could improve access to healthy food, but it may not be equally accessible to all populations—especially those at higher risk for food insecurity. This study aimed to compare the sociodemographic characteristics of families who ordered groceries online versus those who only shopped in-store. Design: We analyzed enrollment survey and 44 weeks of individually-linked grocery transaction data. We used univariate chi-square and t-tests and logistic regression to assess differences in sociodemographic characteristics between households that only shopped in-store and those that shopped online with curbside pick-up (online only or online and in-store). Setting: Two Maine supermarkets. Participants: 863 parents or caregivers of children under 18 years old enrolled in two fruit and vegetable incentive trials Results: Participants had a total of 32 757 transactions. In univariate assessments, online shoppers had higher incomes (P<0.0001), were less likely to participate in WIC or SNAP (P<0.0001), and were more likely to be female (P=0.04). Most online shoppers were 30–39 years old, and few were 50 years or older (P=0.003). After controlling for age, gender, race/ethnicity, number of children, number of adults, income, and SNAP participation, female primary shoppers (OR=2.75, P=0.003), number of children (OR=1.27, P=0.04), and income (OR=3.91 for 186–300% FPL and OR=6.92 for >300% FPL, P<0.0001) were significantly associated with likelihood of shopping online. Conclusions: In this study of Maine families, low-income shoppers were significantly less likely to utilize online grocery ordering with curbside pick-up. Future studies could focus on elucidating barriers and developing strategies to improve access.

Consumer characteristics and e-grocery services: the primacy of the primary shopper

“…Table 1 lists ten studies that have looked into the impact of gender on the adoption of e-grocery shopping. In order to avoid falling prey to selection or confirmation bias, Table 1 only lists studies that Dominici et al [ 1 , p. 2] and Zatz et al [ 2 , p. 4] cite—the idea being that these are the papers that they deemed representative of the state of the literature. 1 I did, however, add the papers by Dominici et al and Zatz et al themselves.…”

“…Looking at Table 1 , the results effectively appear contradictory. As mentioned, [ 1 , 2 ] suggest that this is due to differences in the periods and/or countries studied. While inter-country differences cannot be excluded (as will become clear in Sect.…”

“…3 I illustrate the issue based on the paper by Dominici et al, in Sect. 4 I show that several of the papers cited by [ 1 , 2 ] suffer from similar problems, and in Sect. 5 I perform a systematic literature search.…”

“…Dominici et al [ 1 , p. 2] talk about “conflicting results” but suggest that this is (merely) because the studies “cover a wide timeframe and concern countries located in different geographical areas”. Zatz et al [ 2 , p. 4] emphasise that comparisons “should be interpreted with caution given that [their] study is more recent than prior studies” and in view of “differences in settings and variation in methodology”.…”

“…The remainder of this methodological note unfolds as follows. In the next section I first document the disagreement signalled by [ 1 , 2 ]. In doing so, I focus on gender because this will allow me, in later sections, to demonstrate the methodological problem most intuitively.…”