International Journal of Retail & Distribution ManagementDoes urban form influence grocery shopping frequency? A study from Seattle,Washington, USAJunfeng Jiao Anne Vernez Moudon Adam Drewnowski

Article information:To cite this document:Junfeng Jiao Anne Vernez Moudon Adam Drewnowski , (2016),"Does urban form influence groceryshopping frequency? A study from Seattle, Washington, USA", International Journal of Retail &Distribution Management, Vol. 44 Iss 9 pp. 903 - 922Permanent link to this document:http://dx.doi.org/10.1108/IJRDM-06-2015-0091

Downloaded on: 15 September 2016, At: 20:19 (PT)References: this document contains references to 86 other documents.To copy this document: permissions@emeraldinsight.comThe fulltext of this document has been downloaded 38 times since 2016*

Users who downloaded this article also downloaded:(2016),"Customer experience quality and demographic variables (age, gender, education level, andfamily income) in retail stores", International Journal of Retail & Distribution Management, Vol.44 Iss 9 pp. 940-955 http://dx.doi.org/10.1108/IJRDM-03-2016-0031(2016),"Congruency as a mediator in an IKEA retail setting: Products, services and store image inrelation to sensory cues", International Journal of Retail & Distribution Management, Vol. 44 Iss9 pp. 956-972 http://dx.doi.org/10.1108/IJRDM-03-2016-0035(2016),"Investigating the effects of shyness and sociability on customer impulse buying tendencies:The moderating effect of age and gender", International Journal of Retail & DistributionManagement, Vol. 44 Iss 9 pp. 923-939 http://dx.doi.org/10.1108/IJRDM-12-2014-0166

Access to this document was granted through an Emerald subscription provided byToken:JournalAuthor:A9E47430-5825-45D7-A291-EA0B33B27D4B:

For AuthorsIf you would like to write for this, or any other Emerald publication, then please use our Emeraldfor Authors service information about how to choose which publication to write for and submissionguidelines are available for all. Please visit www.emeraldinsight.com/authors for more information.

About Emerald www.emeraldinsight.comEmerald is a global publisher linking research and practice to the benefit of society. The companymanages a portfolio of more than 290 journals and over 2,350 books and book series volumes, aswell as providing an extensive range of online products and additional customer resources andservices.

Emerald is both COUNTER 4 and TRANSFER compliant. The organization is a partner of theCommittee on Publication Ethics (COPE) and also works with Portico and the LOCKSS initiative fordigital archive preservation.

*Related content and download information correct at time of download.

Dow

nloa

ded

by D

octo

r Ju

nfen

g Ji

ao A

t 20:

19 1

5 Se

ptem

ber

2016

(PT

)

Does urban form influencegrocery shopping frequency?

A study from Seattle,Washington, USA

Junfeng JiaoUniversity of Texas at Austin, Austin, Texas, USA, andAnne Vernez Moudon and Adam DrewnowskiUniversity of Washington, Seattle, Washington, USA

AbstractPurpose – The purpose of this paper is to ascertain how elements of the built environment may ormay not influence the frequency of grocery shopping.Design/methodology/approach – Using data from the 2009 Seattle Obesity Study, the researchinvestigated the effect of the urban built environment on grocery shopping travel frequency in theSeattle-King County area. Binary and ordered logit models served to estimate the impact of individualcharacteristics and built environments on grocery shopping travel frequency.Findings – The results showed that the respondents’ attitude towards food, travel mode, and thenetwork distance between homes and stores exerted the strongest influence on the travel frequencywhile urban form variables only had a modest influence. The study showed that frequent shopperswere more likely to use alternative transportation modes and shopped closer to their homesand infrequent shoppers tended to drive longer distances to their stores and spent more time andmoney per visit.Practical implications – This research has implications for urban planners and policy makersas well as grocery retailers, as the seemingly disparate groups both have an interest in foodshopping frequency.Originality/value – Few studies in the planning or retail literature investigate the influence ofthe urban built environment and the insights from the planning field. This study uses GIS and aplanning framework to provide information that is relevant for grocery retailers and those invested infood distribution.Keywords Geographic information systems, Grocery shopping, Shopping frequency,Urban built environmentPaper type Research paper

IntroductionAs the retail branch of the food system, grocery shopping habits link retailmanagement and marketing, public health, and sustainability. Research concerning thefrequency of grocery shopping has implications for understanding consumerbehaviour and spending tendencies, fresh food intake, and food waste. Whileresearch regarding the effect of the urban environment on grocery shopping has tendedtowards discussions of obesity, policy, and health, studies investigating the consumerbehavioural component of grocery shopping tend towards decisions made in the storeand shopping demographics, but not the built environment. This study investigates International Journal of Retail &

Distribution ManagementVol. 44 No. 9, 2016

pp. 903-922©Emerald Group Publishing Limited

0959-0552DOI 10.1108/IJRDM-06-2015-0091

Received 23 June 2015Revised 7 September 2015

31 March 201631 May 2016

Accepted 8 July 2016

The current issue and full text archive of this journal is available on Emerald Insight at:www.emeraldinsight.com/0959-0552.htm

This research was supported by the Seattle Obesity Study funded by the National Institutes ofHealth. The paper has also greatly benefited from the comments of Professor Neil Towers and thetwo anonymous reviewers.

903

A study fromSeattle,

Washington,USA

Dow

nloa

ded

by D

octo

r Ju

nfen

g Ji

ao A

t 20:

19 1

5 Se

ptem

ber

2016

(PT

)

grocery shopping frequency as it relates to urban form using survey data from Seattle,Washington. Because this research is situated between retail management and thepublic health sector concerned with fresh food access, it has utility for both areas.

This study advances the literature on grocery shopping and the built environmentthrough the exploration of grocery shopping travel frequency in the Seattle-KingCounty area. Synthesising data from multiple King County sources, including the 2009Seattle Obesity Study (SOS), this research uses binary and ordered logit models toestimate how, in addition to social and demographic factors, the form of the urbanenvironment, including residential and job density, availability of at-ground parking,distance between home and store, property values, and restaurant proximity affectsgrocery shopping travel frequency. This research has implications for urban plannersand policy makers as well as grocery retailers, as they seemingly both have an interestin food shopping frequency.

BackgroundGrocery shopping is an important economic activity. According to the Bureau ofLabour Statistics, people living in the Seattle metropolitan area spent 12.8 per cent oftheir average annual expenditures on food in 2011-2012, which is close to the nationalaverage of 12.9 per cent. In comparison, Los Angeles averages 13.6 per cent andSan Francisco, 11.5 per cent – all at the 95 per cent confidence level (BOL Statistics,2014). Grocery shopping is also a very important non-work travel which constitutes asubstantial part of people’s travel activities. According to the 2009 National HouseholdTravel Survey, “Buy goods: groceries/clothing/hardware” was the second mostcommon travel activity in the USA, which was ranked after “Go home” and before“Go to Work” and accounting for 12, 34, and 11 per cent of all reportedUS household trips, respectively. Grocery shopping was identified as the most frequent“Buy goods” activity for American households compared to clothing and hardwareshopping (Krumm, 2012). Food Marketing Institute’s research reveals a nationalaverage of 1.6 trips per week to the supermarket with time and health concernsidentified as barriers to shopping (HartmanGroup, 2014).

Grocery shopping also has an impact on health outcomes. Researchers studied the3,000 households’ grocery shopping behaviours in New Orleans, Louisiana and foundthat produce consumptions, which often associated with body mass index (BMI), werepositively correlated with monthly grocery shopping frequency (Gustat et al., 2015).Researcher also found that people who frequently (at least once per week) shopped atfood co-ops had a significantly better diet quality than those who did not. People whofrequently (at least once per week) shopped at specialty shops or farmers’ markets hadsignificantly lower BMI and waist circumference than those who did not. Food price,quality, proximity, and convenient hours affected people’s grocery shopping frequencyand destination choice (Minaker et al., 2016).

Studies regarding grocery shopping largely fall into economic and public healtharenas, with the built environment considered often as a subset of one or the other.Within the economic and retail arena, studies indicate that individuals’ attitudestowards food stores and established shopping behaviours influence shoppingfrequencies. Operational elements of shopping, such as travel mode ( Jiao et al., 2011,Clifton et al., 2012) and average time and money spent (Ma et al., 2011) correlate withtravel frequency. Socioeconomic and demographic factors also correlate with groceryshopping frequency. Household size, number of workers, income, and age areassociated with more frequent trips (Bawa and Ghosh, 1999; Blaylock, 1989;

904

IJRDM44,9

Dow

nloa

ded

by D

octo

r Ju

nfen

g Ji

ao A

t 20:

19 1

5 Se

ptem

ber

2016

(PT

)

Clifton, 2004; Yoo et al., 2006). The effects of frequent grocery shopping have beenexplored as well, with frequent shoppers spending less money and time at the storesper visit (Bawa and Ghosh, 1999; Beatty, 2008) and increasing fruit and vegetableintake (Reedy et al., 2014).

Public health research focusing on obesity and food access often concerns groceryshopping. Research shows fruit and vegetable intake is influenced by personal opinionsof an area’s food offerings (Reedy et al., 2014; Hollywood et al., 2013). In addition, fruitand vegetable intake increases with additional supermarkets in a given census tract(Morland and Evenson, 2009; Morland et al., 2002). The size of grocery stores may alsoinfluence public health, as obesity rates are lower in areas with supermarkets comparedto those with a high prevalence of fast food and smaller food stores (Morland andEvenson, 2009). The food environment around homes such as the density of largegrocery stores, convenience stores, and fast food restaurants within two and five milesof people’s homes were correlated with people’s fruit, vegetable, and fibre intake(Reitzel et al., 2016).

Urban form has a direct impact on people’s travel mode for both work andnon-work travels (e.g. grocery shopping) (Rajamani et al., 2003). Different travelmodes further influence people’s time and economic travel budget (Milakis et al.,2015b) and impose constraints on people’s ability to carry or move goods(e.g. groceries) (Lee and Moudon, 2006a; Handy, 1996). From the destinationperspective, urban form characteristics include the spatial distribution of routinedestinations, and specially food establishments within the cities (Black et al., 2011;Lamb et al., 2015). Researchers found that the average distance from homes togrocery stores played a significant role in determining people’s grocery shoppingtravel mode. Stores with more at-ground parking were more likely to attractcar-based visits in the Seattle area ( Jiao et al., 2011). This finding was confirmed inSweden and Germany, where people rated that accessibility by car was the mostimportant grocery store attribute (Anne Wiese et al., 2015). Thus that the builtenvironment around people’s homes and grocery stores would affect their groceryshopping travel behaviour.

MethodsTheoretical frameworkUsing findings from previous research, this study classified factors influencinggrocery shopping travel frequency into four categories: socioeconomic status (Bawaand Ghosh, 1999; Kim and Park, 1997; Blaylock, 1989; Frusti et al., 2002), attitudestowards food (Beatty, 2008; Wiig and Smith, 2009; Van Loo et al., 2010), general foodshopping behaviour (Kahn and Schmittlein, 1989; Bawa and Ghosh, 1999; Wilde andRanney, 2000; Handy and Clifton, 2001; Ma et al., 2011; Maxwell et al., 2010), and thecharacteristics of the urban built environment around an individual traveller’s homeand grocery store (Agyemang-Duah et al., 1996; Boarnet and Crane, 2001; Ewing andCervero, 2010; Boarnet, 2011). Socioeconomic status refers to demographiccharacteristics of the sample population, including age, race, and householdincome. Attitudes towards food included measures and perceptions of consumers’expectations whether food should be inexpensive or food should be easy to cook.General food shopping behaviour pertains to consumers’ characteristics such as theaverage money and time spent in the store, as well as measures like travel distanceand mode. Finally, factors comprising the urban built environment included itemslike the amount of at-ground parking and the density of food stores around the home.

905

A study fromSeattle,

Washington,USA

Dow

nloa

ded

by D

octo

r Ju

nfen

g Ji

ao A

t 20:

19 1

5 Se

ptem

ber

2016

(PT

)

Data sourcesThis analysis synthesised data from several sources within the King County area.The 2009 SOS telephone survey provided data on grocery shopping travel and thetravellers’ socioeconomic characteristics. This effort sampled residents of theSeattle-King County area and included 2,001 respondents to a 20-minute survey whichhad 89 questions regarding diet quality, food shopping habits and expenditures, foodinsecurity, perception of home neighbourhood, and stores. The University of WashingtonCentre for Obesity Research (UWCOR) furnished data about individual store food costs.Objective data on the urban built environment came from the King County assessor andinformation regarding residential and employment density, property values, at-groundparking, and food and fitness establishments was supplied and developed by the UrbanForm Lab at the University of Washington. The original food establishment data wereprovided by the Department of Public Health of Seattle and King County, while theoriginal data for fitness facilities came from InfoUSA. The King County GIS Centreprovided road network and other urban transportation infrastructure data. The studywas approved by the University of Washington Institutional Review Board.

MeasurementsGeocoding and store classification by food cost. Of the 2,001 SOS respondents’ homeaddresses, 1,994 were geocoded in GIS. While 1,985 respondents reported the addressof their primary grocery store, the store address, and the related shopping frequencycould only be verified for 1,862 respondents. Geocoding results showed that theserespondents shopped at 207 grocery stores (Figure 1). UWCOR categorised the grocerystores based on the total cost of 100 commonly consumed food items from eightdifferent supermarket chains (Mahmud et al., 2009). Of the 207 stores, 80 were classifiedas low-cost, 110 as medium-cost, and 17 as high-cost stores.

Dependent variablesThe respondents’ travel frequency, or number of reported trips, to the reported primarygrocery store was the dependent variable. As the survey interviewed primary shoppers,or those who completed the most grocery shopping trips in their households, travelfrequency represented grocery shopping behaviour at the household level. Of the1,862-respondent sample, 512 (27 per cent) respondents reported shopping no more thantwice per month, 434 (23 per cent) respondents reported shopping once a week, 787(42 per cent) respondents reported shopping two or three times per week, 92 (5 per cent)and 36 (2 per cent) reported shopping four to six times or more than six times per week.The 946 residents who reported shopping no more than once per week (50.8 per cent) wereclassified as infrequent shoppers. The remaining 916 respondents (49.2 per cent) of thethree groups who reported grocery shopping at least two or three times per week wereclassified as frequent shoppers (Progressive Grocer, 2002). To understand the relationshipbetween the urban built environment and grocery shopping frequency, the researchersbuilt one binary model and one ordered logit model (Huntsinger et al., 2013; Nijland et al.,2013; Arentze et al., 2011). The binary logit model used the aggregated travel frequency(infrequent vs frequent shoppers) as the dependent variable. The ordered logit model usedthe five more detailed travel frequency categories as the dependent variable.

Independent variablesIndependent variables were selected from the following four domains: socioeconomic status,attitude towards food, general food shopping behaviour, and the urban built environment.

906

IJRDM44,9

Dow

nloa

ded

by D

octo

r Ju

nfen

g Ji

ao A

t 20:

19 1

5 Se

ptem

ber

2016

(PT

)

Socioeconomic variables included respondents’ age, gender, education level, householdincome, household size, and the number of children and cars within the household (Hansonand Hanson, 1981; Blaylock, 1989; Benekohal et al., 1994; Bawa and Ghosh, 1999; Frustiet al., 2002; Yoo et al., 2006). Attitude variables included whether the respondents thoughtfood should be inexpensive, easy to cook, and healthy (Beatty, 2008). Shopping behaviourvariables included respondents’ travel mode and travel distance to store, primary grocerystore type by cost of food in the store, and average money and time spent per visit in thestore (Bawa and Ghosh, 1999; Wilde and Ranney, 2000; Yoo et al., 2006).

Urban built environment variables included four different sub-categories as follows(Table I): general neighbourhood environment, measured as residential andemployment density, and residential property values to capture neighbourhoodwealth (Cervero and Kockelman, 1997; Frank and Pivo, 1994; Chen et al., 2008); routineneighbourhood food-related destinations, measured as food stores and restaurantsaround the respondents’ homes and preferred grocery stores (Lee and Moudon, 2006b;

N

Homes

Grocery Stores

King County

Waterbody

Miles52.50

Figure 1.Spatial distribution

of the SOSrespondents’ homes

and primarygrocery stores

907

A study fromSeattle,

Washington,USA

Dow

nloa

ded

by D

octo

r Ju

nfen

g Ji

ao A

t 20:

19 1

5 Se

ptem

ber

2016

(PT

)

McCormack et al., 2008); transportation infrastructure, measured as the density of theroad network, the count of signals and bus stops, the presence of trails, and the amountof parking at ground (Cervero and Radisch, 1996; Greenwald, 2003; Handy and Clifton,2001; Lee and Moudon, 2006b; Steiner, 1998); and traffic conditions, measured as theestimated annual average daily traffic (AADT) on major streets, and the daily count ofbus riders per bus stop (Moudon and Hess, 2000).

Data capture in GISArcGIS Network Analyst and the ESRI StreetMap data were used to ascertain thenetwork distances between homes and respondents’ primary grocery stores. Traveldistance was measured in miles using the shortest driving time distance impedance ofthe ESRI StreetMap data, which considered one-way streets, speed limits, over- andunder-passes, transit-only lanes, etc. Count and density measures of urban builtenvironment variables around each respondent’s home and around the primarygrocery store were derived from the King County parcel GIS data set and summedwithin a half-mile (833 m) buffer. The size of the buffer corresponded to a ten-minutewalking distance, which is a catchment area convention used in other research(Cervero, 1996; Hess et al., 1999; Guo et al., 2007).

AnalysisCorrelation analyses served to test the relationship between the independent variablesselected in each of the four domains. If two variables were highly correlated with

Domain Sub domain Variables Measures (within a 0.5 mi airline buffer)

Builtenvironment

Neighbourhoodenvironmentvariables

Net residentialdensity

Number of units per acre

Net employmentdensity

Estimates ( jobs) per acre

Residential propertyvalue

Assessed property value per residentialunit

Food stores Supermarket Count of placesGrocery storeEthnic food storeConvenience store

Restaurants Traditional restaurant Count of placesEthnic diningFast foodQuick serviceCoffee shopTaverns and pubs

Transportationinfrastructure

Major street density Freeway and arterials in milesOther street density Streets with speed limit ⩽45 mph in milesSignalization Count of traffic signs and signalsParking Count of on ground parking stallsBus stops Count of bus stopsTrail Have trail or not

Trafficconditions

Bus ridership Counts of bus riders per bus stopTraffic volumes Average estimated AADT on major street

Table I.Independentvariables

908

IJRDM44,9

Dow

nloa

ded

by D

octo

r Ju

nfen

g Ji

ao A

t 20:

19 1

5 Se

ptem

ber

2016

(PT

)

each other, the variable with theoretical support and the strongest backing from previousresearch was kept. Bivariate analyses examined the relationship between the uncorrelatedindependent variables and the dependent variable. T-statistical tests were used forcontinuous variables and χ2 tests were used for categorical variables. Only significantvariables were included in the final models.

Binary and ordered logistic regressions were used to test the relationshipbetween independent variables and the likelihood of respondents to be frequentor infrequent shoppers. First, four binary logistic regression models werecreated and tested. Model 1, the base model, only included the respondents’socioeconomic characteristics. In model 2, respondents’ attitudes towards foodand general shopping behaviour were added to the base model. Model 3 addedthe urban built environment variables of the home neighbourhood and model 4 wasthe final binary model, which added the urban built environment variables of theprimary store neighbourhood. In models 2, 3, and 4, variables were added one at atime to the previous models.

Variables in the final binary model were kept in the ordered model to furthertest the relationship between the urban built environment and people’s travelfrequency to grocery stores (Bhat and Guo, 2004; Habib and Miller, 2008; Srinivasanand Bhat, 2005).

ResultsDescriptive analysisOf the 1,862 SOS survey respondents included in the analysis, 62 per cent were femaleand 38 per cent were male. Survey respondents had an average age of 54.45 years andan average household income class of between $50,000 and $75,000. For modellingpurposes, age was treated as a categorical variable and separated into three categoriesas young adult: 18-44 (26 per cent), middle-aged: 45-64 (50 per cent), and older adult:⩾65 (24 per cent) (US Census Bureau, 2009). The average household size was 2.31persons, with an average of 0.22 children between the age of 12 and 18. In total,81 per cent of the sample had at least 1-3 years of college education. The averagenumber of cars per household and per household adult was 1.74 and 1.01, respectively.The sample population was different from the King County population in that it wasolder, more educated, and more female (US Census Bureau, 2009) (Table II).

VariableAll

sampleFrequentshoppers

Infrequentshoppers

KingCounty

Average age 54.53 54.48 54.63 37.00Average household income 50k-75k 50k-75k 35k-50k 70ka

Average household size 2.31 2.31 2.29 2.37Average travel distance(mile) 3.12 2.25 3.99 NAGender (male) 38% 41% 35% 50%No. of children 12-18 yrs old 0.22 0.24 0.19 0.39b

No. of cars per household 1.74 1.79 1.68 1.67No. of cars per household adultc 1.01 1.03 0.98 1.01Education (1-3 years collegeeducation) 81% 83.1% 78.9% 73.5%Notes: aMedian household income; b0-14 years old children (US Census Bureau, 2009); cadult over 18-years-old

Table II.Socioeconomic status

of Seattle ObesityStudy samples and

King County’spopulation

909

A study fromSeattle,

Washington,USA

Dow

nloa

ded

by D

octo

r Ju

nfen

g Ji

ao A

t 20:

19 1

5 Se

ptem

ber

2016

(PT

)

In total, 88 per cent of the 1,862 respondents reported driving to their primary grocerystore with 69 per cent shopping at low-cost grocery stores. In total, 12 per cent of therespondents reported using alternative transportation modes for grocery shopping.Among them, 156 respondents walked to the primary grocery store,62 respondents took public transit, and four respondents rode bicycles.

Respondents’ average grocery store trip length was 33.82 minutes with an averagegrocery bill of $68.71. The average travel distance from home to the primary grocerystore was 3.12 miles. Average residential density around the homes was 4.75 units peracre and there were five food stores (including supermarkets, non-chain grocery stores,ethnic food stores, and convenience stores) and 3,816 at-ground parking stalls within thehalf-mile airline buffer around the homes. Job density and property values around theprimary store averaged 11 jobs per acre and $225,000 per residential unit, respectively.On average, there were 43 different restaurants (which included traditional restaurants,ethnic dining, fast food and quick service restaurants, coffee shops, and tavern/pubs)within the half-mile airline buffer around the primary grocery stores (Table III).

Model resultsFive independent variables were significant in the base model (age, gender, householdincome, car ownership, and the number of 12-18 year old children). In total, 12 variableswere significant in the one-by-one stepwise test process, including the number of foodstores, the amount of at-ground parking, the residential densities around the home, andthe number of restaurants around stores. These variables were included in the finalmodel (Table III).

For the final binary logit model, income and the number of 12-18-year-old childrenwithin the household were positively correlated with the likelihood of being a frequentshopper with an odds ratio of 1.078 and 1.564, respectively. The respondents’ age lostsignificance in the final model and respondents’ gender, and the number of cars peradult member remained insignificant. Thinking food should be inexpensive(OR¼ 0.731), driving to the primary grocery store (OR¼ 0.525), average traveldistance to the store (OR¼ 0.851), and average money and time spent in the store(OR¼ 0.985 and 0.981) were negatively related to the likelihood of being a frequentshopper. Compared to the respondents who shopped at high-cost grocery stores,respondents who shopped at low- or medium-cost stores were less likely to be frequentshoppers (OR¼ 0.436 and 0.478).

Three built environment variables around homes were included in the final binarymodel. Residential density and the number of food stores (grocery store andsupermarket) around homes were not significant in the final binary model. Only theamount of at-ground parking around homes was negatively correlated with the groceryshopping travel frequency (OR¼ 0.860).

Three built environment variables around stores were included in the final binarymodel. They were job density, restaurant proximity, and average property valuearound the stores. Among them, only the average property value around stores wassignificant and negatively correlated with travel frequency (OR¼ 0.655). Both jobdensity and restaurant proximity were insignificant in the final binary model.

For the ordered logit model, the household income and the number of 12-18-year-oldchildren were positively correlated with grocery shopping travel frequency.This confirmed the results of the binary logit model. Being a female or a youngadult (18-44 years old), which were not significant in the binary model, becamepositively correlated with the travel frequency in the ordered logit model.

910

IJRDM44,9

Dow

nloa

ded

by D

octo

r Ju

nfen

g Ji

ao A

t 20:

19 1

5 Se

ptem

ber

2016

(PT

)

A:d

omain

B:v

ariables

C:measures

D:d

escriptiv

estatistics

E:treated

inthemodels

F:one-by

-one

testing

Individu

aland

Socioeconomic

Status

Age

Year

Young

adult:18-44(26%

),middleaged:45-64

(50%

),olderadult:⩾6

5(24%

)Ca

t.Basemodel

Household

income

Incomecategories

Min:1,m

ax:15,mean:2.31,m

edian:2,SD

:1.44

Cont.

Basemodel

Gender

Male/female

Male:724,female:1,179

Cat.

Basemodel

No.of

cars

per

household

adult

Coun

tMin:0,m

ax:4,m

ean:1.01,m

edian:1,SD

:0.54

Cont.

Basemodel

No.of

12-18

year

old

child

ren

Coun

tMin:0,m

ax:4,m

ean:0.22,m

edian:0,SD

:0.55

Cont.

Basemodel

Attitu

detowards

food

Think

ingfood

should

beinexpensive

Yes/no

Yes:986,n

o:603

Cat.

“−”

Food

shopping

behaviour

Driving

togrocerystore

Yes/no

Driving

:1,640,N

on-driving

:222

(walking

:156,

publictransit:62,b

iking:

4)Ca

t.“−

”

Distancefrom

hometo

prim

ary

grocerystore

Networkdistance

inmiles

Min:0.02,max:59.20,m

ean:3.12,m

edian:2.04,

SD:4.40

Cont.

“−”

Storetype

bycost

Low/m

edium/highcost

Low:585,m

edium:1,127,and

high

:190

Cat.

“+”

Average

money

spent

pervisit

Dollar

Min:1,m

ax:350,m

ean:

70.73,median:

50,S

D:

53.85

Cont.

“−”

Average

time

spentpervisit

Minute

Min:4,m

ax:120,m

ean:

34.5,m

edian:

30,S

D:

18.05

Cont.

“−” (con

tinued)

Table III.Independent

variables includedin final models

911

A study fromSeattle,

Washington,USA

Dow

nloa

ded

by D

octo

r Ju

nfen

g Ji

ao A

t 20:

19 1

5 Se

ptem

ber

2016

(PT

)

A:d

omain

B:v

ariables

C:measures

D:d

escriptiv

estatistics

E:treated

inthemodels

F:one-by

-one

testing

Built

environm

ent

around

home

andstore

Residentia

ldensity

around

home

Unit/a

cre

Min:0,m

ax:30,mean:

4.75

median:

3.30,S

D:

4.85

Cont.

“−”

No.of

food

stores

around

home

Coun

tMin:0,m

ax:40,mean:5.03

median:3,SD

:7.02

Cont.

“−”

At-g

roun

dparking

around

home

Coun

tof

atground

parking(log

transformed)

Originalv

alue

(min:0,m

ax:22,076,mean:

3,816,median:2,769,SD

:3581)

Logvalue

(min:0,m

ax:10,mean:8.25,m

edian:7.93,SD:

8.18)

Cont.

“−”

Jobdensity

around

store

Job/acre

Min:0,m

ax:253,m

ean:

11.03median:

4.73,S

D:

24.70

Cont.

“+”

Average

property

value

around

store

Dollar/un

itOriginalv

alue

(min:82k,m

ax:1,015k,

mean:

225k,m

edian:

207k,S

D:80k)L

ogvalue(m

in:

11.32,max:13.83,m

ean:

12.26,median:

12.24,

SD:0.35)

Cont.

“−”

No.of

restaurants

around

store

Coun

tMin:0,m

ax:518.96,mean:42.56median:26.89,

SD:57.60

Cont.

“+”

Table III.

912

IJRDM44,9

Dow

nloa

ded

by D

octo

r Ju

nfen

g Ji

ao A

t 20:

19 1

5 Se

ptem

ber

2016

(PT

)

Similar to the binary model, all six attitude and behaviour variables (thinking foodshould be inexpensive, driving to store, shopping at low- or medium-cost store, averagenetwork distance between homes and stores, and average time or money spent atstores) remained their significance and were negatively correlated with groceryshopping frequency in the final ordered model.

Regarding the built environment variables around the homes, the residential densityaround homes was weekly correlated with grocery shopping travel frequency.The number of at-ground parking around homes was still negatively correlated withgrocery shopping frequency. For variables regarding the built environment around thegrocery store, the average property value around stores was negatively correlated withtravel frequency. Both job density and having a restaurant around stores were notsignificant in the final ordered logit model (Table IV).

Discussion of resultsBoth the binary and ordered logit model showed that the travel frequency to theprimary grocery store was associated with individual socioeconomic characteristics,attitudes towards food, general shopping behaviour, and the urban built environmentaround the home and primary grocery store. With at least two variables significant inthe final regression models, the respondents’ socioeconomic variables clearly affectedtravel frequency. Variables in the attitude and general shopping behaviour categoryexerted a very strong influence on travel frequency. All six variables in this categorywere significant in both models. The built environment variable had a direct, butmodest influence on travel behaviour, which confirmed findings in previous studies(Ewing and Cervero, 2010; Bhat and Guo, 2007; Cao et al., 2009; Milakis et al., 2015a).

Demographic and socioeconomic characteristicsHousehold income and the number of teenage (12-18) children within the householdwere significant in both binary and ordered logit models. Compared to infrequentshoppers, those that shopped frequently lived in higher income households with moreteenage children. The findings about teenage children on grocery shopping travel isnew to the literature and suggests when planning for new grocery stores, both private(grocery store operator) and public sectors (urban planner/government officials) shouldwell consider the increasing demand generated by the teenage population. Similarresults about household income on trip frequency were reported by previous studies(Yoo et al., 2006; Blaylock, 1989; Bawa and Ghosh, 1999; Mortimer and Clarke, 2011;Frusti et al., 2002). Being a young adult (18-44) or a female was significantly correlatedwith more grocery shopping travel in the ordered logit models. Compared to males,females were 19.7 per cent more likely to shop at the grocery stores one or more timesper week, supporting findings in previous literature (Feng et al., 2013; Gentina andBonsu, 2012; Luceri and Latusi, 2012; Frank et al., 2009). The number of cars per adultwas significant in the base model. It lost its significance in both final models afterincluding other car related variables such as: driving to the grocery store or not, and thenetwork distance between homes and primary grocery stores. It’s expected that thevariance explained by car ownership could be explained by these two variables.

Attitude and behaviourThe attitude and behaviour variables showed a very strong influence on groceryshopping frequencies in both final models. Compared to infrequent shoppers,

913

A study fromSeattle,

Washington,USA

Dow

nloa

ded

by D

octo

r Ju

nfen

g Ji

ao A

t 20:

19 1

5 Se

ptem

ber

2016

(PT

)

Binarylogitmodel(n=1,862)

Ordered

logitmodel(n=1,862)

Coefficient

Sign

ificance

Odd

.ratio

(OR)

95%

CIOR

Coefficient

Sign

ificance

95%

CI

Individu

aldemograph

icsandsocioeconomic

Individu

aldemograph

icsandsocioeconomic

Age

(18-44)

0.285

0.086

1.330

0.961

1.840

0.376

0.007

0.104

0.648

Age

(45-64)

0.099

0.485

1.104

0.837

1.455

0.232

0.053

−0.003

0.467

Age

(⩾65)

0.000

0.000

Gender(female)

−0.133

0.245

0.875

0.699

1.096

0.197

0.040

0.009

0.385

Household

income

0.075

0.000

1.078

1.035

1.122

0.089

0.000

0.055

0.124

No.of

12-18yrs.oldchild

ren

0.447

0.000

1.564

1.255

1.948

0.446

0.000

0.271

0.622

No.of

cars

peradult

0.174

0.132

1.190

0.949

1.493

0.152

0.107

−0.033

0.337

Attitu

deandbehaviour

Attitu

deandbehaviour

Thinkingfood

should

beinexpensive

−0.313

0.005

0.731

0.586

0.912

−0.272

0.004

−0.457

−0.086

Driving

toStore

−0.644

0.003

0.525

0.344

0.802

−0.609

0.001

−0.953

−0.265

Storetype(low)

−0.831

0.000

0.436

0.274

0.693

−0.803

0.000

−1.174

−0.431

Storetype(m

edium)

−0.739

0.000

0.478

0.316

0.722

−0.700

0.000

−1.027

−0.373

Storetype(high)

Networkdistance

home_

store(m

ile)

−0.162

0.000

0.851

0.810

0.893

−0.133

0.000

−0.166

−0.099

Average

money

spentat

store($)

−0.015

0.000

0.985

0.982

0.988

−0.016

0.000

−0.018

−0.014

Average

timespentat

store(m

in)

−0.020

0.000

0.981

0.973

0.988

−0.016

0.000

−0.022

−0.010

Hom

eUrban

BuiltEnv

ironment

Hom

eUrban

BuiltEnv

ironment

Residentia

ldensity

around

home

−0.023

0.139

0.978

0.949

1.007

−0.026

0.043

−0.050

−0.001

No.of

food

stores

around

home

0.090

0.236

1.094

0.943

1.269

0.097

0.120

−0.025

0.219

No.of

atground

parkingaround

home(LN

)−0.151

0.000

0.860

0.807

0.916

−0.118

0.000

−0.169

−0.067

Storeurbanbu

iltenvironm

ent

Storeurbanbu

iltenvironm

ent

Jobdensity

around

store

0.000

0.888

0.888

0.995

1.005

0.001

0.522

−0.003

0.006

Average

property

valuearound

store(LN)

−0.422

0.014

0.655

0.467

0.919

−0.393

0.007

−0.677

−0.109

Havearestau

rant

around

store

0.087

0.603

1.091

0.786

1.513

0.069

0.618

−0.202

0.304

Constant

9.061

0.000

Notes

:Bold,

sign

ificant

at0.05

level,ita

lic,categorical

variable,and

LN,n

atural

logtransform

Table IV.Final binary logitand orderedlogit models

914

IJRDM44,9

Dow

nloa

ded

by D

octo

r Ju

nfen

g Ji

ao A

t 20:

19 1

5 Se

ptem

ber

2016

(PT

)

frequent shoppers were less likely to agree that food should be inexpensive and morelikely to shop at the high-cost grocery stores. This suggested they are less economicallyconstrained than infrequent grocery shoppers, who might limit their grocery storevisits due to financial reasons (Clifton, 2004; Darko et al., 2012). Frequent shoppers wereless likely to drive and travelled significantly shorter distances to grocery stores(Table II: 2.55 vs 3.99 miles), which suggested they lived closer to grocery stores andthus might have more choice of foods in their neighbourhood. Further, frequentshoppers spent significantly less money and time per store visit than infrequentgrocery shoppers. This suggested that frequent shoppers were more concerned withthe freshness of the groceries and thus may have better health statuses than infrequentshoppers (Minaker et al., 2016).

As expected, the network distance between homes and stores was negativelycorrelated with travel frequency. Previous studies using the same dataset found thatdrivers travelled significantly farther to the store than non-drivers (3.33 vs 1.52 miles)( Jiao et al., 2011; Hillier et al., 2011). This indicated that shorter distances betweenhomes and stores might increase grocery shopping frequency by modes other thandriving (Handy and Clifton, 2001; Whelan et al., 2002). This finding is interesting forboth policy makers and grocery retailers as they could increase people’s healthy foodexposures by building more urban grocery stores and still potentially reducing relatedcar traffics. The increasing popularity of small urban grocery stores (e.g. Whole Foods,Trader Joe’s) and the shrinking of big box grocery stores (e.g. Walmart, Target) seem tosupport this finding (Tuttle, 2014).

Urban form around the homes and storesUrban form factors had a modest influence on grocery shopping frequency.The number of at-ground parking around homes was negatively correlated withgrocery shopping frequency in both final models. Further tests showed that thevariable was negatively associated with the average distance between homes andstores. This indicated that respondents living in suburban neighbourhoods with lotsof at-ground parking stalls were more likely to make fewer but longer distancegrocery shopping trips. Residents living in urban neighbourhoods with fewerat-ground parking stalls were more likely to make more but shorter trips. Similarfindings about land use mix on general shopping behaviour were reported inprevious studies (Agyemang-Duah et al., 1996; Duncan et al., 2010; Boarnet et al.,2011), in which shopping activities were more likely to be chained with otheractivities (i.e. dining out and entertainment activities) (Dawson and Lord, 2012;Boarnet, 2011). The average property value around stores was negatively correlatedwith grocery shopping frequency in both models. This may point to big box grocerystores mostly locating in low-cost suburbs. Consumers usually do not visit thesestores often, but when they do, they likely spend more time and money in the store tojustify the high-travel cost (Darko et al., 2012; Hillier et al., 2011).

The authors also tested the interaction terms between income, gender, and age withbuilt environment variables around homes and stores. The results showed that incomewas correlated with both the residential density around homes and the at-groundparking around homes. However, when introducing the interaction terms of income anddensity, as well as income and at-ground parking space around homes, to the two finalmodels, none of them were significant. Age and gender were not correlated with anyhome or store built environment variables and the corresponding interaction termswere not significant in either final model.

915

A study fromSeattle,

Washington,USA

Dow

nloa

ded

by D

octo

r Ju

nfen

g Ji

ao A

t 20:

19 1

5 Se

ptem

ber

2016

(PT

)

SummaryOverall, ten variables were significant in both binary and ordered logit models.Summarizing these findings, we found that frequent grocery shoppers were more likelyto be female or young adults. They usually had higher household incomes and moreteenage children in the family. They often shopped at high-cost grocery stores and wereless concerned about food cost. Compared to infrequent shoppers, frequent shoppersusually lived in higher density neighbourhoods with fewer at-ground parking andcloser distances to grocery stores. These urban form characteristics affected theplausible transportation modes available to them and their grocery shopping travelbudgets (measured by time and money). Because of the shorter distance from homes tostores, they were able to visit these stores more often with non-motorisedtransportation modes and also benefited from stronger healthy food exposures thaninfrequent shoppers.

Compared to frequent shoppers, the infrequent shoppers were more likely to be maleor older adults. They usually had lower household incomes and fewer teenage childrenin the household. They tended to shop at low- or medium-cost grocery stores and thinksfood should be inexpensive, which could potentially reduce their healthy foodexposures. In terms of urban form characteristics, they usually lived in low-densityneighbourhoods with more at-ground parking spots around homes and further awayfrom their primary grocery stores. These factors again influenced the plausible travelmodes available to them and affected their grocery shopping time and cost budgets.Due to economic (lower household income) and urban form (longer distance)constraints, they usually made fewer grocery shopping trips. But when they did, theywould drive further away to stores and spend more time and money per visit perhapsas a way to justify the travel cost (Chung and Myers, 1999; Hubley, 2011; Michimi andWimberly, 2010).

The findings inform both public health and transportation planning policies.By allocating smaller grocery stores to lower income neighbourhoods (e.g. food deserts)( Jiao et al., 2012), urban planners and policy makers could reduce the average traveldistances between homes and stores, which in turn would increase local residents’grocery shopping frequency and their healthy food exposures (Minaker et al., 2016).Similar movements have been observed in New York City, Philadelphia, Pittsburgh,Baltimore, and the State of Maryland, where some big box grocery stores have beenreplaced by smaller urban neighbourhood grocery stores (NSOPP Center, 2016).

Meanwhile, by introducing more neighbourhood grocery stores and reducing thenumber of at-ground parking around homes, planners and policy makers couldincrease residents’ probability of using nonmotorized travel modes for groceryshopping (Van and Senior, 2000; Schwanen et al., 2004; Zhang, 2004) and potentiallyreplace grocery shopping trips with local shopping activities (Ewing et al., 1996; Handyand Clifton, 2001). Lastly, by introducing neighbourhood grocery stores, planners, andpolicy makers could also potentially reduce people’s auto dependence for groceryshopping (Handy and Clifton, 2001) and remove potential transportation barriers forthe underserved populations ( Jiao and Dillivan, 2013).

LimitationsThe generalisability of the findings was limited by the following facts. The sampleswere older and included more women than the aggregate King County population,which might not be representative of the food shopper population in the area. Grocerystores were not evenly distributed in the study area, with more grocery stores located in

916

IJRDM44,9

Dow

nloa

ded

by D

octo

r Ju

nfen

g Ji

ao A

t 20:

19 1

5 Se

ptem

ber

2016

(PT

)

the north Seattle area increasing the risk of spatial autocorrelation. Data on therespondents’ grocery shopping travel frequencies and travel times were self-reportedand collected at the individual level rather than from the household level. Furthermore,statistical models were affected by the multicollinearity. A structural equation modelwould provide a better understanding about the causal relationships between variables(Cao et al., 2009; Golob, 2003). Also, no information about possible trip chaining wascollected in the survey, which limited the possibility to explore detailed travelbehaviour. Lastly, characteristics of the urban built environment and urban form in theSeattle-King County area might differ from those of other metropolitan areas. Similarstudies in other regions would help to better understand the relationship between thebuilt environment and people’s grocery shopping travel behaviour.

ConclusionThe study showed that individual travellers’ socioeconomic status, attitude, andgeneral food shopping behaviour (e.g. thinking food should be inexpensive, shopping athigh-cost stores, driving to grocery stores, or money and time spent per store visit), aswell as the urban form surrounding both their homes and their primary grocery stores,were correlated with travel frequency to their primary grocery stores. Among them,people’s attitude and general food shopping behaviour had the strongest influence ongrocery shopping frequency. Driving to stores, thinking food should be inexpensive,and the network distance between homes and stores had a major negative impact ongrocery shopping travel frequency, while the urban built environment factors had amodest influence on travel frequency.

Urban form factors affected the spatial distribution of grocery stores and the relatedtravel modes available for grocery shopping, which together influenced the monetaryand temporal costs of travel for grocery shopping. Specifically the number of at-groundparking stalls around homes and the average property value around the stores werenegatively associated with grocery shopping travel frequency. Frequent shoppers weremore likely to shop at nearby medium- or high-cost grocery stores and usenon-motorised transportation modes. However, infrequent shoppers were more likely todrive significantly longer distances to low-cost grocery stores located in remote areas.

References

Agyemang-Duah, K., Aderson, W.P. and Hall, F.L. (1996), “Trip generation for shopping travel”,Transportation Research Record: Journal of the Transportation Research Board, Vol. 1493,pp. 12-20.

AnneWiese, A.P.S.Z., Waldemar Toporowski, D., Wiese, A., Zielke, S. and Toporowski, W. (2015),“Shopping travel behaviour: influencing factors, shopper types and environmentalconsequences”, International Journal of Retail & Distribution Management, Vol. 43 Nos 4/5,pp. 469-484.

Arentze, T.A., Ettema, D. and Timmermans, H.J. (2011), “Estimating a model of dynamic activitygeneration based on one-day observations: method and results”, Transportation ResearchPart B: Methodological, Vol. 45 No. 2, pp. 447-460.

Bawa, K. and Ghosh, A. (1999), “A model of household grocery shopping behavior”, MarketingLetters, Vol. 10 No. 2, pp. 149-160.

Beatty, T.K.M. (2008), “Expenditure dispersion and dietary quality: evidence from Canada”,Health Economics, Vol. 17 No. 9, pp. 1001-1014.

917

A study fromSeattle,

Washington,USA

Dow

nloa

ded

by D

octo

r Ju

nfen

g Ji

ao A

t 20:

19 1

5 Se

ptem

ber

2016

(PT

)

Benekohal, R.F., Michaels, R.M., Shim, E. and Resende, P.T.V. (1994), “Effects of aging on olderdrivers’ travel characteristics”, Transportation Research Record, Vol. 1438, pp. 91-98.

Bhat, C.R. and Guo, J. (2004), “A mixed spatially correlated logit model: formulation andapplication to residential choice modeling”, Transportation Research Part B:Methodological, Vol. 38 No. 2, pp. 147-168.

Bhat, C.R. and Guo, J.Y. (2007), “A comprehensive analysis of built environment characteristics onhousehold residential choice and auto ownership levels”, Transportation Research Part B:Methodological, Vol. 41 No. 5, pp. 506-526.

Black, J.L., Carpiano, R.M., Fleming, S. and Lauster, N. (2011), “Exploring the distribution of foodstores in British Columbia: associations with neighbourhood socio-demographic factorsand urban form”, Health & Place, Vol. 17 No. 4, pp. 961-970.

Blaylock, J.R. (1989), “An economic model of grocery shopping frequency”, Applied Economics,Vol. 21 No. 6, pp. 843-852.

Boarnet, M. and Crane, R. (2001), “The influence of land use on travel behavior: specification andestimation strategies”, Transportation Research Part A: Policy and Practice, Vol. 35 No. 9,pp. 823-845.

Boarnet, M.G. (2011), “A broader context for land use and travel behavior, and a researchagenda”, Journal of the American Planning Association, Vol. 77 No. 3, pp. 197-213.

Boarnet, M.G., Joh, K., Siembab, W., Fulton, W. and Nguyen, M.T. (2011), “Retrofitting thesuburbs to increase walking: evidence from a land-use-travel study”, Urban Studies,Vol. 48 No. 1, pp. 129-159.

BOL Statistics (Ed.) (2014), Consumer Expenditures for the Seattle-Tacoma-Bremerton Area:2011-2012, US Department of Labour, San Francisco, CA.

Cao, X.J., Mokhtarian, P.L. and Handy, S.L. (2009), “The relationship between the builtenvironment and nonwork travel: a case study of Northern California”, TransportationResearch Part A: Policy and Practice, Vol. 43 No. 5, pp. 548-559.

Cervero, R. (1996), “Mixed land-uses and commuting: evidence from the American HousingSurvey”, Transportation Research Part A: Policy and Practice, Vol. 30 No. 5,pp. 361-377.

Cervero, R. and Kockelman, K. (1997), “Travel demand and the 3Ds: density, diversity, anddesign”, Transportation Research Part D: Transport and Environment, Vol. 2 No. 3,pp. 199-219.

Cervero, R. and Radisch, C. (1996), “Travel choices in pedestrian versus automobile orientedneighborhoods”, Transport Policy, Vol. 3 No. 3, pp. 127-141.

Chen, C., Gong, H. and Paaswell, R. (2008), “Role of the built environment on mode choice decisions:additional evidence on the impact of density”, Transportation, Vol. 35 No. 3, pp. 285-299.

Chung, C. and Myers, S.L. (1999), “Do the poor pay more for food? An analysis of grocery storeavailability and food price disparities”, Journal of Consumer Affairs, Vol. 33 No. 2,pp. 276-296.

Clifton, K., Morrissey, S. and Ritter, C. (2012), “Mode choice and consumer spending: anexamination of grocery store trips”, Transportation Research Board 91st Annual Meeting,Washington, DC.

Clifton, K.J. (2004), “Mobility strategies and food shopping for low-income families: a case study”,Journal of Planning Education and Research, Vol. 23 No. 4, pp. 402-413.

Darko, J., Eggett, D.L. and Richards, R. (2012), “Shopping behaviors of low-income families duringa 1-month period of time”, Journal of Nutrition Education and Behavior, Vol. 45 No. 1,pp. 20-29.

918

IJRDM44,9

Dow

nloa

ded

by D

octo

r Ju

nfen

g Ji

ao A

t 20:

19 1

5 Se

ptem

ber

2016

(PT

)

Dawson, J. and Lord, D. (2012), Shopping Centre Development (RLE Retailing and Distribution),Routledge, New York, NY.

Duncan, M.J., Winkler, E., Sugiyama, T., Cerin, E., Leslie, E. and Owen, N. (2010), “Relationshipsof land use mix with walking for transport: do land uses and geographical scale matter?”,Journal of Urban Health, Vol. 87 No. 5, pp. 782-795.

Ewing, R. and Cervero, R. (2010), “Travel and the built environment: a meta-analysis”, Journal ofthe American Planning Association, Vol. 76 No. 3, pp. 265-294.

Ewing, R., Deanna, M. and Li, S.-C. (1996), “Land use impacts on trip generation rates”,Transportation Research Record: Journal of the Transportation Research Board, Vol. 1518,pp. 1-6.

Feng, J., Dijst, M., Wissink, B. and Prillwitz, J. (2013), “The impacts of household structure on thetravel behaviour of seniors and young parents in China”, Journal of Transport Geography,Vol. 30, pp. 117-126.

Frank, L. and Pivo, G. (1994), “Impacts of mixed use and density on utilization of three modes oftravel: single-occupant vehicle, transit, and walking”, Transportation Research Record:Journal of the Transportation Research Board, Vol. 1466, pp. 44-52.

Frank, L., Kerr, J., Saelens, B., Sallis, J., Glanz, K. and Chapman, J. (2009), “Food outlet visits,physical activity and body weight: variations by gender and race-ethnicity”, BritishJournal of Sports Medicine, Vol. 43 No. 2, pp. 124-131.

Frusti, T., Bhat, C. and Axhausen, K. (2002), “Exploratory analysis of fixed commitments inindividual activity-travel patterns”, Transportation Research Record: Journal of theTransportation Research Board, Vol. 1807, pp. 101-108.

Gentina, E. and Bonsu, S.K. (2012), “Peer network position and shopping behavior amongadolescents”, Journal of Retailing and Consumer Services, Vol. 20 No. 1, pp. 87-93.

Golob, T.F. (2003), “Structural equation modeling for travel behavior research”, TransportationResearch Part B: Methodological, Vol. 37 No. 1, pp. 1-25.

Greenwald, M.J. (2003), “The road less traveled new urbanist inducements to travel modesubstitution for nonwork trips”, Journal of Planning Education and Research, Vol. 23 No. 1,pp. 39-57.

Guo, J., Bhat, C. and Copperman, R. (2007), “Effect of the built environment on motorizedand nonmotorized trip making: substitutive, complementary, or synergistic?”,Transportation Research Record: Journal of the Transportation Research Board,Vol. 2010, pp. 1-11.

Gustat, J., O’Malley, K., Luckett, B.G. and Johnson, C.C. (2015), “Fresh produce consumption andthe association between frequency of food shopping, car access, and distance tosupermarkets”, Preventive Medicine Reports, Vol. 2, pp. 47-52.

Habib, K.M. and Miller, E.J. (2008), “Modelling daily activity program generation consideringwithin-day and day-to-day dynamics in activity-travel behaviour”, Transportation, Vol. 35No. 4, pp. 467-484.

Handy, S.L. (1996), “Understanding the link between urban form and nonwork travel behavior”,Journal of Planning Education and Research, Vol. 15 No. 3, pp. 183-198.

Handy, S.L. and Clifton, K.J. (2001), “Local shopping as a strategy for reducing automobile travel”,Transportation, Vol. 28 No. 4, pp. 317-346.

Hanson, S. and Hanson, P. (1981), “The travel-activity patterns of urban residents: dimensionsand relationships to sociodemographic characteristics”, Economic Geography, Vol. 57 No. 4,pp. 332-347.

HartmanGroup (2014), Food Shopping in America 2014, The Hartman Group, Inc., Bellevue, WA.

919

A study fromSeattle,

Washington,USA

Dow

nloa

ded

by D

octo

r Ju

nfen

g Ji

ao A

t 20:

19 1

5 Se

ptem

ber

2016

(PT

)

Hess, P., Moudon, A., Snyder, M. and Stanilov, K. (1999), “Site design and pedestrian travel”,Transportation Research Record: Journal of the Transportation Research Board, Vol. 1674,pp. 9-19.

Hillier, A., Cannuscio, C.C., Karpyn, A., McLaughlin, J., Chilton, M. and Glanz, K. (2011), “How fardo low-income parents travel to shop for food? Empirical evidence from two urbanneighborhoods”, Urban Geography, Vol. 32 No. 5, pp. 712-729.

Hollywood, L.E., Cuskelly, G.J., O’Brien, M., McConnon, A., Barnett, J., Raats, M.M. and Dean, M.(2013), “Healthful grocery shopping. Perceptions and barriers”, Appetite, Vol. 70, pp. 119-126.

Hubley, T.A. (2011), “Assessing the proximity of healthy food options and food deserts in a ruralarea in Maine”, Applied Geography, Vol. 31 No. 4, pp. 1224-1231.

Huntsinger, L.F., Rouphail, N.M. and Bloomfield, P. (2013), “Trip generation models usingcumulative logistic regression”, Journal of Urban Planning and Development, Vol. 139 No. 3,pp. 176-184.

Jiao, J. and Dillivan, M. (2013), “Transit deserts: the gap between demand and supply”, Journal ofPublic Transportation, Vol. 16 No. 2, pp. 23-39.

Jiao, J., Moudon, A.V. and Drewnowski, A. (2011), “Grocery shopping: how individuals and builtenvironments influence travel mode choice”, Transportation Research Record: Journal ofthe Transportation Research Board, Vol. 2230, pp. 85-95.

Jiao, J., Moudon, A.V., Ulmer, J., Hurvitz, P.M. and Drewnowski, A. (2012), “How to identify fooddeserts: measuring physical and economic access to supermarkets in King County,Washington”, American Journal of Public Health, Vol. 102 No. 10, pp. e32-e39.

Kahn, B.E. and Schmittlein, D.C. (1989), “Shopping trip behavior: an empirical investigation”,Marketing Letters, Vol. 1 No. 1, pp. 55-69.

Kim, B.-D. and Park, K. (1997), “Studying patterns of consumer’s grocery shopping trip”, Journalof Retailing, Vol. 73 No. 4, pp. 501-517.

Krumm, J.Y. (2012), “How people use their vehicles: statistics from the 2009 national householdtravel survey”, Society of Automotive Engineers (SAE) 2012 World Congress, Detroit, MI.

Lamb, K.E., Thornton, L.E., Cerin, E. and Ball, K. (2015), “Statistical approaches used to assess theequity of access to food outlets: a systematic review”, AIMS Public Health, Vol. 2 No. 3,pp. 358-401.

Lee, C. and Moudon, A.V. (2006a), “The 3Ds+R: quantifying land use and urban form correlatesof walking”, Transportation Research Part D: Transport and Environment, Vol. 11 No. 3,pp. 204-215.

Lee, C. and Moudon, A.V. (2006b), “Correlates of walking for transportation or recreationpurposes”, Journal of Physical Activity and Health, Vol. 3, pp. S77-S98.

Luceri, B. and Latusi, S. (2012), “The importance of consumer characteristics and marketstructure variables in driving multiple store patronage”, Journal of Retailing and ConsumerServices, Vol. 19 No. 5, pp. 519-525.

McCormack, G.R., Giles-Corti, B. and Bulsara, M. (2008), “The relationship between destinationproximity, destination mix and physical activity behaviors”, Preventive Medicine, Vol. 46,pp. 33-40.

Ma, Y., Ailawadi, K. L., Gauri, D.K. and Grewal, D. (2011), “An empirical investigation of theimpact of gasoline prices on grocery shopping behavior”, Journal of Marketing, Vol. 75No. 2, pp. 18-35.

Mahmud, N.K., Monsivais, P. and Drewnowski, A. (2009), “The search for affordable nutrient richfoods: a comparison of supermarket food prices in Seattle-King County”, Center for PublicHealth Nutrition, Seattle, WA, available at: http://depts.washington.edu/uwcphn/reports/cphnbrf2.pdf (accessed 9 April 2011).

920

IJRDM44,9

Dow

nloa

ded

by D

octo

r Ju

nfen

g Ji

ao A

t 20:

19 1

5 Se

ptem

ber

2016

(PT

)

Maxwell, K.H., Yinghua, H. and Scott, S. (2010), “Grocery store image, travel distance, satisfactionand behavioral intentions”, International Journal of Retail & Distribution Management,Vol. 38 No. 2, pp. 115-132.

Michimi, A. and Wimberly, M.C. (2010), “Associations of supermarket accessibility with obesityand fruit and vegetable consumption in the conterminous United States”, InternationalJournal of Health Geographics, Vol. 9 No. 1.

Milakis, D., Cervero, R. and VanWee,, B. (2015a), “Stay local or go regional? Urban form effects onvehicle use at different spatial scales: a theoretical concept and its application to theSan Francisco Bay area”, Journal of Transport and Land Use, Vol. 8 No. 2, pp. 59-86.

Milakis, D., Cervero, R., Van Wee, B. and Maat, K. (2015b), “Do people consider an acceptabletravel time? Evidence from Berkeley, CA”, Journal of Transport Geography, Vol. 44,pp. 76-86.

Minaker, L.M., Olstad, D.L., Thompson, M.E., Raine, K.D., Fisher, P. and Frank, L.D. (2016),“Associations between frequency of food shopping at different store types and diet andweight outcomes: findings from the NEWPATH study”, Public Health Nutrition, Vol. 19No. 12, pp. 2268-2277.

Morland, K., Wing, S. and Roux, A.D. (2002), “The contextual effect of the local food environmenton residents’ diets: the atherosclerosis risk in communities study”, American Journal ofPublic Health, Vol. 92 No. 11, pp. 1761-1768.

Morland, K.B. and Evenson, K.R. (2009), “Obesity prevalence and the local food environment”,Health & Place, Vol. 15 No. 2, pp. 491-495.

Mortimer, G. and Clarke, P. (2011), “Supermarket consumers and gender differences relating totheir perceived importance levels of store characteristics”, Journal of Retailing andConsumer Services, Vol. 18 No. 6, pp. 575-585.

Moudon, A.V. and Hess, P.M. (2000), “Suburban clusters – the nucleation of multifamily housingin suburban areas of the central Puget sound”, Journal of the American PlanningAssociation, Vol. 66 No. 3, pp. 243 -264.

Nijland, L., Arentze, T. and Timmermans, H. (2013), “Representing and estimating interactionsbetween activities in a need-based model of activity generation”, Transportation, Vol. 40No. 2, pp. 413-430.

NSOPP Center (2016), “Policy: tax incentives for grocery stores/healthier food vendors in highneed areas”, available at: www.obesitypolicies.org/policies/2011/tax-incentives-grocery-storeshealthier-food-vendors-high-need-areas (accessed 13 March 2016).

Progressive Grocer (2002), “Progressive Grocer annual report”, Nielsen Business Media,New York, NY.

Rajamani, J., Bhat, C., Handy, S., Knaap, G. and Song, Y. (2003), “Assessing impact of urban formmeasures on nonwork trip mode choice after controlling for demographic and level-of-service effects”, Transportation Research Record: Journal of the Transportation ResearchBoard, pp. 158-165.

Reedy, J., Krebs-Smith, S.M., Miller, P.E., Liese, A.D., Kahle, L.L., Park, Y. and Subar, A.F.(2014), “Higher diet quality is associated with decreased risk of all-cause, cardiovasculardisease, and cancer mortality among older adults”, The Journal of Nutrition, Vol. 144No. 6, pp. 881-889.

Reitzel, L.R., Okamoto, H., Hernandez, D.C., Regan, S.D., McNeill, L.H. and Obasi, E.M. (2016),“The built food environment and dietary intake among African-American adults”,American Journal of Health Behavior, Vol. 40 No. 1, pp. 3-11.

Schwanen, T., Dijst, M. and Dieleman, F.M. (2004), “Policies for urban form and their impact ontravel: the Netherlands experience”, Urban Studies, Vol. 41 No. 3, pp. 579-603.

921

A study fromSeattle,

Washington,USA

Dow

nloa

ded

by D

octo

r Ju

nfen

g Ji

ao A

t 20:

19 1

5 Se

ptem

ber

2016

(PT

)

Srinivasan, S. and Bhat, C.R. (2005), “Modeling household interactions in daily in-home andout-of-home maintenance activity participation”, Transportation, Vol. 32 No. 5, pp. 523-544.

Steiner, R. (1998), “Trip generation and parking requirements in traditional shopping districts”,Transportation Research Record: Journal of the Transportation Research Board, Vol. 1617,pp. 28-37.

Tuttle, B. (2014), “Your grocery store may soon be cut in half”, TIME-Every Day Money, 2 June,available at: http://time.com/money/136330/why-your-grocery-store-may-soon-be-cut-in-half/(accessed 13 March 2016).

US Census Bureau (2009), 2006-2008 American Community Survey, Government Printing Office,Washington DC.

Van Loo, E., Caputo, V., Nayga, R.M. Jr, Meullenet, J.F., Crandall, P.G. and Ricke, S.C. (2010),“Effect of organic poultry purchase frequency on consumer attitudes toward organicpoultry meat”, Journal of Food Science, Vol. 75 No. 7, pp. S384-S397.

Van, U.-P. and Senior, M. (2000), “The contribution of mixed land uses to sustainable travel incities”, Achieving Sustainable Urban Form, pp. 139-148.

Whelan, A., Wrigley, N., Warm, D. and Cannings, E. (2002), “Life in a ‘food desert’ ”, UrbanStudies, Vol. 39 No. 11, pp. 2083-2100.

Wiig, K. and Smith, C. (2009), “The art of grocery shopping on a food stamp budget: factorsinfluencing the food choices of low-income women as they try to make ends meet”, PublicHealth Nutrition, Vol. 12 No. 10, pp. 1726-1734.

Wilde, P.E. and Ranney, C.K. (2000), “The monthly food stamp cycle: shopping frequency andfood intake decisions in an endogenous switching regression framework”, AmericanJournal of Agricultural Economics, Vol. 82 No. 1, pp. 200-213.

Yoo, S., Baranowski, T., Missaghian, M., Baranowski, J., Cullen, K., Fisher, J.O., Watson, K.,Zakeri, I.F. and Nicklas, T. (2006), “Food-purchasing patterns for home: a grocery store-intercept survey”, Public Health Nutrition, Vol. 9 No. 3, pp. 384-393.

Zhang, M. (2004), “The role of land use in travel mode choice: evidence from Boston and Hong Kong”,Journal of the American Planning Association, Vol. 70 No. 3, pp. 344-360.

Corresponding authorJunfeng Jiao can be contacted at: jjiao@austin.utexas.edu

For instructions on how to order reprints of this article, please visit our website:www.emeraldgrouppublishing.com/licensing/reprints.htmOr contact us for further details: permissions@emeraldinsight.com

922

IJRDM44,9

Dow

nloa

ded

by D

octo

r Ju

nfen

g Ji

ao A

t 20:

19 1

5 Se

ptem

ber

2016

(PT

)