MEASURES REGISTRY USER GUIDE
Key Concepts in Food Environment Assessment
The conceptual model presented in the previous section (Figure 2) can be used to organize measures for assessing the food environment and help with considerations in choosing the most appropriate measure. Table 1 provides examples of measures that may be used across food environment domains and sub-domains. Section 3 provides information on the types of tools that are used to assess aspects of each of the three types of environments.
Table 1: Food Environment Domains, Sub-Domains, and Examples of Measures
Different types of measures are typically used to assess different types of environmental domains. The following section discusses each environmental domain and highlights the types of measurements commonly used.
Physical Food Environment
Measures to assess the physical environment can be broadly grouped into two kinds of questions and assessment tools:
- Where can people obtain food or particular types of foods in a given area? Geo-spatial analyses such as geographical informational systems (GIS) are used to assess the number, location, and density of stores that offer foods (including full-service grocery stores, convenience stores and corner stores) and restaurants in a given geographic area and their proximity to homes, schools or community venues, and each other.
- What is the environment within stores, restaurants, homes, schools, or community venues? What foods are available, what foods are promoted, and what are the prices of foods? Observational scans or assessments (also called logs, records, and audits) are used to assess food product availability, pricing, placement/merchandising, advertising, and nutrition information.
Geographic Information Systems
Geographic information systems (GIS) are used for integrating and analyzing spatial and geographic data and are generally derived from existing databases that have spatial reference, or are “geolocated,” such as U.S. Census data linked to census track and block groups. For assessing the food environment, GIS tools can be useful in helping to evaluate the accessibility and availability of foods in a geographical area by linking with other data sources that document the existence and types of food outlets in that area. The use of GIS as a measurement tool requires staff who know how to access the data and use the related software.
Data on the existence of food outlets in a geographic area can be determined using several methods including:
- Field work whereby trained data collectors document the existence of a store or restaurant using logs or photographs (often called “ground-truthing”)
- Health, agricultural, tax, or licensing data documenting the existence of food outlets that serve food to the public and are under health department oversight
- Commercially available business databases
In the United States, various data sources for business lists exist, including InfoUSA, Business Analyst, Dun and Bradstreet, and the U.S. Department of Agriculture’s Supplemental Nutrition Assistance Program (SNAP) Retailer Locator. These sources are used to document the types of foods stores in a geographic area (including super stores, supermarkets, grocery stores, and convenience stores) and the types of restaurants that are found (including fast food, chain restaurants, and independent restaurants). By identifying a specific geographic area and the food outlets in the area, the availability of foods in an area can be estimated. In addition to quantifying the number and types of food outlets in an area, GIS data can estimate a person’s distance from a location (for example from home, school, or workplace) to types of food outlets. This distance can be calculated point to point (“as the crow flies”) or by examining street connectivity that may affect access.
Advantages of GIS measures are that they rely on objective data and are relatively easy to derive for large samples by someone with GIS software experience. The preponderance of research on the food environment uses GIS as the tool.12,13 The use of GIS data has some important limitations as well including the assumption that individuals shop or eat in areas that are most proximate to their homes, schools, or worksites. Another important limitation is that the business data that are mapped onto GIS data are often not up to date. Stores and restaurants frequently turn over and expensive ground-truthing is often needed to verify that the stores and restaurants that show up on a business or licensing data base actually exist in the neighborhood. Third, the process of translating a business address into a physical point location, known as geocoding, is subject to errors in accuracy, even under the best of circumstances. The use of parcel boundaries matched to point locations can mitigate error but takes a lot of time, and parcel boundaries are not always available. Finally, although GIS data may show the types of food outlets that are available in a geographic area, no information on the foods that are available within those outlets, or how they are priced or marketed is available through GIS data. Even though the majority of food environment research has used GIS data, the association between what is documented in the environment and obesity-related outcomes has been relatively weak.14
Observational scans have many names, including an observational assessment or environmental scan, log, record, or inventory. For a particular food environment venue such as a retail store, an observational scan could be called a store audit, observational store survey, or store assessment. Observational scans are used to quantitatively assess characteristics of the physical food environment present in any number of venues including stores, restaurants, homes, schools, or other community venues. Observational scans are conducted in both research and practice by various stakeholder groups. Store or restaurant audits (“assessments”) are typically completed by research staff or community stakeholders who have received some training in a specified data collection protocol. Observational scans in a school venue or home may be completed either by an external observer or, for example, by the head of an afterschool program who will document the types of typical snacks offered (school) or a parent (home) who completes a home food inventory. Characteristics of interest captured through environmental scans include food product availability and quality; prices and price discounts; placement of products, merchandising, or shelf space; the presence of food product information or advertisements; or other characteristic specific to a venue, population, health behavior, or health outcome of interest.
In stores, observational scans most frequently document information such as:
- The types of foods that are available (Are fresh fruits and vegetables available in a convenience store?)
- The amount of shelf space or counter space is dedicated to specific foods (What is the ratio of space dedicated to high-fat dairy products relative to space dedicated to lower fat dairy products?)
- The placement of foods (What types of foods are placed near the checkout line?)
- Whether some nutrition information or product-specific advertising is attached to the food (Is there a sign promoting the food as a healthy product)?
In restaurants, observational scans document information such as:
- Types of food products available for dine-in or take-away
- Presence of a drive-through window
- Restaurant size and seating capacity
- Presence of food information, signage, or promotions
- Pricing of individual menu items and combination meals
Sometimes, store or restaurant data collection involves identifying a specific group of food products and their associated prices to determine the price of a typical “market basket” of food. Prices of the “market basket” are assessed to evaluate the accessibility of foods in a neighborhood based on price.
Advantages of store and restaurant audit tools are that data collection and analysis can be relatively straightforward once the protocol is developed and data collectors trained. Data can be entered using paper and pencil or through electronic tool on a smartphone or tablet computer. Audit tools can be very useful tools for community groups as they can easily be used to document food-related factors in their communities and target areas for change. Disadvantages include the labor-intensive nature of the data collection; many stores and restaurants may need to be audited to capture the true essence of the food environment in a neighborhood. In addition, retail environments are not static; therefore, multiple measurements of the same location may be needed to capture the “usual” environment. Care also should be taken in deciding the length of the audit form and the measures included. Parsimony in data collection is important so that the managers of the venue will not feel like the data collection activity is overly disruptive to normal retail activities. In addition, it is always helpful to plan in advance how each data point will be analyzed and used following the data collection effort for efficiency in data management tasks. Finally, an important limitation of observation scan data is that research has determined that food availability in a neighborhood is often not related to dietary intake patterns of neighborhood residents (in other words, the measure lacks construct validity, described in Section 4).2
In homes, observational scans may be in the form of a record, log, or inventory and may be used to document:
- A comprehensive listing of all foods and beverages in the home
- The presence of specific foods in the home, such as fruits and vegetables or desserts and sweets
- The types of beverages available in the home
- Foods and beverages that are on the counter, possibly acting as cues to behavior
In early care and education centers, schools, or community venues, observational scans may be used to document:
- The types of foods or snacks that are available in vending machines, cafeterias, school à la carte lines, school stores, concession stands for events, or given away to students or children
- Food product and meal prices
- Restaurant size and seating capacity
- Placement of vending machines and product placement on service lines
- The presence of food information or signage
Measures of the food environment for early care and education centers, schools, or community venues that children and youth frequent may require different tools than those used for a store or restaurant audit due to the highly individual nature of these venues. Although restaurants and stores likely present food options in similar ways (on a menu or menu board, on shelves, or in coolers), foods in schools and other community venues may have many different ways that food is stored or displayed, requiring different tools and data collection methods. Home, preschool, school, and community venue observational scans may be completed by trained data collectors or by parents, teachers, and other community stakeholders.
Advantages to observational scans in homes, schools, and community venues are that data collection is often simple and data analysis straightforward. Without using a great deal of sophisticated analysis, much can be quickly learned and shared with relevant stakeholders. However, measurement tools to assess foods in homes, schools, and community venues are often quite specific and sometimes limited in the aspect of the environment assessed. For example, many home food inventories focus on only the presence of fruits and vegetables in the home or the presence of soft drinks. Likewise, an environmental observation tool for a school might be limited to foods and beverages in vending machines. Therefore, when choosing the appropriate measurement tool, careful consideration of the needs of the specific project is important.
Measures to assess the social environment can be broadly grouped into three types of purposes that include assessing:
- Social support, role modeling, and social expectations regarding what, when, where, and why to eat
- Policies, practices, or rules about eating behavior within public venues, such as schools and community centers
- Parenting practices and family rules around meal time and foods available to youth
Social Support, Role Modeling, and Social Expectations
Social support is typically assessed as the amount and types of social support that individuals receive from others or offer to others. Social support is typically assessed through a self-report questionnaire. Often, the questionnaire asks about levels or types of perceived support received by multiple referents in the respondent’s social environment. More information on perceived social support is included below.
Children and youth learn through observing others in their environments. Therefore, the adults, peers, family, and friends with whom they interact on a regular basis, as well as the cultural role models that they are exposed to through the Internet, movies, and television, are important elements of their social environment. Collecting data on role modeling and norms can involve collecting data on the eating behavior and food practices of significant others in a youth’s environment. For example, collecting information on the foods and beverages that teachers eat and drink in front of students suggest the role modeling to which youth are exposed. Likewise, collecting dietary data on parent and child dyads may be useful in understanding the food behaviors and dietary intake at the household level. Children are likely to eat foods similar to their parents or other primary caregivers because of accessibility, availability, family food habits, and learned taste preferences. The behaviors and practices of parents and other family members provide important insights into the social environment of youth.
One can also assess role modeling through the eyes of the youth; in other words, how do youth perceive of the behaviors of significant people in their lives? These perceptions are obtained by asking the youth to report on the eating behaviors they see their families and peers engaging in. For example, asking youth to provide their opinions on how healthy their mother’s diet is or asking youth what their friends eat for lunch are examples of youth’s perceptions of modeled behavior.
Collecting detailed information on role modeling behaviors is challenging and can demand significant resources. Collecting dietary intake data on both children and their parents is costly, and expanding beyond that to assess other significant role models is often not practical. In addition, it is difficult to ascertain who are the most significant role models to a young person. Some work has been done to evaluate the role modeling and social norms that occur through the media.15 Linking that exposure to a behavioral or health outcome is difficult because exposure levels will vary greatly in any population and assessing the amount of exposure any one person receives is quite difficult. Finally, asking youth about the behaviors of significant others in their lives is fraught with subjectivity and therefore may not be valid (even though it may be quite interesting!)
Policies, Practices, and Rules About Eating Behaviors Within Venues
Assessing the policies and practices of venues where youth spend time is frequently done with surveys or questionnaires of relevant stakeholders, including principals, food service staff, or wellness coordinators. As an example, the School Health Policy and Practice Survey (SHPPS)16 is a nationwide survey that has been regularly conducted by CDC since 2000 and contains questions on both the policies and practices of schools across a wide area of topics, including nutrition. Typically, the survey is conducted using a phone interview, but it can be completed using a paper and pencil or tablet format. Questions asked range from policy-type questions, such as when foods are available in the school (What times during the school day are the vending machines available?) to more informal teacher and child care providers practices (Are foods ever used as rewards or incentives?).
The advantage of this type of measure is that it is relatively easy to administer and analyze and may produce crucial information for interested stakeholders. In addition, the relationship between school policies and obesity in youth has been found documenting that the measure has relevance to important health outcomes. In an intervention study conducted in 16 middle schools, a significant association was found between school level mean body mass index (BMI) and a seven-item school food practice scale collected from principals assessing such practices as use of food for rewards and incentives and the use of food in classroom fundraisers.17 The disadvantages of these types of questionnaires and surveys are that they rely on self report, typically of a stakeholder who may present a biased view.
Parenting Practices and Family Rules Around Meal Time and Foods Available
Assessing caregivers and parenting practices around food and meal time is typically done with surveys, interviews, or questionnaires of caregivers or their children. Although observations of the home environment have been done, having a data collector in the home, or setting up a camera to record behavior in the home, is often seen as too intrusive. In addition, observational data collection sets up a threat of social desirability where parents may alter their behavior in the presence of an outside evaluator.
Parenting practices around food and meal times may involve collecting data on:
- Family rules enforced at mealtimes (Can the television be on during meal time? Can family members take phone calls during dinner?)
- Eating-related behaviors (Are children expected to try all foods offered or to clean their plate?)
- Parent participation in meals (How often does the family have a meal together? Do the parents eat with their children?)
- Foods and beverages that are present at meal time (Are fruits and vegetables offered at meal time? Are sugar-sweetened beverages offered?)
Mealtime has been shown to be an important aspect of the social environment as well as being significantly related to obesity in family. Positive family meal practices related to foods offered (for example, offering fruits and vegetables but not offering soft drinks at meals) as well as parents’ enforcing rules at meal time (for example, not allowing television watching or phone use during meals) have been associated with reducing obesity risk in both parents and children.18 Several tools have been used and found to be reliable and valid in their ability to assess aspects of the family meal environment.18,19 These data are typically relatively easy to collect and analyze. The largest challenge with these data is social desirability bias, as parents often know the responses that indicate a healthier environment.
Tools to assess the person-centered food environment can be broadly grouped into two areas: (1) assessing individual’s perceptions of their physical environment, and (2) assessing individual’s perceptions of their social environment. However, frequently the same measurement tool is used to assess both the physical and social environment.
Perceptions of the Physical and Social Environment
Measures for this purpose ask respondents their perceptions about their physical and social environments as they relate to food access. Some measures use a phone interview,20 others use a self-administered questionnaire,21 and for young children, a questionnaire may be read to children.22 These surveys and questionnaires may ask about such issues as:
- Perceived availability of foods (Are healthy snacks offered at your afterschool program?)
- Perceived access to foods (Is cost a factor in the foods that you choose at school?)
- Perceived social norms (Do your parents expect you to have healthy eating habits?)
- Perceived social support (Do your friends encourage you to make healthy food choices?)
For example, in the Child and Adolescent Trial for Cardiovascular Health (CATCH), children as young as age 8 years answered questions about their perceptions of support for eating a healthy diet considering parents, teachers, and friends.22
The advantage of these types of measures is that they are relatively easy to administer. Most older children (ages 12 years and older) can complete the questionnaires without help, while younger children may do better if an adult reads the questions to them. Some of the measures are meant to be used as a scale or index, which requires analytic skills for constructing the scale. A disadvantage is that these are self-report assessments, and criterion validity (or the ability to compare a proxy measure to a gold standard) is impossible to assess. However, because perception is often more predictive in explaining behavior than are objectively measured factors, these are important environmental measures to consider.