Physical Activity Environment
MEASURES REGISTRY USER GUIDE
SECTION
3
Key Concepts in Physical Activity Environment Assessment
The built environment can be measured at different levels, using different methods, and in various settings relevant to physical activity. Descriptions of these levels, methods, settings, and other key concepts are presented in this section. The main focus is on measures of environmental characteristics themselves. Related measures of how often a specific setting (e.g., park) is used, and measures of physical activity in a specific setting, such as a park, are discussed later in this section. Some measurement tools capture both environment characteristics and physical activity in the setting.
Levels of Built Environment Attributes
Physical activity environments are complex and can be measured at multiple levels that vary somewhat by the specific settings.
Macro-scale Attributes. "Macro-scale" attributes describe the layout of communities and include the concept of walkability. Walkability refers to the combination of mixed land use, moderate-to-high residential density, and high street connectivity that allows people to walk from home to common destinations, such as shops and schools. Proximity of parks, trails, bicycle networks (e.g., lanes, trails), and public transit can be considered macro-level attributes as well. Geographic information systems (GIS) are often used to capture these attributes when data are available; self-report measures also can assess these variables.
Micro-scale Attributes. "Micro-scale" attributes indicate the design of a setting and can affect the experience of being active or inactive in a given place. Streetscape attributes include presence, quality, and amenities of sidewalks; characteristics of street crossings, such as crosswalks and pedestrian signals; transportation variables, such as transit stops and signage, travel lanes, and speed limits; and aesthetic features of buildings and landscaping. Other micro-scale variables include activity zones and equipment for physical activity in parks and school yards; design of public spaces; characteristics of trail design; and attributes of buildings, including stairs. Micro-scale features are usually measured by direct observation (i.e., audits) and self-report, because they are not often available in GIS.
Social Attributes. A third level of measurement is mostly focused on social dimensions of the environment. Social environments are not well-defined, but commonly studied variables include people in the environment and evidence of people's behavior. Examples are indicators of social disorder such as graffiti, litter, and abandoned buildings; people walking or children playing; maintenance of park or playground equipment; crime; traffic density and speed; and programming and supervision in recreation areas. Social indicators are usually measured by direct observation or self-report.
Box 2. Definitions of walkability
Walkability is a relatively new concept that has been defined in several ways. Definitions from the transportation field often focus on macro-level variables that create an opportunity to walk to destinations,43 whereas other definitions capture macro-scale, micro-scale, and even social features. In the 2016 Surgeon General’s Call to Action on Walking and Walkable Communities, a walkable community was broadly defined as a community where it is safe and easy to walk and where pedestrian activity is encouraged.7,44
Methods of Environmental Assessment
Environmental assessment tools fall into three categories based on their methodology and data collection procedures (see Table 1). Within the Measures Registry, specific measures falling within each of these methods of assessment can be isolated using the check boxes under the “Measure Type” subheading. The Measurement Types are:
- GIS: GIS-based measures, which involve archival data sets that are layered and analyzed within GIS software.
- Environmental Observation: Observational measures, which are obtained using systematic observational methods (audits).
- Questionnaire: Self- or proxy-reports of environmental features, which can be administered by paper-and-pencil questionnaire, computer or online survey, or interview.
Additionally, the Measures Registry includes Electronic Monitor, Record or Log, and Other as modalities of environmental measures. The Electronic Monitor section includes studies that used Global Positioning Systems (GPS) to assess participants’ locations. The Record or Log and Other sections include some observational audit measures as well as direct observation measures of physical activity within specific settings (e.g., System for Observing Play and Recreation in Communities [SOPARC]). These measurement types are not covered in detail here because they are described in the Measures Registry User Guide: Individual Physical Activity.
GIS-based Measures
GIS is a software system (rather than a measure itself) used for integrating and analyzing spatial and geographic data. GIS-based environmental measures are typically derived from existing databases, and these data sources are continuously evolving. Data collected through observations and questionnaires also can be integrated into GIS if they are spatially specific. Data in GIS need to have spatial reference, such as parcel or road database shapefiles. Existing databases are available from multiple sources, such as regional transportation planning agencies, local municipalities, and ongoing surveillance surveys (e.g., American Community Survey, U.S. Census). From this existing information, built environment measures that could be related to physical activity are computed.45,46 For example, road network spatial databases (i.e., shapefiles) can be obtained from many regional transportation planning agencies or local municipalities and used to calculate the number of street intersections within a defined neighborhood or area. The GIS measures in the Measures Registry are simply variable computations performed in GIS that have been used in physical activity research. Many of the publications listed in the “Results” view of the Measures Registry include multiple GIS-derived variables (see Section 6 for suggested process for using the Measures Registry).
Advantages of GIS-based measures are that they rely on (relatively) objective data and are easy to derive for large samples by someone with GIS software expertise. Disadvantages of GIS-based measures are that they are limited to community design (land use) and transportation system information. Data often do not exist for home, school, streetscape, and micro-level (i.e., interior) parks and recreation features. However, data collected in these settings can be integrated into GIS. Using GIS-based measures can be difficult. GIS-based projects often require collaboration with geographers or transportation or public affairs experts. Although GIS-based data on traffic, crime, sidewalks, and crossings exist in some areas, data can vary in quality, completeness, recency, and availability across jurisdictions. The way that data, such as crime data, are recorded across jurisdictions also can vary, making it difficult to compare and pool data across geographic areas. The quality of available GIS data should not be assumed.
Observational Measures
Observational measures, or audit tools, involve systematic observation of an environment, typically by trained research staff or community stakeholders. Good audits tools are accompanied by a manual with well-defined construct definitions and observer training procedures that result in acceptable inter-rater reliability. Audits exist for assessing micro-level streetscape, home, school, park, and recreation center features. In neighborhoods, audits can provide useful information that is not captured in GIS databases, such as availability of crosswalks and duration of pedestrian crossing signals. Audits can be favored over self-reported measures because the former provide more objective, detailed, and place-specific data than self-reported participant or informant perceptions.
Observational measures require substantial time and resources because trained observers must visit each environment being assessed. Although evidence is accumulating to support the validity of conducting observations remotely, such as through Google Earth and Google Streetview,47-49 these methods still require significant time and resources and extensive training of coders. Another limitation of audits is that they have poorer validity for capturing transient and subjective constructs, such as aesthetics or social disorder, because they are typically only conducted at one point in time. Audit measures must be processed and scored, which is a burden on users, and some measures do not have well-developed scoring procedures.
Audit tools serve several important purposes outside of research. They can be used to engage and educate community members about the impacts of environments on health. These tools are also useful for conducting needs assessments to inform local decision making and advocacy. Audit tools can point to specific environmental changes to target, including more modifiable micro-level features (e.g., sidewalk quality) than can be identified through GIS-based measures, which capture constructs that can take decades to change (e.g., street layout). For example, audits are often used by community advocacy groups to identify needed sidewalk and street-crossing improvements.
Questionnaires and Environment Perceptions
Questionnaires are commonly used in physical activity environment assessment and include participant-report and proxy-report tools. Proxy-reports involve the questionnaire being completed by someone other than the participant themselves, such as a parent, key informant, or school staff member. Most questionnaires in the Measures Registry ask participants to report on objective facts, such as presence of sidewalks, proximity to parks, and physical activity equipment in parks, schools, or homes. Objective indicators of aesthetics can even be reported, such as amount of greenery, views, and how well buildings are maintained. Some questionnaires ask participants to evaluate the quality of environments. Examples include perceived safety from traffic, attractiveness of parks, and comfort level while walking in the neighborhood. Self-report questionnaires are the only way to assess perceptive evaluations of the environment, which often do not align with objective measures of the environment.50 One reason for the misalignment is when a participant’s level of awareness of the environment does not reflect actuality (e.g., distance, access to amenities/facilities). Another reason for misalignment relates to the value placed on the environmental attribute, such as safety or aesthetics, which can differ across participants rating the same environment.
Questionnaires range in length, with some using single items and others using multi-item scales to capture various constructs. They frequently use Likert response scales, such as 4- or 5-point scales, with response options ranging from “strongly disagree” to “strongly agree.” Tools that capture the presence or absence of environmental attributes often use tallies or checklists of available resources. Tools that capture value-based perceptions of environmental attributes, such as safety and aesthetics, typically use Likert response scales (e.g., agree vs. disagree). The Neighborhood Environment Walkability Scale (NEWS) is an example of a questionnaire whose scales include reporting of both objective features and evaluative perceptions,51 whereas the sedentary environment module of the Physical and Nutritional Home Environment Inventory is an example of a tally/checklist that can be completed with a good level of objectivity.52
Questionnaires are commonly used because meaningful environmental information can be captured with few resources and little burden to the researcher or practitioner. Questionnaires are useful for capturing information that is difficult to assess when GIS and audit data do not exist or are difficult to capture. Evidence from perceived environment measures has shown that perceptions often have small associations with objective measures, yet the former are important for understanding physical activity over and above objective environment characteristics.53-55 Thus, many studies find it beneficial to include both perceived and objective environment measures. Limitations to questionnaires include participant burden, the potential for low response rates, and floor or ceiling effects. Floor or ceiling effects can occur when the majority of participants respond at the bottom or top of the response scale and thus result in limited variability and power for detecting associations.
Proximity, Accessibility, and Quality of Environmental Attributes
Distinguishing between measures of proximity, accessibility, and quality is important when assessing environments. Proximity refers to the presence, absence, or distance to an environmental feature. For example, having a greater number of parks in the neighborhood and a shorter distance to the nearest park have been associated positively with physical activity in youth.19,56 Accessibility refers to the ease of getting to the feature. For example, a park could be nearby, but a highway or dangerous intersection between the neighborhood and the park reduces accessibility. Accessibility can be affected by the street network, with a connected grid allowing direct access and a disconnected suburban-style network creating a long, indirect pathway. Quality involves rating existing features on attributes such as aesthetics, appeal, condition, ease of use, or safety. The Environmental Assessment of Public Recreation Spaces (EAPRS) tool, for example, can be used to capture ease of access, cleanliness, colorfulness, condition, and comfort of park facilities.57 Similarly, the school audit tool by Jones et al. covers quality attributes such as aesthetics, design, maintenance, and suitability for physical activity of the external grounds of the school.58 GIS-based measures are mainly used to assess proximity through the mixture of land uses in community design and accessibility through the connectivity of the transportation system. Many audit tools and questionnaires capture proximity, accessibility, and quality.
Measurements Using Specific Environments and Physical Activity in Specific Environments
Researchers and practitioners sometimes desire to assess people’s use of specific environments (e.g., whether or how often a park is used), or physical activity levels within specific settings (e.g., how much activity occurs in parks). This information can be used to better understand where physical activity occurs, how exposure to different environments may influence physical activity, or how characteristics of a setting are related to physical activity in that setting. Although measures of use of specific environments and physical activity in specific environments are not directly part of environmental assessment, they are sometimes assessed in conjunction with environmental variables. Specific measures that fall under this category are presented at the end of Section 5, and examples include questionnaires, observational measures, and GPS.
Considerations for Youth
The Measures Registry includes measures relevant for both youth and adults. The “Age” subheading can be used to narrow the number of measures in the “Results” view based on age group. The following factors should be considered when selecting a measure that is specifically relevant for youth.
Environmental correlates unique to youth. Youth-specific measures and evidence should be considered when youth are the focus of study. Some evidence suggests that the community design and transportation environments sometimes have differential associations with physical activity in youth versus adults. For example, greater intersection density and street connectivity have been fairly consistently positively related to walking for transport purposes in adults and youth.19,59,60 However, presence of cul-de-sacs, a street design feature indicative of lower street connectivity, has been positively related to recreational physical activity in youth, likely because these features serve as important low-traffic play areas for youth.19,27,61-64 Recreation environments also may be more important for understanding physical activity in youth versus adults.19,27,61,63 School environments are clearly more important for youth than for adult physical activity.65 Less is known about differences in the importance of environmental attributes in other settings between youth and adults, but in addition to consideration of study purpose, the study population should be kept in mind when selecting appropriate measures.
Self-reporting in youth. Although self-reports have seen vast use in adults, they are used less commonly in children. This is because young children have difficulty with comprehension and/or may not have sufficient knowledge of their environments. No specific age cutpoint has been established for when self-reports become appropriate, but self-reports of young adolescents can produce comparable results to parent reports.66 When assessing environments in younger children, questionnaires can be completed by parents or other proxy reporters.