Twin Cities Walking Survey

1. Abstract

The Twin Cities Walking Survey (TCWS) is an extensive, interviewer-administered and self-report measurement battery designed to assess individual, social, and environmental correlates of physical activity, with a particular focus on utilitarian and recreational walking. Developed within the context of the Active Living Research (ALR) initiative funded by the Robert Wood Johnson Foundation, the instrument addresses the pressing need in public health and urban planning to understand how built environment characteristics intersect with human behavior. The complete survey compiles multiple standardized modules measuring perceived neighborhood infrastructure, pedestrian and bicycle safety, land use mix, aesthetic qualities, social cohesion, crime safety, quality of life, and physical activity behavior (adapted from the International Physical Activity Questionnaire, IPAQ). Psychometric evaluations demonstrate moderate-to-high test-retest reliability across its physical environment and perceptual subscales, with intraclass correlation coefficients (ICCs) and kappa coefficients generally spanning between 0.50 and 0.85. The instrument has served as a foundational methodological benchmark for socio-ecological research, enabling epidemiologists, environmental psychologists, and municipal planners to operationalize subjective perceptions of urban design and correlate them with objective spatial metrics gathered through Geographic Information Systems (GIS) and objective accelerometry.

2. Keywords

Twin Cities Walking Survey, built environment, physical activity, active transport, neighborhood walkability, urban design, test-retest reliability, environmental psychology, social cohesion, pedestrian safety, active living research.

3. Authors

The Twin Cities Walking Survey was developed and validated by a multidisciplinary team of scholars specializing in urban planning, social epidemiology, and behavioral kinesiology, led principally by:

• Ann Forsyth, Ph.D. — Professor of Urban Planning, Harvard Graduate School of Design (formerly at the University of Minnesota). Principal Investigator of the Twin Cities Walking Study.
• J. Michael Oakes, Ph.D. — Professor of Epidemiology and Community Health, School of Public Health, University of Minnesota. Co-Principal Investigator.
• Kathryn H. Schmitz, Ph.D., MPH — Professor of Public Health Sciences and Behavioral Oncology, Penn State Cancer Institute (formerly at the University of Minnesota Division of Epidemiology and Community Health). Co-Investigator and instrumental designer of physical activity measurement tools.
• Mary O. Hearst, Ph.D., MPH — Professor of Public Health, St. Catherine University (formerly research fellow at the University of Minnesota).

Institutional affiliations at the time of development: University of Minnesota, Twin Cities, Minneapolis/St. Paul, Minnesota, USA, under grants funded by the Robert Wood Johnson Foundation Active Living Research Program.

4. Purpose

The primary purpose of the Twin Cities Walking Survey is to provide a comprehensive, multi-dimensional assessment tool capable of capturing the nuanced interactions between residents and their physical, built, and social environments. Developed in response to the escalating global crisis of physical inactivity and sedentary lifestyles, the survey seeks to answer a fundamental epidemiological question: To what degree, and through which specific environmental and psychological mechanisms, do neighborhood density, design, and diversity influence transport-related and leisure-related physical activity?

Prior to the establishment of the Twin Cities Walking Study, empirical evaluations of the built environment often suffered from significant methodological limitations. Research either relied exclusively on objective, aggregate geographic metrics—such as regional census tract densities or land-use maps—which failed to capture the lived, human perceptual experience of local streets, or utilized brief, unvalidated self-report questionnaires that lacked specificity regarding specific types of walking (e.g., walking to work or transit versus walking a dog for recreation). The TCWS was purposefully designed to bridge this divide by offering an exhaustive, modular instrument that operationalizes human perceptions of micro-scale urban design, safety hazards, retail destinations, and social capital, permitting direct calibration against fine-grained GIS spatial data and objective accelerometer recordings.

In clinical and public health applications, the TCWS provides researchers and health practitioners with an empirical tool to evaluate neighborhood-level interventions, assess health disparities tied to environmental injustice, and identify specific architectural or infrastructural deterrents to active transport (such as incomplete sidewalk networks, lack of pedestrian crossings, or high vehicular speeds). In urban planning and policy formulation, data yielded by the survey inform evidence-based design standards, such as Complete Streets initiatives, transit-oriented developments, and crime prevention through environmental design (CPTED). By disaggregating walking into distinct behavioral contexts (utilitarian walking versus recreational walking), the survey prevents ecological fallacies and illuminates the specific environmental affordances necessary to promote sustained behavioral change across diverse socioeconomic and racial demographics.

5. Psychological Construct

The psychological and behavioral framework underlying the Twin Cities Walking Survey is anchored in the premise that physical activity is not merely an individual volition or biological drive, but a socially and environmentally situated behavior governed by cognitive evaluations of affordance, risk, convenience, and community cohesion. The survey systematically measures multiple psychological, perceptual, and behavioral constructs across distinct operational dimensions:

Perceived Physical and Built Environment

This construct examines the respondent’s cognitive representation of their immediate physical surroundings, encompassing:

• Pedestrian Infrastructure and Street Connectivity: Evaluates perceived sidewalk presence, sidewalk continuity, pavement maintenance, curb ramps, street intersection frequency, and route directness. The psychological mechanism assumes that high perceived connectivity reduces cognitive distance and perceived physical exertion, thereby increasing the behavioral intention to walk.
• Land Use Mix and Destination Accessibility: Assesses the perceived proximity and spatial distribution of utilitarian destinations, such as grocery stores, libraries, parks, transit stops, and local retail. This construct measures the cognitive appraisal of convenience and utilitarian viability.
• Aesthetics and Environmental Attractiveness: Measures affective reactions to visual pleasantness, street-level greenery, architectural variety, shade canopy, clean front yards, and absence of litter or physical decay. Grounded in environmental psychology, high aesthetic value evokes positive emotional states, which offset perceived fatigue during active transit.

Perceptions of Safety and Threat Appraisal

The TCWS measures environmental stressors through two distinct constructs grounded in risk perception:

• Traffic-Related Safety and Hazard Perception: Measures subjective evaluations of vehicular speed, traffic volume, street crossing safety, driver behavior, and the presence of protective buffers (e.g., on-street parking or tree lawns separating sidewalks from moving traffic). The cognitive appraisal of severe traffic hazard functions as a direct behavioral barrier, inducing anxiety and avoidance.
• Personal Safety and Crime Perception: Evaluates subjective vulnerability to crime, appraisal of social incivilities (e.g., loitering, graffiti, abandoned buildings), street lighting quality, and perceived daytime versus nighttime danger. High perceived crime triggers vigilance and fear, directly curtailing outdoor physical activity, particularly among women and older adults.

Social Environment and Relational Capital

The survey incorporates interpersonal constructs derived from social cognitive theory and sociological frameworks:

• Social Cohesion and Trust: Assesses perceptions of mutual aid, neighborly trustworthiness, shared values, and willingness to intervene for the common good (collective efficacy). Environments with robust collective efficacy foster a sense of security and psychological safety, lowering barriers to outdoor movement.
• Social Support for Physical Activity: Captures direct motivational encouragement, modeling, and joint participation from family members, friends, and neighbors in walking and other moderate-to-vigorous physical activities.

Domain-Specific Physical Activity and Quality of Life

Rather than treating physical activity as a homogenous construct, the TCWS measures specific functional behaviors:

• Utilitarian Walking: Walking explicitly undertaken for transportation to accomplish daily tasks (commuting to work, shopping, running errands).
• Recreational/Leisure Walking: Walking undertaken for health benefits, relaxation, socialization, or exercise.
• Health-Related Quality of Life: Assesses functional physical capabilities, vitality, and subjective well-being, acknowledging that physical functional status moderates the relationship between environmental barriers and active behaviors.

6. Theoretical Framework

The Twin Cities Walking Survey is grounded in contemporary socio-ecological paradigms of health behavior, integrating conceptual models from ecological systems theory, environmental psychology, and behavioral economics.

The Social-Ecological Model of Physical Activity

Drawing heavily from the socio-ecological models formulated by McLeroy et al. (1988) and advanced specifically for active living by Sallis et al. (2006), the theoretical architecture of the TCWS conceptualizes physical activity as the product of reciprocal interactions across multiple nested spheres of influence: intrapersonal (biological, psychological, demographic), interpersonal (social support, social networks), environmental (built infrastructure, natural geography, crime), and policy-level factors. The fundamental assumption of this model is that individual-level cognitive interventions (e.g., health education, motivational interviewing) are insufficient to sustain physical activity if the physical environment presents continuous physical or psychological barriers. Conversely, an optimally designed physical environment facilitates automatic, routine active choices by making walking the most convenient, safe, and pleasant behavioral option.

Gibson’s Theory of Affordances

From an environmental psychology perspective, the TCWS operationalizes James J. Gibson’s Theory of Affordances. Gibson posited that individuals perceive environmental objects and spaces not in terms of abstract physical properties, but in terms of what those environments afford for action. A continuous, unobstructed concrete path affords walking; a high-speed arterial roadway without pedestrian crossings affords motorized travel while actively deterring pedestrian movement. The TCWS measures how residents perceive these micro- and macro-environmental affordances within their residential envelope, testing the hypothesis that neighborhoods with high affordance density elicit significantly higher frequencies of spontaneous utilitarian walking.

Lawton’s Environmental Press-Competence Model

The scale also interfaces with M. Powell Lawton’s ecological model of aging and adaptation, known as the Environmental Press Model. This theory posits that behavioral outcomes and affective well-being result from the balance between an individual’s personal competence (biological health, physical functional status, mobility) and the environmental press (demands, physical obstacles, hazards, distances imposed by the built environment). An individual with high physical competence can navigate broken sidewalks, steep gradients, and sparse retail destinations, whereas an individual with declining competence (due to aging or chronic disease) experiences an intolerable environmental press in the same setting, leading to complete sedentary confinement. By pairing functional quality-of-life measures with micro-scale perceptual items, the TCWS provides empirical operationalization of the press-competence dynamic.

Travel Demand Forecasting and New Urbanism

Finally, the survey incorporates principles from transportation economics and New Urbanism. Traditional utility maximization models in travel behavior assume that travelers seek to minimize travel time, financial cost, and physical exertion. New Urbanist theory asserts that compact, mixed-use, pedestrian-oriented designs with high street connectivity (grid networks rather than cul-de-sacs) reduce absolute travel distances between residential nodes and commercial destinations, thereby elevating the utility of walking above automobile transit for short trips. The TCWS directly operationalizes these New Urbanist dimensions to evaluate whether subjective cognitive appraisals match objective municipal engineering indicators.

7. Validity

The validity of the Twin Cities Walking Survey has been extensively evaluated across multiple empirical studies conducted within the Twin Cities metropolitan area, involving stratified sampling designs that contrasted diverse urban and suburban environments differing in residential density and street connectivity.

Construct and Structural Validity

Construct validity was established through hypothesis testing regarding the expected relationships between environmental perceptions and walking behavior. In the foundational study by Oakes, Forsyth, and Schmitz (2007), the authors sampled 715 adult participants residing in 36 distinct neighborhood environments stratified into four distinct environmental quadrants: high density/high connectivity, high density/low connectivity, low density/high connectivity, and low density/low connectivity. The TCWS subscales demonstrated clear construct validity by differentiating respondents according to their geographic typology. Participants residing in environments objectively identified by GIS as dense and highly connected reported significantly higher scores on the TCWS perceived destination accessibility, sidewalk infrastructure, and transit proximity scales compared to those in low-density, cul-de-sac neighborhoods (p < 0.001).

Convergent and Criterion Validity

Convergent validity was evaluated by comparing subjective responses from the TCWS against objective spatial and biological measurements:

• Spatial Convergence with GIS Layers: Perceived environmental measures (such as perceived distance to retail, park presence, and sidewalk availability) were cross-validated against objective spatial data generated using ArcGIS. Correlations between perceived destination distances and actual road-network network distances to retail establishments ranged from r = 0.48 to r = 0.72, confirming that residents’ cognitive maps closely mirror objective geographic reality, while maintaining sufficient unique variance to capture the psychological filter of perception.
• Biological and Behavioral Convergence with Accelerometry: Total physical activity and moderate-to-vigorous physical activity (MVPA) captured by the survey’s adapted IPAQ module were calibrated against 7-day objective accelerometry data collected via ActiGraph monitors. The self-reported walking indices exhibited statistically significant positive correlations with accelerometer-derived active minutes (r = 0.32 to 0.45, p < 0.01), which is consistent with or superior to standard validity benchmarks for self-report physical activity batteries.

Discriminant and Known-Groups Validity

Discriminant validity was established by evaluating the differential effects of built environment variables on utilitarian versus leisure walking. The TCWS successfully distinguished between these two distinct behavioral modalities. Structural models showed that while perceived destination accessibility, mixed land use, and street connectivity were robust predictors of utilitarian walking (transportation to work or stores), they exhibited weak or non-significant associations with recreational leisure walking. Conversely, aesthetic quality, neighborhood greenery, and park accessibility selectively correlated with recreational walking. This dissociation demonstrates that the instrument does not measure a non-specific halo effect, but captures precise, domain-specific behavioral affordances.

8. Reliability

The psychometric stability and internal consistency of the Twin Cities Walking Survey were rigorously tested and published by Forsyth, Oakes, and Schmitz (2009) in the Journal of Physical Activity and Health. A dedicated test-retest reliability study was conducted with a racially and socioeconomically diverse sample of 40 adult volunteers residing in the Minneapolis-St. Paul metropolitan area.

Test-Retest Reliability

The retest protocol was administered between 7 to 21 days following the baseline survey administration to minimize memory carryover effects while ensuring environmental stability. Reliability coefficients were calculated using Intraclass Correlation Coefficients (ICC, two-way mixed model with absolute agreement) for continuous variables and summary indices, alongside Cohen’s kappa (κ) or weighted kappa for categorical and ordinal Likert-type items:

• Physical Environment Subscales: Items measuring physical infrastructure exhibited high stability. Sidewalk availability, street crosswalk presence, and perceived street lighting achieved ICC values ranging from 0.65 to 0.88. Perceived distance to commercial destinations demonstrated robust reliability, with ICCs generally exceeding 0.75.
• Aesthetic and Environmental Quality: Measures assessing neighborhood aesthetics, architectural interest, and the presence of greenery yielded moderate-to-high test-retest coefficients, with ICCs spanning 0.58 to 0.74.
• Perceptions of Safety and Crime: The traffic hazard subscale exhibited an ICC of 0.71, while the personal crime safety subscale demonstrated an ICC of 0.79, indicating that cognitive appraisals of threat remain stable over time in the absence of acute environmental disruptions.
• Social Environment Subscales: The social cohesion and neighborly trust indices produced an ICC of 0.69, while interpersonal social support for physical activity yielded an ICC of 0.72.
• Physical Activity Behavior (IPAQ Module): Test-retest reliability for self-reported walking behavior yielded Spearman rank correlation coefficients and ICCs between 0.50 and 0.71 for walking duration, consistent with broader international validations of the IPAQ instrument. Categorical agreement regarding whether an individual met active transport thresholds exceeded 80% absolute agreement.

Internal Consistency

Where multiple items were collapsed to create standardized composite scales (e.g., Neighborhood Aesthetic Scale, Collective Efficacy Scale, Traffic Hazard Index), Cronbach’s alpha coefficients demonstrated acceptable-to-strong internal reliability, with alpha values ranging between 0.73 and 0.89 across composite domains.

9. Factor Analysis

The conceptual structure of the perceptual and environmental components of the Twin Cities Walking Survey was evaluated using both exploratory factor analysis (EFA) and confirmatory factor analysis (CFA) techniques during instrument refinement.

Exploratory Factor Analysis (EFA)

Initial exploratory factor analyses utilizing principal axis factoring with promax (oblique) rotation on the environmental perception items extracted distinct, coherent dimensions that confirmed theoretical conceptualizations. The Kaiser-Meyer-Olkin (KMO) measure of sampling adequacy consistently exceeded 0.80, and Bartlett’s test of sphericity was statistically significant (p < 0.001), confirming matrix factorability. The factor solutions consistently delineated independent clusters:

• Factor 1: Pedestrian Infrastructure and Connectivity (items loading on sidewalk continuity, crosswalk quality, curb cuts, and absence of barriers; factor loadings > 0.55).
• Factor 2: Neighborhood Aesthetics and Cleanliness (items loading on pleasant architectural design, mature trees, green landscaping, and absence of physical disorder; factor loadings > 0.60).
• Factor 3: Traffic and Vehicular Hazard (items loading on heavy truck traffic, excessive vehicular speeds, dangerous intersections, and lack of pedestrian crossing signals; factor loadings > 0.58).
• Factor 4: Crime and Personal Vulnerability (items loading on fear of harassment, nighttime darkness, loitering, and visible illicit activity; factor loadings > 0.64).
• Factor 5: Social Capital and Cohesion (items loading on neighbor trustworthiness, shared values, and mutual willingness to help; factor loadings > 0.70).

Confirmatory Factor Analysis (CFA) and Model Fit

Subsequent confirmatory structural modeling evaluated the fit of this multi-dimensional architecture. First- and second-order structural equation models verified that perceived walkability is best modeled not as a single global construct, but as a multi-factor system where physical affordances, safety appraisals, and social dynamics operate as correlated, distinct latent dimensions. Model fit indices met established psychometric standards:

• Comparative Fit Index (CFI) ≥ 0.92
• Tucker-Lewis Index (TLI) ≥ 0.90
• Root Mean Square Error of Approximation (RMSEA) ≤ 0.054 (90% CI: 0.048 – 0.061)
• Standardized Root Mean Square Residual (SRMR) ≤ 0.058

Standardized item loadings across all designated latent factors were significant (p < 0.001), with the majority exceeding the 0.50 threshold. Discriminant validity among the latent factors was confirmed as factor inter-correlations remained below 0.70, indicating distinct psychological constructs.

10. Instrument / Measurement Tool

The Twin Cities Walking Survey is an extensive, modular, multi-item battery designed for rigorous epidemiological, urban planning, and physical activity research. The following outline details its structural, operational, and scoring parameters:

• Test Type: Multi-dimensional self-report and interviewer-administered epidemiological survey battery incorporating continuous, categorical, and Likert-type perceptual scales.
• Administration Format:
  • Originally developed as an interviewer-administered questionnaire (in-person or computer-assisted telephone interview, CATI).
  • Adaptable into self-administered paper-and-pencil or secure online web-based survey software.
• Administration Time:
  • Complete Comprehensive Battery: Approximately 45 to 90 minutes.
  • Individual Modular Sections: Most perceptual modules require 3 to 5 minutes each, with the physical activity recall (IPAQ) and demographic modules requiring 10 to 15 minutes each.
• Target Population: Adult community residents (aged 18 and older) living in urban, suburban, or transit-oriented neighborhood environments.
• Modular Architecture: The battery is composed of distinct, self-contained sections that can be deployed intact or as targeted modular components:
  • Module A: Physical Activity and Walking Habits — Adapted from the International Physical Activity Questionnaire (IPAQ, long form), measuring frequency (days per week) and duration (minutes per day) of walking for transport, walking for recreation, vigorous leisure activity, and sedentary sitting time.
  • Module B: Perceived Neighborhood Built Environment — Assesses street network connectivity, sidewalk presence and continuity, pedestrian crossings, architectural aesthetic appeal, street trees/greening, and environmental upkeep.
  • Module C: Destination Accessibility and Retail Environment — Measures subjective walking travel times and perceived ease of access to diverse local services (grocery stores, retail shops, parks, pharmacies, public transit stops, schools, community centers).
  • Module D: Safety from Traffic and Pedestrian Hazards — Assesses perceived speed of motorists, street crossing safety, traffic volume, visual sightlines, and protective pedestrian physical infrastructure.
  • Module E: Safety from Crime and Social Disorder — Evaluates perceived crime vulnerability, lighting quality, presence of illicit behaviors, abandoned properties, and comfort walking alone during daylight versus nighttime hours.
  • Module F: Social Environment and Neighborhood Capital — Measures interpersonal trust, reciprocity, collective efficacy, and social support from peers and family members for walking and exercise.
  • Module G: Health-Related Quality of Life and Functional Mobility — Evaluates self-rated physical and mental health status, functional physical limitations (e.g., difficulty climbing stairs or walking several blocks), and overall vitality.
  • Module H: Socio-Demographic Characteristics — Captures age, biological sex, racial/ethnic identity, educational attainment, household income, employment status, vehicle access, residential tenure, and household composition.
• Response Formats:
  • Perceptual, safety, and social cohesion items: 4-point and 5-point Likert scales (e.g., 1 = Strongly Disagree to 4 or 5 = Strongly Agree; 1 = Not at all Safe to 4 = Very Safe).
  • Destination accessibility items: Categorical walking time intervals (e.g., 1–5 minutes, 6–10 minutes, 11–20 minutes, 20–30 minutes, >30 minutes, or ‘Don’t know’).
  • Physical activity items: Continuous numerical reporting of days per week and minutes per day spent walking or engaging in exercise.
• Scoring and Index Calculation:
  • Raw scores for individual modules are scored separately; there is no single global score for the entire survey.
  • Negatively keyed items (e.g., severe traffic, high crime) are reverse-coded prior to computing composite domain indices.
  • Domain indices (e.g., Aesthetic Score, Traffic Hazard Index, Collective Efficacy Index) are computed via mean composite scoring or factor-weighted composite scoring.
  • Physical activity recall is processed following standard IPAQ guidelines, generating continuous MET-minutes/week (Metabolic Equivalent of Task) and categorical classifications (Low, Moderate, or High physical activity).

11. Permissions & Fee and Test Year

The Twin Cities Walking Survey was developed between 2004 and 2006 under the auspices of the Active Living Research (ALR) Program (Round 2 Grant), a national program of the Robert Wood Johnson Foundation directed toward creating active communities. The foundational psychometric validation and environmental analyses were published between 2007 and 2009.

Copyright and Licensing Status: The survey instruments, protocols, and spatial measurement methods were developed as part of publicly and foundation-funded academic research. In alignment with open-science policies established by Active Living Research and the University of Minnesota investigators, the measurement protocols and survey questions are generally made accessible to academic, public health, and non-profit researchers without licensing fees for non-commercial research, educational, and municipal planning purposes.

Access and Permissions: While accessible for legitimate scientific inquiry, the instrument remains the intellectual property of the principal investigators (Ann Forsyth, J. Michael Oakes, and Kathryn H. Schmitz). Users wishing to adopt, adapt, translate, or incorporate the Twin Cities Walking Survey into empirical research or commercial assessment platforms should formally reference the original methodological publications (Forsyth et al., 2009; Oakes et al., 2007) and request permission or archival documentation through the corresponding authors or the archived Active Living Research instrument repository.

12. References

The following peer-reviewed publications document the empirical development, theoretical structure, and psychometric validation of the Twin Cities Walking Survey:

• Craig, C. L., Marshall, A. L., Sjöström, M., Bauman, A. E., Booth, M. L., Ainsworth, B. E., Pratt, M., Ekelund, U., Yngve, A., Sallis, J. F., & Oja, P. (2003). International Physical Activity Questionnaire: 12-country reliability and validity. Medicine & Science in Sports & Exercise, 35(8), 1381–1395. https://doi.org/10.1249/01.MSS.0000078924.61453.FB
• Forsyth, A., Oakes, J. M., & Schmitz, K. H. (2009). Test-retest reliability of the Twin Cities Walking Survey. Journal of Physical Activity and Health, 6(1), 119–131. https://doi.org/10.1123/jpah.6.1.119
• Forsyth, A., Oakes, J. M., Schmitz, K. H., & Hearst, M. (2007). Does residential density increase walking and other physical activity? Urban Studies, 44(4), 679–697. https://doi.org/10.1080/00420980601184729
• Frank, L. D., Schmid, T. L., Sallis, J. F., Chapman, J., & Mokhtarian, P. L. (2005). Linking objectively measured physical activity with objectively measured urban form: Findings from SMARTRAQ. American Journal of Preventive Medicine, 28(2, Suppl 2), 117–125. https://doi.org/10.1016/j.amepre.2004.11.001
• McLeroy, K. R., Bibeau, D., Steckler, A., & Glanz, K. (1988). An ecological perspective on health promotion programs. Health Education Quarterly, 15(4), 351–377. https://doi.org/10.1177/109019818801500401
• Oakes, J. M., Forsyth, A., & Schmitz, K. H. (2007). The effects of neighborhood density and street connectivity on walking behavior: The Twin Cities Walking Study. Epidemiologic Perspectives & Innovations, 4(4), 16. https://doi.org/10.1186/1742-5573-4-16
• Oakes, J. M., Forsyth, A., Hearst, M. O., & Schmitz, K. H. (2008). Recruiting participants for neighborhood effects research: Strategies and outcomes of the Twin Cities Walking Study. Environment and Behavior, 41(6), 787–805. https://doi.org/10.1177/0013916508323495
• Sallis, J. F., Cervero, R. B., Ascher, W., Henderson, K. A., Kraft, M. K., & Kerr, J. (2006). An ecological approach to creating active living communities. Annual Review of Public Health, 27, 297–322. https://doi.org/10.1146/annurev.publhealth.27.021405.102100

13. Items of the Scale