Inventory of the physical environment domains and subdomains measured by neighborhood audit tools: A systematic literature review

Journal of Environmental Psychology 36 (2013) 179e189

Contents lists available at ScienceDirect

Journal of Environmental Psychology journal homepage: www.elsevier.com/locate/jep

Review

Inventory of the physical environment domains and subdomains measured by neighborhood audit tools: A systematic literature review Jen Nickelson*, April R. Wang, Qshequilla P. Mitchell, Kate Hendricks, Angelia Paschal The University of Alabama, College of Human Environmental Sciences, Department of Health Science, Box 870311, Tuscaloosa, AL 35487-0311, USA

a r t i c l e i n f o

a b s t r a c t

Article history: Available online 13 August 2013

The purpose of this review was to inventory and document existing neighborhood physical environment audit tools and the domains and subdomains measured by these tools. At total of 31 articles met inclusion criteria. We identified 20 major domains and 291 subdomains. Audit instruments most commonly assessed Streets/Traffic, Safety, Land Uses, and Physical Disorder domains. Least commonly assessed domains were Barriers, Neighborhood Identification/Legibility, Steepness, Views/Enclosure, and Ethnic Identification. Within a domain, between 1 and 36 subdomains were assessed. This review will help neighborhood auditors identify instruments that measure domains and subdomains most relevant to their study. This information may also be used to develop customized audit tools that capture those physical environmental characteristics of neighborhoods that auditors are most interested in. Ó 2013 Elsevier Ltd. All rights reserved.

Keywords: Neighborhood Environment Audit Instrument Measurement Review

1. Introduction Quality of life is clearly linked to the physical environment in which people live (World Health Organization, 1997). The physical environment includes both the natural and built environments (Secretary’s Advisory Committee, 2010). The physical environment can include particular settings, such as homes, worksites, and schools; and it can include the neighborhoods and communities in which people live, work, and play. Neighborhoods are geographic and social units that can have profound effects on quality of life (Leung, Gregorich, Laraia, Kushi, & Yen, 2010). However, the nature of the relationship between neighborhood conditions and residents’ well being (and the mediating and moderating factors at play) remains unclear (Parsons et al., 2010). Characterizing neighborhoods and the specific physical features of neighborhoods that may contribute to residents’ quality of life is a complex and difficult undertaking (Parsons et al., 2010). To do so, neighborhoods must be defined by both specific boundaries and then by the physical characteristics within those boundaries. Residents living within the same neighborhood have varying definitions of neighborhood boundaries (Sastry, Pebley, & Zonta, 2002), which exacerbates the difficulty relating neighborhood

* Corresponding author. Tel.: þ1 205 348 3715; fax: þ1 205 348 2956. E-mail addresses: [email protected], [email protected] (J. Nickelson), [email protected] (A.R. Wang), [email protected] (Q.P. Mitchell), [email protected] (K. Hendricks), [email protected] (A. Paschal). 0272-4944/$ e see front matter Ó 2013 Elsevier Ltd. All rights reserved. http://dx.doi.org/10.1016/j.jenvp.2013.07.009

conditions to quality of life. The U. S. Census typically defines neighborhoods in terms of census tracts, which reflect prominent physical features of neighborhoods as well as social and ethnic divisions (Kohen, Brooks-Gunn, Leventhal, & Hertzman, 2002). Census tracts can vary in size, but are usually so large that smaller areas within census tracts must be sampled for research purposes. To maintain a consistent, reproducible sampling area, studies often define their units of analysis as street segments or block faces (e.g., a section of a street bounded by two intersections or a dead end). Sometimes specific settings within a neighborhood are studied, such as housing units, schools, parks, or playgrounds. Schaefer-McDaniel et al. (2010) reviewed the most common methods employed to document the physical characteristics of neighborhoods. Three major approaches used are: (1) resident surveys that give subjective accounts of the perceived environment, (2) administrative data including those derived by censuses, crime reports, etc., and (3) direct observation by outside raters (including by use of audit instruments). Geographic information systems (GIS) are also being employed to allow features of the environment derived by any of the above methods to be mapped to specific neighborhood locations. Each of these approaches has advantages and disadvantages (Schaefer-McDaniel et al., 2010). Although direct observation is subject to its own limitations and observer bias (see Schaefer-McDaniel et al. for a discussion of methodological rigor), it does overcome some of the limitations of subjective and administrative data. For example, direct observation methods are not subjected to residents’ social desirability bias and can identify neighborhood characteristics (such as the presence of

180

J. Nickelson et al. / Journal of Environmental Psychology 36 (2013) 179e189

trash) that are not captured by administrative data (SchaeferMcDaniel et al., 2010). Many of the direct observation methods employed are derived from the work of Raudenbush and Sampson (1999), Caughy, O’Campo, and Patterson (2001), and Pikora et al. (2002). The systematic social observation method developed by Raudenbush and Sampson (1999) involves driving through neighborhoods while one researcher videotapes and two others record their observations in a log. This method involves coding observations after driving through the neighborhood. Caughy’s and Pikora’s groups created audit instruments that allow trained observers to objectively rate components of the neighborhood environment while they are physically present in the neighborhood. Neighborhood audit instruments traditionally have been developed and used by social scientists and urban planners to examine community needs and assets, and their application to public health issues also has been well recognized (Dannenberg et al., 2003). For example, social scientists have used neighborhood audit instruments to examine the role of the physical environment on residents’ perceptions of safety and crime (see, for example, Perkins, Meeks, & Taylor, 1992). Theoretically, neighborhoods can be designed to minimize crime and promote safety, and audit instruments have been used to assess the relationship between neighborhood design and crime (e.g., Minnery & Lim, 2005). Much of the attention to the influence of the physical environment on health has been directed toward the walkability or bike ability of neighborhoods, and a variety of instruments have been developed to assess these features (e.g., Pikora et al., 2002). Neighborhood environment audit tools have been developed or revised to meet particular needs. Some instruments assess both the social and physical environment; some address only the physical environment. Some assess the environment for qualities that might influence a particular health or social issue (such as pregnancy or crime), behavior (such as walking or bicycling), or population (such as African Americans or older adults). Because audit instruments are used for an array of purposes, numerous neighborhood characteristics or domains have been assessed. For researchers trying to select or develop an instrument to meet their own specific needs, a complete listing of available audit instruments and domains assessed would be helpful. A domain can be defined as a broad category of similar environmental characteristics, such as land use or physical disorder. The individual items comprising a domain, such as types of residential land use and presence of graffiti, can be called subdomains. Domains and subdomains are conceptually similar to Pikora et al.’s (2002) elements and items, respectively. Schaefer-McDaniel et al. (2010) have recently reviewed the literature on neighborhood observations, focusing on methodological rigor, geographical boundaries, and the relationship between neighborhood characteristics and residents’ health. Whereas their review does provide domains assessed, this is not the focus of their paper, and they do not limit their review to observations using audit instruments. A comparison of audit tools that have been used to assess characteristics of the built environment that may be associated with physical activity was developed by Day (n.d.) and is posted on the Active Living Research website (http://www. activelivingresearch.org/files/AuditToolsComparisonTable.pdf). This is a useful comparison with many domains and subdomains provided, but it is limited to five tools that are concerned only with the physical activity environment. Other reviews (e.g., Moudon & Lee, 2003) also focus exclusively on the physical activity environment, whereas investigators may be interested in physical environment factors that are related to other behaviors or quality of life issues that may not be affected by physical activity. Therefore, the purpose of this review is to fill the gap in the literature by providing a complete inventory of domains and subdomains assessed by existing neighborhood physical environment audit instruments.

2. Methods 2.1. Search procedure We searched the following social and medical sciences databases for publications in peer-reviewed journals related to direct observations of neighborhood physical environments: Pubmed, CINAHL Plus, PsycINFO, Health Technology Assessments, International Bibliography of the Social Sciences, Criminal Justice Abstracts, Social Services Abstracts, and Sociological Abstracts. Key search terms included the following methodological terms: audit, scan, assessment, observation, checklist, inventory, measure, rating, windshield survey, direct observation, social observation, systematic observation, and systematic social observation; and the following neighborhood terms (using both spellings, neighborhood and neighbourhood): neighborhood/neighbourhood, built environment, neighborhood/neighbourhood environment, community environment, urban environment, and rural environment. 2.2. Inclusion and exclusion criteria The inclusion criteria used to determine a publication’s relevance to this study are as follows: (1) measures include direct observations of neighborhood physical (i.e. as opposed to social) environments, (2) the unit of analysis measured was a segment of a street (e.g., block, block face, or street segment as opposed to a housing unit or playground), (3) measures were original (including modifications of existing measures that made them essentially a different instrument), (4) the article was published in an Englishlanguage peer-reviewed journal between 1990 and June 2010, and (5) studies were conducted in developed countries, as defined by the United Nations (2012), as developed and developing countries may have varying physical environment characteristics and needs (see, for example, Konteh, 2009). Studies that used geographic information system (GIS) or administrative data without also using a neighborhood audit instrument were excluded. Measures in the form of participatory surveys and measures of the social environment were also excluded (unless they also included direct observations of the physical environment). In the event that a publication included measures that were not clearly delineated within the manuscript, we contacted the author using the contact information provided within the publication. If the author could not be reached for, or was unwilling to provide, further clarification of the measures, the publication was excluded. If a measure was described multiple times in the literature without modification to the instrument, only the publication that most fully described the instrument development and psychometric testing was included. The initial literature search generated a total of 11,565 citations. To create a more manageable number of articles to review, duplicate citations were eliminated and citations were systematically searched for exclusion criteria and specific disciplines for which environment may be a common term (e.g., microbiology, climatology, etc.). Articles that clearly did not meet inclusion criteria were excluded, resulting in 2374 citations. From these, titles were assessed for inclusion in this review, resulting in 391 citations. Each of these citations’ abstracts and, if necessary, full articles, was reviewed to determine if they met inclusion criteria. A total of 87 publications were reviewed in their entirety. Of these, 31 (Table 1) were included in this literature review. 2.3. Data abstraction Data abstraction involved two phases. In the first phase, an Excel file was developed based upon a similar one designed by Day (n.d.).

J. Nickelson et al. / Journal of Environmental Psychology 36 (2013) 179e189

181

Table 1 Summary of articles included in review. Citation

#

Instrument or study name (if available)

Andresen et al., 2008 Boarnet, Day, Alfonzo, Forsyth, & Oakes, 2006 Brownson et al., 2004 Brownson et al., 2004 Caughy et al., 2001 Clifton, Livi Smith, & Rodriguez, 2007 Cunningham, Michael, Farquhar, & Lapidus, 2005 Evenson et al., 2009 Ewing, Handy, Brownson, Clemente, & Winston, 2006 Foltete & Piombini, 2007 Franzini et al., 2009 Franzini, Caughy, Nettles, & O’Campo, 2008 Furr-Holden et al., 2010 Hoehner, Ivy, Brennan Ramirez, Handy, & Brownson, 2007 Kamphuis et al., 2008 King, 2008 Laraia et al., 2006 Leung et al., 2010 Loukaitou-Sideris, Liggett, Iseki, & Thurlow, 2001 McMillan, 2007 Michael, Keast, Chaudhury, Day, Mahmood, & Sarte, 2009 Minnery & Lim, 2005 Parsons et al., 2010 Perkins et al., 1992 Pikora et al., 2002 Schaefer-McDaniel, 2009 Seymour, Reynolds, & Wolch, 2010 Suminski, Heinrich, Poston, Hyder, & Pyle, 2007 Wright & Kloos, 2007 Zenk et al., 2007 Zhu & Lee, 2008

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31

African American Health (AAH) Neighborhood Assessment Scale Irvine-Minnesota Inventory St. Louis Audit Tool e Analytic Version St. Louis Audit Tool e Checklist Version Neighborhood Brief Observation Tool Pedestrian Environmental Data Scan (PEDS) Senior Walking Environmental Assessment Tool (SWEAT) PIN3 Neighborhood Audit Instrument Field Manual Urban Design Measures for NYC N/A Healthy Passages N/A Neighborhood Inventory for Environmental Typology (NIfETy) Active Neighborhood Checklist Victorian Lifestyle and Neighbourhood Environments Study (VicLANES) Neighborhoods and Senior Health (NASH) Neighborhood Attributes Inventory Cohort Study of Young Girls’ Nutrition, Environment, and Transitions (CYGNET) N/A Safe Routes to School (SR2S) Senior Walking Environmental Audit Tool-Revised (SWEAT-R) Crime Prevention Through Environmental Design (CPTED) Scale St. Michael’s Neighborhood Observation Data Collection Tool Block Environmental Inventory (BEI) Systematic Pedestrian and Cycling Environmental Scan Instrument (SPACES) N/A Systematic Pedestrian and Cycling Environmental Scan for Alleys Instrument (SPACES for Alleys) Block Walk Method (BWM) Housing Environment Rating Scale-Neighborhood Quality (HERS-NQ) Healthy Environments Partnership Neighborhood Observational Checklist (NOC) N/A

With Day’s table as a starting point, two coders (AW and QM) independently reviewed each of the 31 included instruments to create a list of domains and subdomains. Domains represented the broad categories of environmental characteristics included in an instrument, such as land use and physical disorder. Subdomains represented the specific items, or questions, addressed in a measure to characterize the domain, such as the types of residential land use and the presence of graffiti. New domains and subdomains were added as needed based on the contents of each individual measure of neighborhood physical environments. The process of assigning labels to each of the domains and subdomains used a qualitative data analysis procedure, where themes (domains and subdomains) were identified, and codes (labels or names) were assigned to themes (Grbich, 1999). When possible, the codes were informed by existing labels found in the literature (starting with Day’s terminology); however, because different authors may have referred to similar domains and subdomains with different terms, a single code was selected based on the context, frequency of terms used, or the term’s application to a general audience (instead of just to researchers interested in the physical activity environment, for example). If a subdomain did not definitively fit within an existing domain, a new domain was created. The process was meant to identify all themes captured by neighborhood audit instruments while making the list as parsimonious as possible. To help limit the list, a classification system was created for the type of data collected: amount (A), condition (C), location (L), presence (P), types (T), or a check mark (U). For example, instead of having two separate subdomains for amount of heat lamps and presence of heat lamps, we had a single heat lamp subdomain that could be marked with one or more of the classification codes. An A denoted that the instrument specifically inquired about the quantity of an item, a C referred to the state of repair, and a P denoted the existence of a given item on the street segment. A T referred to detailed questions about variance in type;

for example, if an instrument required the observer to delineate specific types of commercial land uses present, such as movie theaters and/or grocery stores, a T would be coded for the commercial land use subdomain. A U meant that the audit instrument assessed the subdomain in general, but it did not explicitly assess the amount, condition, location, presence, or types of the particular subdomain. During the first phase, coders independently abstracted information from the instruments to include in the Excel file. Inclusion of a subdomain within an audit instrument was denoted in the Excel file by an A, C, L, P, T, or U according to the classification system described above. When audit instruments included observation of the social environment, resident perceptions of the neighborhood environment, or different units of analysis (e.g., housing units), only the information pertaining to raters’ direct observation of the physical environment in neighborhood street segments or block faces (per inclusion criteria) was extracted. For example, the Block Environmental Inventory (Perkins et al., 1992) has four sections, but coders extracted data only from Section 1 (excluding the audit of people), which assesses block-level characteristics. Coders also did not extract data from Sections 2e4, which assess property-level characteristics. Coders discussed and resolved discrepancies until a consensus was reached. The second phase of data abstraction was conducted to establish inter-rater reliability between coders during phase 2 and between consensus achieved during phase 1 and phase 2. In this phase, two coders (QM and KH) first independently abstracted information from approximately 10% (n ¼ 3) of the articles, using the process described for phase 1. The coders then checked the manner in which they were operationalizing the labels to ensure congruence in coding and assessment label meaning. After coming to agreement, they independently abstracted information from the remaining articles. Inter-rater reliability between coders was calculated as a percentage of the total number of possible

182

J. Nickelson et al. / Journal of Environmental Psychology 36 (2013) 179e189

subdomains for which coders agreed on subdomain identification and classification (Table 2). Cohen’s Kappa also was estimated, using SPSS (IBM SPSS, Version 21). Kappa values range from 0 to 1.0, with larger values representing better reliability. Kappa values may be interpreted as follows: 0.0e0.2 ¼ slight agreement, 0.21e 0.40 ¼ fair agreement, 0.41e0.60 ¼ moderate agreement, 0.61e 0.80 ¼ substantial agreement, and 0.81e1.0 ¼ almost perfect agreement (Landis & Koch, 1977). Discrepancies were discussed and consensus reached. The consensus reached during the second phase was compared to that obtained during the first phase to establish inter-rater reliability between phases for both subdomain identification and classification. When there was disagreement between phases, an independent coder (JN) ultimately determined data coding to create the final inventory of domains and subdomains (Appendix A). 3. Results Table 1 provides a listing of the 31 neighborhood physical environment audit instruments identified by our review, listed in alphabetical order by the last name of the first author. Consistent with the methodology used by Schaefer-McDaniel et al. (2010), instruments were numbered to assist the reader in linking them with other findings reported in this paper. In addition, the name of the instrument or, alternately, the name of the study was provided when it was available. Twenty major domains were identified (Table 3). Audit instruments assessed between 3 (study number 11) and 19 (study number 2) domains (mean ¼ 11.0, median ¼ 12, mode ¼ 13). Audit instruments most commonly assessed Streets/Traffic, Safety, Land

Uses, and Physical Disorder domains. Least commonly assessed domains were Barriers, Neighborhood Identification/Legibility, Steepness, Views/Enclosure, and Ethnic Identification. Within a domain, between 1 and 36 subdomains were assessed (mean ¼ 14.6, median ¼ 13.5, mode ¼ 17). We identified 291 subdomains (Table 4). The Amenities for Outdoor Public Spaces domain measures convenience and comfort features found in outdoor public spaces. It delineates 16 subdomains. Of these, the most commonly measured are bus/transit stops (studies numbered 2e4, 6e8, 13e14, 16, 19, 21, 23, 29e30), street furniture (studies numbered 2e4, 6, 7, 9, 13e14, 16, 20, 23, 27, 29), playground and sports equipment (studies numbered 3, 4, 14, 18, 23, 27, 30), trash bins (studies numbered 3e6, 16, 23, 27), bench and/or covered shelter at transit stop (studies numbered 2, 3, 6, 8, 14, 21), outdoor dining areas (studies numbered 2e4, 9, 21, 23), and public telephones (studies numbered 3e5, 16, 19, 23). The Architecture/Building Characteristics domain measures building attributes that are permanent, or not easily changed. It includes 20 subdomains. The most popularly measured are buildings/residential units with a front porch (studies numbered 2, 5, 8, 10, 13, 16, 17, 21, 23), buildings/residential units with a front yard (studies numbered 5, 7, 8, 13, 15e17, 21, 23), interesting and/or varied architecture/design (studies numbered 2e4, 6, 10e12, 14, 25), building height (studies numbered 2, 7, 10, 15, 21, 23, 27), buildings/residential units with some form of decoration (studies numbered 5, 8, 13, 16, 17, 23, 30), and properties/residential units with a border (studies numbered 5, 8, 13, 16, 17, 23, 30). The Barriers domain measures obstacles in the environment that may be difficult to overcome for, or off-putting to, pedestrians and

Table 2 Inter-rater reliability. Instrument #

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 a

Inter-rater reliability for coders during phase 2

Inter-rater reliability for consensus between phase 1 and phase 2

Inter-rater reliability in subdomain identification

Inter-rater reliability in subdomain classification

Inter-rater reliability in subdomain identification

Inter-rater reliability in subdomain classification

% Agreementa

Kappa (95% CI)

% Agreementa

Kappa (95% CI)

% Agreementa

Kappa (95% CI)

% Agreementa

Kappa (95% CI)

100.0 100.0 99.7 99.7 100.0 100.0 100.0 99.7 100.0 99.7 100.0 100.0 100.0 99.7 99.7 99.7 100.0 99.7 99.3 99.7 100.0 100.0 99.0 100.0 100.0 100.0 98.6 99.7 99.7 99.3 100.0

1.0 1.0 0.99 0.99 1.0 1.0 1.0 0.99 1.0 0.97 1.0 1.0 1.0 0.99 0.97 0.99 1.0 0.98 0.94 0.98 1.0 1.0 0.98 1.0 1.0 1.0 0.94 0.96 0.98 0.98 1.0

99.0 98.6 99.7 99.0 99.3 99.3 99.3 99.3 99.7 98.3 100.0 99.7 99.3 99.3 99.7 98.6 99.7 99.0 99.3 99.3 100.0 100.0 97.9 100.0 99.0 100.0 97.6 99.0 99.7 98.3 100.0

0.97 0.96 0.99 0.98 0.98 0.98 0.97 0.97 0.97 0.93 1.0 0.96 0.98 0.98 0.97 0.95 0.99 0.95 0.94 0.93 1.0 1.0 0.96 1.0 0.96 1.0 0.90 0.87 0.98 0.96 1.0

90.0 98.6 97.6 100.0 100.0 100.0 100.0 100.0 100.0 99.0 100.0 100.0 100.0 100.0 100.0 100.0 94.5 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0

0.35 0.90 0.95 1.0 1.0 1.0 1.0 1.0 1.0 0.91 1.0 1.0 1.0 1.0 1.0 1.0 0.82 1.0 1.0 1.0 1.0 1.0 1.0 1.0 1.0 1.0 1.0 1.0 1.0 1.0 1.0

88.7 94.2 97.6 100.0 100.0 98.3 99.3 99.7 99.7 99.0 100.0 100.0 99.7 100.0 100.0 100.0 93.1 100.0 100.0 100.0 100.0 100.0 99.0 99.7 100.0 100.0 99.3 100.0 100.0 99.7 100.0

0.28 0.86 0.92 1.0 1.0 0.98 0.96 0.99 0.97 0.91 1.0 1.0 0.93 1.0 1.0 1.0 0.82 1.0 1.0 1.0 1.0 1.0 0.98 0.97 1.0 1.0 0.97 1.0 1.0 0.99 1.0

(1.0, 1.0) (1.0, 1.0) (0.98, 1.01) (0.98, 1.01) (1.0, 1.0) (1.0, 1.0) (1.0, 1.0) (0.96, 1.01) (1.0, 1.0) (0.91, 1.03) (1.0, 1.0) (1.0, 1.0) (1.0, 1.0) (0.96, 1.01) (0.91, 1.03) (0.96, 1.01) (1.0, 1.0) (0.94, 1.02) (0.85, 1.02) (0.93, 1.02) (1.0, 1.0) (1.0, 1.0) (0.95, 1.00) (1.0, 1.0) (1.0, 1.0) (1.0, 1.0) (0.88, 1.00) (0.86, 1.04) (0.93, 1.02) (0.97, 1.00) (1.0, 1.0)

(0.94, 1.01) (0.93, 1.00) (0.98, 1.01) (0.95, 1.00) (0.96, 1.01) (0.96, 1.01) (0.94, 1.01) (0.93, 1.00) (0.93, 1.02) (0.83, 1.03) (1.0, 1.0) (0.89, 1.03) (0.95, 1.01) (0.95, 1.01) (0.91, 1.03) (0.91, 1.00) (0.96, 1.01) (0.88, 1.02) (0.86, 1.02) (0.86, 1.01) (1.0, 1.0) (1.0, 1.0) (0.93, 0.99) (1.0, 1.0) (0.92, 1.00) (1.0, 1.0) (0.82, 0.97) (0.73, 1.00) (0.93, 1.02) (0.92, 0.99) (1.0, 1.0)

(0.26, 0.52) (0.84, 0.96) (0.91, 0.99) (1.0, 1.0) (1.0, 1.0) (1.0, 1.0) (1.0, 1.0) (1.0, 1.0) (1.0, 1.0) (0.90, 0.92) (1.0, 1.0) (1.0, 1.0) (1.0, 1.0) (1.0, 1.0) (1.0, 1.0) (1.0, 1.0) (0.74, 0.91) (1.0, 1.0) (1.0, 1.0) (1.0, 1.0) (1.0, 1.0) (1.0, 1.0) (1.0, 1.0) (1.0, 1.0) (1.0, 1.0) (1.0, 1.0) (1.0, 1.0) (1.0, 1.0) (1.0, 1.0) (1.0, 1.0) (1.0, 1.0)

(0.13, 0.42) (0.80, 0.93) (0.88, 0.96) (1.0, 1.0) (1.0, 1.0) (0.96, 1.01) (0.91, 1.00) (0.97, 1.01) (0.90, 1.02) (0.81, 1.01) (1.0, 1.0) (1.0, 1.0) (0.88, 0.98) (1.0, 1.0) (1.0, 1.0) (1.0, 1.0) (0.73, 0.91) (1.0, 1.0) (1.0, 1.0) (1.0, 1.0) (1.0, 1.0) (1.0, 1.0) (0.96, 1.00) (0.92, 1.03) (1.0, 1.0) (1.0, 1.0) (0.94, 1.01) (1.0, 1.0) (1.0, 1.0) (0.98, 1.01) (1.0, 1.0)

% Agreement ¼ total number of subdomains (291)  number of subdomains for which there was disagreement, divided by total number of subdomains, 100.

0 1 0 0 0 0 0 2 0 1 1 0 0 2 6 0 2 1 1 1 10 2 3 0 0 9 5 3 7 1 1 0 10 2 13 1 8 2 0 7 0 4 S-D ¼ subdomains. a Total number (%) of instruments assessing domain.

30 29

2 1 0 0 0 3 1 4 0 0 0 6 1 0 1 0 0 0 3 0 11 0 0 0 0 0 0 2 2 0 0 0 2 0 1 2 0 0 0 3 0 6

28 27

3 2 1 0 0 10 1 1 0 2 1 4 0 5 0 2 2 1 3 1 15 0 0 0 0 0 0 0 1 0 0 0 7 0 1 1 0 0 0 1 0 5

26 25

0 1 0 3 0 6 4 1 0 3 5 4 0 2 5 0 0 1 8 1 13 0 0 1 0 0 0 1 1 1 1 0 3 3 2 0 0 0 0 6 0 10

24 23

8 7 0 2 0 8 2 6 2 5 1 13 5 8 3 10 2 0 8 0 16 0 0 0 0 0 1 0 1 0 0 1 0 1 3 0 0 0 0 3 0 7

22 21

4 2 0 1 0 10 4 3 0 2 6 4 6 2 11 1 0 1 9 0 15 1 1 0 1 0 3 1 1 0 0 2 3 0 2 2 0 0 0 7 0 11

20 19

3 0 0 0 0 3 0 0 0 2 0 4 0 1 1 0 0 0 4 0 13 1 0 4 1 0 7 0 0 0 0 2 7 2 0 2 0 0 0 4 0 9

18 17

1 5 0 0 0 8 0 8 2 2 2 9 1 5 1 8 0 0 3 0 13 4 4 0 1 0 4 1 0 1 2 6 5 3 5 6 2 0 0 5 0 14

16 15

0 2 0 1 0 0 1 1 0 2 2 0 0 1 3 0 0 1 4 1 10 5 2 0 1 0 7 2 1 0 2 3 6 3 0 9 0 0 1 12 0 13

14 13

2 5 0 0 0 5 2 4 0 0 0 15 4 5 1 1 1 0 7 0 12 0 1 3 0 0 2 0 2 0 0 0 8 0 1 0 0 0 0 0 0 4

12 11

0 2 3 0 0 2 0 0 0 0 0 6 0 0 0 0 0 0 1 0 3 0 8 8 0 0 4 4 0 1 2 0 0 1 0 2 0 0 0 1 0 9

10 9

2 0 0 0 0 0 3 0 0 0 0 0 3 2 0 0 1 0 0 2 7 2 4 0 4 0 6 0 9 1 2 3 5 1 5 4 1 0 0 6 0 14

8 7

2 2 0 1 0 5 2 1 0 3 4 1 0 2 10 1 0 1 5 0 14 6 2 0 6 0 6 1 2 0 4 10 0 0 1 15 0 0 1 10 1 13

6 5

3 5 0 0 0 10 1 7 2 2 1 4 2 5 0 14 0 0 4 0 13 11 1 4 3 0 8 7 0 0 3 4 11 8 5 2 12 2 0 8 0 15

4 3

12 2 4 6 0 8 7 0 0 3 4 11 8 5 8 13 2 0 10 0 15 6 8 8 1 0 9 12 2 1 2 1 3 7 3 5 1 1 1 1 2 19

2 1

0 0 0 0 0 6 1 4 0 1 0 10 0 6 0 3 2 0 3 0 9 16 20 10 11 9 14 17 13 2 12 17 23 13 25 24 21 3 1 36 4

Study number of neighborhood audit instrument (see Table 1) # of S-D

Table 3 Domains identified, number of subdomains in each domain, and instruments assessing each domain.

Amenities for outdoor public spaces Architecture/building characteristics Barriers Cycling environment Ethnic identification Land uses Landscaping/nature features Maintenance/appearance Neighborhood identification/legibility Parking & driveways Pedestrian environment Physical disorder Recreational uses/public spaces Safety Sidewalks Signs/messages Smell/pollution/noise Steepness Streets/traffic Views/enclosure Number of domains assessed

31

20 23 9 14 1 25 22 23 9 21 19 25 18 26 23 14 10 9 29 7

N (%)a

(64.5) (74.2) (29.0) (45.2) (0.3) (80.6) (71.0) (74.2) (29.0) (67.7) (61.3) (80.6) (58.1) (83.9) (74.2) (45.2) (32.3) (29.0) (93.5) (22.6)

J. Nickelson et al. / Journal of Environmental Psychology 36 (2013) 179e189

183

cyclists. It encompasses 10 subdomains. Those measured most often are highway (studies numbered 2e4, 10e12, 18), railroad track (studies numbered 2e4, 10e12, 18), and bridge (studies numbered 3, 4, 11, 12, 18). The Cycling Environment domain measures features that either hinder or promote an environment conducive to cycling. It is composed of 11 subdomains. Of these, the most frequently measured are bicycle parking facilities (studies numbered 2, 3, 6, 8, 15, 16, 23, 25), bicycle lanes (studies numbered 3, 4, 7, 8, 20, 21, 23), “share the road” sign (studies numbered 3, 4, 6, 8, 14), and other bicyclist friendly traffic sign (studies numbered 3, 4, 6, 8). The Ethnic Identification domain measures the presence of businesses and institutions with displays, signage, and murals tailored to a specific ethnic group. It consists of 9 subdomains. Given that audit instrument 30 exclusively measures these 9 items within this domain, no one subdomain is measured more often than another. The Land Uses domain measures the functional purpose of land segments. It contains 14 subdomains. Of these, the most often measured are commercial retail uses (studies numbered 1e8, 10, 12e14, 16e19, 21, 23, 25, 27, 29, 30), residential land uses (studies numbered 1e8, 10, 13, 14, 16e19, 21, 23, 25, 27, 30), institutional buildings (studies numbered 1e8, 13, 17, 18, 21, 23, 27, 30), industrial/manufacturing uses (studies numbered 1, 2, 5, 6, 8, 10, 14, 17, 18, 21, 25, 27, 30), and public/civic buildings (studies numbered 2e 5, 7, 10, 14, 17e19, 21, 23). The Landscaping/Nature Features domain measures the characteristics of surrounding scenery. It includes 17 subdomains. The most popularly measured are street trees (studies numbered 2e7, 9, 10, 14, 15, 20, 21, 23e25, 28, 30), landscape features (studies numbered 3, 4, 10, 13, 25, 28), lake/pond (studies numbered 2e4, 16, 21), open field/open space (studies numbered 2e4, 10, 27), public art (studies numbered 2, 9, 13, 14, 30), and stream/river/canal/creek (studies numbered 2e4, 10, 21). The Maintenance/Appearance domain primarily measures the current condition of buildings in addition to building attributes that are not permanent. It delineates 13 subdomains. Those most often measured are maintenance of buildings/lots (studies numbered 2, 5e7, 13, 17, 20, 21, 24, 26, 29, 31), condition of grounds (studies numbered 5, 8, 15, 17, 21, 23, 25, 28e30), condition of public recreational spaces (studies numbered 1, 5, 8, 12, 17, 21e23, 29, 30), and condition of residential buildings (studies numbered 1, 5, 8, 12, 17, 23, 29, 30). The Neighborhood Identification/Legibility domain measures unique features that give definition to a particular neighborhood. It contains 2 subdomains. The neighborhood monuments/ markers/banners (studies numbered 2, 5, 8, 10, 17, 23, 30) subdomain is measured more frequently than the indication of neighborhood/block uniformity (studies numbered 5, 16, 17, 23, 24) subdomain. The Parking & Driveways domain predominantly measures appropriate places to leave a vehicle while it is not in operation. It encompasses 13 subdomains. Those most commonly measured are parking lots/structures (studies numbered 1e4, 7, 10, 14, 19, 21, 23e 25, 27, 30), on-street parking (studies numbered 3, 4, 6e8, 14, 16, 19, 21, 23, 27, 31), driveways (studies numbered 3, 4, 6, 15, 25), and parking restrictions (studies numbered 7, 8, 15, 23, 25). The Pedestrian Environment domain measures features that either hinder or promote an environment conducive to pedestrian activity. It is composed of 17 subdomains. Of these, the most popularly measured are traffic/pedestrian signal system (studies numbered 3e8, 14e17, 20e21, 25), crosswalks (studies numbered 3e4, 6e8, 14, 16, 20e21, 25), pedestrian friendly traffic sign (studies numbered 3e4, 6, 8, 18, 21), pedestrianized street (studies numbered 3e4, 6, 15, 21, 25), difficulty for walking (studies

184

J. Nickelson et al. / Journal of Environmental Psychology 36 (2013) 179e189

Table 4 Subdomains included under each domain. Amenities for outdoor public spaces

Architecture/building characteristics

Barriers

Cycling environment

Ethnic identification

 Bench and/or covered shelter at transit stop  Bus headway (timing between buses)  Bus stop/transit stop signage  Bus stops/transit stops  Equipment rental stand  Heat lamps  Newspaper dispenser  Outdoor dining areas  Playground, sports equipment  Public restrooms  Public telephones  Sports stands/seating  Street furniture  Trash bins  Vending machines  Water fountains (for drinking)

 Average height of borders (e.g. fence)  Building accent colors  Building access  Building height  Building setbacks  Building shapes  Buildings  Buildings that have garage doors facing the street  Buildings with identifiers  Buildings with windows facing the street  Buildings/residential units with front porches  Buildings/residential units with front yard  Buildings/residential units with some form of decoration  Fire escapes  Historic buildings  Interesting, varied architecture/design  Prominence of garage doors when viewing front of building  Properties/residential units with a border (e.g. fence)  Vertical-mixed use (different land uses on different floors of the building)  Windows/blank walls at street level

 Ability to overcome the specified barrier  Barriers (general)  Bridge  Drainage ditches  Highway (elevated or below ground)  Impassable land use (e.g. gated community, major industrial complex, etc.)  Railroad track  River  Road with 6 or more lanes  Tunnel

 Attractiveness for cycling  Bicycle lanes  Bicycle parking facilities  Continuity of bicycle route  Crossing aids for bicyclists, non-specified  Demarcation of bicycle lanes  Difficulty for cycling  Obstructions in bicycle lanes  Other bicyclist friendly traffic sign  Perceived safety while biking  "Share the road" sign

 Business featuring an explicit display of colors, murals, or symbols oriented toward African Americans  Business or institution featuring an explicit display of colors, murals, or symbols oriented toward Latinos  Business or institution with “African,” “Caribbean,” or “African American” in the name?  Business or institution with a “Mexican,” “Latino,” “Cuban,” or a Spanish name or surname in the name or a name in Spanish?  Business or institution with a sign or advertisement in Spanish on the building or property  Business or with a sign or advertisement indicating that they sell African or Caribbean goods or provide services specifically for African Americans  Business or institution with a sign or advertisement indicating that they sell Latino or Mexican goods or provide services specifically for Latinos  Sayings/symbols/murals of African American identity or pride  Sayings/symbols/murals of Mexican or Latino identity or pride

Land uses

Landscaping/nature features

Maintenance/appearance

Neighborhood identification/legibility

Parking and driveways

 Agricultural land/ranch/farming  Commercial/retail uses  Educational uses  Industrial/manufacturing uses  Institutional buildings  Integration of land use  Non-residential land uses (general)  Predominant land use  Predominant type of residential housing  Public recreational spaces  Public/civic buildings  Residential land uses  Transportation facilities  Undeveloped land

 Desert  Forest/woods  Fountain/reflecting pool  Height of trees  Lake/pond  Landscape features  Landscaped open space  Mountains/hills  Nature features  Ocean  Open field/open space  Private yard/garden  Public art  Small planters  Stream/river/canal/ creek  Street trees  Waterfront

 Buildings under renovation  Chipped exterior paint or broken fixtures  Clothes drying  Condition of commercial buildings  Condition of grounds  Condition of industrial buildings  Condition of institutional buildings  Condition of public recreational spaces  Condition of residential buildings  Condition of undeveloped property  Maintenance of buildings, lots  Municipal maintenance-related facilities  Newly built buildings

 Indication of neighborhood/block uniformity  Neighborhood monuments/markers/banners

 Driveways  Need to walk through parking lots to enter buildings  No parking/stopping sign  Off-street parking  On-street parking  Parking lots/structures  Parking restrictions  Parking spaces  Parking violations  Predominant form of parking  Predominant use of parking structure on ground floor  Resident parking sign

J. Nickelson et al. / Journal of Environmental Psychology 36 (2013) 179e189

185

Table 4 (continued) Pedestrian environment

Physical disorder

Recreational uses/public spaces

Safety

Sidewalks

 Attractiveness for walking  Crossing aids for pedestrians, non-specified  Crosswalks  Difficulty for walking  Length of crosswalks  Pedestrian cut through  Pedestrian friendly traffic sign  Pedestrian street buffers  Pedestrian street distance from curb  Pedestrian street material  Pedestrian street obstructions  Pedestrianized street  Perceived convenience for pedestrian crossing  Perceived safety while walking  Street markings for pedestrian crossings  Traffic/pedestrian signal system  Traffic/pedestrian signal system length of time

      

 Accessibility of recreational facilities/public spaces  Availability of recreational facilities/public spaces  Beach  Facilities for handicap accessibility  Golf course  Harbor/marina  Institutional yard  Outdoor pool  Park/playground  Playing/sports field  Plaza/square/courtyard  Public garden  Sports track

 Bars on windows  Security bars/gratings  Pull-down metal security blinds  Gate on property  Security fencing  High mesh fencing with barbed wire or spiked tops  Low (<6 ft) security fencing  Surveillance cameras  Security devices  Presence of neighborhood watch/neighborhood block club signs  Block Home or Safe Haven signs  Security warning signs  No tresspassing sign  Beware of dog sign  Visibility from surrounding buildings  Visibility of street and outside lot from 1st floor windows  Police cars  Illegal/unlicensed taxis  Chemical storage barrels  Hazardous liquids  Outdoor lighting  Dogs, stray animals  Animals (other than dogs, stray animals)  Hiding places in street  Perceived safety

 Alternative path buffers  Alternative path distance from curb  Alternative path material  Alternative path obstructions  Alternative paths  Completeness of sidewalks  Continuity of alternative path  Curb cuts  Distance from curb  Features that provide protection from sun/rain/snow  Items in sidewalk buffer zone  Length of alternative path  Lighting covering the sidewalk  Marking for multi-use on/near alternative path  Sidewalk buffers  Sidewalk continuity  Sidewalk material/decorative paving  Sidewalk obstacles/obstructions  Sidewalk shading (e.g. shade from trees)  Sidewalk under construction/being repaired  Sidewalk width  Sidewalks  Width of alternative path  Width of sidewalk buffer zone

     

    

    

Abandoned building(s) Abandoned car(s) Blood Boarded-up buildings Boarded-up windows Broken glass Buildings with broken windows Burned buildings Cigarette butts, tobacco paraphernalia Condoms Empty beer, liquor bottles Eviction notice Evidence of graffiti that has been painted over Graffiti Litter No dumping sign Old, beat-up vehicles Piles of garbage or dumped materials on street Police tape/outlines Shell cases Syringes, drug paraphernalia Vacant lots Vandalism

Signs

Smell/pollution/noise

Steepness

Streets/traffic

Views/enclosure

 Alcohol billboards/signs advertising  Billboards  Cultural messages/events  Drug-free zone sign  Empty sign posts  Fast food billboards/signs advertising  For sale’ and/or ‘for rent’ signs  Help wanted sign  Home-based business sign  Inspirational/educational sayings  Neighborhood/social message/event  Other entertainment/event sign  Physical activity messages, events/billboard  Place to post personal notices/signs  Political message/event  Presence of athletic event sign  Religious messages/events  Sign with a health message  Signs  Tobacco billboards/signs advertising  Unreadable sign/billboard

 Air pollution  Noise levels  Unpleasant smell



 Alley  Alternative routes  Availability of alternative transportation modes  Curb  Curb extension  Curb height  Expressway  Freeway over/underpass  Intersection  Length of street segment  Marked lanes  Median  No commercial vehicles’ sign  Noise level of street  Obstructions in the shoulder  Posted speed limit  Potholes  Road connectivity  Road curve warning sign  Road material  Road obstructions  Road under construction/being repaired  Semis  Shoulder continuity  Shoulder width  Shoulders or wide outside lanes  Special speed zone  Street cleaning sign  Street segments  Street traffic  Traffic calming measures  Turn lane  Vehicle travel lanes  Vehicles  Way-finding aids  Width of street segment

   

Grade/steepness/slope

Attractiveness of views Degree of enclosure Open views/long sight lines Views

186

J. Nickelson et al. / Journal of Environmental Psychology 36 (2013) 179e189

numbered 16, 25, 27, 31), and perceived safety while walking (studies numbered 2, 6, 16, 22). The Physical Disorder domain measures characteristics that render the surrounding environment unsightly and potentially unsafe. It incorporates 23 subdomains. The most often measured are graffiti (studies numbered 1e5, 8, 11e14, 16e21, 23e30), litter (studies numbered 1e5, 7, 8, 11, 12, 14, 16e19, 21, 23, 25e28, 30), abandoned buildings(s) (studies numbered 2e5, 8, 12e14, 16e21, 23, 26), vacant lots (1, 3, 4, 13, 18e20, 23, 24, 27, 29, 30), empty beer/ liquor bottles (studies numbered 1, 3, 4, 11e13, 17, 18, 23, 29, 30), abandoned car(s) (studies numbered 1, 3, 4, 11, 12, 17, 23, 24, 29, 30), broken glass (studies numbered 1, 3, 4, 13, 14, 18, 21, 25, 26, 30), and syringes/drug paraphernalia (studies numbered 1, 3, 4, 12, 13, 17, 23, 29). The Recreational Uses/Public Spaces domain measures destinations and facilities open to the public. It consists of 17 subdomains. Those most frequently measured are park/playground (studies numbered 2e5, 8, 9, 13, 14, 16e18, 21e24, 30), playing/sports field (studies numbered 2e4, 13, 14, 16, 18, 21, 23, 30), public garden (studies numbered 2, 9, 16, 21, 23, 24), outdoor pool (studies numbered 3, 4, 13, 14, 23), and plaza/square/courtyard (studies numbered 2, 9, 10, 13, 21). The Safety domain principally measures physical indicators of the level of protection from harm one might expect. It includes 25 subdomains. Of these, the most often measured are outdoor lighting (studies numbered 2e4, 6e9, 13, 15e17, 19e22, 24, 25, 27, 28), security warning signs (studies numbered 1, 3e5, 8, 13, 16, 17, 23, 30), presence of neighborhood watch/neighborhood block club signs (studies numbered 1, 3e5, 16, 17, 23, 24, 30), no trespassing sign (studies numbered 3e5, 8, 17, 23, 24, 30), bars on windows (studies numbered 2, 7, 16, 21, 23, 30), beware of dog sign (studies numbered 3e5, 8, 23, 30), and security bars/gratings (studies numbered 1, 5, 12, 17, 23, 30). The Sidewalks domain measures the characteristics of paved walking paths. It encompasses 24 subdomains. The most commonly measured are sidewalks (studies numbered 2e4, 6e8, 10, 13e14, 16e18, 20, 21, 23, 26, 28, 30, 31), alternative paths (studies numbered 3, 4, 6, 8, 14e16, 18, 21, 23, 25), sidewalk width (studies numbered 3, 6e7, 10, 14, 16, 19e21, 31), sidewalk buffers (studies numbered 3, 6e8, 14e15, 21, 25, 31), sidewalk obstacles/ obstructions (studies numbered 3, 6e7, 14, 16, 21, 29, 31), sidewalk continuity (studies numbered 3, 6e7, 14, 16, 21, 28), and curb cuts (studies numbered 6e7, 14, 16, 21). The Signs/Messages domain measures visible signs and advertisements. It incorporates 21 subdomains. Those most commonly measured are signs (studies numbered 3e4, 7, 13, 16, 21, 27), cultural messages/events (studies numbered 3e5, 17, 23, 30), ‘for sale’ and/or ‘for rent’ signs (studies numbered 1, 5, 16, 17, 23, 30), neighborhood/social message/event (studies numbered 3e5, 17, 23, 30), alcohol billboards/signs advertising (studies numbered 1, 3e5, 30), billboards (studies numbered 2, 3, 8, 23, 27), political message/ event (studies numbered 3e5, 17, 23), and tobacco billboards/signs advertising (studies numbered 1, 3e5, 30). The Smell/Pollution/Noise domain measures noise and air pollution, including undesirable odors. It delineates 3 subdomains. Of these, the noise levels (studies numbered 3, 4, 9, 13, 23, 27, 31) subdomain is measured most often. The Steepness domain measures incline. It has only 1 subdomain, which is measured frequently. This subdomain is grade/steepness/ slope (studies numbered 2, 6, 7, 14, 15, 21, 25, 27, 31). The Streets domain measures various features of paved vehicular paths. It includes 36 subdomains. The most popularly measured are street segments (studies numbered 1, 3, 5e7, 14, 16, 17, 21, 23e26, 28, 30, 31), traffic calming measures (studies numbered 3, 4, 6e8, 13e18, 20, 21, 23, 25, 28), vehicle travel lanes (studies numbered 3,

6e8, 10, 14e16, 20, 21, 23e25), median (studies numbered 3, 4, 6, 8, 14, 16, 20, 21, 24, 25), posted speed limit (studies numbered 3e6, 8, 14, 16, 17, 20, 23), street traffic (studies numbered 1, 6, 7, 11, 19, 22, 23, 29, 30), curb extension (studies numbered 3, 15, 18, 20, 21, 25), intersection (studies numbered 6, 19, 23, 24, 27, 30), width of street segment (studies numbered 4, 13, 19, 20, 24, 27), and freeway over/ underpass (studies numbered 2e4, 6, 18). The Views/Enclosure domain measures the existence, extent, and appeal of the range of sight. It consists of 4 subdomains. The most commonly measured of these are open views/long sight lines (studies numbered 2, 9, 27) and views (studies numbered 9, 15, 25). 4. Discussion Many direct observation audit instruments measure the physical environment of neighborhoods. These have been developed or revised to meet the needs of those using them. Because audit instruments are used for any number of purposes, numerous domains have been assessed. For researchers trying to select or develop a customized instrument to meet their own specific needs, a more complete listing of available audit instruments and domains assessed was needed. Other reviews of physical environment audit instruments were incomplete (Day, n.d.), focused solely on the physical activity environment (Day, n.d.; Moudon & Lee, 2003), or were not limited to audit instruments (Schaefer-McDaniel et al., 2010). The present review attempted to fill the gap in the existing literature by identifying the domains and subdomains assessed by neighborhood environment audit tools. We identified 20 domains and nearly 300 subdomains, in contrast to the approximately 22 domains identified by Day (n.d.) and approximately 200 subdomains identified by Moudon and Lee (2003). We retained most of the domains and domain names provided by Day, but did exclude, combine, or change the name of some domains. For example, we excluded Day’s People domain because we were not interested in the social environment for this study. We combined Noise and Smell/Pollution, because we believed these all represented different forms of pollution. We changed Health Support to Signs/Messages, because the subdomains represented by this domain were not always healthsupport related. Inter-rater reliability between coders during phase 2 and between phase 1 and phase 2 was generally high, in part because the total number of subdomains is high. Between coders, more discrepancies were found in how to classify a subdomain (e.g., A, C, L, P, T, or U) than in identifying the existence of a subdomain; but once consensus was reached, there was improved agreement between phase 1 and phase 2 in subdomain classification. We believed that properly identifying the subdomain was more important than properly classifying it, especially as instruments varied in how they measured particular concepts. Sometimes the classification was clear. For example, the PEDS instrument (Clifton et al., 2007) assesses various road attributes, including “condition of road.” Coders assigned a C to the street segment subdomain under the Streets/Traffic domain. Sometimes more deliberation was required. For example, coders translated measures of amount (A) to also imply presence (P) of the attribute being measured; therefore, if coders assigned an A, they typically did not also assign a P. However, in some cases, a P was also assigned. For example, PEDS (Clifton et al.) assesses off-street parking lot spaces with response options of (a) 0e5, (b) 6e25, or (c) 26þ. An A was clearly an appropriate classification for the off-street parking subdomain under the Parking and Driveways domain. However, this instrument also instructs the rater: “If none go to Q17.” This “if none” implies an extra layer of “presence.” Therefore, coders also assigned a P to this

J. Nickelson et al. / Journal of Environmental Psychology 36 (2013) 179e189

subdomain for this particular instrument. When coders did not have the actual instrument, but relied only on the article to identify subdomains, classifying the type of measurement was especially challenging (see, for example, Foltete & Piombini, 2007). The lowest inter-rater reliability between phase 1 and phase 2 was for the African American Health Neighborhood Assessment Scale (Andresen et al., 2008). This occurred because the article described several different audit instruments, which caused confusion among the coders as to which instrument should be coded. The varying subdomain names, the way they were measured, and the availability of the actual instrument itself made our task challenging and lent it to varying interpretations. Whereas our inter-rater reliability was high, others may review these instruments and interpret them differently. Furthermore, inconsistencies in terminology became apparent during the course of teasing out the various subdomains and placing each within the most appropriate domain. Different audit instruments often used different words and phrases to describe the same concept. For example, many instruments (Brownson et al., 2004; Clifton et al., 2007; Evenson et al., 2009; Michael et al., 2009) refer to the space between a street and a sidewalk as a sidewalk buffer; but this space is sometimes called a verge (Kamphuis et al., 2008; Pikora et al., 2002). A single code was selected based on either the frequency of terms used or the term’s application to a general audience; and, if a subdomain did not definitively fit within an existing domain, a new domain was created. It was necessary to derive a single term or phrase for the purpose of identifying all of the possible domains and subdomains, and we would recommend using consistent terminology when possible to allow for comparisons between studies. The present review provides suggested standard terminology, but we would encourage neighborhood auditors to not limit themselves to what currently exists if alternate domains and subdomains need to be assessed. Schaefer-McDaniel et al. (2010) note that one standard neighborhood assessment approach may not be appropriate given the variety of research topics, settings, etc. Similarly, standard terminology, while useful, may not fully capture the environmental variable being assessed in varied settings. The placement of each subdomain within the most appropriate domain was challenging at times given some overlap between domains. In these instances, the coder’s discernment was paramount to the proper placement of the subdomain. For example, the subdomain buildings under renovation could justifiably be placed under the Architecture/Building Characteristics domain as well as the Maintenance/Appearance domain, under which it currently resides. Building renovations are both a distinguishing feature of a building at the time of renovation, hence its appropriateness in being labeled as a building characteristic, in addition to being a definitive aspect of the building’s current appearance and state of maintenance. However, because building renovations are not a permanent building feature indicative of the Architecture/Building Characteristics domain, it speaks most to the building’s current appearance and state of maintenance. One solution to this dilemma is to create duplicate subdomains under multiple domains. For example, instead of creating a Steepness domain with a single grade/steepness/slope subdomain, we could have placed this subdomain under the domains of Cycling Environment, Pedestrian Environment, and Streets/traffic (as steepness might influence the bike ability or walkability of a neighborhood or the safety of a street segment). We included presence of neighborhood watch/neighborhood block club signs under the Safety domain, but it might have also gone under the Signs/Messages domain. We cannot ignore the fact that the instruments we reviewed were used in the context of specific assumptions or theories about the environment that may have informed the original domain and

187

subdomain selection and naming. The process of creating an inventory of domains and subdomains measured by neighborhood physical environment audit tools could essentially decontextualize the concepts and in some ways make them less meaningful. It is only when the subdomains and domains are organized with some theory of how these characteristics affect people that their meaning becomes clear. As able, our coders considered the context when abstracting data, which may have also influenced coding and the placement of subdomains within particular domains. For example, we included graffiti as a subdomain under Physical Disorder because researchers have traditionally viewed it as a physical indicator of social disorder and crime (Raudenbush & Sampson, 1999). However, if we changed our assumptions about graffiti, we might have placed it under Ethnic Identification or Signs/Messages. We included an array of signs and messages under the Signs/Messages domain that, in a given context, might be placed under different domains. Occasionally, the context was vague, and having a general Signs/Messages catch-all category was useful. For example, the Irvine-Minnesota Inventory (Boarnet et al., 2006), designed to assess environmental characteristics that might influence physical activity, had an “Other Features” category under which amount of billboards was assessed. Without knowing how billboards might affect physical activity, we placed billboards under the Signs/Messages domain. However, the “share the road” sign subdomain (e.g., Brownson et al., 2004) had a clear and consistent link to bicycling safety, and so we placed it under the Cycling Environment domain. Since we were interested in parsimony at the time we were abstracting data, we selected what we believed was the one best domain under which to include subdomains, taking into consideration the context and typical use in the literature or the term’s application to a general audience (instead of just to researchers interested in the physical activity environment, for example). Because of the possible overlap of subdomains among different domains, neighborhood auditors using this review to create customized audit instruments might find useful subdomains under domains they might not have otherwise been interested in assessing. In many cases, subdomains could be viewed as indicators of a specific construct or factor (domain), as in the case of Physical Disorder. In these cases, researchers theoretically should be able to tally a score for all of the items within the domain and determine the degree to which the domain exists. Internal consistency or scale reliability for these domains should be high, because the subdomains or items within the domain should all be indicators of the same factor. In other cases, subdomains are more representative of the variety of items that could fall within a particular category (domain), as in the case of Land Use. Rather than tally the items, researchers would determine the degree to which each item exists within the category. Internal consistency reliability for these domains would be low, because the subdomains within the domain are not necessarily related or similar. As an example of this, the internal consistency reliability (Cronbach’s a, measured in 2 waves) for the NIfETy Physical Disorder subscale (sample items: broken windows, eviction signs, trash in street, graffiti) was 0.71 and 0.77; but it was 0.27 and 0.27 for the Dwelling/Edifice subscale (sample items: attached homes, single family/detached homes, liquor stores, churches) (Furr-Holden et al., 2010). Our review provides the reader with an inventory of domains and subdomains and the instruments that measure them, but we are not suggesting that all domains have high internal consistency reliability. Neighborhood auditors using this review to create customized audit instruments should make their own determinations about the relation between the subdomains and domains they choose to study based upon their own research questions, hypotheses, and theories driving their research. Inclusion criteria limited audit instruments to those that assessed a block face or street segment as the unit of analysis

188

J. Nickelson et al. / Journal of Environmental Psychology 36 (2013) 179e189

because we needed a uniform unit of analysis to accurately document components measured by these instruments. We excluded property-level audit instruments that looked solely at one component of a neighborhood environment, such as an apartment building (e.g., Kuo & Sullivan, 2001), park (e.g., Saelens et al., 2006), pedestrian access way (e.g., Cozens & Love, 2009), walking route (e.g., Hooker, Cirill, & Wicks, 2007), etc. We also excluded portions of included instruments that assessed property-level data, such as those subdomains assessed by Sections 2e4 of the Block Environmental Inventory (Perkins et al., 1992). Although some of these property-level data might have been observed by the auditor from the street level, to include them in this case would have required we include property-level data from other instruments that had been excluded based on our specified unit-of-analysis inclusion criterion. We also excluded neighborhood audit instruments that used illdefined, or abstract, definitions of a neighborhood, which would have made comparisons difficult. For example, we excluded one instrument that assessed an entire community, with no specification as to what comprised an entire community (Goldberg, Rudd, & Dietz, 1999); we excluded another that assessed the neighborhood block that was visible from school grounds (Limbos & Casteel, 2008). Similarly, we excluded instruments or portions of instruments that assessed the social environment. These excluded instruments or portions of instruments may assess neighborhood characteristics that, being observed from the street, may influence the overall sense of the neighborhood and ultimately impact quality of life; but for this review, boundaries needed to be set in order to undertake the task of documenting these domains and subdomains. Readers are encouraged to consider other instruments or portions of instruments that may assess alternate units of analysis and other aspects of the neighborhood environment. Despite the plethora of items assessed in existing neighborhood physical environment audit instruments, the field is not yet saturated. For example, the location of a driveway can present a potential pedestrian safety hazard. If the driveway must cut through an existing sidewalk in order to reach the front or side of a home, it is a danger to pedestrians using the sidewalk. However, if the driveway is located behind the home and accessed through a roadway so that it does not traverse a sidewalk, it presents less danger to pedestrians using the sidewalk. Yet, no audit instrument included in this review assessed the location of driveways or identified driveways as a sidewalk obstruction. As researchers continue to identify neighborhood physical environmental components that may influence quality of life, the number of domains and subdomains will continue to grow. 5. Conclusions The neighborhood environment clearly influences quality of life. To study these influences, components of the environment must be measured. This review provided an inventory of existing neighborhood physical environment audit instruments, and the domains and subdomains they measure. This review will help neighborhood auditors identify those instruments that measure domains most relevant to their study. This information may also be used to develop customized audit tools that capture those physical environmental characteristics of neighborhoods that auditors are most interested in. Appendix A. Supplementary data Supplementary data related to this article can be found at http:// dx.doi.org/10.1016/j.jenvp.2013.07.009.

References Andresen, E. M., Malstrom, T. K., Wolinsky, F. D., Schootman, M., Miller, J. P., & Miller, D. K. (2008). Rating neighborhoods for older adult health: Results from the African American health study. BMC Public Health, 8(35). http://dx.doi.org/ 10.1186/1471-2458-8-35. Boarnet, M. G., Day, K., Alfonzo, M., Forsyth, A., & Oakes, M. (2006). The IrvineMinnesota inventory to measure built environments: Reliability tests. American Journal of Preventive Medicine, 30(2), 153e159. http://dx.doi.org/10.1016/ j.amepre.2005.09.018. Brownson, R. C., Hoehner, C. M., Brennan, L. K., Cook, R. A., Elliott, M. B., & McMullen, K. M. (2004). Reliability of 2 instruments for auditing the environment for physical activity. Journal of Physical Activity and Health, 1, 191e208. Caughy, M. O., O’Campo, P. J., & Patterson, J. (2001). A brief observational measure for urban neighborhoods. Health & Place, 7(3), 225e236. Clifton, K. J., Livi Smith, A. D., & Rodriguez, D. (2007). The development and testing of an audit for the pedestrian environment. Landscape & Urban Planning, 80(1e 2), 95e110. http://dx.doi.org/10.1016/j.landurbplan.2006.06.008. Cozens, P., & Love, T. (2009). Manipulating permeability as a process for controlling crime: Balancing security and sustainability in local contexts. Built Environment, 35(3), 346e365. Cunningham, G. O., Michael, Y. L., Farquhar, S. A., & Lapidus, J. (2005). Developing a reliable senior walking environmental assessment tool. American Journal of Preventive Medicine, 29(3), 215e217. http://dx.doi.org/10.1016/ j.amepre.2005.05.002. Dannenberg, A. L., Jackson, R. L., Frumkin, H., Schieber, R. A., Pratt, M., Kochtitzky, C., et al. (2003). The impact of community design and land-use choices on public health: A scientific research agenda. American Journal of Public Health, 93, 1500e1508. Day, K. (n.d.). Audit tools for research on built environment features tied to active living. Retrieved 08.08.12 from http://www.activelivingresearch.org/files/ AuditToolsComparisonTable.pdf. Evenson, K. R., Sotres-Alvarez, D., Herring, A. H., Messer, L., Laraia, B. A., & Rodriguez, D. A. (2009). Assessing urban and rural neighborhood characteristics using audit and GIS data: Derivation and reliability of constructs. International Journal of Behavioral Nutrition and Physical Activity, 6(44). http://dx.doi.org/ 10.1186/1479-5868-6-44. Ewing, R., Handy, S., Brownson, R. C., Clemente, O., & Winston, E. (2006). Identifying and measuring urban design qualities related to walkability. Journal of Physical Activity and Health, 3(S1), S223eS240. Foltete, J. C., & Piombini, A. (2007). Urban layout, landscape features, and pedestrian usage. Landscape and Urban Planning, 81, 225e234. http://dx.doi.org/10.1016/ j.landurbplan.2006.12.001. Franzini, L., Caughy, M. O., Nettles, S. M., & O’Campo, P. (2008). Perceptions of disorder: Contributions of neighborhood characteristics to subjective perceptions of disorder. Journal of Environmental Psychology, 28(1), 83e93. http:// dx.doi.org/10.1016/j.jenvp.2007.08.003. Franzini, L., Elliott, M. N., Cuccaro, P., Schuster, M., Gilliland, J., Grunbaum, J. A., et al. (2009). Influences of physical and social neighborhood environments on children’s physical activity and obesity. American Journal of Public Health, 99(2), 271e278. Furr-Holden, C. D. M., Campbell, K. D. M., Milam, A. J., Smart, M. J., Ialongo, N. A., & Leaf, P. J. (2010). Metric properties of the neighborhood inventory for environmental typology (NIfETy): An environmental assessment tool for measuring indicators of violence, alcohol, tobacco, and other drug exposures. Evaluation Review, 34(3), 159e184. http://dx.doi.org/10.1177/0193841X10368493. Grbich, C. (1999). Qualitative research in health. London: Sage Publications. Hoehner, C. M., Ivy, A., Brennan Ramirez, L., Handy, S., & Brownson, R. C. (2007). Active neighborhood checklist: A user-friendly and reliable tool for assessing activity friendliness. American Journal of Health Promotion, 21, 534e537. Goldberg, J., Rudd, R. E., & Dietz, W. (1999). Using 3 data sources and methods to shape a nutrition campaign. Journal of the American Dietetic Association, 99(6), 717e722. Hooker, S. P., Cirill, L., & Wicks, L. (2007). Walkable neighborhoods for seniors: The Alameda County experience. Journal of Applied Gerontology, 26(2), 157e181. http://dx.doi.org/10.1177/0733464807299997. Kamphuis, C. B. M., Giskes, K., Kavanagh, A. M., Thornton, L. E., Thomas, L. R., van Lenthe, F. J., et al. (2008). Area variation in recreational cycling in Melbourne: A compositional or contextual effect? Journal of Epidemiology and Community Health, 62, 890e898. http://dx.doi.org/10.1136/jech.2007.067116. King, D. (2008). Neighborhood and individual factors in activity in older adults: Results from the neighborhood and senior health study. Journal of Aging and Physical Activity, 16(2), 144e170. Kohen, D. E., Brooks-Gunn, J., Leventhal, T., & Hertzman, C. (2002). Neighborhood income and physical and social disorder in Canada: Associations with young children’s competencies. Child Development, 73(6), 1844e1860. Konteh, F. H. (2009). Urban sanitation and health in the developing world: Reminiscing the nineteenth century industrial nations. Health & Place, 15, 69e78. Kuo, F. E., & Sullivan, W. C. (2001). Environment and crime in the inner city: Does vegetation reduce crime? Environment and Behavior, 33(3), 343e367. Landis, J. R., & Koch, G. G. (1977). The measurement of observer agreement for categorical data. Biometrics, 33, 159e174. Laraia, B. A., Messer, L., Kaufman, J. S., Dole, N., Caughy, M., O’Campo, P., et al. (2006). Direct observation of neighborhood attributes in an urban area of the U.S. south: Characterizing the social context of pregnancy. International Journal of Health Geographics, 5(11). http://dx.doi.org/10.1186/1476-072X-5-11.

J. Nickelson et al. / Journal of Environmental Psychology 36 (2013) 179e189 Leung, C. W., Gregorich, S. E., Laraia, B. A., Kushi, L. H., & Yen, I. H. (2010). Measuring the neighborhood environment: Associations with young girls’ energy intake and expenditure in a cross-sectional study. International Journal of Behavioral Nutrition and Physical Activity, 7(52). http://dx.doi.org/10.1186/1479-5868-7-52. Limbos, M. A. P., & Casteel, C. (2008). Schools and neighborhoods: Organizational and environmental factors associated with crime in secondary schools. Journal of School Health, 78(10), 539e544. http://dx.doi.org/10.1111/j.1746-1561.2008.00341.x. Loukaitou-Sideris, A., Liggett, R., Iseki, H., & Thurlow, W. (2001). Measuring the effects of built environment on bus stop crime. Environment and Planning B: Planning and Design, 28, 255e280. http://dx.doi.org/10.1068/b2642r. McMillan, T. E. (2007). The relative influence of urban form on a child’s travel mode to school. Transportation Research Part A, 41, 69e79. http://dx.doi.org/10.1016/ j.tra.2006.05.011. Michael, Y. L., Keast, E. M., Chaudhury, H., Day, K., Mahmood, A., & Sarte, A. F. I. (2009). Revising the senior walking environmental assessment tool. Preventive Medicine, 48, 247e249. http://dx.doi.org/10.1016/j.ypmed.2008.12.008. Minnery, J. R., & Lim, B. (2005). Measuring crime prevention through environmental design. Journal of Architectural and Planning Research, 22(4), 330e341. Moudon, A. V., & Lee, C. (2003). Walking and bicycling: An evaluation of environmental audit instruments. American Journal of Health Promotion, 18(1), 21e37. Parsons, J. A., Singh, G., Scott, A. N., Nisenbaum, R., Balasubramaniam, P., Jabbar, A., et al. (2010). Standardized observation of neighbourhood disorder: Does it work in Canada? International Journal of Health Geographics, 9(6). Perkins, D. D., Meeks, J. W., & Taylor, R. B. (1992). The physical environment of street blocks and resident perceptions of crime and disorder: Implications for theory and measurement. Journal of Environmental Psychology, 12, 21e34. Pikora, T. J., Bull, F. C., Jamrozik, K., Knuiman, M., Giles-Corti, B., & Donovan, R. J. (2002). Developing a reliable audit instrument to measure the physical environment for physical activity. American Journal of Preventive Medicine, 23(3), 187e194. Raudenbush, S. W., & Sampson, R. J. (1999). Systematic social observation of public spaces: A new look at disorder in urban neighborhoods. American Journal of Sociology, 10, 603e651. Saelens, B. E., Frank, L. D., Auffrey, C., Whitaker, R. C., Burdette, H. L., & Colabianchi, N. (2006). Measuring physical environments of parks and playgrounds: EAPRS instrument development and inter-rater reliability. Journal of Physical Activity and Health, 3(Suppl. 1), S190eS207.

189

Sastry, N., Pebley, A. R., & Zonta, M. (2002). Neighborhood definitions and the spatial dimension of daily life in Los Angeles. Paper presented at the 2002 annual meetings of the Population Association of America, Atlanta, Georgia. Schaefer-McDaniel, N. (2009). Neighborhood stressors, perceived neighborhood quality, and child mental health in New York City. Health & Place, 15, 148e155. http://dx.doi.org/10.1016/j.healthplace.2008.03.007. Schaefer-McDaniel, N., O’Brien, M., O’Campo, P., & Gearey, W. (2010). Examining methodological details of neighbourhood observations and the relationship to health: A literature review. Social Science & Medicine, 70, 277e292. Secretary’s Advisory Committee on National Health Promotion and Disease Prevention Objectives for 2020. (2010). Healthy people 2020: An opportunity to address societal determinants of health in the U.S.. Retrieved from http:// healthypeople.gov/2020/about/advisory/SocietalDeterminantsHealth.pdf. Seymour, M., Reynolds, K. D., & Wolch, J. (2010). Reliability of an audit tool for systematic assessment of urban alleyways. Journal of Physical Activity and Health, 7, 214e223. Suminski, R. R., Heinrich, K. M., Poston, W. S. C., Hyder, M., & Pyle, S. (2007). Characteristics of urban sidewalks/streets and objectively measured physical activity. Journal of Urban Health: Bulletin of the New York Academy of Medicine, 85(2), 178e190. http://dx.doi.org/10.1007/s11524-007-9251-x. United Nations. (2012). Composition of macro geographical (continental) regions, georgraphical sub-regions, and selected economic and other groupings. Retrieved from http://unstats.un.org/unsd/methods/m49/m49regin. htm#developed. World Health Organization. (1997). WHOQOL: Measuring quality of life. Retrieved from http://www.who.int/mental_health/media/68.pdf. Wright, P. A., & Kloos, B. (2007). Housing environment and mental health outcomes: A levels of analysis perspective. Journal of Environmental Psychology, 27(1), 79e 89. http://dx.doi.org/10.1016/j.jenvp.2006.12.001. Zenk, S. N., Schulz, A. J., Mentz, G., House, J. S., Gravlee, C. C., Miranda, P. Y., et al. (2007). Inter-rater and testeretest reliability: Methods and results for the neighborhood observational checklist. Health & Place, 13(2), 452e465. http:// dx.doi.org/10.1016/j.healthplace.2006.05.003. Zhu, X., & Lee, C. (2008). Walkability and safety around elementary schools. American Journal of Preventive Medicine, 34(4), 282e290. http://dx.doi.org/ 10.1016/j.amepre.2008.01.024.