Built environment and transit use empirical research database

This data set comprises descriptive and quantitative information extracted from prior empirical studies exploring the built environment and transit use. The data contained in this database was compiled for the purpose of conducting meta-analysis of the relationship between the built environment and public transport (transit) use. The dataset includes coded data for 1,662 data points, deriving from 146 studies.

Data extracted for each of the priors includes study invariant characteristics as well as observational data. The identification, eligibility, extraction and coding framework used to compile the database follows the PRISMA-P framework (The PRISMA-P Group 2015). The review protocol is available on figshare (1).
The scope of the coded data includes:
- Data point identification
- Data point typology
- Dependent variable
- Sample attributes
- Study design
- Independent built environment variables
- Independent (external) variables
- Model specification
- Parameter estimates
- Measures of fit

Database contents:

1. BE-TR_Database: Complete database for 146 studies, including descriptive and quantitative information for reported and omitted variables.
2. Metadata: Index of column headers describing attributes and corresponding levels.
3. Metadata_BE: Description of coding of variables into analysis subgroups
4. Metadata_notes: Assumptions and formulae pertaining to derivations and imputations of missing data
5. Priors_Reference_List: Citation information for records included in this meta-analysis

Version 5 of the built environment and transit use meta-analysis database contains one substantive change from prior versions:
5.1 Updated 'correlation types' for derivation of common effect sizes. Correlation types are outined in '4. Metadata_Notes

Version 3&4 of the built environment and transit use meta-analysis database contains the following substantive and structural updates from prior versions:
3.1 Introduced 30 new studies
3.2 Introduction of dichotomous variable to identify control for 'regional accessibility' and 'self-selection' (coding section: "study design")
3.3 Updated estimation formulae for logistic regression (refer to 1.5 Metadata_notes equations 7a)
3.4 Distinguished 'pseudo coefficients' (logistic methods) from regular coefficients of determination
3.5 Consolidated database and meta-data into single spreadsheet
3.6 Consolidated 'Data_All', 'Data_Eligible' (meta-analysis ), 'into single spreadsheet 'BE-TR_Database'


Notes
1 - Aston, Laura; Currie, Graham; Delbosc, Alexa; O'Hare, Tyler; Kamruzzaman, MD; Teller, David (2019): Empirical built environment and transit use literature since 2000: Comprehensive literature review protocol and results. figshare. Dataset. Available on figshare at https://doi.org/10.26180/5c3fd9990600c