Skip to main content
. 2021 Sep 22;65:127354. doi: 10.1016/j.ufug.2021.127354

Fig. 1.

Fig. 1

Percent of respondents who selected each concern about visiting urban green space (UGS) (A) and important feature of UGS (G), and coefficients of binomial regressions predicting differences in these responses across gender, race/ethnicity, and NYC borough (B-F and H-L). Plots show the odds ratios of the given variables compared to their associated baseline group (e.g. Male : Female is comparing the relative likelihood of selecting the response for people who identified as male, compared to the baseline group Female). If a variable's coefficient is not sufficiently different from the baseline group, the odds ratio takes a value of one (shown as a vertical black line for reference). Statistically significant differences (those where a variable's 95 % confidence bounds do not overlap one) are shown in darker shades. Upper limits of confidence intervals are capped at 8 for ease of viewing; confidence intervals that reach 8 have higher upper limits (see Appendix D Table 1). Model responses are shown for the five most frequently selected concerns (B-F) and features (H-L). Results for other responses can be found in Appendix D Table 1.