Table 6.

Summary of factors associated with self-protective behaviors

Variable	Metric	CB	SD	WB	Variable	Metric	CB	SD	WB
Income	Significant	✓	✓	✓	Countryside	Significant	✗	✗	✗
	Sign	+	+	+		Sign	NA	NA	NA
Non-White	Significant	✗	✗	✓	No access to open air	Significant	✗	✓	✗
	Sign	NA	NA	+		Sign	NA	–	NA
Male	Significant	✗	✓	✗	Not working	Significant	✗	✗	✓
	Sign	NA	–	NA		Sign	NA	NA	+
Florida	Significant	✗	✓	✗	Tele-working	Significant	✓	✓	✓
	Sign	NA	–	NA		Sign	+	+	+
New York	Significant	✗	✗	✗	Not working pre-Covid	Significant	✗	✓	✗
	Sign	NA	NA	NA		Sign	NA	+	NA
Texas	Significant	✗	✓	✗	Lost income	Significant	✓	✓	✓
	Sign	NA	–	NA		Sign	+	+	+
56 or older	Significant	✗	✗	✓	Effectiveness of SD	Significant	✓	✓	✓
	Sign	NA	NA	+		Sign	+	+	+
Pre-existing condition	Significant	✓	✓	✓	Area infection rate	Significant	✗	✓	✓
	Sign	+/–	–	+		Sign	NA	–	–
Semi-urban/residential	Significant	✗	✗	✓	Benefits	Significant	✓	✓	✓
	Sign	NA	NA	–		Sign	+	+	+

This table summarizes the findings from linear probability models examining the association between socio-demographic characteristics and three different self-protective behaviors. Detailed tables with coefficient estimates, standard errors, significance levels and other controls can be found in the Appendix. “CB” stands for changed behaviors, “SD” stands for increased social distancing behaviors, and “WM” stands for increased hand washing or mask wearing