. 2020 Aug 4;76(4):1161–1185. doi: 10.1007/s10640-020-00489-y

Table 1.

Parameter estimates from a four-class latent class regression model identifying parameters for the ORVal recreation demand model identifying patterns of demand-shift observed under lockdown

Parameter	Latent class
Class membership
$τ_{h}$	0	0.559 (0.375)	− 0.086 (0.403)	0.111 (0.386)
Probability $(π_{h})$	0.209	0.365	0.192	0.234
Demand-shift params $({\tilde{β}}_{Td}^{h})$
Strict rules
Weekday	− 0.310*** (0.028)	− 0.154*** (0.017)	− 0.127*** (0.031)	0.060** (0.030)
Weekend	− 0.001 (0.023)	0.204*** (0.020)	0.082*** (0.025)	0.474*** (0.025)
Relaxed rules
Weekday	− 0.646*** (0.027)	− 0.369*** (0.015)	− 0.325*** (0.028)	− 0.265*** (0.026)
Weekend	− 0.559*** (0.024)	− 0.141*** (0.012)	− 0.154*** (0.021)	− 0.048*** (0.022)
$σ_{k}$	0.511*** (0.054)	0.285*** (0.024)	1.402*** (0.142)	0.356*** (0.042)

Parameter

Latent class

Class 1

Class 2

Class 3

Class 4

Class membership

τ_{h}

0.559

(0.375)

− 0.086

(0.403)

0.111

(0.386)

Probability

(π_{h})

0.209

0.365

0.192

0.234

Demand-shift params

({\tilde{β}}_{Td}^{h})

Strict rules

Weekday

− 0.310***

(0.028)

− 0.154***

(0.017)

− 0.127***

(0.031)

0.060**

(0.030)

Weekend

− 0.001

(0.023)

0.204***

(0.020)

0.082***

(0.025)

0.474***

(0.025)

Relaxed rules

Weekday

− 0.646***

(0.027)

− 0.369***

(0.015)

− 0.325***

(0.028)

− 0.265***

(0.026)

Weekend

− 0.559***

(0.024)

− 0.141***

(0.012)

− 0.154***

(0.021)

− 0.048***

(0.022)

σ_{k}

0.511***

(0.054)

0.285***

(0.024)

1.402***

(0.142)

0.356***

(0.042)

Table reports the coefficient estimate with the standard error below in brackets. Coefficients significant at the 90% level are highlighted with *, those at the 95% level with ** and those at 99% at ***