Skip to main content
. Author manuscript; available in PMC: 2016 Jan 1.
Published in final edited form as: Appl Econ. 2015 Jan;47(5):431–448. doi: 10.1080/00036846.2014.972546

Table 4.

Loglog and Complementary Loglog Models under Random Misclassification

Variable Loglog Cloglog
Outcome: Tobacco user
Constant -2.139 (0.160) *** -1.006 (0.111) ***
Gender
Male 0.540 (0.081) *** 0.402 (0.049) ***
Age (Excluded: 25-50 years)
20-25 years 0.083 (0.157) 0.100 (0.086)
50-65 years -0.346 (0.092) *** -0.302 (0.069) ***
Over 65 years -1.153 (0.140) *** -0.881 (0.100) ***
Race/Ethnicity (Excluded: Non Hispanic White)
Mexican American -0.496 (0.110) *** -0.391 (0.073) ***
Other Hispanic -0.257 (0.114) ** -0.321 (0.076) **
Non Hispanic Black -0.024 (0.093) -0.065 (0.059)
Mixed or other race/ethnicity 0.281 (0.157) * 0.237 (0.107) **
Education (Excluded: Less than 9th grade)
College graduate or above -1.078 (0.138) *** -0.716 (0.111) ***
Some college or AA degree -0.403 (0.100) *** -0.290 (0.111) ***
High school graduate/ GED or equivalent -0.062 (0.094) -0.053 (0.111)
Marital status (Excluded: Married)
Widowed 0.548 (0.162) *** 0.416 (0.113) ***
Divorced 0.892 (0.103) *** 0.574 (0.079) ***
Separated 0.903 (0.189) *** 0.624 (0.126) ***
Never married 0.463 (0.103) *** 0.257 (0.069) ***
Living with a partner 0.725 (0.148) *** 0.517 (0.093) ***
Body structure (Excluded: Normal)
Underweight 0.703 (0.299) ** 0.603 (0.181) **
Overweight -0.156 (0.085) * -0.089 (0.059)
Obese -0.478 (0.096) *** -0.248 (0.061) ***
Other
Early initiator 0.808 (0.101) *** 0.597 (0.068) ***
Consume Alcohol 0.775 (0.115) *** 0.516 (0.078) ***
Number of (additional) smokers at home 1.807 (0.215) *** 1.553 (0.128) ***
Pregnant -0.308 (0.446) -0.018 (0.462)

Conditional probability of identifying a tobacco user as a nonuser
Self-reported data 0.109 (0.027) *** 0.048 (0.022) **
Biochemical measure 0.105 (0.030) *** 0.042 (0.022) *

Conditional probability of identifying anonuser as a user
Self-reported data 0.003 (0.008) 0.030 (0.009) ***
Biochemical measure 0.001 (0.008) 0.029 (0.009) ***

Number of observations 5,051 5,051
Log likelihood -3955.19 -3930.43
***

p<0.01;

**

p<0.05;

*

p<0.10