Skip to main content
. Author manuscript; available in PMC: 2017 Mar 1.
Published in final edited form as: Tob Control. 2015 Feb 19;25(2):181–187. doi: 10.1136/tobaccocontrol-2014-051979

Table 2.

GEE results testing the changes in smokers’ reactions and attentional orientation to health warning labels (HWLs) from pre (2011) to post (2013) implementation of new HWLs and plain packaging.

Outcome variables n/N b Weighted
prevalence
estimates a
Survey year main effects
(2013 vs 2011)

Pre
(2011)
Post
(2013)
B (SE) OR (95% CI)

Notice 1499/2159 3.23 3.40 .15 (.05)** --
Read 1501/2164 2.33 2.28 .00 (.04) --
Cognitive reactions 1502/2163 1.82 1.95 .11 (.02)*** --
Forgo 1500/2159 1.23 1.28 .01 (.02) --
Avoid (Yes vs No) 1504/2169 13.3 33.9 -- 3.06 (2.50-3.75)***
Attentional orientation
 (HWL first vs Branding first)
1504/2169 29.1 64.4 -- 4.19 (3.52-4.99)***

All estimates in the table adjusted for age, sex, income, education, cigarettes per day, past year quit attempts, survey mode (phone vs web) and wave of recruitment;

B, regression coefficients; SE, standard errors;

OR, odds ratios; CI, confidence interval;

--, Not applicable;

*

p<.05;

**

p<.01;

***

p<.001;

n, number of unique individuals; N, number of person-wave observations;

Score range: Notice =1-5, Read =1-5, Cognitive reactions=1-3.7, Forgo=1-4;

a

, all figures refer to mean scores except those in bold which refer to percentages

b

, variation in n is due to missing data on one or more independent variables included in the models;