Skip to main content
. 2015 Dec 17;2:6–11. doi: 10.1016/j.conctc.2015.12.001

Table 3.

Univariate and multivariate analyses of associations between “Knowledge” and demographic factors.


Univariate analysis
Multivariate analysisa
Factor N [%] Mean (SD) Mean difference (95% confidence interval) P Mean difference (95% confidence interval) P
Age
<45 18 [18] 75 (14.7) Reference Reference
45–64 68 [67] 72 (14.6) −3 (−10 to 5) 0.67 −2 (−10 to 6) 0.86
≥65 15 [15] 70 (16.0) −5 (−15 to 5) 0.38# −3 (−14 to 8) 0.60#
Gender
Male 17 [17] 72 (13.7) Reference Reference
Female 84 [83] 73 (15.0) 1 (−7 to 8) 0.88 −1 (−9 to 8) 0.89
Education
Elementary school 18 [18] 69 (14.5) Reference Reference
High school 20 [20] 76 (17.4) 7 (−3 to 16) 0.37 6 (−4 to 16) 0.50
University 62 [61] 73 (13.5) 4 (−4 to 11) 0.37# 3 (−5 to 12) 0.46#
Missing 1 [1]
Living alone or not
Cohabitant 70 [69] 74 (12.7) Reference Reference
Alone 30 [30] 70 (17.9) −4 (−11 to 2) 0.17 −4 (−11 to 2) 0.19
Missing 1 [1]
Randomized study or not
Not randomized 22 [22] 77 (16.8) Reference Reference
Randomized 79 [78] 71 (14.0) −6 (−13 to 2) 0.12 −6 (−13 to 1.3) 0.11
Total number of patients in the study with information on knowledge 101 72 (14.7)

Part A. Phase specific questions excluded. Total number of questions included 14 (out of 20). For each question, correct answers are assigned a score of 100, unsure and incorrect answers are assigned a score of 0.# Test for linear trend. a Adjusted for all other listed factors. a Adjusted for all listed factors.

Linear regression was used to estimate mean differences and confidence intervals. All reported p-values are two-sided and based on the Wald test.