Skip to main content
. Author manuscript; available in PMC: 2014 Jun 1.
Published in final edited form as: Ann Epidemiol. 2013 Jun;23(6):355–360. doi: 10.1016/j.annepidem.2013.03.014

Table 3.

Mediation Analysis With The Analytic Sample Size, The Baseline Hazards Of Television Hours On

Putative Mediator
(Olkin & Finn P Value)
N
Less trusting 3136
People are not fair 2826
Feel afraid in neighborhood*** 8002
People are not helpful 2843
Less general happiness** 9309
Less satisfied with hobbies** 2151
Less satisfied with friends* 2156
Unfair to bring a child into world*** 4954
Cannot control bad things in life** 683
Should not plan much in life*** 683
Less confidence in military*** 3705
Less confidence in organized religion*** 3688
Less confidence in the federal government 3699
Less confidence in the scientific community*** 3561
Government should not pay for medical care 1940
Government should do less 1907
Government should not help the poor 1938
***

P <.001,

**

P <.01,

*

P <.05

a

This is the analytic sample of respondents who report good or excellent self- rated health only

b

Television viewing hazard ratio from the baseline regression model (television hours as the independent variable and mortality hazards as the dependent variable) but no mediator added

c

Television viewing coefficient in the baseline regression model (b above) with the mediator variable added

d

Coefficient of the mediator within model c above