How are Total Effects on Target and Direct Effects on Target calculated in BayesiaLab?

Answer

Both the Total and Direct Effects are the derivative of their corresponding Effect Function computed at the a-priori mean value (delta = 0).

The Total Effect Function is estimated by using the Mean Value Analysis (based in the MinXEnt) to go through the variation domain of a variable in order to measure its impact on the Target mean.

Adding 1 to the mean value of Corresponds (green curve) increases the target node Pleasure_(4) by 0.78

Total Effects are the derivatives of the Total Effect functions, taken at the a-priori mean values of the variables,

the Standardized value normalizing the effect by taking into account the ration between the standard deviation of the variable and the one of the Target.

Node

Value/Mean

Standardized Total Effects

Total Effects

Pleasure

6.05

0.96

0.8

Corresponds

5.76

0.95

0.79

Easy to wear

6.5

0.83

0.79

Intensity

3.07

-0.19

-0.55

The Direct Effect Function is estimated by using the Mean Value Analysis (based in the MinXEnt) to go through the variation domain of a variable in order to measure its impact on the Target mean, while holding fixed the probability distributions of all the variables, except:

The Not-Observable variables belonging to the Class “Factor”

The variables belonging to the Class “Non_Confounder” .

Adding 1 to the mean value of Corresponds (green curve) while holding fixed all the marginal probability distributions of all the other variables increases the target node Pleasure_(4) by 0.27

Direct Effects are the derivatives of the Direct Effect functions, taken at the a-priori mean values of the variables,

the Standardized value normalizing the effect by taking into account the ration between the standard deviation of the variable and the one of the Target.