Regressions with Dummy

Variables and Path Analysis

• Dummy (dichotomous) variables: Variables with
only two categories (yes vs no, tall vs low, etc)
• Dummy variables can be arranged into 2 or
more categories with (k-1) rules; k = categories.
• Y = a + b1X1 + b2X2 + DRace + e
or
• Y = a + b1X1 + b2X2 + DJateng + DJabar +
DJatim
 Rules
• Dummy Variables:
– 1 : included group
– 0: excluded group
• It then compares the included group (1)
with the excluded group (0) on the basis of
“degree” form (% of (1) is higher (lower)
compared to (0)
•
 Ex (1)
• IPK = a + b1SosEk + b2WT + D.Math
• D.Math= 1 Math: A; 0 otherwise
• IPK = 0.21 + 0.1 SosEk + 0.2WT +
0.3D.Math
• the GPA of students with A Math Score is
30% higher compared to students who
had math score B and lower
 Ex (2)
• Income = a + b1X1 + b2X2 + DJateng +
DJabar + DJatim
• Income = 0.21 + 0.1 X1 + 0.2 X2 + 0.2
DJateng – 0.1 DJabar + 0.3DJatim
• Compared to Income of DKI, the income of
DJateng is higher by 20% and the income
of DJabar is 10% lower
 Intervening Variables in Path
Analysis
• Intervening (Mediating) Variables:
Variable(s) that mediate the relationship
between independent and dependent
variable
• E.g.:
– The impact of Studying hours to GPA is
indirectly affected by IQ
– The effect of budgetary participation to
managerial performance is mediated by
budgetary slack
 Path Analysis
Wealth
0.673 -0.135

Income Saving

0.474

• X2 = a + b1X1
• Y = a + b2X1 + b3 X2
• bn = standardized estimated parameter
 X2
0.673 -0.135

X1 Y

0.474

• Indirect Effects: b1 x b3 = 0.673*-0.135 = -

0.09
• Direct Effects: b2 = 0.474
• Total Effects: b2 + (b1 x b3) = 0.384
 X2

b1 b3
0.70 0.800 b5
0.20
X1 X4 Y
b2 b4
0.5 0.70
X3

b6
0.05
 Full and partial mediation,
consider:

Where C is total effect and c’ is called direct effect

 Full vs partial mediation
• Step 1: Show that the causal variable is correlated with the outcome. Use Y as the
criterion variable in a regression equation and X as a predictor (estimate and test path c in
the above figure). This step establishes that there is an effect that may be mediated.
• Step 2: Show that the causal variable is correlated with the mediator. Use M as the
criterion variable in the regression equation and X as a predictor (estimate and test
path a). This step essentially involves treating the mediator as if it were an outcome
variable.
• Step 3: Show that the mediator affects the outcome variable. Use Y as the criterion
variable in a regression equation and X and M as predictors (estimate and test path b). It is
not sufficient just to correlate the mediator with the outcome because the mediator and the
outcome may be correlated because they are both caused by the causal variable X. Thus,
the causal variable must be controlled in establishing the effect of the mediator on the
outcome.
• Step 4: To establish that M completely mediates the X-Y relationship, the effect of X on Y
controlling for M (path c') should be zero. The effects in both Steps 3 and 4 are estimated
in the same equation.

• The mediating effect can calculated using Z Sobel test:

z-value = a*b/SQRT(b2*sa2 + a2*sb2)
•