Notice: On April 23, 2014, Statalist moved from an email list to a forum, based at statalist.org.
[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
st: binary mediation command
From
Pina Valle <[email protected]>
To
[email protected]
Subject
st: binary mediation command
Date
Wed, 15 Jun 2011 13:18:44 -0400
I am trying to test mediation with a dichotomous outcome, and I have looked around and found a command in STATA called binary_mediation. However, there isn't really any indication in the notes I found on whether the mediation is significant. Here is an example of my output along with the commands that I have used:
. quietly bootstrap r(indir_1) r(tot_ind) r(dir_eff) r(tot_eff), ///
> reps(500): binary_mediation, dv (evercoh) iv (hhstruct) mv (adrel) ///
> cv (race income parenteduc agew1 respeduc respinc)
.
. estat bootstrap, percentile bc
Bootstrap results Number of obs = 7314
Replications = 500
command: binary_mediation, dv(evercoh) iv(hhstruct) mv(adrel) cv(race income parenteduc agew1 respeduc respinc)
_bs_1: r(indir_1)
_bs_2: r(tot_ind)
_bs_3: r(dir_eff)
_bs_4: r(tot_eff)
------------------------------------------------------------------------------
| Observed Bootstrap
| Coef. Bias Std. Err. [95% Conf. Interval]
-------------+----------------------------------------------------------------
_bs_1 | .00601173 -.0000315 .00163333 .0031361 .0097957 (P)
| .003367 .0100647 (BC)
_bs_2 | .00601173 -.0000315 .00163333 .0031361 .0097957 (P)
| .003367 .0100647 (BC)
_bs_3 | .10223182 -.000063 .01316529 .0764666 .128446 (P)
| .0778659 .1297743 (BC)
_bs_4 | .10824356 -.0000945 .01314362 .0834475 .1347804 (P)
| .0835382 .1354965 (BC)
------------------------------------------------------------------------------
(P) percentile confidence interval
(BC) bias-corrected confidence interval
.
. binary_mediation, dv(evercoh) mv(adrel) iv(hhstruct) ///
> cv(race income parenteduc agew1 respeduc respinc)
Logit: adrel on iv (a1 path)
Logistic regression Number of obs = 7314
LR chi2(7) = 409.52
Prob > chi2 = 0.0000
Log likelihood = -4414.5505 Pseudo R2 = 0.0443
------------------------------------------------------------------------------
adrel | Coef. Std. Err. z P>|z| [95% Conf. Interval]
-------------+----------------------------------------------------------------
hhstruct | .1206886 .0270068 4.47 0.000 .0677563 .1736209
race | -.169737 .0276822 -6.13 0.000 -.2239932 -.1154808
income | .0525551 .0238988 2.20 0.028 .0057143 .0993959
parenteduc | -.0627959 .0256909 -2.44 0.015 -.1131491 -.0124426
agew1 | .3029958 .017249 17.57 0.000 .2691884 .3368033
respeduc | -.0576789 .019877 -2.90 0.004 -.0966371 -.0187208
respinc | .0497683 .0238852 2.08 0.037 .0029541 .0965825
_cons | -3.79803 .2948662 -12.88 0.000 -4.375958 -3.220103
------------------------------------------------------------------------------
Logit: dv on iv (c path)
Logistic regression Number of obs = 7314
LR chi2(7) = 418.40
Prob > chi2 = 0.0000
Log likelihood = -4854.404 Pseudo R2 = 0.0413
------------------------------------------------------------------------------
evercoh | Coef. Std. Err. z P>|z| [95% Conf. Interval]
-------------+----------------------------------------------------------------
hhstruct | .1888678 .0246015 7.68 0.000 .1406498 .2370858
race | -.0719781 .0265371 -2.71 0.007 -.1239898 -.0199664
income | .0100539 .0221388 0.45 0.650 -.0333374 .0534452
parenteduc | -.0148689 .0240379 -0.62 0.536 -.0619824 .0322446
agew1 | -.0632524 .0157713 -4.01 0.000 -.0941636 -.0323411
respeduc | -.251559 .0187565 -13.41 0.000 -.2883211 -.2147968
respinc | -.1222125 .0223661 -5.46 0.000 -.1660492 -.0783758
_cons | 2.046847 .2767139 7.40 0.000 1.504498 2.589197
------------------------------------------------------------------------------
Logit: dv on mv & iv (b & c' paths)
Logistic regression Number of obs = 7314
LR chi2(8) = 460.19
Prob > chi2 = 0.0000
Log likelihood = -4833.5111 Pseudo R2 = 0.0454
------------------------------------------------------------------------------
evercoh | Coef. Std. Err. z P>|z| [95% Conf. Interval]
-------------+----------------------------------------------------------------
adrel | .3428757 .0532173 6.44 0.000 .2385717 .4471796
hhstruct | .1814979 .024678 7.35 0.000 .1331299 .2298658
race | -.0599992 .026681 -2.25 0.025 -.112293 -.0077055
income | .0064338 .0221953 0.29 0.772 -.0370683 .0499358
parenteduc | -.0100817 .0241264 -0.42 0.676 -.0573685 .0372051
agew1 | -.0853616 .0161993 -5.27 0.000 -.1171116 -.0536116
respeduc | -.2490846 .0188213 -13.23 0.000 -.2859737 -.2121954
respinc | -.1265046 .0224421 -5.64 0.000 -.1704904 -.0825189
_cons | 2.154015 .2779551 7.75 0.000 1.609233 2.698797
------------------------------------------------------------------------------
Indirect effects with binary response variable evercoh
indir_1 = .00601173 (adrel, binary)
total indirect = .00601173
direct effect = .10223182
total effect = .10824356
c_path = .10680445
proportion of total effect mediated = .05553895
ratio of indirect to direct effect = .05880491
Binary models use logit regression
It seems as if about 6% (0.055) of effect of family structure on engagement in a cohabiting union is mediated by engagement in an adolescent relationship. And I would assume it was significant as a result of the first part of the output above that has the indirect effect (bs_1), as the confidencen interval includes 0. I just wanted to see if I was on the right track with this. Any help would be greatly appreciated.
Thanks.
Pina
*
* For searches and help try:
* http://www.stata.com/help.cgi?search
* http://www.stata.com/support/statalist/faq
* http://www.ats.ucla.edu/stat/stata/