Hello,

I am running a moderation analysis looking to see if Race (4 categories) moderates the relationship between one continuous and one categorical variable. Here is my code for the regression analysis with moderation and the output. The relationship between cohesion and loneliness is significantly different (stronger) for Puerto Ricans as compared to Whites (my comparison group) (p= 0.016). I ran an Adjusted Wald Test with testparm. The categories are equal to zero but the test is not significant. Does this mean that the addition of the moderation does not add to the model or is it ok? Meaning, are my results meaningful (statistically). Theoretically this makes sense to me, but I am uncertain if my data supports this. Any thoughts on the CI? Is this too large? It's survey data at the community level so the CIs are big in many of the analysis I perform on this dataset.

Thank you for any insight you may have.

svy, subpop(finsamp): regress loneliness_score i.agecat i.GENDER i.immigrant i.edu i.income_cv2 i.relationship i.employment_cat i.tot_org lived_hood_c i.own_d i.screendep c.cohesion_score##i.RE5
eststo model8

testparm c.cohesion_score#i.RE5

Here are the results:

Number of strata = 10 Number of obs = 1,427
Number of PSUs = 151 Population size = 1,430.1868
Subpop. no. obs = 1,224
Subpop. size = 1,220.4961
Design df = 141
F( 26, 116) = 5.68
Prob > F = 0.0000
R-squared = 0.2286

--------------------------------------------------------------------------------------
| Linearized
loneliness_score | Coef. Std. Err. t P>|t| [95% Conf. Interval]
---------------------+----------------------------------------------------------------
agecat |
2 | .0106775 .0792043 0.13 0.893 -.145904 .167259
3 | .0464823 .0806891 0.58 0.565 -.1130344 .2059991
4 | -.0616393 .0803887 -0.77 0.445 -.2205623 .0972836
|
GENDER |
Female | -.0274727 .0294611 -0.93 0.353 -.0857152 .0307698
1.immigrant | -.0385613 .0632635 -0.61 0.543 -.1636289 .0865062
|
edu |
HS grad | -.0153531 .0583026 -0.26 0.793 -.1306133 .0999071
some college | -.0230631 .0615131 -0.37 0.708 -.1446703 .0985441
college grad | -.0483493 .0619744 -0.78 0.437 -.1708683 .0741698
|
income_cv2 |
2 | -.0896955 .0578998 -1.55 0.124 -.2041594 .0247683
3 | -.0938808 .0625916 -1.50 0.136 -.21762 .0298584
4 | -.188409 .0730158 -2.58 0.011 -.3327562 -.0440617
|
1.relationship | -.103569 .0392778 -2.64 0.009 -.1812185 -.0259194
|
employment_cat |
1 | .0597762 .0760285 0.79 0.433 -.090527 .2100793
2 | .0214649 .0399409 0.54 0.592 -.0574954 .1004253
|
tot_org |
1 | -.0785157 .0505309 -1.55 0.122 -.1784118 .0213804
2 | -.0204454 .0523854 -0.39 0.697 -.1240077 .0831169
|
lived_hood_c | -.0012079 .0014861 -0.81 0.418 -.0041458 .00173
|
own_d |
Own | .004368 .0420274 0.10 0.917 -.0787173 .0874533
1.screendep | .5169033 .0886463 5.83 0.000 .3416556 .6921509
cohesion_score | .0217501 .0728157 0.30 0.766 -.1222016 .1657018
|
RE5 |
NHB | .3508753 .2625977 1.34 0.184 -.1682624 .870013
Mexican | .4093082 .3232419 1.27 0.208 -.2297188 1.048335
Puerto_Rican | 1.238435 .5203541 2.38 0.019 .2097308 2.26714
|
RE5#c.cohesion_score |
NHB | -.1250218 .0867501 -1.44 0.152 -.2965207 .0464771
Mexican | -.1405104 .1060462 -1.32 0.187 -.3501566 .0691357
Puerto_Rican | -.4452275 .1831064 -2.43 0.016 -.8072162 -.0832387
|
_cons | 1.443502 .2573488 5.61 0.000 .9347407 1.952262
--------------------------------------------------------------------------------------

. eststo model8

.
. testparm c.cohesion_score#i.RE5

Adjusted Wald test

( 1) 1.RE5#c.cohesion_score = 0
( 2) 2.RE5#c.cohesion_score = 0
( 3) 3.RE5#c.cohesion_score = 0

F( 3, 139) = 2.14
Prob > F = 0.0974

.
end of do-file