madelyn.plating
New member
- Joined
- Apr 7, 2021
- Messages
- 1
Hello, I am working on a bachelors project, and I am needing help understanding how to interpret effect size. I am using STATA to run my tests, and I am using OLS regressions. I have a sample size of 30 (on the low end). None of my IV are significantly significant. My DV is an index score which might be my difficulty in interpreting my data. hpsmedian is my IV I am interested in and everything else is a control variable. I also ran a test for effect size, but that is also confusing to me given the differences in confidence intervals stata spits out. It could be a problem with my code. If anyone can lend me any help with this it would change my life.
Source | SS df MS Number of obs = 30
-------------+---------------------------------- F(5, 24) = 0.29
Model | .086027227 5 .017205445 Prob > F = 0.9117
Residual | 1.40577277 24 .058573866 R-squared = 0.0577
-------------+---------------------------------- Adj R-squared = -0.1387
Total | 1.4918 29 .051441379 Root MSE = .24202
------------------------------------------------------------------------------
changeinhdi | Coef. Std. Err. t P>|t| [95% Conf. Interval]
-------------+----------------------------------------------------------------
hpsmedian | .0878233 .1092055 0.80 0.429 -.1375657 .3132123
intensity | -.0646888 .0970185 -0.67 0.511 -.2649252 .1355476
promkept | -.0158435 .1044076 -0.15 0.881 -.2313302 .1996431
rebel | -.0088158 .0401453 -0.22 0.828 -.0916716 .07404
pko | .0435967 .0961426 0.45 0.654 -.1548319 .2420254
_cons | .3782594 .0958405 3.95 0.001 .1804543 .5760646
------------------------------------------------------------------------------
* esize twosample changeinhdi, by(hpsmedian) all
Effect size based on mean comparison
Obs per group:
hpsmedian==0 = 15
hpsmedian==1 = 16
---------------------------------------------------------
Effect Size | Estimate [95% Conf. Interval]
--------------------+------------------------------------
Cohen's d | -.2026539 -.9072415 .5053938
Hedges's g | -.1973596 -.8835398 .4921904
Glass's Delta 1 | -.2000718 -.9047937 .5116509
Glass's Delta 2 | -.2051563 -.9099594 .506353
Point-Biserial r | -.1041389 -.4244398 .2526572
---------------------------------------------------------
Source | SS df MS Number of obs = 30
-------------+---------------------------------- F(5, 24) = 0.29
Model | .086027227 5 .017205445 Prob > F = 0.9117
Residual | 1.40577277 24 .058573866 R-squared = 0.0577
-------------+---------------------------------- Adj R-squared = -0.1387
Total | 1.4918 29 .051441379 Root MSE = .24202
------------------------------------------------------------------------------
changeinhdi | Coef. Std. Err. t P>|t| [95% Conf. Interval]
-------------+----------------------------------------------------------------
hpsmedian | .0878233 .1092055 0.80 0.429 -.1375657 .3132123
intensity | -.0646888 .0970185 -0.67 0.511 -.2649252 .1355476
promkept | -.0158435 .1044076 -0.15 0.881 -.2313302 .1996431
rebel | -.0088158 .0401453 -0.22 0.828 -.0916716 .07404
pko | .0435967 .0961426 0.45 0.654 -.1548319 .2420254
_cons | .3782594 .0958405 3.95 0.001 .1804543 .5760646
------------------------------------------------------------------------------
* esize twosample changeinhdi, by(hpsmedian) all
Effect size based on mean comparison
Obs per group:
hpsmedian==0 = 15
hpsmedian==1 = 16
---------------------------------------------------------
Effect Size | Estimate [95% Conf. Interval]
--------------------+------------------------------------
Cohen's d | -.2026539 -.9072415 .5053938
Hedges's g | -.1973596 -.8835398 .4921904
Glass's Delta 1 | -.2000718 -.9047937 .5116509
Glass's Delta 2 | -.2051563 -.9099594 .506353
Point-Biserial r | -.1041389 -.4244398 .2526572
---------------------------------------------------------