Table 4.
HLM examining anxiety as a mediator of the relationship between co-twin's sex and disordered eating in all twins.
| Model | β | Standard Error | df | t | p |
|---|---|---|---|---|---|
| Step 1: Co-twin's Sex → Disordered Eating | -.05 | .02 | 526.48 | -3.38 | .001 |
| Covariates | |||||
| Twin Sex | -.13 | .02 | 531.72 | -8.15 | <.0001 |
| Assessment Method (Laboratory vs. Home) | .03 | .02 | 406.52 | 1.63 | .11 |
| BMI | .03 | .00 | 536.07 | 7.28 | <.0001 |
| Step 2: Co-twin's Sex → Trait Anxiety | -1.26 | .46 | 563.03 | -2.76 | <.01 |
| Covariates | |||||
| Twin Sex | -2.30 | .47 | 562.97 | -4.90 | <.0001 |
| Assessment Method (Laboratory vs. Home) | .58 | .66 | 460.05 | .88 | .38 |
| BMI | .13 | .11 | 568.42 | 1.22 | .22 |
| Step 3a: Trait Anxiety → Disordered Eating | -.03 | .02 | 491.66 | -2.15 | .03 |
| Covariates | |||||
| Twin Sex | -.10 | .02 | 507.08 | -6.52 | <.0001 |
| Assessment Method (Laboratory vs. Home) | .02 | .02 | 380.56 | 1.55 | .12 |
| BMI | .02 | .00 | 487.48 | 7.28 | <.0001 |
| Step 4a: Co-twin's Sex and Trait Anxiety → Disordered Eating | -.03 | .02 | 491.66 | -2.15 | .03 |
| Covariates | |||||
| Twin Sex | -.10 | .02 | 507.08 | -6.52 | <.0001 |
| Assessment Method (Laboratory vs. Home) | .02 | .02 | 380.56 | 1.55 | .12 |
| BMI | .02 | .00 | 487.48 | 7.28 | <.0001 |
Note: BMI = Body Mass Index; Sex variables were effect coded as 1 for males and -1 for females; Assessment Method was effect coded as 1 for home assessment and -1 for laboratory assessment.
a
Estimates for steps three and four are identical because these parameters are calculated within the same equation.