allows for predictor variables to be added in sequence to determine the total direct effect and, by subtraction, the indirect
effect of all predictor variables (Cohen et al., 2003).
Not exact matches
A common goal for a statistical research project is to investigate causality, and in particular to draw a conclusion on the
effect of changes in the values
of predictors or independent
variables on response or dependent
variables.
In place
of reporting probit coefficients, we report the marginal
effect of the
variable on the award probability, which is the change in the award probability due to each
predictor separately, with other
variables evaluated at their mean values.
To assess the potential
effect of missing data (ie, ignorable vs informative missing data), a pattern - mixture analysis was implemented using 2 - tailed tests.51 We defined patterns using a binary completer status
variable, which was entered as a
predictor in the RRM and MMANOVA.
To put the
effect sizes for the hypothesized associations on wave 6 reckless driving into perspective, we re-ran the final model using logistic regressions (for the connections between the wave 6 indicators and the wave 6 latent
variables) to obtain odds ratios (OR) for the indirect
effects of wave 1
predictors on the individual wave 6 reckless driving items.
For the logit - based analyses and the t tests
of differences in means, 1 - tailed tests
of significance were conducted (α =.05) because we had an a priori prediction about the direction
of the
effect for each
predictor variable.
To test this potential indirect
effect, we used a non-parametric Monte Carlo simulation method, in which the indirect
effect obtained from the a (the link between the
predictor variable and the indirect
effect variable) and b (the link between the indirect
effect variable and the dependent
variable, controlling for the remaining
predictors) paths in a series
of regression analyses is simulated k number
of times using the slopes and standard errors obtained from the data (we used k = 50,000).
We also report whether there were significant indirect
effects on 18 - month outcomes
of any demographic baseline
predictors mediated through the baseline latent
variables.
A third model with the same
predictors and random
effects was performed with the FaceReader measure
of disgust as dichotomous outcome
variable.
When one tests for the presence
of a moderational
effect with multiple regression, one examines whether an interaction between two
variables (one independent
variable and a moderator) is a significant
predictor of an outcome
variable, after controlling for the
effect of the two
predictors.
For example, if one were interested in whether the association between a parenting
variable (e.g., father psychological control; Holmbeck, Shapera, & Hommeyer, in press) and an outcome (e.g., school grades) is moderated by group status (e.g., spina bifida vs. an able - bodied comparison sample), one would test the interaction
of psychological control and group as a
predictor of school grades after controlling for the parenting and group main
effects.