China
Africa
Explore chapters and articles related to this topic
Working with continuous outcome variables
Published in Ewen Harrison, Pius Riinu, R for Health Data Science, 2020
In base R form, it produces an ANOVA table which includes an F-test. This so-called omnibus test tells you whether there are any differences in the comparison of means of the included groups. Again, it is important to plot carefully and be clear what question you are asking.
Structural Models for Comparing Dependent Means and Proportions
Published in Jason T. Newsom, Longitudinal Structural Equation Modeling, 2015
The specification of the latent difference score model for three or more time points (Figure 3.2c) follows a fairly simple extension of the model for two waves. T−1 difference score factors, ηk, are specified at each time point except the first, with the path between the factor and the observed variable at the corresponding time point set equal to 1. Each observed variable, yt, is predicted by the observed variable at the prior time point, yt−1, with the path set equal to 1. The disturbance ζt associated with each endogenous variable is set equal to 0. Correlations among exogenous variables are estimated. The difference factor means could be used to compute the sum of squares in the same manner as shown for the contrast coding model in Equation (3.5). An omnibus test can also be obtained by setting the difference factor means equal to 0, with the fit representing a test of the null hypothesis that all means are equal.6
Development and Validation of an Arabic Version of the Drug Abuse Screening Test-10 (DAST-10) among Saudi Drug Abusers
Published in Journal of Psychoactive Drugs, 2022
Hussam Aly Sayed Murad, Nawaf Abdulaziz AlHarthi, Marwan Abdulrhman Bakarman, Zohair Jamil Gazzaz
Table 1 shows the participants’ sociodemographic characteristics (N = 460) and the use pattern for users (n = 360). The Poisson regression analysis showed a well-fitted model. The Goodness-of-Fit table provided a value equal to 0.871 for the Pearson chi-square indicating a small violation (0.129) of the equidispersion assumption. The Omnibus Test table showed a p-value < 0.001 indicating a statistically significant overall model. The tests of the Model Effects table displayed non-statistically significant values equal to 0.466, 0.548, 0.357, 0.139, 0.865, 0.669, and 0.339 for all the independent variables including gender, age, marital status, education level, age of start of use, duration of use and frequency of use, respectively.
Structural equation modeling of pedestrian behavior at footbridges in Ghana
Published in International Journal of Injury Control and Safety Promotion, 2022
Thomas Kolawole Ojo, Anthony Baffour Appiah, Abena Obiri-Yeboah, Atinuke Olusola Adebanji, Peter Donkor, Charles Mock
The omnibus test is conducted using the absolute fit indices employed in an SEM to test whether S (Kelloway, 1995). The GFI determines the amount of variance and covariance in the sample variance matrix that is predicted by the 2008), which is affected by sample size. The GFI is represented in equation: 2008). The AGFI adjusts the GFI for model complexity with degrees of freedom (Hooper et al., 2008) As in GFI, the AGFI falls between 0–1 and it is also sensitive to the sample size (Hooper et al., 2008). AGFI is calculated using Eq. (5):
An Evaluation of Occupational Behavior in Individuals with and without Attention Deficit/Hyperactivity Disorder
Published in Human Performance, 2018
Gregory A. Fabiano, Kevin F. Hulme, Sandro M. Sodano, Abigail Caserta, Karen Hulme, Gina Stephan, Alyssa C. Smyth
Study results were analyzed by comparing dependent measures collected during the three components of the LABOR assessment: (a) job application, (b) job interview, and (c) job performance. For each of these three segments of the LABOR assessment, group was entered as a between-subjects, fixed factor in a multivariate analysis of variance. Following a statistically significant omnibus test, univariate effects were examined. Chi-square tests were used to examine whether there were differences in the proportions of the groups on the driving history variables. Table 3 presents the descriptive statistics and intercorrelations among all the study variables for the total sample; Table 4 presents the same information broken by the participant group.