The Performance of Multivariate Methods for Two-Group Comparisons with Small Samples and Incomplete Data

Abstract

In intervention studies having multiple outcomes, researchers often use a series of univariate tests (e.g., ANOVAs) to assess group mean differences. Previous research found that this approach properly controls Type I error and generally provides greater power compared to MANOVA, especially under realistic effect size and correlation combinations. However, when group differences are assessed for a specific outcome, these procedures are strictly univariate and do not consider the outcome correlations, which may be problematic with missing outcome data. Linear mixed or multivariate multilevel models (MVMMs), implemented with maximum likelihood estimation, present an alternative analysis option where outcome correlations are taken into account when specific group mean differences are estimated. In this study, we use simulation methods to compare the performance of separate independent samples t tests estimated with ordinary least squares and analogous t tests from MVMMs to assess two-group mean differences with multiple outcomes under small sample and missingness conditions. Study results indicated that a MVMM implemented with restricted maximum likelihood estimation combined with the Kenward–Roger correction had the best performance. Therefore, for intervention studies with small N and normally distributed multivariate outcomes, the Kenward–Roger procedure is recommended over traditional methods and conventional MVMM analyses, particularly with incomplete data.     

Vocational Preference and P–O Fit: Reflections on Holland's Theory of Vocational Choice

It is 40 years since Holland first proposed his theory of vocational choice. In that period it is probably true to say that it has attracted more research than any other vocational guidance measure. This paper looks at the central concept in his theory (congruence) and other areas of vocational guidance, namely "fit". It also looks at current research issues relating to Holland's theory. Particular attention is paid to the research papers that examine the overlap between the six (RAISEC) personality types of Holland and the consensually accepted "Big Five" factor view of personality theorists. Finally a summary of the major theoretical, empirical, and methodological problems in the area is discussed.     

Assessing inattention and impulsivity in children during the Go/NoGo task

Behavioural performance in the Go/NoGo task was compared with caregiver and teacher reports of inattention and hyperactivity‐impulsivity in 1,151 children (N=557 boys; N=594 girls) age 9–10 years old. Errors of commission (NoGo errors) were significantly correlated with symptom counts of hyperactivity‐impulsivity, while errors of omission (Go errors) were significantly correlated with symptom counts for inattention in both caregiver and teacher reports. Cross‐correlations were also evident, however, such that errors of commission were related to inattention symptoms, and errors of omission were related to hyperactivity‐impulsivity. Moreover, hyperactivity‐impulsivity and inattention symptoms were highly intercorrelated in both caregiver (r=.52) and teacher reports (r=.70), while errors of commission and omission were virtually uncorrelated in the Go/NoGo task (r=.06). The results highlight the difficulty in disentangling hyperactivity‐impulsivity and inattention in questionnaires, and suggest that these constructs may be more clearly distinguished in laboratory measures such as the Go/NoGo task.