Decompositions and biplots in three-way correspondence analysis

In this paper correspondence analysis for three-way contingency tables is presented using three-way generalisations of the singular value decomposition. It is shown that in combination with Lancaster's (1951) additive decomposition of interactions in three-way tables, a detailed analysis is possible of the deviations from independence. Finally, biplots are shown to produce powerful graphical representations of the results from three-way correspondence analyses. An example from child development is used to illustrate the theoretical developments.     

Some uniqueness results for PARAFAC2

Whereas the unique axes properties of PARAFAC1 have been examined extensively, little is known about uniqueness properties for the PARAFAC2 model for covariance matrices. This paper is concerned with uniqueness in the rank two case of PARAFAC2. For this case, Harshman and Lundy have recently shown, subject to mild assumptions, that PARAFAC2 is unique when five (covariance) matrices are analyzed. In the present paper, this result is sharpened. PARAFAC2 is shown to be usually unique with four matrices. With three matrices it is not unique unless a certain additional assumption is introduced. If, for instance, the diagonal matrices of weights are constrained to be non-negative, three matrices are enough to have uniqueness in the rank two case of PARAFAC2. The authors are obliged to Richard Harshman for stimulating this research, and to the Associate Editor and reviewers for suggesting major improvements in the presentation.     

Effects of component length and of the transitions among components in multiple schedules

Pigeons received equal variable-interval reinforcement during presentations of two line-orientation stimuli while five other orientations appeared in extinction. Component duration was 30 seconds for all orientations and the sequence was arranged so that each orientation preceded itself and each other orientation equally often. The duration of one component (0°) was shortened to 10 seconds and the other (90°) was lengthened to 50 seconds. All animals showed large increases in response rate in the shortened component and this increase was recoverable after an interpolated condition in which all components were again 30 seconds in duration. This effect was replicated in a second experiment in which component duration was changed from 150 seconds to 50 seconds and 250 seconds. An examination of local contrast effects during the first experiment showed that the shortened component produced local contrast during subsequent presentations of the lengthened component, just as would a component associated with more frequent reinforcement. When the presentation sequence was changed so that the lengthened component was always followed by the shortened component, response rates generally increased during the lengthened component. When the sequence was arranged so that the shortened component always preceded the longer component, response rate decreased in the former. These effects, as well as the increases in response rate following change in component length, seem not to be the product of local contrast effects among components.     

The following schedule of reinforcement as a fundamental determinant of steady state contrast in multiple schedules

Two experiments investigated whether steady-state interactions in multiple schedules depend exclusively on the following schedule of reinforcement. Experiment 1 used a four-component multiple schedule in which two components were associated with the same constant schedule of reinforcement, and where rate of reinforcement was varied in the component that followed one of these. Contrast effects were reliable only in the component that preceded the point of reinforcement variation, although some contrast did occur otherwise. In those instances where contrast other than the following-schedule effect did occur, it was accounted for by the effect of the preceding schedule, an effect for which there were consistent individual differences among subjects, and which varied with component duration. Experiment 2 used a three-component schedule, in which reinforcement rate was varied in the middle component. The results were consistent with Experiment 1, as the following-schedule effect was the only consistent effect that occurred, although an effect of the preceding schedule did occur for some subjects under some conditions, and was especially evident early in training. The conclusion from both experiments is that there is no general effect of relative rate of reinforcement apart from the sum of the effects of the preceding and following schedules, and that the following-schedule effect is the fundamental cause of steady-state interactions.     

Concise formulas for the standard errors of component loading estimates

Concise formulas for the asymptotic standard errors of component loading estimates were derived. The formulas cover the cases of principal component analysis for unstandardized and standardized variables with orthogonal and oblique rotations. The formulas can be used under any distributions for observed variables as long as the asymptotic covariance matrix for sample covariances/correlations is available. The estimated standard errors in numerical examples were shown to be equivalent to those by the methods using information matrices.The author is indebted to anonymous reviewers for the corrections and suggestions on this study, which have led to improvements of earlier versions of this article.     

Significance of Childhood Conduct Problems to Later Development of Conduct Disorder Among Children With ADHD: A Prospective Follow-Up Study

This study investigates whether low to moderate levels of childhood oppositional defiant disorder (ODD) and conduct disorder (CD) behaviors contribute to the development of clinically diagnosed CD in adolescence, in children with attention deficit hyperactivity disorder (ADHD). Participants were 207 White boys (ages 6-12) with ADHD free of conduct disorder diagnoses. Parent and teacher ratings were obtained. Participants were assessed at mean age 18 by clinicians blind to childhood status. A non-ADHD group (recruited in adolescence) was also studied. ODD behavior ratings did not predict CD in adolescence, whereas CD behavior ratings did. No single ODD or CD behavior predicted adolescent outcome. ADHD probands with very low ratings (Not at all, Just a little) by parents and teachers on all CD behaviors were still at significantly increased risk for CD in adolescence, compared to non-ADHD controls. The same relationships were found between childhood ODD and CD behaviors, and antisocial personality disorder in adulthood (mean age, 25). We conclude that childhood ADHD is a developmental precursor of later antisocial disorder, even in the absence of comorbid ODD or CD in childhood. However, low levels of CD-type problems are not innocuous, because they predict later CD among children with ADHD without comorbid CD.     

Are immigrants in the nursing industry at increased risk of bullying at work? A one‐year follow‐up study

The purpose of this study was to explore whether (a) immigrant health care workers (HCW) are more at risk of bullying at work than Danish staff members, (b) this association is increased by previous exposure to bullying and (c) immigrants experience more bullying from supervisors, colleagues and clients/residents. We analyzed cross-sectional baseline data from 5,635 health care students of whom 10.4% were immigrants, and conducted a prospective analysis by following 3,109 of these respondents during their first year of employment. More than a third of the respondents had previous experiences with bullying. The baseline analyses showed that immigrants are more at risk of being bullied during both their theoretical education and trainee periods than their Danish co-students. At follow-up we found that 9.1% of the total cohort had been exposed to bullying at work during their first year of employment, hereof 1.8% frequently. "Non-Western" immigrants had a significantly higher risk of exposure to bullying at work during follow-up than the Danish respondents independent of previous experience with bullying. Danish and immigrant health care workers were more exposed to bullying from co-workers than from supervisors with no statistically significant difference between the Danes and the immigrant groups. Both "Western" and "non-Western" respondents were more at risk of bullying from clients/residents than the Danish respondents.     

认知负荷与编码方式影响小学数学学业不良生的前瞻记忆及其成分

通过两个实验探讨认知负荷对小学数学学业不良生前瞻记忆及其成分的影响, 以及是否可以通过执行意向编码提高其前瞻记忆及其成分的成绩。实验1结果显示数学学业不良生前瞻记忆、前瞻成分及回溯成分正确率显著低于数学学优生, 高认知负荷条件前瞻记忆、前瞻成分及回溯成分的正确率显著低于低认知负荷条件; 实验2结果显示执行意向编码条件下, 数学学业不良生前瞻记忆成绩显著好于标准编码条件, 前瞻成分和回溯成分的正确率高于标准编码条件。结果表明数学学业不良生前瞻记忆及其成分表现差于数学学优生; 无论认知负荷高低, 执行意向编码均可通过提升前瞻成分和回溯成分改善数学学业不良生前瞻记忆表现。     

Timing of Stressful Life Events Affects Stability and Change of Neuroticism

Neuroticism is a predictor of many health problems. To study the determinants of within‐subject change in neuroticism, three hypotheses were tested: (i) subjects who experienced stressful life events (SLEs) show an increase in neuroticism; (ii) high baseline neuroticism moderated this effect; and (iii) recent SLEs had a greater impact on neuroticism than distant SLEs. Data came from the Finnish Twin Cohort. Neuroticism data were collected in 1975 and 1981 and SLEs data in 1981 (n = 21 085). By entering baseline neuroticism as a predictor for neuroticism at follow‐up, the outcome measure was change in neuroticism. Changes in neuroticism were predicted from SLE indices or their interaction with baseline neuroticism. Timing of SLEs was taken into account by distinguishing recent from distant SLEs. To control for confounding by shared genes and environments, both within‐twin pair and between‐twin pair effects were tested for monozygotic and dizygotic twin pairs separately. Neuroticism's six‐year stability was high (r = .58, p < .001). Exposure to SLEs modestly increased neuroticism (βs > .55, ps < .001), unconfounded by shared genes. This effect was not moderated by high baseline neuroticism. Recent SLEs (.09 < βs < .15) had more impact than distant SLEs (.03 < βs < .11; ps < .01). In conclusion, the findings strongly supported a model of environmentally driven SLEs causing dynamic fluctuations around a person's set point of neuroticism. Copyright © 2013 John Wiley & Sons, Ltd.     

The effect of different distractions on remembering delayed intentions

In safety-critical domains, frequently intentions need to be delayed until an ongoing task is completed. Research using the delay-execute paradigm showed that interruptions during the delay cause forgetting. However, staff members often handle an initial distraction not by interrupting the ongoing task but by acknowledging the distraction or multitasking. In Experiments 1a and 1b, we observed that, compared to a no distraction condition, multitasking significantly decreased remembering of intentions and interrupting decreased remembering even further. In Experiment 2, interruptions with context change reduced remembering of intentions compared to uninterrupted delays, and at the same time, interruptions without context change improved memory performance compared to uninterrupted delays. However, improved memory performance resulted in decreased interrupting task performance. Theoretically, the results support the contextual cueing mechanism of delay-execute tasks. Considering safety-critical domains, multitasking, interruptions and context changes can contribute to forgetting of tasks.