The tight coupling between category and causal learning

Prior knowledge plays a central role in causal induction, helping to explain how people are capable of identifying causal relationships from small amounts of data. Bayesian inference provides a way to characterize the influence that prior knowledge should have on causal induction, as well as an explanation for how that knowledge could itself be acquired. Using the theory-based causal induction framework of Griffiths and Tenenbaum (2009), this chapter reviews recent work exploring the relationship between prior knowledge and causal induction, highlighting some of the ways in which people’s expectations about causal relationships differ from approaches to causal learning in statistics and computer science.

This study explored the relationships of senior high school EFL students' motivation, strategy use, and English achievement. The first two attempts of the study were to investigate how well the senior high school EFL students' motivation and strategy use can predict students' English achievement respectively. The third attempt was to explore the causal relationship among senior high school EFL students' motivation, strategy use, and English achievement.Totally 132 senior high EFL school boy students took part in the study. Two questionnaires-Motivation Questionnaire (Chen, 2008; Lai, 2008) and Strategy Inventory for Language Learning (SILL, version 7.0)-were used for data collecting.With SPSS for Windows Version 17.0, data analysis, including item analysis, exploratory factor analysis (EFA), reliability analysis, and validity analysis, was used. Furthermore, Structural equation modeling (SEM) was adopted through second-order confirmatory factor analysis (CFA) in LISREL 8.8 to confirm the causal relationships among senior high school EFL students' motivation, strategy use, and English achievement.The causal relationships between motivation and English achievement and between strategy use and English achievement were both proved. Motivation could account for 39% of the total variance toward English achievement while strategy use could account for 19% of the total variance toward English achievement. Intrinsic Motivation and Utilitarian Purpose took up the most variance toward English achievement, whereas Metacognitive Strategies was the only individual factor contributing to the total variance toward English achievement. The relationship among senior high school EFL students' motivation, strategy use, and English achievement was a full mediating model with a reciprocal relationship between strategy use and English achievement.

The purpose of this study is that analyzing the causal relationship between Infra, Process and Performance of companies which are executing the Supply Chain Quality Management(SCQM) with their subcontractors and partners. Korean Standards Association(KSA) provides the Supply Chain Quality Management Model and Quality Collaboration Index for 4 years, but a few study has investigated the critical variables and their causal relationship to organizational performance. Therefore we examine the SCQM model and related index and choose the quality, human resource and risk management processes for identifying the path to organizational performance. In addition, exploratory factor analysis is conducted for figuring out the major factors among the 3 processes. Structural Equation Model are successively used for determining which characteristics of the infra and processes are the most critical variables to performance. The data was collected from KSA and composed of 52 companies and 346 their partners. The result shows that risk management process has no significant effect on the organizational performance and pre-production process collaboration.

Existing models of causal induction primarily rely on the contingency between the presence and the absence of a causal candidate and an effect. Yet, classification of observations into these four types of covariation data may not be straightforward because (a) most causal candidates, in real life, are continuous with ambiguous, intermediate values and because (b) effects may unfold after some temporal lag, providing ambiguous contingency information. Although past studies suggested various reasons why ambiguous information may not be used during causal induction, the authors examined whether learners spontaneously use ambiguous information through a process called causal assimilation. In particular, the authors examined whether learners willingly place ambiguous observations into one of the categories relevant to the causal hypothesis, in accordance with their current causal beliefs. In Experiment 1, people's frequency estimates of contingency data reflected that information ambiguous along a continuous quantity dimension was spontaneously categorized and assimilated in a causal induction task. This assimilation process was moderated by the strength of the upheld causal hypothesis (Experiment 2), could alter the overall perception of a causal relationship (Experiment 3), and could occur over temporal sequences (Experiment 4).

This study adopted the concept of customer relationship management to investigate the cause and effect among caused-related marketing, relationship selling behavior, brand image management, brand trust, brand relationship and brand value. This study used the Linear Structural Relation Model (LISREL) to test the causal relationship among caused-related marketing, relationship selling behavior, brand image management, brand trust, brand relationship, and brand value. This study posted questionnaires using quota sampling to the customers of four areas at department stores in Taiwan. Empirical results verify that there exist positive relationships with the proposed hypotheses: different caused-related, relationship selling behavior and brand image management affect diverse categories of brand value in light of different types of brand trust and brand relationship. Consequently, the managerial implications of this study can provide the appropriate strategies for department stores or service industries to develop and execute efficient customer relationship activities.

Objective Combined with susceptibility weighted imaging (SWI) and proton magnetic resonance spectroscopy (1H-MRS) to study the hippocampus in the first-episode depression patients for exploring the causal relationship. Methods 20 first-episode depression patients and 20 healthy volunteers underwent SWI scanning and hippocampal 1H-MRS image, analysis and comparison the changes between the venous system and metabolism of two groups. Results Proton spectroscopy analysis showed that compared with the control group, bilateral hippocampal NAA/Cr(left hippocampus: depression group 0.95±0.77, control group 1.78±0.82; right hippocampus: depression group 1.12±0.74, control group 1.91±0.80) and Cho/Cr(left hippocampus: depression group 1.08±0.83, control group 1.75±0.73; right hippocampus: depression group 0.91±0.68, control group 1.95±0.74)in the depression group was significantly lower, the difference was significant(NAA/Cr: t value in left and right were -3.14, -3.49, P 0.05). Magnetic sensitive scan showed depression group compared with the control group was significantly reduced in lateral vein diameter, length and branches (t value were 3.23, 4.56, 2.66, P<0.05). Conclusion The SWI sequence and 1H-MRS imaging methods can become a new research idea about the central nucleus, and this idea may explain the causal relationship between the hippocampus in depression patients of vascular abnormalities and hippocampal neuronal metabolic abnormalities. Key words: Susceptibility weighted imaging; Proton magnetic resonance spectroscopy; Hippocampus; Depressive disorder

This paper focuses on the casual relationship between export, import and Gross Domestic Product (GDP) for Bhutan using annual data from 1980 to 2005. The Granger causality test and Co-integration Models are employed taking care of stochastic properties of the variables. The co-integration analysis suggests that there is a long-run equilibrium relationship. The results of Granger causality test shows that there is a causal relationship between the examined variables. The causal nexus is unidirectional from export to import and GDP, and GDP to import only. Here export led growth is empirically proven in Bhutan.

Amodel based on Bayesian belief network is put forward to solve the innovation performance problems of the urban subway construction safety in China. After a systematic analyzing of the influence factors on the technological innovation performance, 13 variables were chosen in the light of implementation factors, interference factors, mediation factors and control factors, by taking the application of building information model as a case study. The causal relationship between these variables was constructed by the cause and effect diagram of experts' prior knowledge, combined with the correlation analysis method to determine the Bayesian belief network system structure. Sample data was got through a structural questionnaire survey. It was fit by NETICA software to get the probability density of each node. The application results show that this method can achieve a certain quantitative analysis of the innovation performance of the subway construction safety technology, and has better compliance compared with actual engineering situation.

Structural-equation modeling is used to test causal relationships between narcotics addiction and the associated cost-support activities of property crime and drug dealing across four critical periods of the addiction career. It is argued that structural-equation methodology yields greater insight into the causal dynamics of such activities than the typical methodologies of comparing means and proportions. Using structural coefficients of longitudinal models to infer causal relationships and stability, it is found that (1) dealing is often a predictor of future narcotics involvement; (2) narcotics use, property crime, and dealing are mutually interrelated during periods of elevated narcotics use and are not related during periods of reduced narcotics use; (3) dealing appears to be the most stable of the three variables, although narcotics use and property crime show appreciable stability as well; and (4) property crime and dealing appear to be inversely related to a moderate extent. The relevance of these findings to the understanding of the economic behavior of addicts during the initiation and cessation of addiction, relapse, and “maturing out” is discussed.

One challenge when inferring the strength of cause-effect relations from time series data is that the cause and/or effect can exhibit temporal trends. If temporal trends are not accounted for, a learner could infer that a causal relation exists when it does not, or even infer that there is a positive causal relation when the relation is negative, or vice versa. We propose that learners use a simple heuristic to control for temporal trends-that they focus not on the states of the cause and effect at a given instant, but on how the cause and effect change from one observation to the next, which we call transitions. Six experiments were conducted to understand how people infer causal strength from time series data. We found that participants indeed use transitions in addition to states, which helps them to reach more accurate causal judgments (Experiments 1A and 1B). Participants use transitions more when the stimuli are presented in a naturalistic visual format than a numerical format (Experiment 2), and the effect of transitions is not driven by primacy or recency effects (Experiment 3). Finally, we found that participants primarily use the direction in which variables change rather than the magnitude of the change for estimating causal strength (Experiments 4 and 5). Collectively, these studies provide evidence that people often use a simple yet effective heuristic for inferring causal strength from time series data. (PsycINFO Database Record

Financial barriers can affect timely access to maternal health services. Health insurance can influence the use and quality of these services and potentially improve maternal and neonatal health outcomes. We conducted a systematic review of the evidence on health insurance and its effects on the use and provision of maternal health services and on maternal and neonatal health outcomes in middle- and low-income countries. Studies were identified through a literature search in key databases and consultation with experts in healthcare financing and maternal health. Twenty-nine articles met the review criteria of focusing on health insurance and its effect on the use or quality of maternal health services, or maternal and neonatal health outcomes. Sixteen studies assessed demand-side effects of insurance, eight focused on supply-side effects, and the remainder addressed both. Geographically, the studies provided evidence from sub-Saharan Africa (n = 11), Asia (n = 9), Latin America (n = 8), and Turkey. The studies included examples from national or social insurance schemes (n = 7), government-run public health insurance schemes (n = 4), community-based health insurance schemes (n = 11), and private insurance (n = 3). Half of the studies used econometric analyses while the remaining provided descriptive statistics or qualitative results. There is relatively consistent evidence that health insurance is positively correlated with the use of maternal health services. Only four studies used methods that can establish this causal relationship. Six studies presented suggestive evidence of over-provision of caesarean sections in response to providers' payment incentives through health insurance. Few studies focused on the relationship between health insurance and the quality of maternal health services or maternal and neonatal health outcomes. The available evidence on the quality and health outcomes is inconclusive, given the differences in measurement, contradictory findings, and statistical limitations. Consistent with economic theories, the studies identified a positive relationship between health insurance and the use of maternal health services. However, more rigorous causal methods are needed to identify the extent to which the use of these services increases among the insured. Better measurement of quality and the use of cross-country analyses would solidify the evidence on the impact of insurance on the quality of maternal health services and maternal and neonatal health outcomes.

Depression is a heterogeneous mental illness with symptoms including sadness, helplessness, anhedonia, fatigue and suicidality. Based on its increasing prevalence, depression is predicted to be the second major cause of disability worldwide by 2020. Despite this fact, the aetio-pathophysiology of depression remains poorly understood. Human clinical and experimental studies have linked the inflammatory cytokine network with depressive symptoms, leading to the cytokine hypothesis of depression. However, the direction of the causal relationship between depression and the activation of the immune system, as well as the involved mediators and pathways, have not yet been clarified. Valid translational animal models are essential to investigate the hypothesis postulating that peripheral and central immune-activation, especially increased TNF, is induced by stress and effects neurobiological changes leading to depression. Several studies in rodents have investigated the effects of a single TNF application on behavior, but not of a prolonged TNF increase, although this would correspond more closely to the low, prolonged increase in plasma TNF observed in human depression. This PhD project provides insight into the differential effects of a prolonged increase of TNF in the periphery versus specific brain regions on behaviors that can be translated to symptoms or domains of depression. The main focus was placed on the reactivity to negative stimuli, as assessed by fear conditioning-expression and avoid-escape behavior, and on the reactivity to positive stimuli, as assessed by saccharin versus water consumption and preference. Repeated injection as well as continuous delivery of TNF in the periphery were investigated, and chronic elevation of TNF expression in specific brain regions, namely periventricular regions, amygdala and hippocampus were achieved using viral vectors. This project demonstrated in mice that peripheral increases in TNF resulted in moderate sickness, an increased reactivity to negative stimuli and a reduced reactivity to reward, effects that co-occurred with splenomegaly and increases in plasma IL-B6 and IL-B10 while not leading to glial activation. Central increase in TNF led to increased reactivity to negative stimuli when localized to periventricular regions of hippocampus, while increased amygdala TNF led to reduced reactivity to saccharin and did so in the absence of both sickness in terms of body weight loss and changes in the plasma levels of cytokines. These findings provide important new insights into the causal inter-relationships between peripheral TNF, central TNF and depression-relevant behavior and will inform research and development relevant to anti-inflammatory strategies of antidepressant treatment.

The use of mobile phones and the risk of brain tumors among children and adolescents

The effect of confounding factors cannot be ignored in real world causal discovery tasks. A common cause is a general confounder between two variables. In this paper, we propose using intrinsic dimension estimation as a necessary condition to determine a possible common cause for two variables. Simulated application showed that the proposed method worked well for both linear and non-linear functions. Testing using different types of noise showed that it generally worked well for different types of added noise. In particular, it worked better than a kernel-based conditional independence test for Poisson noise. Testing of how the estimated intrinsic dimension is affected by different types of distributions showed that the estimated dimension is nearly not affected by the type of distribution. Simulation of mixed pattern showed that the proposed method can still tell a possible common cause when it is mixed with causal relationship. Finally, experiments using variables from the CauseEffectPairs dataset showed that the proposed method can give correct inferred results for real world data.

Mean arterial blood pressure was measured for 2 days in conscious dogs to estimate the frequency of ultradian fluctuations and to compare the magnitudes of the ultra- and circadian fluctuations. The ultradian fluctuation had an average frequency of 0.38 +/- 0.04 cycle/h and 60% more power than the circadian rhythm. The ultradian fluctuation was consistently more powerful during daylight than nighttime hours, but there was no consistent change in the frequency with time of day. Electromyogram (EMG) activity was also measured to test a possible causal relationship between the two variables. There were both circ- and ultradian fluctuations in EMG activity, and the peak cross-spectral power was also in these frequency bands. The coherence function was also high, but the phase angle between the two variables was so variable, both in sign and in magnitude, that it was judged unlikely that there was tight coupling between them.