Causal Indicators Research Articles

An approach to modelling concrete bridge condition deterioration using a statistical causal relationship based on inspection data

While physical models of concrete structures may give a fundamental understanding of structural deterioration mechanisms, their application is more or less limited in reflecting actual scenarios as a result of inevitable theoretical assumptions, whereby it is often difficult to cover complex structural behaviour and environment in the real world. Uncertainties and fuzziness, along with complexity, add more difficulty to the estimation of structural deterioration by physical models, which are typically described by quantitative mathematical equations. To complement physical models, an approach based on a statistical causal relationship, combining logics and statistics, is established for modelling structural condition deterioration. The relationship includes three kinds of basic variables in terms of causal factors, consequences and indicators, and it is represented by a set of conditional probabilities among those variables. Level two inspection data on concrete bridge slabs, as an example, is applied to establish the relationship. On the basis of the probabilistic causal relationship, prediction and diagnosis of structural conditions are implemented. To take advantage of this approach in dealing with qualitative issues such as human factors, the influence of inspectors' judgement on structural condition rating is addressed, while the effects of weather conditions on inspectors' judgement are also illustrated.

Interpretational confounding is due to misspecification, not to type of indicator: Comment on Howell, Breivik, and Wilcox (2007).

R. D. Howell, E. Breivik, and J. B. Wilcox (2007) have argued that causal (formative) indicators are inherently subject to interpretational confounding. That is, they have argued that using causal (formative) indicators leads the empirical meaning of a latent variable to be other than that assigned to it by a researcher. Their critique of causal (formative) indicators rests on several claims: (a) A latent variable exists apart from the model when there are effect (reflective) indicators but not when there are causal (formative) indicators, (b) causal (formative) indicators need not have the same consequences, (c) causal (formative) indicators are inherently subject to interpretational confounding, and (d) a researcher cannot detect interpretational confounding when using causal (formative) indicators. This article shows that each claim is false. Rather, interpretational confounding is more a problem of structural misspecification of a model combined with an underidentified model that leaves these misspecifications undetected. Interpretational confounding does not occur if the model is correctly specified whether a researcher has causal (formative) or effect (reflective) indicators. It is the validity of a model not the type of indicator that determines the potential for interpretational confounding.

Organizational use of the internet

Purpose – To contribute to the development of a theoretically grounded measure of organizational internet use (OIU) by conceptualizing and operationalizing it as a theoretical construct, and by empirically assessing its reliability and validityDesign/methodology/approach – The focal construct OIU was conceptualized as an abstract collective object with three components, forming an index with formative, causal indicators. A multi‐method research design – including a cross‐sectional drop‐and‐collect survey among small technology‐based firms (STBFs) in Germany and an observational study of web sites – was applied to assess empirically the theoretically developed construct OIU. This was achieved by using the component‐based partial least squares (PLS) structural modeling technique using PLS‐Graph.Findings – The empirical assessment of the scale, applied to the international business domain, proved to be reliable and valid in the structural model and across assessment methods.Research limitations/implications ...

Validating a Measure Across Groups

ABSTRACT This study tests the effects of causal indicators on the factorial invariance of a measure across groups. If a measure lacks invariance, MIMIC models (Multiple Indicators/Multiple Causes) can be used to identify the source of invariance. Using the Caregiver Well-Being Scale, two phases of multi-group modeling with post-care-givers and non-caregivers were analyzed. The first phase tested the scale's validity and reliability using multi-group modeling without controlling for the demographics differences between the groups. The second phase analyzed a MIMIC model with multi-group modeling. Results indicate that when a MIMIC model is not used, the measure is unequally reliable across groups (unequivalent structure). By controlling for the demographic differences using MIMIC modeling, analysis indicates that the scale is factorially invariant. Implications on the use of MIMIC models in scale development are discussed.

Being Inconsistent About Consistency: When Coefficient Alpha Does and Doesn't Matter

One of the central tenets of classical test theory is that scales should have a high degree of internal consistency, as evidenced by Cronbach's α, the mean interitem correlation, and a strong first component. However, there are many instances in which this rule does not apply. Following Bollen and Lennox (1991), I differentiate between questionnaires such as anxiety or depression inventories, which are composed of items that are manifestations of an underlying hypothetical construct (i.e., where the items are called effect indicators) and those such as Scale 6 of the Minnesota Multiphasic Personality Inventory (Hathaway & McKinley, 1943) and ones used to tap quality of life or activities of daily living in which the items or subscales themselves define the construct (these items are called causal indicators). Questionnaires of the first sort, which are referred to as scales in this article, meet the criteria of classical test theory, whereas the second type, which are called indexes here, do not. I discuss the implications of this difference for how items are selected, the relationship among the items, and the statistics that should and should not be used in establishing the reliability of the scale or index.

The Quality of Life Scale (QOLS): Reliability, Validity, and Utilization

The Quality of Life Scale (QOLS), created originally by American psychologist John Flanagan in the 1970's, has been adapted for use in chronic illness groups. This paper reviews the development and psychometric testing of the QOLS. A descriptive review of the published literature was undertaken and findings summarized in the frequently asked questions format. Reliability, content and construct validity testing has been performed on the QOLS and a number of translations have been made. The QOLS has low to moderate correlations with physical health status and disease measures. However, content validity analysis indicates that the instrument measures domains that diverse patient groups with chronic illness define as quality of life. The QOLS is a valid instrument for measuring quality of life across patient groups and cultures and is conceptually distinct from health status or other causal indicators of quality of life.

Causal Variables, Indicator Variables and Measurement Scales: An Example from Quality of Life

SummaryThere is extensive literature on the development and validation of multi-item measurement scales. Much of this is based on principles derived from psychometric theory and assumes that the individual items form parallel tests, so that simple weighted or unweighted summation is an appropriate method of aggregation. More recent work either continues to promulgate these methods or places emphasis on modern techniques centred on item response theory. In fact, however, clinical measuring instruments often have different underlying principles, so adopting such approaches is inappropriate. We illustrate, using health-related quality of life, that clinimetric and psychometric ideas need to be combined to yield a suitable measuring instrument. We note the fundamental distinction between indicator and causal variables and propose that this distinction suffices to explain fully the need for both clinimetric and psychometric techniques, and identifies their respective roles in scale development, validation and scoring.

Performing Feeling Without Fear

Performing Feeling Without Fear

Assessment of exposure to political violence and other potentially traumatizing events. A critical review.

This paper focuses on the common use of internal-consistency reliability, test-retest, and interrater correlations based on counts of events, events sampling, and factor-analytic techniques in assessment of exposure to political violence and other potentially traumatizing events. The author attributes the continued use of these strategies to a tendency among researchers to identify items from conventional events lists as effect indicators. Through a discussion of four alternative measurement models, the rationale is provided for the proposition that exposure to political violence and similar constructs should be treated as composite variables with causal indicators, rather than as latent variables with effect indicators.

Single or multiple familial cognitive risk factors in schizophrenia?

The fact that relatives of patients with schizophrenia display subtle cognitive abnormalities suggests genetic transmission of an underlying cognitive endophenotype. It was examined to what extent the cognitive abnormalities that discriminate patients and relatives from controls do so independently of each other, and independent of IQ. Neuropsychological measures were assessed in 50 patients with schizophrenia, 50 first-degree relatives of patients with schizophrenia, and 50 healthy controls. The assessment focused on episodic memory, attentional span, simple and complex speed of information, and semantic memory. Factor analysis of the cognitive test results yielded four factors reflecting speed, episodic memory, working memory, and semantic fluency. Performance of the relatives was intermediate to that of the patients and the controls after adjustment for age, sex, educational level, and IQ. For both patients and relatives, speed of information processing, working memory, and episodic memory independently discriminated from control performance, with a similar pattern in the order of the size of the effects. The results suggest the existence of more than one familial cognitive risk factor for schizophrenia. Independent familial cognitive risk factors may represent separate causal influences or separate indicators of risk related to the same genetic mechanism.

A tetrad test for causal indicators.

The authors propose a confirmatory tetrad analysis test to distinguish causal from effect indicators in structural equation models. The test uses "nested" vanishing tetrads that are often implied when comparing causal and effect indicator models. The authors present typical models that researchers can use to determine the vanishing tetrads for 4 or more variables. They also provide the vanishing tetrads for mixtures of causal and effect indicators, for models with fewer than 4 indicators per latent variable, or for cases with correlated errors. The authors illustrate the test results for several simulation and empirical examples and emphasize that their technique is a theory-testing rather than a model-generating approach. They also review limitations of the procedure including the indistinguishable tetrad equivalent models, the largely unknown finite sample behavior of the test statistic, and the inability of any procedure to fully validate a model specification.

A tetrad test for causal indicators.

A tetrad test for causal indicators.

Viewpoint – Export performance measurement: reflective versus formative indicators

While the conceptualization and operationalization of export performance has been receiving increased attention in the literature, practically all multi‐item measures of export performance available are based on reflective (effect) indicators. Explores an alternative measurement approach based on formative (causal) indicators and examines the implications of constructing an index rather than a scale of export performance. The aim is to assist researchers in choosing the appropriate measurement model for their needs when conducting substantive export performance studies.

Modeling Acts of Aggression and Dominance as Wife Abuse and Exploring Their Adverse Health Effects

This study, based on data from the Canadian Violence Against Women Survey, determined the effectiveness of modeling of aggression and dominance as causal indicators of forms of wife abuse and their effects on women's health, including physical injuries, psychopathology, alterations to the psyche, anger, alcohol and drug use, and subsequent general health status. When income, education, and age were controlled, abused wives who developed psychopathology or used drugs or medications to cope had poorer longterm health, regardless of when the abuse occurred. Health problems were influenced by the frequency of abuse, whether it occurred during pregnancy, and the nature of the abuse, as well as the severity of physical injuries sustained. Abuse during pregnancy resulted in poorer outcomes. The initial estimated model relied on severity weights for the of physical aggression reported in the literature. Through model revision, new estimates of the relative severity of various of physical aggression in relation to their likelihood of causing injury are provided. Key Words: health, psychological abuse, sexual abuse, structural equation model, violence, wife abuse. The Conflict Tactics Scales (CTS; Straus, 1979) frequently have been used to measure the occurrence of wife abuse. Despite widespread use, the scales have been criticized for failing to consider the context in which abuse occurs, differences in severity, and its consequences (Brush, 1993; Dobash, Dobash, Wilson, & Daly, 1992; Straus, 1990a). Straus (1990b) acknowledged that frequency-weighted scales do not reflect the differences in severity of violence inherent in the hierarchical structure of the violent acts (p. 539). For example, a woman who had been slapped six times would score higher than a woman who was severely beaten on three occasions. As a result of these and other drawbacks, Straus (1990b) introduced additional methods of scoring the scales. He indicated that the physical aggression items could be weighted by their frequency of occurrence, by the product of their frequency times a severity weight, or by using Guttman scales. None of these approaches, however, resolves all the problems associated with the confounding of severity and frequency. For example, creating three subscales for minor, severe, or very violence implicitly suggests that all within each scale are of equal severity. Guttman scales account for the hierarchical nature or severity of only if the original ordering of items is correct, but we have little empirical evidence that, for example, being hit with is more severe than being hit with a fist. Additionally, such scaling precludes consideration of the frequency of each act. The most promising approach, the severity-weighted scale, also has limitations. Specifically, there is an underlying assumption that the are intercorrelated and arise from a common cause. Yet, when Hayduk and Avakame (1990/1991) subjected this type of measurement structure to the test provided by multiple-indicator structural equation modeling (Hayduk, 1996; Ratner, Bottorff, Johnson, & Hayduk, 1996), they concluded that no single dimension underlay three of the CTS items: having something thrown; being pushed, grabbed, or shoved; and being slapped. (Other CTS items were not included in their model.) Recently, Straus, Hamby, Boney-McCoy, and Sugarman (1996) published a revised version of the CTS (the CTS2). The revised scale, now 39 items, distinguishes between minor and severe of physical assault and adds two scales measuring sexual coercion and physical injury. The work presented here is based on conceptualizations of physical aggression that underlie both versions of the CTS and introduces an approach to incorporating notions of severity that may resolve some of the problems associated with the approaches described above. The present study analyzed data collected in a cross-sectional survey and estimated a structural equation model that specified the relationships among various forms of wife abuse and their health consequences. …

Factor analysis, causal indicators and quality of life.

Exploratory factor analysis (EFA) remains one of the standard and most widely used methods for demonstrating construct validity of new instruments. However, the model for EFA makes assumptions which may not be applicable to all quality of life (QOL) instruments, and as a consequence the results from EFA may be misleading. In particular, EFA assumes that the underlying construct of QOL (and any postulated subscales or 'factors') may be regarded as being reflected by the items in those factors or subscales. QOL instruments, however, frequently contain items such as diseases, symptoms or treatment side effects, which are 'causal indicators'. These items may cause reduction in QOL for those patients experiencing them, but the reverse relationship need not apply: not all patients with a poor QOL need be experiencing the same set of symptoms. Thus a high level of a symptom item may imply that a patient's QOL is likely to be poor, but a poor level of QOL need not imply that the patient probably suffers from that symptom. This is the reverse of the common EFA model, in which it is implicitly assumed that changes in QOL and any subscales 'cause' or are likely to be reflected by corresponding changes in all their constituent items; thus the items in EFA are called 'effect indicators.' Furthermore, disease-related clusters of symptoms, or treatment-induced side-effects, may result in different studies finding different sets of items being highly correlated; for example, a study involving lung cancer patients receiving surgery and chemotherapy might find one set of highly correlated symptoms, whilst prostate cancer patients receiving hormone therapy would have a very different symptom correlation structure. Since EFA is based upon analyzing the correlation matrix and assuming all items to be effect indicators, it will extract factors representing consequences of the disease or treatment. These factors are likely to vary between different patient subgroups, according to the mode of treatment or the disease type and stage. Such factors contain little information about the relationship between the items and any underlying QOL constructs. Factor analysis is largely irrelevant as a method of scale validation for those QOL instruments that contain causal indicators, and should only be used with items which are effect indicators.

Causal indicators in quality of life research.

Quality of Life (QOL) questionnaires contain two different types of items. Some items, such as assessments of symptoms of disease, may be called causal indicators because the occurrence of these symptoms can cause a change in QOL. A severe state of even a single symptom may suffice to cause impairment of QOL, although a poor QOL need not necessarily imply that a patient suffers from all the symptoms. Other items, for example anxiety and depression, can be regarded as effect indicators which reflect the level of QOL. These indicators usually have a more uniform relationship with QOL, and therefore a patient with poor QOL is likely to have low scores on all effect indicators. In extreme cases it may seem intuitively obvious which items are causal and which are effect indicators, but often it is less clear. We propose a model which includes these two types of indicators and show that they behave in markedly different ways. Formal quantitative methods are developed for distinguishing them. We also discuss the impact of this distinction upon instrument validation and the design and analysis of summary subscales.

The use of causal indicators in covariance structure models: some practical issues.

In conventional representations of covariance structure models, indicators are defined as linear functions of latent variables, plus error. In an alternative representation, constructs can be defined as linear functions of their indicators, called causal indicators, plus an error term. Such constructs are not latent variables but composite variables, and they have no indicators in the conventional sense. The presence of composite variables in a model can, in some situations, result in problems with identification of model parameters. Also, the use of causal indicators can produce models that imply zero correlation among many measured variables, a problem resolved only by the inclusion of a potentially large number of additional parameters. These phenomena are demonstrated with an example, and general principles underlying them are discussed. Remedies are described so as to allow for the evaluation of models that contain causal indicators.

The use of causal indicators in covariance structure models: Some practical issues.

The use of causal indicators in covariance structure models: Some practical issues.

Conventional wisdom on measurement: A structural equation perspective.

The applicability of 5 conventional guidelines for construct measurement is critically examined: (a) Construct indicators should be internally consistent for valid measures, (b) there are optimal magnitudes of correlations between items, (c) the validity of measures depends on the adequacy with which a specified domain is sampled, (d) within-construct correlations must be greater than between-construct correlations, and (e) linear composites of indicators can replace latent variables. A structural equation perspective is used, showing that without an explicit measurement model relating indicators to latent variables and measurement errors, none of these conventional beliefs hold without qualifications. Moreover, a causal indicator model is presented that sometimes better corresponds to the relation of indicators to a construct than does the classical test theory effect indicator model.

Brief Communication

The applicability of the ICD E code as a causal indicator of nonfatal injuries has been criticized. New codes have been developed to replace the ICD codes. We compared the coding reliability of the ICD E and place vs. the Nordic mechanism (M) and place codes. The mean accuracy (76 vs. 70%) (p less than 0.002) and the intercoder reliability (84 vs. 69%) (p less than 0.001) were better for the E than M code. The accuracy of the place codes was the same (83%). A short training improved (p less than 0.001) the accuracy of all four codes. The replacement of the E code with the M code would not improve the reliability of data on causes of injuries.