Rolf Steyer

Rolf Steyer
Friedrich Schiller University Jena | FSU · Department of Psychology

Professor

About

112
Publications
94,677
Reads
How we measure 'reads'
A 'read' is counted each time someone views a publication summary (such as the title, abstract, and list of authors), clicks on a figure, or views or downloads the full-text. Learn more
4,236
Citations
Additional affiliations
October 1996 - present
Friedrich Schiller University Jena
Position
  • Professor (Full) for Research Methodology
Description
  • My main research interest is in the theory of causality and the theory of latent variables.
April 1982 - September 1994
Universität Trier
Position
  • PostDoc Position

Publications

Publications (112)
Chapter
Full-text available
We provide an introduction to revised latent state-trait (LST-R) theory. Subsequently, we present longitudinal structural equation models for analyzing variability and change across time that are based on LST-R theory. We cover latent state, latent state-trait, latent change score, and latent growth curve models as well as models with method factor...
Article
This study examined resistance of item parameters of a personality questionnaire to faking. We proposed seven construct-irrelevant item attributes (CIIAs), each composing of two opposite values, representing degrees of adherence to the respective CIIAs (e.g., based on the time frame embedded in the stem, items were divided into past and present). W...
Article
We present a probit latent state model for dichotomous items with one latent state variable for each occasion of measurement and one latent item-effect variable for each item, except for a reference item. All latent variables are well defined in terms of conditional probabilities. The model offers the possibility to include explanatory variables fo...
Chapter
This chapter discusses adjustment when covariates are not perfectly reliable. It starts with reviewing a theoretical framework that applies to fallible and latent covariates. This framework allows for deriving conditions under which adjustment has to be based on the latent covariate and conditions under which adjustment has to be based on the falli...
Article
Full-text available
Background: The project "Quality Assurance in Ambulatory Psychotherapy in Bavaria" (QS-PSY-BAY) focuses on the quality assurance of outpatient psychotherapy (OPT) in Germany in terms of symptom reduction and cost reduction under naturalistic conditions. In this study, we examined the effectiveness of psychotherapy in terms of pre-post cost reducti...
Data
Multiple studies examined health-care costs before and after outpatient psychotherapy in Germany. The sample size varied between N = 47 and N = 666 (see Table S1 in Supplementary Material). Long-term psychoanalytic psychotherapy was most often examined. All studies reported cost reductions. But, the amount of reduction varied from study to study (e...
Article
We present a framework for estimating average and conditional effects of a discrete treatment variable on a continuous outcome variable, conditioning on categorical and continuous covariates. Using the new approach, termed the EffectLiteR approach, researchers can consider conditional treatment effects given values of all covariates in the analysis...
Article
Full-text available
Anhand zweier Beispiele wird zunächst gezeigt, dass und warum die Analyse von Mittelwerten und ihren Unterschieden zwischen Gruppen zu kausalen Fehlschlüssen führen kann. Danach werden die Grundbegriffe der Theorie individueller und durchschnittlicher kausaler Effekte eingeführt. Anhand eines Beispiels zur nonorthogonalen Varianzanalyse wird dann d...
Article
Engl. title: Typical patterns of depressive disorders during outpatient psychotherapy and their prediction Objectives: Are there typicalpatterns of outpatient psychotherapy among depressed patients? What characterizes patients with different patterns? Methods:WeexaminedN=548patientswithprimarydepressivedisordersusinganaturalistic design. Using a l...
Article
Are there typical patterns of outpatient psychotherapy among depressed patients? What characterizes patients with different patterns? We examined N= 548 patients with primary depressive disorders using a naturalistic design. Using a latent-state-mixture model and depression measures at baseline, therapy end and 1-year follow-up we found a total of...
Article
Wu and Browne present a paper that is of high quality in a number of aspects, and it raises an interesting question: How to evaluate a covariance structure model (CSM)? Browne has made seminal contributions to this issue before, e.g., the development of the ADF method (Browne, 1984) and his contribution to the RMSEA statistic (Browne & Cudeck, 1993...
Chapter
Classical test theory (CTT) has been developed to quantify measurement error and to solve related problems such as correcting observed dependencies between variables (e.g., correlations) for attenuation due to measurement errors. Basic concepts of CTT are true score and measurement error variables. These concepts are defined as specific conditional...
Article
Full-text available
Mediation analysis, or more generally models with direct and indirect effects, are commonly used in the behavioral sciences. As we show in our illustrative example, traditional methods of mediation analysis that omit confounding variables can lead to systematically biased direct and indirect effects, even in the context of a randomized experiment....
Article
Objective: Because premature discontinuation of psychotherapy limits the effectiveness of the interventions, in a naturalistic design we examined the prevalence, predictors, and outcome of premature discontinuation. Methods: The sample included N = 584 patients with various mental disorders. Risk factors were identified using regression analysis...
Article
Full-text available
We present a revision of latent state-trait (LST-R) theory with new definitions of states and traits. This theory applies whenever we study the consistency of behavior, its variability, and its change over time. States and traits are defined in terms of probability theory. This allows for a seamless transition from theory to statistical modeling of...
Article
Full-text available
Conventionally, multilevel analysis of covariance (ML-ANCOVA) has been the recommended approach for analyzing treatment effects in quasi-experimental multilevel designs with treatment application at the cluster-level. In this paper, we introduce the generalized ML-ANCOVA with linear effect functions that identifies average and conditional treatment...
Article
Due to the treatment costs, extensions of the standard therapy duration are a matter of critical examination. This study investigates which factors characterize patients with treatment extensions in the German health system and how eff ective these extensions are for a reduction of the patients' symptoms. We analysed a disorder heterogeneous sample...
Article
Full-text available
Objective: Two patient-focused long-term research projects performed in the German outpatient psychotherapy system are focused on in this article. The TK (Techniker Krankenkasse) project is the first study to evaluate a quality assurance and feedback system with regard to its practical feasibility in German routine care. The other study ("Quality...
Article
Due to the treatment costs, extensions of the standard therapy duration are a matter of critical examination. This study investigates which factors characterize patients with treatment extensions in the German health system and how effective these extensions are for a reduction of the patients' symptoms. We analysed a disorder heterogeneous sample...
Chapter
Campbell and Fiske (1959) proposed multitrait-multimethod (MTMM) designs for the validation of measurement instruments. In these designs each of several constructs (traits) are measured with the same set of methods. According to Campbell (1959),discriminant validity is supported if the trait under investigation can be distinguished from other trait...
Article
Full-text available
We present a 3-step approach to defining latent growth components. In the first step, a measurement model with at least 2 indicators for each time point is formulated to identify measurement error variances and obtain latent variables that are purged from measurement error. In the second step, we use contrast matrices to define the latent growth co...
Article
I argue that the trait and network theories of personality are not necessarily contradictory. If appropriately formalized, it may turn out that network theory incorporates traits as part of the theory. I object the opinion that if a trait is a cause of behaviour, then it is necessarily an entity operating in the minds of individuals. Finally, I arg...
Article
Full-text available
This study investigated the changes in attachment characteristics of patients undergoing inpatient group psychotherapy in routine care. We collected data from 265 consecutively recruited patients and 260 non-clinical control persons using self-report measures of attachment, depression, and socio-demographic characteristics. The effects of treatment...
Article
Full-text available
General Psychotherapy (GPT; Grawe, 1997) is a research-informed psychotherapy that combines cognitive-behavioral and process-experiential techniques and that assumes motivational clarification and problem mastery as central mechanisms of change. To isolate the effect of motivational clarification, GPT was compared to a treatment that proscribed mot...
Article
An old issue in psychological assessment is to what extent power and speed each are measured by a given intelligence test. Starting from accuracy and response time data, an approach based on posterior time limits (cut-offs of recorded response time) leads to three kinds of recoded data: time data (whether or not the response precedes the cut-off),...
Article
Full-text available
General Psychotherapy (GPT; Grawe, 1997) is a research-informed psychotherapy that combines cognitive-behavioral and process-experiential techniques and that assumes motivational clarification and problem mastery as central mechanisms of change. To isolate the effect of motivational clarification, GPT was compared to a treatment that proscribed mot...
Article
Full-text available
Method effects often occur when constructs are measured by different methods. In traditional multitrait-multimethod (MTMM) models method effects are regarded as residuals, which implies a mean method effect of zero and no correlation between trait and method effects. Furthermore, in some recent MTMM models, traits are modeled to be specific to a ce...
Article
The true intra-individual change model is generalized by defining individual method effects. This allows the analysis of non-congeneric test-retest variables assumed to measure a common, possibly (temporally) transient, attribute. Temporal change in the attribute between different times of measurement is modelled by the true-change variable. Indivi...
Article
Method effects often occur when different methods are used for measuring the same construct. We present a new approach for modelling this kind of phenomenon, consisting of a definition of method effects and a first model, the "method effect model", that can be used for data analysis. This model may be applied to multitrait-multimethod data or to lo...
Article
Full-text available
Am Beispiel der Differenzierung von existentieller Schuld und Mitleid werden die Möglichkeiten von Strukturgleichungsmodellen zur Validitätsprüfung aufgezeigt. Die Ergebnisse einer Fragebogenerhebung, an 340 Personen sind mit der Annahme vereinbar, daß die verwendeten sprachlichen Items zur differenzierten Erhebung von existentieller Schuld und Mit...
Article
How is affective change rated with positive adjectives such as good related to change rated with negative adjectives such as bad? Two nested perfect and imperfect forms of dynamic bipolarity are defined using latent change structural equation models based on tetrads of items. Perfect bipolarity means that latent change scores correlate -1. Meaningf...
Article
Full-text available
Although both individual and average causal effects are defined in Rubin's approach to causality, in this tradition almost all papers center around learning about the average causal effects. Almost no efforts deal with developing designs and models to learn about individual effects. This paper takes a first step in this direction. In the first and...
Article
Full-text available
11.1 Introduction Is it possible to feel good and bad at the same time? A simple question, it seems, and one which should have a simple answer. However, this question is neither simple nor clear as the intensive and ongoing discussions at least since Wundt (1896) reveal (see, e.g. Watson, 1988). One reason why this question is not so simple is beca...
Chapter
Psychological interventions such as, for example, therapeutic treatments or educational training programs do not aim at ephemeral but primarily at permanent changes in behavior, feelings, attitudes etc., that is, at changes which are not situation-dependent, but rather consistent across situations. Similarly, changes of interest for developmental p...
Presentation
Contents: - Rubins potential outcomes, individual and average causal effects - Expected outcomes - Pearls method of computing causal effects - Ignorability vs. unconfoundedness - The logic of causality beyond the experimental setting - Examples
Article
Full-text available
During the last two decades, Structural Equation Modeling (SEM) has evolved from a statistical technique for insiders to an established valuable tool for a broad scientific public. This class of analyses has much to offer, but at what price? This paper pro- vides an overview on SEM, its underlying ideas, potential applications and current software....
Article
The new impulse given in the last decade to the theory of indi- vidual and average causal efiects is mostly due to the approach developed by Steyer and others and resulted in valuable theoretical results such as, for ex- ample, the link between unconfoundedness and causal unbiasedness. This ap- proach has also allowed for the development of difiere...
Article
Full-text available
The theory of individual and average causal effects presented in a previous paper is extended introducing conditioning on covariates. From a causal modeling point of view, there are two purposes of including covariates in a regression: (a) to study the condi-tional average causal effects of X on Y given the values z of the (possibly multi-dimension...
Article
Classical test theory (CTT) has been developed to quantify measurement error and to solve related problems such as correcting observed dependencies between variables (e.g., correlations) for attenuation due to measurement errors. Basic concepts of CTT are true score and measurement error variables. These concepts are defined as specific conditional...
Article
Obwohl das im letzten Kapitel dargestellte mathematische Gerüst, das keinerlei restriktive Annahmen enthält, zur Definition der True-Score- und der Fehlervariablen sowie der Reliabilität genügt, reicht es nicht aus, um die Varianzen und die Reliabilität der Testwertvariablen Yi aus empirisch schätzbaren Kennwerten berechnen zu können, Genausowenig...
Article
In diesem Kapitel geben wir eine kurze Einführung in das Nominalskalenmodell der Repräsentationstheorie des Messens. Dabei setzen wir die Kenntnis der Anhänge A — D voraus. Obwohl es innerhalb anderer Meßtheorien ebenfalls nominalskalierte Größen gibt (s. z.B. das Latent-Class-Modell im Rahmen der Probabilistischen Testtheorie; s. z.B. Langeheine,...
Article
Nachdem im vorigen Kapitel die wichtigsten Bestandteile und Begriffe des Nominalskalenmodells der Repräsentationstheorie des Messens dargestellt wurden, wollen wir in diesem Kapitel Beispiele für verschiedene Arten seiner Anwendung behandeln.
Article
Nachdem in Kap. 16 eine Einführung in das Rasch-Modell gegeben wurde, sollen nun die dort behandelten Inhalte vertieft werden. Zunächst werden wir die Annahme der Rasch-Homogenität einführen, ohne dabei schon auf die latente Personvariable ξ Bezug zu nehmen. Aus der Rasch-Homogenität folgt ohne weitere Annahmen die Existenz der theoretischen Größen...
Article
In den letzten beiden Kapiteln wurde in die wichtigsten Ideen des Nominalskalenmodells eingeführt. Im folgenden sollen nun die dabei verwendeten Begriffe präzisiert, die aufgestellten Behauptungen in einer exakten Sprache formuliert und bewiesen werden. Dies scheint uns für eine Vertiefung des Verständnisses der „Logik“ des Nominalskalenmodells nüt...
Article
In diesem Kapitel wird ein weiteres Modell eingeführt, innerhalb dessen die Bestimmung der Reliabilität und damit die Abschätzung der Zuverlässigkeit der durch eine Variable Yi repräasentierten Testwerte möglich wird. Wir zeigen Möglichkeiten, Folgerungen aus den Modellannahmen zu testen und wie die theoretischen Größen innerhalb des Modells aus de...
Article
In diesem Kapitel wird mit dem Rasch-Modell (nach dem dänischen Mathematiker Georg Rasch) ein stochastisches Meßmodell der Item-Response-Theorie (IRT) vorgestellt, das zur Analyse von Variablen, die nur 2 verschiedene Werte annehmen können, verwendet werden kann. Im Falle solcher zweiwertiger (dichotomer) Variablen sind das Modell essentiell τ-äqui...
Article
In Kap. 16 wurde das Rasch-Modell als ein Modell der Item-Response-Theorie vorgestellt. Da in Modellen der Item-Response-Theorie die Beziehungen zwischen den Lösungswahrscheinlichkeiten einzelner Items und den Ausprägungen einer latenten Variablen spezifiziert werden, eignen sich diese Modelle in besonderer Weise zur Selektion von Items für ein psy...
Article
Nachdem im vorangegangenen Kapitel die theoretischen Grundlagen dargestellt wurden, die eine Beurteilung der Zuverlässigkeit einer Messung erlauben, sollen diese nun am Beispiel eines Fragebogens zur Erhebung der Angst als Zustand (State-Angst) illustriert werden. Die verwendeten Daten stammen aus einer Untersuchung von Steyer et al. (1990) mit dem...
Article
Das im vorangegangenen Kapitel eingeführte Modell τ-kongenerischer Variablen soll nun an einem Beispiel illustriert werden. Wie in Kap. 10 liegt ein Fragebogen zugrunde, diesmal allerdings zur Erhebung der Angst als Eigenschaft (Trait-Angst). Die verwendeten Daten stammen wieder aus der Untersuchung von Steyer et al. (1990), bei der das State-Trait...
Article
In Kap. 10 wurden die wichtigsten Ideen des Modells essentiell τ-äquivalenter Variablen eingeführt. Wie bei den bisher behandelten Meßmodellen sollen nun die dabei verwendeten Begriffe präzisiert, die aufgestellten Behauptungen in einer exakten Sprache formuliert und bewiesen werden. Dies scheint uns für eine Vertiefung des Verständnisses der. „Log...
Article
Bisher wurden nur die einfachsten Modelle der Repräsentationstheorie des Messens dargestellt, Dabei könnte der Eindruck entstanden sein, daß man bei diesem Ansatz immer von Paarvergleichen ausgeht und generell bestenfalls Messungen auf Ordinalskalenniveau erreichen kann. Beides trifft jedoch nicht zu, denn innerhalb der Repräsentationstheorie gibt...
Article
In diesem Kapitel geben wir eine kurze Einführung in das Ordinalskalenmodell, das ebenso wie das vorher behandelte Nominalskalenmodell der Repräsentationstheorie des Messens zuzuordnen ist. Wir setzen dabei wiederum die Kenntnis der Anhänge A – D voraus.
Article
Wie beim Nominalskalenmodell sollen auch für das Ordinalskalenmodell der Reprasentationstheorie des Messens die im Einführungskapitel dargestellten Grundbegriffe anhand einiger Beispiele illustriert werden. Wir gehen zunächst wieder davon aus, daß die empirischen Relationen durch Paarvergleiche hergestellt werden. Andere Ausgangspunkte werden ansch...
Article
Die Darstellung des Ordinalskalenmodells in den letzten beiden Kapiteln soll nun durch die formalen Definitionen der zentralen Begriffe und die exakte Formulierung der Behauptungen und ihrer Beweise vertieft werden. Zunächst werden wir die Annahmen des Ordinalskalenmodells einführen. Es handelt sich dabei um die Anforderungen an die beiden Relation...
Article
Wie bei den bisher behandelten Meßmodellen soll im folgenden das Verständnis des Modells τ-kongenerischer Variablen vertieft werden, indem wir die in Kap. 13 eingeführten Begriffe präzisieren, die Behauptungen in einer exakten Sprache formulieren und beweisen. Nach der Definition τ-kongenerischer Variablen werden die Existenz-, Eindeutigkeits-, Bed...
Article
In diesem Kapitel geben wir eine Einführung in die Grundbegriffe stochastischer Meßmodelle und der Klassischen Theorie Psychometrischer Tests (KTT). Nach einigen vorbereitenden Überlegungen werden die einzelnen Bestandteile klassischer Meßstrukturen und damit die Begriffe True -Scare -Variable, Fehlervariable und Reliabilität eingeführt, die in den...
Article
Psychological assessment is a complex field comprising multi-tasks, from the design, administration, evaluation and interpretation of data-collection procedures to the treatment evaluation and following-up. Hence, psychological assessment is concerned with the assessment instruments and procedures, from their construction to their translation into...
Chapter
Nachdem wir in den vorangegangenen Kapiteln verschiedene Meßmodelle im Detail behandelt haben, ist es angebracht, sich das sich nun bietende Gesamtbild noch einmal aus etwas größerer Distanz anzusehen. Wir betrachten zunächst die Rolle von Meßmodellen im Wissenschaftsprozeß, ihren Stellenwert im Rahmen einer deduktivistischen Methodologie, und schl...
Article
Full-text available
In diesem Bericht wird in einem einführenden ersten Teil begründet, warum die Latent-State-Trait-Theorie (LST-Theorie) auch für die wehrpsychologische Eignungsdiagnostik von Bedeutung ist. Das Kernargument dabei ist, dass auch in diesem Bereich der Diagnostik damit gerechnet werden muss, dass die Testergebnisse nicht nur messfehlerbehaftet sind, so...
Article
Full-text available
We reformulate the theory of individual and average causal effects developed by Neyman, Rubin, Holland, Rosenbaum, Sobel, and others in terms of probability theory. We describe the kind of random experiment to which the theory refers, define individual and average causal effects, and study the relation between these concepts and the condi-tional ex...
Article
The first year under our joint editorship has now passed, and so as Editors-in-Chief we thought it appropriate in this Editorial to mention various changes and developments concerning the journal that have taken place over the past year and their effects for the future. The new format of the European Journal of Psychological Assessment (EJPA) has b...
Chapter
Full-text available
In this paper we present a class of structural equation models in which the true intraindividual change scores between two or more occasions of measurement are the values of the endogenous latent variables. This specification allows to examine the predictors of interindividual differences in true intraindividual change between time points. The mode...
Article
We consider regression models with discrete units and a discrete treatment variable. In this framework, individual and average causal effects as well as causal unbiasedness of conditional expected values E(Y | X = x) and of their differences were defined in a previous paper where it was also noted that a hypothesis of causal unbiasedness is not emp...
Article
Full-text available
Latent state–trait (LST) theory is a generalization of classical test theory designed to take account of the fact that psychological measurement does not take place in a situational vacuum. The basic concepts of latent state–trait theory (LST theory) are introduced. The core of LST theory consists of two decompositions: (a) the decomposition of any...
Article
The European Journal of Psychological Assessment (EJPA) was launched as the official organ of the European Association of Psychological Assessment (EAPA), which itself was founded in 1990. The main purpose of the EJPA is to gather important papers allowing theoretical and methodological reflection on the development of psychological assessment and...
Article
We consider the large sample Wald test for a nonlinear null hypothesis on a multidimensional parameter in a statistical model. The Wald statistic is known to be asymptotically chi-square distributed under the null hypothesis, provided that the Jacobian of the restriction function describing the null hypothesis has full rank. However, there may be ‘...
Article
Full-text available
It is shown how to specify a structural equation model in such a way that the true intraindividual change scores between two occasions of measurement are the values of an endogenous latent variable in the model. This makes possible to explain (and/or study the correlates of) interindividual differences in intraindividual change. An empirical exampl...
Article
Grawe and Braun (1994) proposed a computerized procedure for evaluating therapy effects, which is based on difference scores. However, in this procedure a post-pre-difference which is due only to measurement error may look as big as a post-pre-difference based on perfectly reliable measurements. Therefore, we propose two simple alternative indices...
Article
Following a discussion of the concepts of variability-sensitivity and change-sensitivity and an overview of some methods for estimating them, the basic assumptions of a latent state-trait model for variables with ordered categories are described. An application of the model is then presented, aimed at estimating the variability-sensitivity of the i...
Chapter
The main concern of causal modeling in the tradition of Neyman, Rubin, Holland, Rosenbaum, Sobel, and others is to make sure (e.g., by randomized assignment of observational units to experimental conditions) that the difference between the expectations of a (dependent) variable Y in a treatment and a control condition is equal to the average of the...