Wintersemester 2015/2016, Kurs, Sprache: Englisch, Thema: Methods of evaluation research

Topic: Analysis of conditional and average total causal treatment effects

Why we need a theory of causal effects

Concepts of probability theory

Basic ideas of the theory of causal effects

The core of the theory of causal effects

Causality conditions (sufficient conditions for unbiasedness), randomization, and covariate selection

First example: nonorthogonal analysis of variance

Intercept function, effect functions, and average effects in the example of nonorthogonal analysis of variance

Estimating intercept function, effect functions, and average effects via SPSS and its limitations

Analyzing the data of nonorthogonal analysis of variance with SPSS ANOVA (Typ I, II, III sums of squares) and why these analyses yields wrong results

The hypothesis of no treatment effects via SPSS using the R^{2}-difference test

Analysis of the data of nonorthogonal analysis of variance with EffectLiteR

Second example: The Kirchmann study on the treatment effects on depression and its analysis with EffectLiteR

Third example: The Klauer study on the training of inductive reasoning

EffectLiteR analysis of the Klauer data with a continuous covariate

Main hypotheses in EffectLiteR and the various conditional and average treatment effects

EffectLiteR analysis of the Klauer data with a continuous covariate and a qualitative covariate

Um die Videos anzuschauen, benötigen Sie den VLC media player, den Sie hier herunterladen können. Die Videos werden bereitgestellt über die Digitale Bibliothek Thüringen (dbt). Die grün verlinkten Videos und Materialien sind kostenfrei abrufbar. Klicken Sie auf den grünen Link, um das Video zu sehen oder die Datei herunterzuladen! Um alle Materialien und Videos abrufen zu können, müssen Sie sich einloggen.

Literature

Causal effects

Campbell, D. T. & Stanley, J. C. (1963). Experimental and Quasi-Experimental Designs for Research on Teaching. In N. L. Gage (Ed.), Handbook of research on teaching. Chicago: Rand McNally.

West, S. G., Biesanz, J. C. & Pitts, S. C. (2000), Causal inference and generalization in field settings. Experimental and quasi-experimental designs. In H. T. Reis and C. M. Judd (eds.), Handbook of research methods in social and personality psychology. Cambridge University Press.

Steyer, R. (2003). Wahrscheinlichkeit und Regression. Berlin: Springer. (Kapitel 15 - 17)

Steyer, R. (2004). Was wollen und was können wir durch empirische Kausalforschung erfahren? In E. Erdfelder & J. Funke (Hrsg.), Allgemeine Psychologie und deduktivistische Methodologie (pp.127-147). Göttingen: Vandenhoek und Ruprecht.

Steyer, R. (2005). Analyzing Individual and Average Causal Effects via Structural Equation Models. Methodology-European Journal of Research Methods in the Behavioral and Social Sciences, 1, 39-54.

Steyer, R. & Partchev, I. (2006). Manual for EffectLite: A Program for the Uni- and Multivariate Analysis of Unconditional, Conditional and Average Mean Differences Between Groups.

Pohl, S., Steyer, R. & Kraus, K. (2008). Modelling method effects as individual causal effects. Journal of the Royal Statistical Society. Series A, 171, 41--63.

Steyer, R., Partchev, I., Kröhne, U., Nagengast, B., & Fiege, C. (in preparation). Probability and Causality.

Intercept function and conditional-effect functions in the nonorthogonal ANOVA Example

Parameterization of the intercept function and conditional-effect functions in this example

Analysis of conditional effects in the nonorthogonal ANOVA Example with the Linear Regression program of SPSS

Point estimation of the conditional effects based on this data analysis

Limitations of the analysis conditional treatment effects via Linear Regression: No standard errors of conditional effects, no average treatment effects

The distinction between fixed and stochastic regressors.

Data analysis with (nonorthogonal) ANOVA (SPSS): Type I, II, III, and IV of decomposing the sum of squares. All of them do not test the hypothesis that the average treatment effect is zero.

Hypotheses that are tested as the so-called main effects with Typ I, II, and III.

Summarizing the basic concepts and equations in the analysis of conditional and average effects.

EffectLiteR analysis of the Klauer data with a qualitative and a quantitative covariate

Model equation and linearity assumption for the regression of the outcome variable on the quantitative covariate in each cell

Meaning of the four main hypotheses in terms of (a) expected values or effects, (b) the g-functions, and (c) the coefficients of the g-functions

Adjusted expected values

Conditional effects given values of the qualitative covariate

Conditional effects given values of the qualitative covariate and the treatment variable

Conditional effects given values of the qualitative covariate and the quantitative covariate

Conditional expected values of the outcome variable under treatment and under control given values of the qualitative covariate and the quantitative covariate

EffectLiteR analysis of the Klauer data with a latent covariate and a latent outcome variable. Models of essentially tau-equivalent and tau-congeneric variables

Testing the model with a goodness-of-fit test and the RMSEA

Conditional effects given estimated values of the latent covariates

EffectLiteR analysis of the Klauer data with two latent covariates and a latent outcome variable

EffectLiteR analysis of the Klauer data with a latent covariate, a method factor, and a latent outcome variable