G. Rasch discusses the statistical methods of psychometrics and their application in psychology, emphasizing the need for models that can separate individual and stimulus parameters. He presents three models for reactions to tests, focusing on the separability of parameters and the ability to check model adequacy independently of these parameters. In the first model, the number of misreadings in oral reading tests follows a Poisson distribution, and parameters are separated into individual and text factors. The second model applies to intelligence tests, where the probability of a correct answer is determined by individual and stimulus parameters. Both models allow for the estimation of parameters independently and the checking of model adequacy. The third model, not discussed here, is dynamic. Rasch argues that models with separable parameters allow for the estimation of individual and stimulus parameters without assuming a specific distribution for the population. He also discusses the importance of transitivity in comparisons and the necessity of equal dimensions for parameters in transitive comparisons. The paper concludes that the model (4.2) allows for the measurement of individual and stimulus parameters, with the parameters being linearly related. The paper emphasizes the importance of statistical models in psychology and the need for models that can separate individual and stimulus parameters to enable accurate comparisons and measurements.G. Rasch discusses the statistical methods of psychometrics and their application in psychology, emphasizing the need for models that can separate individual and stimulus parameters. He presents three models for reactions to tests, focusing on the separability of parameters and the ability to check model adequacy independently of these parameters. In the first model, the number of misreadings in oral reading tests follows a Poisson distribution, and parameters are separated into individual and text factors. The second model applies to intelligence tests, where the probability of a correct answer is determined by individual and stimulus parameters. Both models allow for the estimation of parameters independently and the checking of model adequacy. The third model, not discussed here, is dynamic. Rasch argues that models with separable parameters allow for the estimation of individual and stimulus parameters without assuming a specific distribution for the population. He also discusses the importance of transitivity in comparisons and the necessity of equal dimensions for parameters in transitive comparisons. The paper concludes that the model (4.2) allows for the measurement of individual and stimulus parameters, with the parameters being linearly related. The paper emphasizes the importance of statistical models in psychology and the need for models that can separate individual and stimulus parameters to enable accurate comparisons and measurements.