The paper by G. RASCH discusses the statistical methods in psychometrics and addresses criticisms from Skinner and Zubin regarding the limitations of group-centered methods in understanding individual behavior. RASCH proposes three models for reactions to attainment and intelligence tests, which are designed to handle data in a way that allows for the separation of personal and stimulus parameters. These models are based on the assumption that the number of misreadings follows a Poisson distribution and that the parameters can be estimated independently. RASCH then applies these models to two specific types of tests: oral reading and intelligence tests. For oral reading, the model assumes that the number of misreadings is influenced by both the individual's ability and the difficulty of the text. For intelligence tests, the model suggests that the probability of a correct answer depends on the individual's ability and the item's difficulty, with responses being independent between items and individuals. The paper also explores the theoretical properties of these models, including the separability of parameters and the conditions for sufficient statistics. RASCH emphasizes the importance of independence in both static and dynamic situations, where responses are not influenced by preceding or subsequent responses from the same or different individuals. Finally, RASCH discusses the principles of comparison in psychology, arguing that well-defined comparisons between individuals and stimuli require unique solutions for the parameters. He concludes that transitivity, a fundamental property of measurement, implies that the dimensions of the parameters must be equal. This leads to the introduction of l-dimensional measurement of personal and stimulus parameters, which can be used to establish a reference point and measure the distance between points on a scale.The paper by G. RASCH discusses the statistical methods in psychometrics and addresses criticisms from Skinner and Zubin regarding the limitations of group-centered methods in understanding individual behavior. RASCH proposes three models for reactions to attainment and intelligence tests, which are designed to handle data in a way that allows for the separation of personal and stimulus parameters. These models are based on the assumption that the number of misreadings follows a Poisson distribution and that the parameters can be estimated independently. RASCH then applies these models to two specific types of tests: oral reading and intelligence tests. For oral reading, the model assumes that the number of misreadings is influenced by both the individual's ability and the difficulty of the text. For intelligence tests, the model suggests that the probability of a correct answer depends on the individual's ability and the item's difficulty, with responses being independent between items and individuals. The paper also explores the theoretical properties of these models, including the separability of parameters and the conditions for sufficient statistics. RASCH emphasizes the importance of independence in both static and dynamic situations, where responses are not influenced by preceding or subsequent responses from the same or different individuals. Finally, RASCH discusses the principles of comparison in psychology, arguing that well-defined comparisons between individuals and stimuli require unique solutions for the parameters. He concludes that transitivity, a fundamental property of measurement, implies that the dimensions of the parameters must be equal. This leads to the introduction of l-dimensional measurement of personal and stimulus parameters, which can be used to establish a reference point and measure the distance between points on a scale.