Criação de Escalas Quantitativas para Estudos de Eventos Mentais: Uma Revisão da Lei do Julgamento Comparativo

Abstract

Louis Leon Thurstone (1887-1955) was a pioneer in developing scales for non-metric stimuli, creating an experimental model centered on the subject without the need for physical units of comparison. This model, known as the Law of Comparative Judgment, relies on probabilistic measures rather than discrete ones, making it a powerful tool for quantifying psychological attributes through interval scales. Thurstone proposed that sensory organs operate probabilistically, infl uenced by an inherent noise within the organism, which can be either internally or externally induced. This noise impacts the hypothetical psychological continuum, causing the observer’s response to be not absolute but based on information received at the moment. His model functions analogously to a statistical sample. The Law of Comparative Judgment is structured around three fundamental premises: (i) each stimulus initiates a discriminative process represented on a psychological continuum; (ii) momentary fl uctuations, both internal and among diff erent observers, can cause a stimulus to be perceived variably, characterized by a normal distribution; (iii) the mean (μ) and standard deviation (σ) of this distribution represent the scale value and its discriminative dispersion, respectively. One of the most widely adopted experimental methods derived from Thurstone’s model is the pairwise comparison protocol. In this protocol, the observer’s task is to discriminate between two stimuli presented simultaneously. In each trial, the observer must judge which of the two stimuli is superior in a specifi ed dimension. The process involves comparing each possible pair of stimuli, with each pair being repeated a specifi c number of times. To organize and analyze the results, a matrix is constructed to capture the diff erences between all stimuli in the pairwise comparison. The fi nal value assigned to each stimulus is determined by the average diff erence between that stimulus and all others in the set. The comparative judgment equation provides a method for transforming these diff erences from all pairwise comparisons into scalar values, whether metric or non-metric. This protocol not only off ers a systematic and rigorous approach for measuring psychological attributes but also allows for a detailed and quantitative analysis of the observer’s perceptions and preferences. Faced with the challenge of conducting pairwise comparisons under various experimental conditions, Thurstone developed a statistical approach to transform data from a simple ranking task into a binary matrix of comparisons. This method allows for the application of the comparative judgment equation, overcoming the challenges of traditional psychophysical methods related to the duration and complexity of experimental sessions. The ranking method, by simplifying data collection, has proven highly eff ective in obtaining interval scales in various psychological and social measurements, particularly in contexts of limited experimental control. This effi ciency suggests that the method could be adapted for large-scale studies and outside laboratory environments, enhancing research in applied fi elds of psychology and related areas.Keywords: Law of Comparative Judgment, ranking order procedure, interval scale; quantitative measure, psychophysics.Creación de Escalas Cuantitativas para Estudios de Eventos Mentales: Una Revisión de la Ley del Juicio ComparadoResumenLouis Leon Thurstone (1887-1955), interesado en cuantifi car actitudes, creencias, preferencias y opiniones, creó una ecuación conocida como la Ley del Juicio Comparativo. El modelo de escala