Algumas diretrizes estatísticas para a avaliação do erro do método na mensuração de variável quantitativa

Abstract

INTRODUÇÃO: o erro do método deve ser quantificado para fundamentar a validade da pesquisa e sua aplicação clínica. A determinação da confiabilidade na identificação dos pontos de referência para a obtenção de medidas constitui pré-requisito fundamental para uma interpretação segura e utilização correta dos dados obtidos. Cada medida observada pode ser analisada pelo seu valor real e pela sua componente de erro. A confiabilidade é verificada a partir de medidas repetidas. OBJETIVO: o propósito deste trabalho é descrever um método estatístico que permita que o pesquisador verifique se as mensurações que realizou num certo momento são repetidas por ele num momento posterior. Este método confirma a não ocorrência do erro sistemático e permite uma avaliação da ocorrência do erro casual, que está sempre presente em qualquer forma de mensuração. Apresenta a aplicação deste método estatístico em dados referentes a duas medidas cefalométricas e aponta uma sintaxe para resolução do erro de método a partir do uso do programa SPSS® para microcomputador.

Evaluation of two techniques to measure the available space in the mandibular dental arch using the method error

“…The measurement of quantitative variables is a very common procedure in Orthodontic research (1) and clinical practice, since it is an important requirement for model analysis and for correct diagnosis and planning of orthodontic treatment. One key factor in model analysis is the assessment of the available space in the dental arch, which corresponds to the measurement of the space on the alveolar crest bone, in millimeters, from the mesial surface of the mandibular first permanent molars (2).…”

“…The measurement error is defined as the difference between the value obtained in the process of measurement and the true value of the magnitude measurement (1). Two types of error can occur in measurements executed on an object: systematic error, also known as bias, and random error (12), also called casual.…”

“…The systematic error is due to factors which, by having an effect on an instrument, on the measurement process, or even on the operator, affect the results in the same way. This means that the value which is found overstates or understates the true value of magnitude (1). Random error is defined as an error in which, due to competition between a number of factors, the difference between the measure and the measured magnitude may be positive or negative.…”

“…Random error is defined as an error in which, due to competition between a number of factors, the difference between the measure and the measured magnitude may be positive or negative. The greater the number of measurements taken, the more clearly the sum of these differences tends to zero and becomes subject to a law of probability distribution, known as the Normal Distribution or Gauss-Laplace Curve (1). Even in the simplest of experiments, it is impossible to control all of the factors involved.…”

Evaluation of two methods for mixed dentition analysis using the method error

“…The systematic error tends to overestimate or underestimate the true value of the magnitude. The random error results from unpredictable factors, such as the limitations of the equipment, the measurement procedure, or a variety of other factors, often making it impossible to accurately identify its source 21,23 .…”