Psychometric Validity Analysis of Logical–Mathematical Intelligence

Abstract

This study is a validation and instrument-development research whose primary goal is to examine and ensure the psychometric quality of a Logical–Mathematical Intelligence (LMI) questionnaire for Senior High School students. LMI, as a fundamental aspect of scientific reasoning, requires a well-established measurement tool. The adapted instrument consists of 24 items using a five-point Likert scale and was constructed based on Gardner’s five core LMI dimensions (numerical ability, logical reasoning, problem solving, pattern recognition, and deductive/inductive thinking). Data were collected online from 317 high school students through convenience sampling. Analysis was performed using the Rasch Model (Winsteps 3.73). The results indicate that the instrument demonstrates adequate unidimensionality (raw variance explained 33.3%) and excellent reliability (Cronbach’s Alpha 0.89; Item Reliability 0.96). All items were found to be fit and functioning properly, with strong discriminative power (Item Separation 4.98), indicating that the instrument can classify item difficulty into seven distinct levels. These findings explicitly confirm that the adapted LMI questionnaire meets international psychometric standards and is valid for mapping the logical–mathematical reasoning abilities of high school students.