Intraclass correlation coefficient and its application to quality tests of measurement instrument

• Published in: 2011 International Conference on Computer Science and Service System (CSSS) pp. 3302 - 3305
• Main Authors: Yan Wang, Shouying Zhao
• Format: Conference Proceeding
• Language: Chinese; English; Japanese
• Published: IEEE 01-06-2011
• Subjects: Analysis of variance; Computational modeling; Correlation; Indexes; Instruments; Intraclass correlation; Pearson product moment correlation; Psychology; Reliability; the model of one-way random effects
• ISBN: 9781424497621; 1424497620
• DOI: 10.1109/CSSS.2011.5974773

Correlation coefficients are dominant indexes of measurement instrument quality. Among the varieties of correlation coefficients Pearson product moment correlation coefficient is the most popular. That is why intraclass correlation attracts relatively less relation. But in certain conditions, Pearson product moment correlation coefficient is not effective and intraclass correlation takes advantage in stead. Intraclass correlation coefficient plays an important role in testing the quality of measurement instruments in behavior sciences, psychological measurement, behavioral genetics and medical sciences. In case of the constructs to measure is two or more aspects of the same concept, to test the reliability and the validity of the measurement intraclass correlation will exert its prevailing advantages. If correlation between unorder pairs groups of interest, Pearson product moment correlation will be of no sense in this case and intraclass correlation coefficient functions properly. Under the frame of ANOVA, three basic models of intraclass correlation are classified, as are models of one-way random effects, two-way random effects and two-way mixed-effect. This study will focus on the model of one-way random effects and its application in measurement instrument quality test.