Open AccessCalibration for Instrument Argumentation Skills on the Subject of Fluid Statics Using Item Response Theory
Author(s) -
Harry Affandy,
Dewanta Arya Nugraha,
Scundy Nourma Pratiwi,
C. Cari
Publication year - 2021
Publication title -
journal of physics. conference series
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.21
H-Index - 85
eISSN - 1742-6596
pISSN - 1742-6588
DOI - 10.1088/1742-6596/1842/1/012032
Subject(s) - reliability (semiconductor) , test (biology) , calibration , item response theory , statistics , argumentation theory , psychology , sample (material) , value (mathematics) , range (aeronautics) , computer science , mathematics education , mathematics , psychometrics , engineering , paleontology , power (physics) , physics , philosophy , chemistry , epistemology , chromatography , quantum mechanics , biology , aerospace engineering
This study aims to calibrate the instrument of argumentation skills on the subject of Fluid Statics using item response theory (IRT). The blueprint instrument developed based on the aspects and sub-aspects of argumentation skills, totaling 20 items. The scoring uses the Partial Credit Model (PCM) principle in 4 categories. Subjects involved in the study were 206 undergraduate students at two different universities. The validation of the instrument using expert judgment techniques, which consisted of linguists, physicists, and educational measurement experts. Instrument calibration for argumentation skills is carried out by (1) analysis of item fit with PCM by testing goodness of fit items, (2) testing prerequisite analysis, (3) reliability, (4) difficulty level, and (5) Standard Error of Measurement (SEM). The test developed is valid with Aiken V index values in the range 0.93 to 1. The analysis of IRT assumptions uses two parameters, the adequacy test of the sample and the factor analysis. The Chi-Square value in the Bartlett test is 2202.67 with df 190, and a p-value value (Sig.) <0.05 indicates that the sample size used has met the unidimensional requirements. The instrument developed has met the criteria for scoring the Polytomus items, based on a) the MNSQ INFIT value is in the range of 0.70 to 1.30, and b) the difficulty index test is in the range -2.00 <b <2.00. The results of the reliability estimate on the test equipment obtained a reliability value of 0.86, which means that the reliability of the test is in the high category. The graph of the information function and SEM converge on the ability scales of -2.8 and +1.5, meaning that the skills test kit argues it is suitable for testees in the low to moderate ability range.