<![CDATA[Penelitian ini bertujuan untuk mengukur self-evaluation mahasiswa pendidikan matematika terhadap materi perkuliahan analsis real melalui pembelajaran online. Penelitian ini menggunakan pendekatan kuantitatif dengan pengambilan sampel berdasarkan pertimbangan tertentu (purposive sampling) yaitu mahasiswa Program Studi Pendidikan Matematika FKIP Universitas Pattimura yang mengikuti mata kuliah analisis real sebanyak 106 responden. Data diperoleh dengan menyebarkan angket self evaluation melalui google form dan berupa angket respon dengan metode skala Likert. Analisis awal dilakukan untuk mengetahui validitas dan reliabilitas instrumen melalui perhitungan Exploratory Factor Analysis (EFA) untuk mengetahui validitas konstruk instrumen dan Cronbach’s Alpha untuk mengetahui reliabilitas. Selanjutnya dilakukan analisis Item Response Teory (IRT) politomus Rating Scale Model (RSM) untuk mengukur self-evaluation mahasiswa. Hasil analisis EFA menggunakan program R menghasilkan 7 faktor laten dari 30 item indikator dengan nilai Kaiser-Meyer-Olkin (KMO) adalah 0. 79 dan MSA tiap item menunjukkan hasil > 0.5, loading faktor tiap item > 0.3 dan estimasi reliabilitas 0.877. Hasil perhitungan IRT politomus RSM menunjukkan model fit untuk setiap item dan diperoleh hasil IIF = 94.99%, TIF = 94.89%. Kategori respon mahasiswa untuk self evaluation terhadap materi perkuliahan analisis real melalui pembelajaran online dengan θ = -1.55 sampai 1.89 yaitu kategori mudah, sedang dan sulit. Berdasarkan hasil analisis maka instrumen dapat digunakan untuk mengukur self evaluation. Penelitian ini berkonstribusi terhadap implementasi praktek penskalaan pada instrumen penelitian pengukuran data kuantitatif. Rating Scale Model: Student self-evaluation through online learning Abstract: This study aims to measure the self-evaluation of mathematics education students on real analysis lecture material through online learning. This research uses a quantitative approach with sampling based on certain considerations (purposive sampling), namely student of Mathematics Education Study Program, FKIP Pattimura University who take real analysis courses as many as 106 respondents. Data was obtained by distributing self-evaluation questionnaires through google form and in the form of a response questionnaire with the Likert scale method. Initial analysis was carried out to determine the validity and reliability of the instrument through the calculation of Exploratory Factor Analysis (EFA) to determine the construct validity of the instrument and Cronbach's Alpha to determine reliability. Furthermore, Item Response Theory (IRT) analysis of the polytomus Rating Scale Model (RSM) was carried out to measure student self-evaluation. The results of the EFA analysis using the R program resulted in 7 latent factors from 30 indicator items with a Kaiser-Meyer-Olkin (KMO) value of 0. 79 and MSA for each item showing results > 0.5, factor loading for each item > 0.3 and a reliability estimate of 0.877. The results of the RSM polynomial IRT calculation show a fit model for each item and the results obtained IIF = 94.99%, TIF = 94.89%. Student response categories for self-evaluation of real analysis lecture material through online learning with θ = -1.55 to 1.89 are easy, moderate and difficult categories. Based on the results of the analysis, the instrument can be used to measure self-evaluation. The research contributes to implementation of scaling practices in quantitative data measurement research instruments. ]]>
