Claim Missing Document
Check
Articles

Found 3 Documents
Search

Kualitas butir bank soal statistika (Studi kasus: Instrumen ujian akhir mata kuliah statistika Universitas Terbuka) Agus Santoso; Kartianom Kartianom; Gulzhaina K. Kassymova
Jurnal Riset Pendidikan Matematika Vol 6, No 2: November 2019
Publisher : Program Studi Pendidikan Matematika Program Pascasarjan Universitas Negeri Yogyakarta

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.21831/jrpm.v6i2.28900

Abstract

Penelitian ini bertujuan untuk mendeksripsikan kualitas butir soal ujian akhir semester mata kuliah statistika ekonomi yang dikembangkan oleh Universitas Terbuka (UT) sebagai dasar dalam mengembangkan bank soal yang terkalibrasi menggunakan pendekatan Teori Respons Butir. Penelitian ini merupakan penelitian deskriptif kuantitatif. Sumber data penelitian ini adalah pola jawaban mahasiswa UT yang telah mengikuti ujian akhir semester (UAS) mata kuliah statistika ekonomi selama enam masa ujian, dengan ukuran sampel sebanyak 23334 mahasiswa. Hasil penelitian ini menunjukkan bahwa butir-butir soal ujian akhir semester mata kuliah statistika ekonomi yang dikembangkan UT: (1) terbukti valid secara konstruk, yakni hanya mengukur satu faktor dominan, yaitu kemampuan statistika ekonomi; (2) memiliki kehandalan yang baik dengan nilai koefisien reliabilitas empiris lebih dari 0,70 (koefisien reliabilitas empiris = 0,7335); (3) dari 140 butir soal yang dikalibrasi terdapat 108 butir soal (25 butir soal berkualitas baik atau tanpa revisi dan 83 butir soal berkualitas kurang baik atau perlu revisi) yang layak disimpan dalam bank soal, sedangkan 32 butir soal berkualitas tidak baik; dan (4) mampu memberikan informasi akurat terkait kemampuan statistika ekonomi mahasiswa pada level kemampuan yang tinggi (-1,3 sampai +4,0). Quality of statistical test bank items (Case study: Final exam instrument of statistics courses in Universitas Terbuka) AbstractThis study aims to determine the quality of final semester test items of economic statistics course that was developed by Universitas Terbuka (UT) as a basis for developing calibrated item banks using Item Response Theory. This research uses a quantitative descriptive approach. The researcher investigates the answer pattern of the final semester exam (UAS) in the economic statistics course during six periods of the final exams. The sample size in this study was 23334 students. The results of this study indicate that the final semester exam items of economic statistics courses developed by UT: (1) proved to construct valid, i.e. only measure one dominant factor, namely the ability of economic statistics; (2) has good reliability with empirical reliability coefficient values more than 0.70 (empirical reliability coefficient = 0.7335); (3) of the 140 items calibrated there are 108 items (25 items of good quality or without revision and 83 items of poor quality or need to be revised) that are worth keeping in the question bank, while 32 items of quality are not good; and (4) able to provide accurate information related to students' economic statistical abilities at a high level of ability (-1.3 to +4.0)
Prediction model of teacher candidate student graduation status: Decision Tree C4.5, Naive Bayes, and k-NN Kartianom Kartianom; Arpandi Arpandi; Gulzhaina K. Kassymova; Oscar Ndayizeye
Ekspose: Jurnal Penelitian Hukum dan Pendidikan Vol 21, No 2 (2022)
Publisher : Institut Agama Islam Negeri (IAIN) Bone

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.30863/ekspose.v21i2.3407

Abstract

This study aims to determine the prediction model of the graduation status of prospective teacher students at IAIN Bone in terms of attributes, accuracy levels, and differences in the level of accuracy produced in the attributes of decision tree C4.5, Naïve Bayes, and k-NN data mining algorithms. This research uses a quantitative approach by adopting the Data Mining method. This research was conducted at IAIN Bone. The data collection process in this study used documentation techniques in the form of data on alumni of the Tarbiyah Faculty of IAIN Bone. The data analysis used was a descriptive analysis using decision tree C4.5, Naive Bayes, and k-NN data mining algorithms assisted by the RapidMiner application.  The results of this study show that (1) model prediction of the graduation status of prospective teacher students in IAIN Bone in terms of attributes generated in the Decision Tree C4.5 and Naïve Bayes data mining algorithms  consist of gender, age, Semester 1 IP, Semester 2 IP, Semester 3 IP, Semester 4 IP, and GPA, while the attributes produced in   k-NN data mining algorithm  consists of gender, regional origin, number of siblings, age, IP Semester 1, IP Semester 2, IP Semester 3, IP Semester 4, and GPA; (2) model prediction of graduation status of iain bone teacher candidate students in terms of the accuracy rate generated in the Decision Tree C4.5 data mining algorithm  of 93.90%, Naïve Bayes by 90.24%, and k-NN of 92.07%; and (3) there was no significant difference between the accuracy rate produced by decision tree's data mining algorithm.  C4.5 and Naïve Bayes (p-value = 1.00); Decision Tree C4.5 and k-NN (p-value = 1.00); as well as Naïve Bayes and k-NN (p-value = 1.00) in predicting the graduation status of iain bone teacher candidate students.
Blended learning as an effective method for school and university teachers of Kazakhstan Ainur Kaliaskarova; Turakhan Zhundybaeva; Mochamad Bruri Triyono; Gulzhaina K. Kassymova
Jurnal Pendidikan Vokasi Vol 12, No 3 (2022)
Publisher : ADGVI & Graduate School of Universitas Negeri Yogyakarta

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.21831/jpv.v12i3.48115

Abstract

In this article, the author compares different definitions of the term blended learning given by foreign and local scientists and gives his interpretation. The article analyzes world scholars' works on blended learning, determining the advantages and disadvantages of this method. The online survey method was used to find out how far Kazakhstani teachers are familiar with the blended learning method, whether they use it or not, and understand if it is effective. The survey results showed that school and university teachers use blended learning in their lessons and will use it in the future. According to the questionnaire, the benefits of this method are that it encourages teachers to be creative, increases students' interest, and helps them to study independently. Significant shortcomings of blended learning are technical problems and poor Internet connection. Also, teachers need more time to prepare for their lessons. It is challenging for them. Although there are disadvantages to blended learning, the benefits outweigh the disadvantages. Therefore, today this method is effective.