Pengaruh Panjang Tes dan Ukuran Contoh pada Pendugaan Parameter Model IRT

Siti Hasanah; Dewi Zulyani Pomalingo; Anang Kurnia; Alona Dwinata

doi:10.30998/sap.v8i3.19663

Pengaruh Panjang Tes dan Ukuran Contoh pada Pendugaan Parameter Model IRT

Siti Hasanah^(1*), Dewi Zulyani Pomalingo⁽²⁾, Anang Kurnia⁽³⁾, Alona Dwinata⁽⁴⁾

(1)
(2) Statistika dan Sains Data, IPB University
(3) Statistika dan Sains Data, IPB University
(4) Pendidikan Matematika, Universitas Maritim Raja Ali Haji
(*) Corresponding Author

Abstract

IRT (Item Response Theory) can explain the relationship between a person's response to an item and a latent variable (ability). The number of items or exercise questions given by an elementary to high school teacher is usually not very large. The same goes for the number of students, who are usually divided into different classes, ranging from 30 to 35 students per class. This study was conducted in June 2023 and aims to determine the influence of sample size and test length on the estimation of parameters in the item response theory (IRT) 1 PL model with small-sized data. The parameters analyzed through the IRT 1 PL model include item parameters, namely difficulty, and ability parameters. The data were obtained from generation of simulation results. Data generation simulations with several scenarios were used to construct the IRT 1 PL model, based on test lengths (10, 15, 20) and sample sizes (32, 64). The research results are indicated by the closeness of the estimated parameter ? (ability) to the true parameter, which increases as the test length increases. When the test length is considered constant, the closeness of the estimated difficulty parameter increases with an increase in the sample size in small-scale scenarios.

Keywords

Item Response Theory, IRT 1 PL, Sample Size, Test Length, Logistic Parameter

Full Text:

PDF (Indonesian)

References

B. Sumintono, “Model Rasch untuk Penelitian Sosial Kuantitatif,” in Kuliah umum di Jurusan Statistika, Surabaya, 2014. [Online]. Available: https://core.ac.uk/display/162013734?utm_source=pdf&utm_medium=banner&utm_campaign=pdf-decoration-v1

N. A. Thompson, “Ability Estimation with Item Response Theory,” Assesment Syst. Corp., pp. 1–11, 2009, [Online]. Available: www.assess.com

R. Budiarti, “Pengaruh Sample Size (N) dan Test Length (n) terhadap Item Parameter Estimate dan Examinee Parameter Estimate, suatu Studi Simulas,” J. Math. Its Appl., vol. 12, no. 1, pp. 25–36, Jul. 2013, doi: 10.29244/jmap.12.1.25-36.

M. Tarigan and F. Fadillah, “Analisis Item Response Theory Raven’s Coloured Progressive Matrices pada Sampel Anak Usia Dini,” PSIKODIMENSIA, vol. 20, no. 2, pp. 158–169, Dec. 2021, doi: 10.24167/psidim.v20i2.3101.

S. DS, “Pengaruh Panjang Tes dan Ukuran Sampel terhadap Kekekaran Estimasi Parameter pada Teori Respon Butir (Item Response Theory),” Cakrawala Pendidik., vol. 3, 2006, doi: 10.21831/cp.v1i3.8593.

F. Fajrianthi, W. Hendriani, dan B. G. Septarini, “Pengembangan Tes Berpikir Kritis dengan dengan Pendekatan Item Response Theory,” J. Penelit. dan Eval. Pendidik., vol. 20, no. 1, pp. 45–55, Jun. 2016, doi: 10.21831/pep.v20i1.6304.

M. C. Edwards, “An Introduction to Item Response Theory Using the Need for Cognition Scale,” Soc. Personal. Psychol. Compass, vol. 3, no. 4, pp. 507–529, 2009, doi: 10.1111/j.1751-9004.2009.00194.x.

Hambleton, Swaminathan, and Rogres, Item Response Theory. Boston MA: Kluwer Inc, 1985.

I. Falani dan S. A. Kumala, “Kestabilan Estimasi Parameter Kemampuan pada Model Logistik Item Response Theory Ditinjau dari Panjang Tes,” SAP (Susunan Artikel Pendidikan), vol. 2, no. 2, Dec. 2017, doi: 10.30998/sap.v2i2.2028.

S. A. Purba, “Estimasi Parameter Data Berdistribusi Normal Menggunakan Maksimum Likelihood Berdasarkan Newton Raphson,” J. Sains Dasar, vol. 9, no. 1, pp. 16–18, Feb. 2021, doi: 10.21831/jsd.v9i1.38564.

Rahayu, W., “Metode Estimasi Parameter dan Meter Equating pada Ukuran Sampel Kecil Berdasarkan Item Respons Theory,” Prosiding Semirata2015, pp. 315–324.

DOI: http://dx.doi.org/10.30998/sap.v8i3.19663