Pengontrolan Vektor Rata-Rata dengan Menggunakan R-Chart
Abstract
Abstract. One of the factors that determine customer satisfaction is customer perception of service quality. One of the control charts that do not require assumptions from the data distribution of a process is the Run-Control Chart (abbreviated as R-Chart) so that this control chart is included in the nonparametric control chart. This research will discuss the use of R-Chart based on Mahalanobis depth data for the control process that involves multivariate quality measurement. This control chart will be used to control the average vector of the four quality characteristics as a form of quality improvement efforts at the DKI Jakarta BPS library. Observation data measuring 180 with five service variables used as historical data. Process data (actual) generated as many as 100 pieces with the first 50 observations representing the in control process and the next 50 observations representing the out of control process where only the mean of the first variable (X1) was shifted by 1.5 sigma. The conclusion in this study is that it can be said that the R-chart with CL = 0.5 and LCL = 0.00273 gives an out of control signal because the R value for data no 77 and no 99 is smaller than LCL.
Keywords: Control Chart, Nonparametric, Mahalanobis data depth, R-Chart.
Abstrak. Salah satu faktor yang menentukan kepuasan pelanggan adalah persepsi pelanggan mengenai kualitas jasa pelayanan. Diagram kontrol yang tidak memerlukan asumsi dari distribusi data suatu proses salah satunya adalah Run Control Chart (disingkat R-Chart) sehingga diagram kontrol ini termasuk pada diagram kontrol nonparametrik. Dalam penelitian ini akan dibahas penggunaan R-Chart berdasarkan data depth Mahalanobis untuk proses kontrol yang melibatkan pengukuran kualitas multivariat. Diagram kontrol ini akan digunakan untuk mengontrol vektor rata-rata keempat karakteristik mutu sebagai wujud upaya perbaikan mutu di perpustakaan BPS DKI Jakarta. Data pengamatan berukuran 180 dengan empat variabel jasa pelayanan dijadikan data historis. Data proses (aktual) dibangkitkan sebanyak 100 buah dengan 50 pengamatan pertama representasi dari proses in control dan 50 pengamatan berikutnya merepresentasikan dari proses out of control dimana hanya rata-rata variabel pertama (X1) yang bergeser sebesar 1.5 sigma.  Kesimpulan pada penelitian ini adalah dapat dikatakan bahwa R-chart dengan GP = 0.5 dan BKB = 0.00273 memberikan sinyal out of kontrol karena nilai R untuk data no 77 dan no 99 nilainya lebih kecil dar BKB.
Kata Kunci: Control Chart, Nonparametric, Mahalanobis data depth, R-Chart.
Keywords
Full Text:
PDFReferences
Andriana, D. S. (2007). Peta Kendali Nonparametric Berdasarkan Jumlah Ranks.
Hamurkaroglu, C., Mert, M., & Saykan, Y. (2004). Nonparametric Control Charts Based On Mahalanobis Depth. Hacettepe Jurnal of Mathematics and Statistics, 57-67.
Hussain, I (2020). Outlier Detection using Nonparametric Depth-Based Techniques in Hydrology International Journal of Advanced Computer Science and Applications, Vol. 11, No. 9, 2020 www.ijacsa.thesai.org
Liu, R. (1990). On a notion of data depth based on random simplices. The Annals of Statistics, 405-414.
Mongomery, D. C. (2001). Introduction to statistical Quality Control 4th Edition. New York: Jon Wiley & Sons.
Montgomery, D. C. (2009). Introduction to Statistical Quality Control Sixth Edition. New York: John Wiley & Sons.
Rahmadina, R. P., Ayu, R., Rizky, F., & Rahayu, S. P. (2019). Deteksi Missing Value, Deteksi Outlier, Uji Normal Multivariat, Uji Bartlett, dan Uji T2 Hotelling pada Data Pengaruh Kadar Enzim Alanine Transaminase, Enzim Gamma Glutamyl Transferase, dan Molekul Creatinine terhadap Ketahanan Hidup Pasien Kanker Hepatoc. Analisis Multivariate.
Salam, M. I. (2009). Penentuan Batas-Batas Bagan Kontrol Dispersi Multivariat VV dengan Metode Bootstrap. Tesis S2 Statistika Pasca Sarjana UNPAD.
DOI: http://dx.doi.org/10.29313/.v7i1.26316
