Descriptive, Diagnostic, dan Clustering Analytics Stres Kerja Karyawan Pada Perusahaan EPC Menggunakan Metode CRISP-DM Untuk Rekomendasi Pengendalian Manajemen Stres Kerja

Hakim, Emral (2024) Descriptive, Diagnostic, dan Clustering Analytics Stres Kerja Karyawan Pada Perusahaan EPC Menggunakan Metode CRISP-DM Untuk Rekomendasi Pengendalian Manajemen Stres Kerja. Masters thesis, Institut Teknologi Sepuluh Nopember.

[thumbnail of 6032212013-Master_Thesis.pdf] Text
6032212013-Master_Thesis.pdf - Accepted Version
Restricted to Repository staff only until 1 April 2026.

Download (8MB) | Request a copy

Abstract

Organisasi yang berkomitmen mempertahankan Sertifikat International Organization for Standardization (ISO) 14001:2018 ataupun Sistem Keselamatan dan Kesehatan Kerja (SMK3) harus melakukan internal audit sebagai langkah identifikasi dan kontrol manajemen risiko. Salah satu internal audit yang dilakukan berhasil menemukan risiko kerja di area Human Resources Department (HRD), yaitu lonjakan jumlah pengunduran diri karyawan akibat stres kerja dan sudah mendominasi sejak kuartal ke-4 2021 - kuartal ke-1 2023. Hal ini mengakibatkan produktivitas perusahan menurun dan sulit untuk memenuhi target perusahaan. Pengukuran dan evaluasi tingkat stres kerja didapat melalui pengisian kuesioner online dengan acuan Peraturan Menteri Ketenagakerjaan Republik Indonesia Nomor 5 Tahun 2018 Tentang Keselamatan dan Kesehatan Kerja Lingkungan Kerja kepada seluruh karyawan dibawah direksi dengan 242 responden yang mengisi. Data yang diperoleh memiliki tiga tingkatan stres kerja yaitu stres kerja ringan, sedang, dan berat. Selanjutnya eksplorasi data stres kerja melalui analisis deskriptif dan diagnostik mengikuti pola Cross Industry Standard Process for Data Mining (CRISP-DM) dengan model clustering k-modes. Jumlah optimal cluster diukur menggunakan teknik elbow. Pada tahap deployment yaitu merancang rekomendasi pengendalian stres kerja karyawan untuk setiap cluster dengan pendekatan Socialization, Externalization, Combination, Internalization (SECI). Analisis deskriptif menemukan dua faktor teratas yang mempengaruhi tingkat stres kerja di EPC yaitu tuntutan kualitas pekerjaan dan ikut bertanggungjawab terhadap hasil kerja orang lain. Lebih spesifiknya, tingkat stres kerja karena karyawan dituntut melebihi kemampuan yang dimiliki namun pada saat yang bersamaan harus membantu orang lain menyelesaikan permasalahan. Selain itu, berdasarkan analisis Cluster ditemukan 2 segmen optimal. Segmen 1 memiliki centroid stres sedang untuk keseluruhan faktor stres kerja. Segmen-2 memiliki centroid stres ringan untuk keseluruhan faktor stres kerja. Rekomendasi pengendalian stres kerja segmen-1 adalah mempersiapkan program peningkatan kompetensi karyawan dan memperbaiki sistem pengukuran kinerja. Pengendalian stres kerja segmen-2 pemantauan tahunan melalui survei stres kerja karyawan.
=================================================================================================================================
Organizations that are committed to maintaining the International Organization for Standardization (ISO) 14001:2018 Certificate or Occupational Safety and Health System (SMK3) must carry out internal audits as a risk management identification and control step. One of the internal audits carried out succeeded in finding work risks in the Human Resources Department (HRD) area, namely a spike in the number of employee resignations due to work stress and has dominated since the 4th quarter of 2021 - the 1st quarter of 2023. This has resulted in company productivity decreasing and it is difficult to meet company targets. Measurement and evaluation of work stress levels was obtained by filling out an online questionnaire with reference to Regulation of the Minister of Manpower of the Republic of Indonesia Number 5 of 2018 concerning Occupational Safety and Health in the Work Environment for all employees under the directors with 242 respondents filling in. The data obtained has three levels of work stress, namely light, medium and heavy work stress. Next, explore work stress data through descriptive and diagnostic analysis following the Cross Industry Standard Process for Data Mining (CRISP-DM) pattern with the k-modes clustering model. The optimal number of clusters is measured using the elbow technique. At the deployment stage, namely designing recommendations for controlling employee work stress for each cluster using the Socialization, Externalization, Combination, Internalization (SECI) approach. Descriptive analysis found the top two factors that influence the level of work stress in EPC, namely demands for work quality and taking responsibility for other people's work results. More specifically, the level of work stress is because employees are required to exceed their capabilities but at the same time have to help other people solve problems. Apart from that, based on Cluster analysis, 2 optimal segments were found. Segment 1 has a moderate stress centroid for overall work stress factors. Segment-2 has a light stress centroid for the overall work stress factor. The recommendation for controlling work stress in segment-1 is to prepare a program to increase employee competency and improve the performance measurement system. Controlling work stress segment-2 annual monitoring through employee work stress surveys.

Item Type: Thesis (Masters)
Uncontrolled Keywords: Analisis Klaster, Keselamatan dan Kesehatan Kerja (K3), K-Modes, Pengendalian Stres Kerja, Clustering Analytic, Occupational Safety and Health, Work Stress Control
Subjects: H Social Sciences > HA Statistics > HA29 Theory and method of social science statistics
H Social Sciences > HD Industries. Land use. Labor > HD30.2 Knowledge management.
H Social Sciences > HD Industries. Land use. Labor > HD61 Risk Management
Divisions: Interdisciplinary School of Management and Technology (SIMT) > 61101-Master of Technology Management (MMT)
Depositing User: Emral Hakim
Date Deposited: 06 Feb 2024 06:55
Last Modified: 06 Feb 2024 06:55
URI: http://repository.its.ac.id/id/eprint/106287

Actions (login required)

View Item View Item