Claim Missing Document
Check
Articles

Perbandingan Exponential Smoothing With Trend dan Linear Trend Line pada Peramalan Penjualan Topsin 500 SC Hardari, Wafika Putri; Widhiarso, Wahyu
Journal of Industrial and Mechanical Engineering Vol 3 No 2 (2025): Journal of Industrial and Mechanical Engineering
Publisher : Department of Industrial Engineering, Universitas Jenderal Soedirman.

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.20884/1.jimien.2025.3.2.16990

Abstract

Suatu perusahaan dalam kondisi ideal diharapkan dapat meramalkan penjualan dengan akurasi tinggi sehingga produksi, persediaan, dan distribusi dapat dikelola secara optimal. Hal ini akan membantu PT. XYZ menjaga ketersediaan produk, meminimalkan risiko kelebihan atau kekurangan stok, dan menghindari biaya tambahan akibat ketidakseimbangan antara permintaan dan persediaan. Namun kenyataannya, metode peramalan penjualan yang digunakan PT. XYZ masih bersifat manual dan belum mampu menangkap perubahan permintaan secara akurat dalam menghadapi fluktuasi tajam pada penjualan Topsin 500 SC. Ketidakakuratan dalam peramalan tersebut menyebabkan perusahaan menghadapi risiko overstock. Penelitian ini bertujuan untuk mengevaluasi hasil peramalan dari dua metode peramalan yang lebih efektif dalam meramalkan penjualan Topsin 500 SC dan memverifikasi akurasi metode peramalan terpilih. Penelitian ini menggunakan metode Exponential Smoothing with Trend dan Linear Trend Line. Peta tracking signal digunakan untuk memverifikasi metode peramalan terbaik yang telah dipilih. Berdasarkan hasil perhitungan menunjukkan MAD, MSE, dan MAPE terkecil berada pada metode Linear Trend Line yang memiliki MAD sebesar 1.529,314, nilai MSE sebesar 5.743.558 dan nilai MAPE sebesar 10,956%. Metode Linear Trend Line terverifikasi menjadi metode yang paling efektif dan memiliki tingkat akurasi peramalan di masa mendatang
JIKA RESPONDEN MENIPU, APAKAH HASIL PENGUKURAN OLEH SKALA MASIH ANDAL? Widhiarso, Wahyu
MANASA Vol. 1 No. 1 (2012): Juni, 2012
Publisher : Atma Jaya Catholic University of Indonesia

Show Abstract | Download Original | Original Source | Check in Google Scholar

Abstract

The aim of this study is to explore the impact of faking responses on measurement reliability of psychological scales. Experiments were conducted on 200 subjects who were asked to complete Five Factors Personality Scale (BFI-44) with two different instructions. First instruction asked subjects to respond within a neutral condition, whereas second instruction asked subjects to respond to the scale when they are participating in employment selection (motivated conditions). Comparisons of alpha coefficient between these conditions suggest that faking responses does not interfere measurement reliability. Pure score proportion on observed score was unaffected by presence of faking responses.However, using the state-trait model, it is found that untainted score proportion on observed scores are more related to state score compared to the trait score as main measurement target. Results of separated analysis model toward personality factors shown that Agreeableness and Openness are factors which were relatively resistant to faking responses. Trait score proportion of these factors still large although faking responses were presence.