REGRESSION ANALYSIS OF PRODUCTIVITY USING MIXED EFFECT MODEL
ABSTRACT: Production plants of
a company are located in several areas that spread across Middle and East Java.
As the production process employs mostly manpower, we suspected that each
location has different characteristics affecting the productivity. Thus,
the production data may have a spatial
and hierarchical structure. For fitting a linear regression using the ordinary
techniques, we are required to make some assumptions about the nature of the
residuals i.e. independent, identically and normally distributed. However, these
assumptions were rarely fulfilled especially for data that have a spatial and
hierarchical structure. We worked out the problem using mixed effect model.
This paper discusses the model construction of productivity and several
characteristics in the production line by taking location as a random effect.
The simple model with high utility that satisfies the necessary regression
assumptions was built using a free statistic software R version 2.6.1.
Author: Siana Halim, Indriati
N Bisono
Journal Code: jptindustrigg070008
Artikel Terkait :
Jp Teknik Industri gg 2007
- ROBUST-HYBRID GENETIC ALGORITHM FOR A FLOW-SHOP SCHEDULING PROBLEM (A Case Study at PT FSCM Manufacturing Indonesia)
- FUZZY MODELING APPROACH AND GLOBAL OPTIMIZATION FOR DUAL RESPONSE SURFACE
- A REPLENISHMENT POLICY FOR ITEM WITH PRICE DEPENDENT DEMAND AND DETERIORATING UNDER MARKDOWN POLICY
- SERVICE QUALITY CONTROL MEASUREMENT: QUALITOMETRO II METHOD
- DEALING WITH KANO MODEL DYNAMICS: STRENGTHENING THE QUALITY FUNCTION DEPLOYMENT AS A DESIGN FOR SIX SIGMA TOOL
- DESIGN FOR SIX SIGMA
- Running-in as an Engineering Optimization