Modeling and Evaluating

Regression Analysis Report

Introduction

This report is first going to construct the best model to track maintenance cost of the existing production line for Aberdour Foods plc together with examining the structure of the maintenance cost. Then the assumptions and limitations of the selected model will be discussed respectively. Last, we will have a glance at the related guidance for potential use of the machines by future operation.

Model evaluation

Simple plotting

The two scatter plots above indicate a positive linear relationship between maintenance costs and machine hours, as well as Bulk bins. To compare, the machine hours seems to have a slightly stronger linear relationship with maintenance costs. This result suggests that the machine hours might be the better predictor.

Correlation matrix

Maintenance Costs Machine Hours Bulk Bins

Maintenance Costs 1

Machine Hours 0.865551 1

Bulk Bins 0.815458 0.908056 1

The correlation analysis suggests that the machine hours has a stronger relationship with the dependent variance compare to the other independent variance which again proved to be the better estimator.

Simple regression

Regression Analysis: Maintenance Costs(￡0） versus Machine Hours

The regression equation is

Maintenance Costs(￡0） = 68.2 + 0.278 Machine Hours

Predictor Coef SE Coef T P

Constant 68.16 12.19 5.59 0.000

Machine Hours 0.27849 0.02763 10.08 0.000

S = 8.42900 R-Sq = 74.9% R-Sq(adj) = 74.2%

Analysis of Variance

Source DF SS MS F P

Regression 1 7215.3 7215.3 101.55 0.000

Regression Statistics

Multiple R 0.865551

R Square 0.749178

Adjusted R Square 0.741801

Standard Error 8.429005

Observations 36

Looking into the Analysis of Variance, the P-value 0.000 (< 0.005) makes the whole model...