Unavoidable conceptual assumptions for regression analysis

• Main Author: Daoud, Jamal Ibrahim
• Format: Article
• Language: English
• Published: IDOSI Publications 2013
• Subjects: HA Statistics
• Online Access: http://irep.iium.edu.my/30018/1/15_%281%29.pdf; http://irep.iium.edu.my/30018/

Statistical Analysis has become an indispensable tool for most of researches. Regression analysis is one of most widely known and used statistical tools for analysing multifactor data. Nowadays, it is hard to find statistical analysis without involvement of regression analysis. It is powerful tool and at the same time it is easy and clearly showing and describing the relationship between different variables associated in a certain relationship. One of the important issues when using the regression analysis, are the assumptions made on the model based on the sample drawn from a population. To manufacture a good device, the components of the device should be manufactured in such a perfect way so that it will yield the maximum satisfactory properties, as well as working in harmony with other components which maximizes the reliability of the device. Then the device can be produced in mass production. The issue with assumptions on regression model is similar to the device reliability. The sample based model has to be verified and then used to make inferences on the population from which the sample has been obtained. In this paper the general concepts will be demonstrated without going in details on computations and calculations.