How to Use Confidence Intervals in Selecting a Suitable Time-Dependent Distribution Model for the Process

Purpose: This paper proposes the possibility of using confidence intervals for the mean and variance in selecting a suitable time-dependent distribution model for the process. Methodology/Approach: The approach describes a characteristic under consideration by the distribution, the location, the dispersion, and the shape, all of which are functions of time. The values of the characteristics under consideration are determined by taking samples from the process flow. Time-dependent distribution models are classified into four groups based on whether the location and dispersion moments remain constant or vary over time. Findings: The paper explains a method for creating random samples, along with the calculation, presentation, and interpretation of the confidence intervals for the mean and variance in choosing the appropriate time-dependent distribution model of the process. Research Limitation/Implication: The methods described in this document pertain exclusively to continuous quality characteristics. They can be applied to analyse processes across various industrial and economic sectors. The presented procedures assume that all instantaneous distributions are normal. Originality/Value of paper: Confidence intervals can improve decision-making when selecting a suitable time-dependent distribution model.