Process Capability and Data Contamination
Purpose: The paper centres on process capability and its relation to data contamination. Process capability may be distorted due to imprecise data. The paper analyses to what extent capability changes reflect problems in data so that the changes can be attributed to data sampling rather than the t...
Main Authors: | , |
---|---|
Format: | Article |
Language: | English |
Published: |
Technical University of Kosice
2017-11-01
|
Series: | Kvalita Inovácia Prosperita |
Subjects: | |
Online Access: | https://www.qip-journal.eu/index.php/QIP/article/view/910 |
_version_ | 1797745532083372032 |
---|---|
author | Filip Tošenovský Josef Tošenovský |
author_facet | Filip Tošenovský Josef Tošenovský |
author_sort | Filip Tošenovský |
collection | DOAJ |
description |
Purpose: The paper centres on process capability and its relation to data contamination. Process capability may be distorted due to imprecise data. The paper analyses to what extent capability changes reflect problems in data so that the changes can be attributed to data sampling rather than the true performance of the process. This is important because it is usually much simpler to increase the precision of data sampling than the process itself.
Methodology/Approach: The paper has two major parts. In part one, effect of data contamination on the observed process characteristic is analysed. The effect is analysed using data obtained from simulated random drawings and the chi-squared test. In the other part, reaction of capability to data contamination is observed. The capability is measured by a univariate capability index.
Findings: Regarding the sensitivity of the index to contamination, it is different depending on the capability before the contamination. This leads to conclusions about when the company using the index should focus more on the way the data is measured, and when it should focus more on improving the process in question. The analysis shows that if the company is used to high levels of capability and records its drop, it is worth analysing its measurement system first, as the index is at higher levels more sensitive to data contamination.
Research Limitation/implication: The study concerns a single univariate index, and the contamination is modelled with only several probability distributions.
Originality/Value of paper: The findings are not difficult to detect, but are not known in practice where companies do not realize that problems with their process capability may sometimes lie in the data they use and not in the process itself.
|
first_indexed | 2024-03-12T15:25:41Z |
format | Article |
id | doaj.art-55f6f84a857b401fb3a3fb46580a5b8d |
institution | Directory Open Access Journal |
issn | 1335-1745 1338-984X |
language | English |
last_indexed | 2024-03-12T15:25:41Z |
publishDate | 2017-11-01 |
publisher | Technical University of Kosice |
record_format | Article |
series | Kvalita Inovácia Prosperita |
spelling | doaj.art-55f6f84a857b401fb3a3fb46580a5b8d2023-08-10T13:37:30ZengTechnical University of KosiceKvalita Inovácia Prosperita1335-17451338-984X2017-11-0121310.12776/qip.v21i3.910Process Capability and Data ContaminationFilip Tošenovský0Josef Tošenovský1Dept. of Quality Management VŠB-Technical University of OstravaDept. of Quality Management VŠB-Technical University of Ostrava Purpose: The paper centres on process capability and its relation to data contamination. Process capability may be distorted due to imprecise data. The paper analyses to what extent capability changes reflect problems in data so that the changes can be attributed to data sampling rather than the true performance of the process. This is important because it is usually much simpler to increase the precision of data sampling than the process itself. Methodology/Approach: The paper has two major parts. In part one, effect of data contamination on the observed process characteristic is analysed. The effect is analysed using data obtained from simulated random drawings and the chi-squared test. In the other part, reaction of capability to data contamination is observed. The capability is measured by a univariate capability index. Findings: Regarding the sensitivity of the index to contamination, it is different depending on the capability before the contamination. This leads to conclusions about when the company using the index should focus more on the way the data is measured, and when it should focus more on improving the process in question. The analysis shows that if the company is used to high levels of capability and records its drop, it is worth analysing its measurement system first, as the index is at higher levels more sensitive to data contamination. Research Limitation/implication: The study concerns a single univariate index, and the contamination is modelled with only several probability distributions. Originality/Value of paper: The findings are not difficult to detect, but are not known in practice where companies do not realize that problems with their process capability may sometimes lie in the data they use and not in the process itself. https://www.qip-journal.eu/index.php/QIP/article/view/910capability indexdata contaminationindex sensitivity |
spellingShingle | Filip Tošenovský Josef Tošenovský Process Capability and Data Contamination Kvalita Inovácia Prosperita capability index data contamination index sensitivity |
title | Process Capability and Data Contamination |
title_full | Process Capability and Data Contamination |
title_fullStr | Process Capability and Data Contamination |
title_full_unstemmed | Process Capability and Data Contamination |
title_short | Process Capability and Data Contamination |
title_sort | process capability and data contamination |
topic | capability index data contamination index sensitivity |
url | https://www.qip-journal.eu/index.php/QIP/article/view/910 |
work_keys_str_mv | AT filiptosenovsky processcapabilityanddatacontamination AT joseftosenovsky processcapabilityanddatacontamination |