Assignable causes of variation and statistical models: another approach to an old topic

Adler, Y.; Shper, V.; Maksimova, Olga V.

doi:10.1002/qre.1207

Cited by 6 publications

(10 citation statements)

References 8 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…More specific to the kind of data that we analysed is the issue that ‘technical fixes’ of over‐dispersion do not address the fundamental question of why over‐dispersion exists . Regarding control charts in general, the fundamental soundness of applying statistics in quality improvement in the traditional way is being challenged . Bearing all that in mind, we believe that it is of major benefit for wide adoption of a statistical method if – in addition to theoretical soundness and good performance in realistic simulations – it is simple and thus understandable to a wide audience.…”

Section: Discussionmentioning

confidence: 99%

“…58 Regarding control charts in general, the fundamental soundness of applying statistics in quality improvement in the traditional way is being challenged. 59,60 Bearing all that in mind, we believe that it is of major benefit for wide adoption of a statistical method ifin addition to theoretical soundness and good performance in realistic simulationsit is simple and thus understandable to a wide audience. A useful practical test of that is if the method can be implemented in an electronic spreadsheet only with formulas (i.e.…”

Section: Discussionmentioning

confidence: 99%

See 1 more Smart Citation

Outlier Detection for Healthcare Quality Monitoring – A Comparison of Four Approaches to Over‐Dispersed Proportions

Vidmar

Blagus

2013

Quality & Reliability Eng

View full text Add to dashboard Cite

bOutlier detection among over-dispersed proportions is important in healthcare quality monitoring. We had previously introduced control limits for double-square-root chart on the basis of prediction intervals from regression-through-origin and compared our approach to common outlier detection tests. In this study, we develop our approach further by adjusting the confidence level (in the spirit of Chauvenet's criterion and Bayesian thinking) and transforming the chart into an asymmetric funnel plot. We compare it to Laney's approach (p'-chart adapted for cross-sectional data), Spiegelhalter's approach (funnel plots based on multiplicative or additive regression models) and Carling's median rule. Comparisons are performed on simulated and real data. The simulations comprise 'small' (<0.2; highly right-skewed) and 'large' (>0.5; symmetrically distributed) proportions, drawn in samples of size 10-100 from lognormal distribution either without outliers or with one outlier added. The real data comprise hospital readmissions from the UK (used by Laney and Spiegelhalter) and business indicators of healthcare quality for Slovenian hospitals. In the simulations, Spiegelhalter's approach tended to yield very high false alarm rates, except the multiplicative version in very small samples. Laney's approach produced fewest false alarms but could not detect the outlier in very small samples among small proportions, and regardless of sample size among large proportions. Median rule performed similarly. Our approach performed the best overall, although it is slightly less liberal than median rule for small proportions, it appears to be the only generally useful approach for large proportions.

show abstract

Section: Discussionmentioning

confidence: 99%

Section: Discussionmentioning

confidence: 99%

Outlier Detection for Healthcare Quality Monitoring – A Comparison of Four Approaches to Over‐Dispersed Proportions

Vidmar

Blagus

2013

Quality & Reliability Eng

View full text Add to dashboard Cite

show abstract

“…Вот некоторые редкие примеры таких исследований. В 2011 г. была опубликована работа «Особые причины вариаций и статистические модели: новый подход к старой теме» [13]. Авторы (один из которых является соавтором данной статьи) предложили разделить особые причины вариаций на те, которые вызваны вмешательством, имеющим ту же самую функцию распределения (ФР), что и исходный процесс, и те, которые порождены иной ФР.…”

Section: Abstract: Shewhart Control Charts Non-normality Capability I...unclassified

“…Поэтому видится целесообразным ввести в рассмотрение различные типы особых (специальных) причин. В работе [13] авторы предложили ввести два типа особых причин вариаций. Слегка изменив формулировки работы [13], авторы предлагают следующие определения:…”

Section: различные типы нестабильности процессовunclassified

See 1 more Smart Citation

Shewhart control charts – A simple but not easy tool for data analysis

Shper,

Sheremetyeva,

Smelov

et al. 2024

Izv. vysš. učebn. zaved., Čern. metall.

View full text Add to dashboard Cite

Shewhart control charts (ShCCs) are a powerful and technically simple tool for process variability analysis. However, simultaneously, they cannot be fully algorithmized and require deep process knowledge together with additional data analysis. ShCCs are well known, though, and the number of papers is great, as well as standards on ShCCs work in most countries, there are some serious obstacles for their effective application which are not being discussed in either educational or scientific literature. Just these problems are being considered in this paper. We analyzed two sides of standard assumption about data normality. First, we discuss the widely-spread misconception that measurement data are always distributed according Gauss law. Then, it is shown how the deviation from normality may impact the method of ShCCs’ constructing and interpreting. Using a specific process data, we debate on right and wrong ways to build ShCC. Further, the paper describes two new definitions of assignable causes of variation: not changing (I-type) and changing (X-type) the system. At the end, we discuss how the work with ShCCs should be organized effectively. It is outlined that creating and analyzing ShCCs is always a system question of interaction between the process and the person who tries to improve this process.

show abstract

The importance of time order with Shewhart control charts

Shper

Adler

2017

Quality & Reliability Eng

View full text Add to dashboard Cite

In Phase I analysis, data are used retrospectively to check process stability and to establish limits that will be later used in Phase II for the prospective monitoring. It is clear that the results of Phase I analysis are of crucial importance for the right interpretation of control chart during Phase II. Many of recent papers devoted to Phase I analysis discuss the accuracy of limits found from Phase I data and the impact of these limits' accuracy on the probability of Type I or II errors. Although this is an interesting issue, we think that its importance is overestimated because there is another problem that escapes the attention of many researchers. This is the problem of the order of points in Phase I or the problem of data randomness. We review the role of data succession on right application of simple Shewhart control chart and hope to draw the attention of the statistical community to this problem because it may affect not only Phase I issues but also many other sides of statistical process control. Besides, we suggest a new test for data randomness and give a preliminary estimate of how it works. We also state our thoughts about the possible causes of neglecting the data order and suppose ideas about future researches.

show abstract

Assignable causes of variation and statistical models: another approach to an old topic

Cited by 6 publications

References 8 publications

Outlier Detection for Healthcare Quality Monitoring – A Comparison of Four Approaches to Over‐Dispersed Proportions

Outlier Detection for Healthcare Quality Monitoring – A Comparison of Four Approaches to Over‐Dispersed Proportions

Shewhart control charts – A simple but not easy tool for data analysis

The importance of time order with Shewhart control charts

Contact Info

Product

Resources

About