Change-Point Analyzer is a shareware software package for analyzing time ordered data to determine whether a change has taken place. It detects multiple changes and provides both confidence levels and confidence intervals for each change. The results are clearly displayed in table form and supplemented by easy to interpret plots. It can be used with all types of data: pass/fail, individual values, averages, ranges, standard deviations, normal, nonnormal and ill-behaved data like bioburden and particulate counts. It can be used to trend complaints, inventory turns, sales, particulate counts and just about anything else.
Change-Point Analyzer is simple to use, even for someone with no experience with control charting. It can be learned in about 15 minutes by following the tutorials provided below. It automatically verifies assumptions, checks for outliers and guides the user in handling such special cases. It even comes with a Excel Add-In making it possible to perform the analysis directly from Excel. For those in FDA regulated industries, it comes complete with a validation package.
Change-Point Analyzer is an important problem-solving tool which can be used to complement real-time control charts. Analyzing control charting data using Change-Point Analyzer will help to better isolate the time of a change, help to identify more subtle changes missed by the control chart, and expose false detections.
When performing a one-time analysis on historical data, a change-point analysis is preferred to control charting, especially when you are dealing with large datasets. When performing such analyses, a change-point analysis has numerous advantages over a control chart.
Change-Point Analyzer was developed by Dr. Wayne Taylor. It utilizes state of the art techniques including CUSUM charts and bootstrap analysis.
- Detects multiple changes
- For each change:
- Provides a confidence level that change occurred
- Estimates the time of the change and provides a corresponding confidence interval
- Handles all types of data: pass/fail, individual values, averages, ranges, standard deviations, normal, nonnormal, counts, and ill-behaved data like bio and particulate counts.
- More powerful than control charts at detecting sustained changes.
- Better characterizes changes including detection of multiple changes, providing associated confidence levels, and providing confidence intervals for the times of the changes.
- Reduces the number of false detections when dealing with large data-sets.
- Robust to outliers
- Flexible: The same procedure works for all types of data.
- Simpler to use and easier to interpret
- Change-Point Analyzer received ZDNet’s four-star rating.
Applications of change-point analysis include trending of process and quality data, problem solving by identifying the start of the problem, health and medicine, environment, climate change, fraud detection, defense and security, infectious diseases, purchasing transactions and much more. Below are articles using change-point analysis from a few of our customers.
Health and Medicine
- The Impact of a Celebrity Promotional Campaign on the use of Colon Cancer Screening: The Katie Couric Effect – Peter Cram, A. Mark Fendrick, John Inadomi, Mark E. Cowen, Daniel Carpenter, Sandeep Vijan – Archives of Internal Medicine (July 14, 2003), Vol. 163, 1601-1605. A change-point analysis was performed of colon screenings with the primary change-point corresponding to Katie Couric’s March 2000 Today Show colorectal cancer awareness campaign on colonoscopy rates.
- A Case for Revisiting Peer Review: Implications for professional self-regulation and quality improvement – Terry E. Hill , Peter F. Martelli, Julie H. Kuo – Plos One – Quality improvement in healthcare has often been promoted as different from and more valuable than peer review and other professional self-regulation processes. Our objectives were to (1) evaluate the relative contributions of professional accountability and quality improvement interventions to an observed decrease in population mortality and (2) explore the organizational dynamics that potentiated positive outcomes.
- Movement-Related Changes in Synchronization in Human Basal Ganglia – Michael Cassidy, Paolo Mazzone, Antonio Oliviero, Angelo Insola, Pietro Tonali, Vincenzo Di Lazzaro and Peter Brown – Brain (2002 June), 125 (Pt 6), 1235-46. Brain is a prestigious neuroscience journal. A change-point analysis was used to analyze the reaction of the brain track to certain stimulus allowing comparisons between treated and untreated patients with Parkinson’s disease.
- 300-Hz Subthalamic Oscillations in Parkinson’s Disease – G. Foffani, A. Priori, M. Egidi, P. Rampini, F. Tamma, E. Caputo, K. A. Moxon, S. Cerutti and S. Barbieri – Brain (2003 August 21), 126, 2153-2163. Brain is a prestigious neuroscience journal. A change-point analysis was used to analyze changes in the 300-Hz subthalamic oscillations in response to a motion stimulus so that the resulting changes could be compared before and after the administration of levodopa.
- Was King John of England bipolar?: a medical history using mathematical modeling – Gillespie, Janet, Ph.D. Thesis, University of St Andrews. Using primary historical sources and published analyses, bipolar symptoms were identified and their temporal relationship to the ICD-10 compliant CPA periods evaluated. The pattern of changes in King John’s activity is highly suggestive of bipolar disorder with primary historical sources describing synchronous bipolar behavior. Change Point Analysis merits greater consideration when analyzing time-based data, as does the use of activity as an objective marker of human behavior.
- Technical Proficiency in Hand-Assisted Laparoscopic Colon and Rectal Surgery – Determining How Many Cases Are Required to Achieve Mastery, Rajesh Pendlimari, MBBS; Stefan D. Holubar, MD; Eric J. Dozois, MD; David W. Larson, MD; John H. Pemberton, MD; Robert R. Cima, MD, MA. Change-Point Analysis (CUSUM) was used to define the number of cases required to effect improvement in operative time. Cases before and after the change point were considered as being in the “learning period” and “skilled period.”
- A quasi-experimental test of an intervention to increase the use of thiazide-based treatment regimens for people with hypertension, Carol M Ashton, Myrna M Khan, Michael L Johnson, Annette Walder, Elizabeth Stanberry, Rebecca J Beyth, Tracie C Collins, Howard S Gordon, Paul Haidet, Barbara Kimmel, Anna Kolpakchi, Lee B Lu, Aanand D Naik, Laura A Petersen, Hardeep Singh and Nelda P Wray. Statistical process control charts, change point analyses, and before-after analyses were used to estimate the intervention’s effects.
- Estimation of the distribution of change-points with application to fMRI data, John A D Aston and Claudia Kirchz. Using the developed methods, a large study of resting-state fMRI data is conducted to determine whether the subjects undertaking the resting scan have
non-stationarities present in their time courses.
Process and Quality Data
- Variability Detection by Change-Point Analysis, Seo-Won Chang and Yong-Ik Byun, Department of Astronomy, Yonsei University, Seoul, Korea. We confirm that the CPA is a powerful method to determine whether a change has taken place in time series dataset. From our re-analysis of MMT transit survey data, we found previously unknown evidence about stellar variability, including a total of 606 flare events, 18 eclipsing-like features, and 3 transit-like features.
Environment and Climate Change
- An event-state depiction algorithm using CPA methods with continuous feed data, Philip Sallis and Sergio Hernández. Wind velocity data is sampled in real-time. The continuous data feed is processed by a change-point analysis algorithm, which has been derived for this purpose to predict damaging wind gusts.
- Urban Wind Speed Analysis in Global Climate Change Perspective: Karachi as a Case Study, Muhammad A. Hussain, Muhammad J. Iqbal, Safeeullah Soomro It is now well known that coastal urban local climate has been showing changing pattern due to global climate change. This communication attempts to explore the fluctuating pattern of urban average monthly wind speed during the past 50 years (1961-2010). Change-point analysis has confirmed the change in the pattern of observed average monthly wind speed data near 1992.
- Observed Abrupt Changes in Minimum and Maximum Temperatures in Jordan in the 20th Century, Mahmoud M. Smadi. Major change points in the mean minimum (night-time) and mean maximum (day-time) temperatures occurred in 1957 and 1967,
- Abrupt changes in air temperature and precipitation – do they matter for organic matter? Temnerud, Johan and Weyhenmeyer, Gesa (2007). For air temperature, we found significant changes in 1930 and 1989 and for precipitation in 1920, 1979, and 1998. We conclude that increases in air temperatures can make ecosystems more sensitive to further changes in precipitation.
- Temperature and Precipitation Changes in Târgu-Mures (Romania) from Period 1951-2010, O. Rusz. CUSUM charts indicate occurs of changes points at 1988, 2005, 2009 (mean temperature) respectively at 1989, 2004 (precipitation), and at 1968, 1992 (daily temperature range).
- Climatic controls on phytoplankton biomass in a sub-tropical estuary, Florida Bay, USA., Henry O. Briceño and Joseph N. Boyer. We performed statistical analysis and subdivided the bay into six zones having unique biogeochemical characteristics. Significant shifts in the drivers were identified in all the chlorophyll a time-series. Chlorophyll a concentrations closely follow global forcing and display a generalized declining trend on which seasonal oscillations are superimposed, and it is only interrupted by events of sudden increase triggered by storms which are followed by a relatively rapid return to pre-event conditions trailing again the long-term trend.
- Do phytoplankton communities evolve through a self-regulatory abundance–diversity relationship? Shovonlal Roy. A change-point analysis demonstrates that the diversity of non-toxic phytoplankton increases with the increase of toxic species up to a certain level. However, for a massive increase of the toxin-producing species the diversity of phytoplankton at species level reduces gradually.
- Evaluation of a Denitrification Wall to Reduce Surface Water Nitrogen Loads, Casey A. Schmidt * and Mark W. Clark. A change point analysis was performed on N concentration and load to determine statistically significant changes in these values at an α level of 0.05 with the Change-Point Analyzer software (Taylor Enterprises, Inc.).
- Analysis of Annual Sums of Precipitation in Serbia, Gorica Stanojević . This paper analyses the series of annual sums of precipitation for 27 stations in Serbia for the period 1951-2010. The advantage of applying the method of cumulative sums and “bootstrapping” in the analysis of time series in relation to the method of linear trend has been emphasized.