ASQ

Exclusive Content & Downloads from ASQ

Building Empirical Models for Data from Factorial Designs with Time Series Responses Toward Fraud Prevention and Detection

Summary: [This abstract is based on the author's abstract.]

Techniques for empirical model building and analysis for data arising from a factorial experiment with time-series responses in each cell is presented. Two graphical techniques are used: plotting contrast series and plotting the cumulative sums of contrast series. The techniques are demonstrated on the experimental data by Matsumura and Tucker. They examine the joint effects on fraud committed and fraud detected of four experimental factors. The two graphical techniques suggest that the four factors might affect both the mean and trend over time of the response series. Therefore, an empirical model that formally characterizes and summarizes these effects using a two-stage regression procedure is developed. Selecting suitable representation of the regression lime allows the multiresponse experiment to be treated as two single-response experiments. This facilitates analysis and interpretation. The empirical model and trendanalysis provided the following findings: (1) interactions among the four factors are important and care must be taken when interpreting the effect of a factor in isolation; (2) factors affect not only the mean responses but also changes in strategies over time. Including both a mean and trend component in the empirical model allows for improved identification of the cause and nature of a factor's effect.

Anyone with a subscription, including Site and Enterprise members, can access this article.


Other Ways to Access content:

Join ASQ

Join ASQ as a Full member. Enjoy all the ASQ member benefits including access to many online articles.

  • Topics: Design of Experiments, Data Quality
  • Keywords: Multiresponse experiment,Time series,Design of experiments (DOE),Data analysis,Factorial experiments,Cumulative sum control chart (CUSUM)
  • Author: Hau, Ian; Matsumura, Ella Mae; Tucker, Robert R.
  • Journal: Quality Engineering