SSI 1.0: The Sequential Sampling Investigator

Written by David Legg
Department of Renewable Resources
University of Wyoming
Laramie, Wyoming
82071
USA

What is the SSI?

The Sequential Sampling Investigator, or SSI, is a computer program that will calculate sequential sampling stop values, via Wald's Sequential Probability Ratio Test method OR Iwao's Confidence Interval method, assuming the normal distribution holds for both. What is more, it will also classify the values of an input file, using a sequential sampling model, to see if they are 'high' or 'low' relative to a threshold value. Finally, it will provide the error rate and average sample number for values in any input file. This last feature is particularly useful when evaluating a sequential sampling model to see if it performs satisfactorily. Performances are often evaluated by constructing average sample number 'curves', operating characteristic 'curves', and/or proportion of incorrect classification 'curves'.

Why I Wrote this Program

Sequential sampling models that are constructed with Iwao’s confidence interval method or Wald's sequential probability ratio test–based method may not perform at theoretically–stipulated levels. The reasons why are well known (Nyrop and Simmons 1984, Weatherill and Glazebrook 1986, Fowler and Lynch 1987, Binns 1994, Nyrop et al. 1999, Binns et al. 2000, Legg and Chen 2000). To find out how specific sequential sampling models do perform, I constructed the Sequential Sampling Investigator (SSI). To be sure, there are other computer programs that will perform this task, and their code can be found in Nyrop and Binns (1991 and 1992), and Binns (1994).

In addition to finding out how specific sequential sampling models actually perform, the Sequential Sampling Investigator can do more. One of the things it can do is, given a data file taken right out of the field (so to speak), it will randomly select individual values, without replacement, and feed them into a selected sequential sampling model, so the model can conduct the sequential test, value by value, until one of three things happens: 1) classification is made, 2) maximum number of samples is reached without classification, without assessing all values in the sample, or 3) final value in the sample is fed into the sequential sampling model and still no classification is made. The SSI makes random selections without replacement by creating a random permutation of the values in the data set, so there is no need to randomly sort the data prior to using the SSI. Finally, once you are satisfied with the performance of a particular sequential samplng model, the Sequential Sampling Investigator can generate an ASCII file of sequential sampling stop values so it can be copied into spread sheets, word processors, etc., and printed for distribution to and for use by practitioners.

Click here for examples of using the SSI.

Conditions of Distribution and Limitation of Liability

All computer programs distributed by the University of Wyoming are on an AS IS basis without warranty. The University of Wyoming shall have no liability or responsibility to cooperator or any other person or entity with respect to liability, loss, or damage caused or alleged to be caused directly or indirectly by programs distributed by the University of Wyoming, including but not limited to interruption of service, loss of business, or anticipatory profits or consequential damages resulting from use or operation. The University of Wyoming shall not be liable for any damages caused by delay in delivering or furnishing programs.

Before You Download the SSI....

· You will need a copy of WINZIP to extract the program and data files

· You will need to decide on the drive and subdirectory (folder) in which to place the program and data files

Before You can Run the SSI....

· You will need to extract the program and data file (do this with WINZIP)

To Download the SSI....

· There are 11 executable files. The names of these are as follows: SPRTCLAS.EXE, BRUN45.EXE, SSI.EXE, MAINSS.EXE, SPRTNORM.EXE, SPRTLINE.EXE, SPRTTEST.EXE, IWAONORM.EXE, IWAOLINE.EXE, IWAOCLAS.EXE, IWAOTEST.EXE.

· There is also a practice data file, DATA.HBI

· Click here to download the compressed file, ssi.zip

To Run the SSI....

· In Windows®, click on START,

· Click on RUN,

· Select ssi.exe

References Cited

Binns, M. R. 1994, pp. 137-174 In Pedigo, L. P., and Buntin, G. D. (Eds.), Handbook for sampling methods for arthropods in agriculture, CRC Press, Boca Raton.

Binns, M. R., Nyrop, J. P, and van der Werf, W. 2000, Sampling and monitoring in crop protection. CABI, New York.

Fowler, G. W., and Lynch, A. M. 1987, Environmental Entomology, 16, 345.

Legg, D. E., and Chen, B. M. 2000, American Entomologist, 46, 250.

Nyrop, J. P., and Binns, M. 1991, pp. 67-132 In Pimentel, D. (Ed.), Handbook of pest management in Agriculture vol. 2, CRC, Boca Raton.

Nyrop, J. P, and Binns, M. R. 1992, J. Economic Entomology, 85, 1253.

Nyrop, J. P., Binns, M., and van der Werf, W. 1999, Phytopathology, 89, 1104.

Nyrop, J. P., and Simmons, G. A. 1984, Environmental Entomology, 13, 1459.

Weatherill, G. B., and Glazebrook, K. D. 1986, Sequential methods in statistics, 3rd ed., Chapman and Hall, New York.