Research Reports from the Department of Operations
Document Type
Report
Publication Date
11-15-1956
Abstract
An information-collection situation is said to arise when a sample is to be taken from some universe, observations are to be made on the elements of the sample, and action is to be undertaken on the basis of these observations with the expectation of return. An information-collection procedure in such a situation is specified by assigning values to the following two quantities: n : the number of elements in the sample, l : the resources expended in making an observation on one element of the sample. An information-collection procedure is considered optimal, in this paper, if the expected return from its use is no smaller than the expected return from the use of any other admissible information-collection procedure. This optimization problem differs from the ones usually discussed in sampling theory, in that observational error is treated as a random variable. In most works on sampling theory, either the possibility of observational errors is ignored, or observational errors are assumed to cancel one another out. The inclusion of l as an explicit element of the information-collection procedure, in this paper, permits the consideration of observational error in the following sense. The planner may increase l, at his discretion, weighing the cost of such an increase against the advantages from the resulting expected decrease in observational error. Information-collection situations are classified in two independent ways - quantitative vs. qualitative, and collective vs. distributive. This leads to a four-fold classification of information-collection situations, viz.: Qualitative - Distributive, Quantitative - Distributive, Qualitative - Collective, Quantitative - Collective. The main body of the paper consists of a detailed examination of each of these four sub-classes of information-collection situations, including a mathematical treatment of the optimization question and examples.
Keywords
Operations research, Information theory, Sampling (Statistics), Statistical services, Mathematical optimization, Observation (Scientific method)
Publication Title
Technical Memorandums from the Department of Operations, School of Management, Case Western Reserve University
Issue
Technical memorandum no. 1
Rights
This work is in the public domain and may be freely downloaded for personal or academic use
Recommended Citation
Yaspan, Arthur J.; Halbert, Michael H.; and Ackoff, Russell L., "An exploratory study on the consideration of observational errors in the design of information collection procedures" (1956). Research Reports from the Department of Operations. 201.
https://commons.case.edu/wsom-ops-reports/201