Optimal Data Quality (Q3704716)
From MaRDI portal
!
This is the item page for this Wikibase entity, intended for internal use and editing purposes. Please use the normal view instead:
scientific article; zbMATH DE number 3932177
| Language | Label | Description | Also known as |
|---|---|---|---|
| default for all languages | No label defined |
||
| English | Optimal Data Quality |
scientific article; zbMATH DE number 3932177 |
Statements
1985
0 references
benefit-cost analysis
0 references
sample size
0 references
Morgenstern's hypothesis
0 references
data quality
0 references
loss function
0 references
data use
0 references
risk
0 references
Numerical examples
0 references
Optimal Data Quality (English)
0 references
0.7361800670623779
0 references
0.7212023735046387
0 references
0.7208782434463501
0 references
The following commonly accepted hypotheses concerning needed data quality are formally defined and analyzed in the context of decision-theoretic models: 1. The more rudimentary the use of the data, the less quality is needed (Morgenstern's hypothesis). 2. The needed data quality increases as the magnitude of the consequences of data errors increases. 3. As the probability that a decision uses data increases, the needed data quality increases. It is shown that the truth of the hypotheses depends on the nature of the loss function and of the decision problem. Further, two hypotheses in case of two-action and two-state problems are investigated: 4. If the probability of data use is high enough to justify collecting any data, then it justifies collecting high quality data. 5. Estimating the probability of data use for determining how much data quality is needed incurs less risk by over-estimating the probability than underestimating it. Numerical examples indicate that the practical implications of hypotheses 1-5 can be very sensitive to the characteristics of the data use considered in these hypotheses.
0 references