Historical Test Data Research Articles

Verification of a mathematical method for the estimation of the acute ingestion hazard of detergent preparations

The hazard posed by the accidental ingestion of preparations has traditionally been assessed either by carrying out an acute oral toxicity test on animals or by extrapolation from data on tested preparations of similar composition. The EC Directive on the Classification, Packaging and Labelling of Dangerous Preparations (CPL Directive; EEC, 1988) requires that a preparation is classified by a mathematical procedure known as the Conventional Method. This method can result in preparations being classified as “Harmful” by ingestion even though they are not considered to be a significant ingestion hazard on the basis of the results of tests on similar preparations or surveys of accidents with marketed products. This paper proposes an alternative mathematical method, the Proportionality Method, for assessing the acute ingestion hazard of detergent preparations. Based on historical animal test data, the Proportionality Method predicted acute ingestion hazard classification better than the Conventional Method. The number of preparations which would be labelled unnecessarily is reduced, while the safety to the consumer is not compromised by underclassification. The Proportionality Method could reduce the number of animal tests needed to overturn incorrect classifications obtained by the Conventional Method by providing a more rational estimate of the acute oral ingestion hazard.

Computerized adaptive testing in instructional settings

Item response theory (IRT) has most often been used in research on computerized adaptive testing (CAT). Depending on the model used, IRT requires between 200 and 1,000 examinees for estimating item parameters. Thus, it is not practical for instructional designers to develop their own CAT based on the IRT model. Frick improved Wald's sequential probability ratio test (SPRT) by combining it with normative expert systems reasoning, referred to as an EXSPRT-based CAT. While previous studies were based on re-enactments from historical test data, the present study is the first to examine how well these adaptive methods function in a real-time testing situation. Results indicate that the EXSPRT-I significantly reduced test lengths and was highly accurate in predicting mastery. EXSPRT is apparently a viable and practical alternative to IRT for assessing mastery of instructional objectives.

New Life for the 555"Pool, San Joaquin Basin, California: A Study in Synergy"

The 555' Pool, discovered in 1957 and unitized in 1961, is located on the northwestern nose of the Buena Vista Hills anticline in the western San Joaquin basin, Kern County, California. The trap is a combination of stratigraphic pinch-out and anticlinal rollover with approximately 750 ft of closure. Two distinct productive intervals in the upper Miocene Stevens Sand are present, namely, the massive sand zone, comprising thick (< 1,000 ft) turbidite channel sands, and the Interbeds zone, comprising interbedded sands and shales about 500 ft thick. Cumulative oil production for the Massive and Interbeds zones is 44.9 MMBO (44% recovery of OOIP) and 4.2 MMBO (35% recovery of OOIP), respectively. Workover results in the middle to late 1980s led to incremental improvement in production rates and the realization that substantially more reserves may be located behind pipe within the field limits. A multidisciplinary task-force reviewed all available historical core, log, drill-stem test, and production data in addition to workover results and gas identification logs. This review resulted in an improved reservoir description, better geological reservoir characterization, and the identification of possible reserves not accessed by then-current completions. Reserve estimates were made by simultaneous solution using material balance and volumetric methods.more » This study projected a significant increase in the remaining recoverable reserves. New workover wells have resulted in substantial improvement in production rates and an amended interpretation of the location of the oil band.« less

Application of the Bayes Equation to Predicting Reactor System Reliability

Reliability predictions for systems exhibiting few, if any, failures require the use of all available information. The Bayes equation incorporates prior engineering information with test data to provide statistically improved posterior estimates. Classical results agree with those obtained from the Bayes equation by using no prior information. For the case of failure-on-demand, this is equivalent to assuming a 50% mean failure probability for the prior information--hardly an appropriate estimate for a reliable system such as a reactor scram system. The method of Bayes conjugates applied to the cases of aging failure and failure-on-demand yields formulas for calculating mean, standard deviation, and confidence values. Various methods for incorporating prior information are possible. For example, calculating scram failure probabilities by incorporating prior information obtained from fault tree analysis of a scram system with historical test data indicates a mean scram failure probability of approx. 8 x 10/sup -6/ per demand.