Empirical credibility

Validation of models is desired but can be difficult to achieve. Models are empirically validated by examining how output data (predictions) compare with observed data (such comparisons, of course, must be conducted on data sets that have not been used to create or specify the model). However, model validations conducted in this manner are difficult given limitations on data sources. As an alternative approach, model credibility can be assessed by a careful examination of the subcomponents of the model and inputs. One should ask the question Does the selection of input variables and the way they are processed make sense Also, confidence in the model may be augmented by peer reviews and the opinion of the scientific community. Common faults and shortcomings are... 

With the exception of STO-3G and both MP2 models, all models (including semi-empirical models) provide a credible account of relative CH bond energies. In terms of mean absolute error, BP and B3LYP models with the 6-311+G basis set are best and Hartree-Fock 3-21G and 6-3IG models, local density 6-3IG models and semi-empirical models are worst. More careful scrutiny turns up sizeable individual errors which may in part be due to the experimental data. For example, the best of the models appear to converge on a CH bond dissociation for cycloheptatriene which is 35-37 kcal/mol less than that in methane (the reference compound) compared with the experimental estimate of 31 kcal/mol. It is quite possible that the latter is in error. The reason for the poor performance of MP2 models, with individual errors as large as 16 kcal/mol (for cycloheptatriene) is unclear. The reason behind the unexpected good performance of all three semi-empirical models is also unclear. 

Bell IR. 2003. Translating complexity science into empirical tests for homeopathy. In Improving the success of homeopathy 4 Bridging the credibility gap. International conference 3-4 April 2003, London. The Royal London Hemeopathic Hospital, University College London Hospitals, pp 9-15. 

The kinetic study assists in the development of a credible reaction mechanism which describes all aspects of the reaction - not just the kinetics [ 1 ]. The complete exercise involves empirical and theoretical considerations which run in parallel they are complementary and feedback between them is essential [2]. Aspects (i) and (ii) above were covered in the previous chapter, and we now focus first on the derivation of the rate law (rate equation) from a mechanistic proposal (the mechanistic rate law) for comparison with the experimental finding. In simple cases, the derivation is usually straightforward but can be mathematically challenging for complex reaction mechanisms. Once derived, the mechanistic rate law is compared with the experimental, and the quality of the agreement is one test of the applicability of the mechanism. Different mechanisms may lead to the same rate law (they are kinetically equivalent), and, whilst agreement between mechanistic and experimental rate laws is required, this alone is not a sufficient proof of the validity of the mechanism [3-7]. We conclude the chapter by working through several case histories. 

Four qualitative criteria or hallmarks for judging the quality of data have been described— appropriateness, accuracy, integrity and transparency. Attentiveness to each hallmark is critical to the defensibility and credibility of exposure assessments. Thus, exposure assessment should follow the main scientific desiderata of any applied science empirical testing, data modelling, documentation and reproducibility of results, explicit reporting of assumptions, limitations and uncertainty, peer review and an open debate about underlying theories and models. Transparency is key to the success of all of these steps. 

A key element in the procedure of time scaling is the use of a value or values for n in the equation C" x t = k. If empirical data for exposure durations other than the AEGL-specified exposure periods are available to quantify the exposure concentration-exposure duration relationships for a health-effect endpoint, including lethality, the value of n should be derived using the method of calculation described in this section. It is believed that empirically derived values of n are scientifically more credible than a default value of n = 1 (Haber s rule) or attempting to derive an alternate value of n. 

As a distinct and credible literature of process development for bulk drugs and fine chemicals has come into being and grows, statements of applicable empirical wisdom are appearing with a modicum of organization (2—6) and the field should one day become amenable to independent study (it is not currently taught formally an3rwhere). 

The selection of any NDT method for a particular inspection will be due to some extent on the probability of detection (POD) for that method. The POD is a measure of the reliability of the method to detect a particular type of flaw size [13,14]. POD curves are created from empirical studies to show the reliability of a method for detection of features or defects. A useful curve for PODs is a 90% POD with 95% confidence, referred to as a 90/95 POD. Figure 15.8 is a notional POD curve that shows how one method may perform more reliably than another for the detection of particular flaw sizes. POD curves can be expensive to develop experimentally because they require test samples with known defect sizes and numerous inspection tests by different operators to develop credible statistics. Model-assisted-based methods are being developed to reduce this burden [15]. 

We consider the first update (/ = 1). By the approach in step 3, we estimate the rate of DU-failures by either Xdu.i or Xdu.i- In the following, assume that we have chosen to use the Bayesian estimate Xdu.i-(If we use the empirical estimate, we get the same formulas). Next, determine the 90% confidence (or credibility) interval for Xdu.i We then calculate the ratio XDU.0/Xdu.i - This ratio indicates the fractional change in failure rate and thus the allowed change of the test interval. By using eq. (1), an updated test interval t can now be calculated as ... 

Each new plant that fabricates, processes, or otherwise handles fissionable materials undergoes a safety analysis prior to startup. One part of the safety analysis is to assess the potential consequences of a postulated woi t credible criticality accident. To perform this assessment, the total energy release—expressed as total number of fissions from the criticality excursion—must be estimated or assumed. This paper presents the results of study for development of an empirical model to estimate energy release from a criticality accident. 

As far back as the Babylonian Empire, a strong foundational leadership centering on justice moved society toward safety initiatives. From Hammurabi s Code addressing the liability of homebuilders in an eye for an eye justice system, the concern for safety was viewed as a just duty. Persons placed in leadership roles had a duty originating from a virtue of justice, to consider the safety of those around them. In modern studies of leadership, justice is considered a foundational virtue of leaders. Leaders must display a sense of justice that is viewed as moral by those that they lead. Safety is indeed one of those principles that come from the virtue of justice. Leaders must display a sincere effort for subordinate and peer safety or risk moral decay of their credibility. 

---