Member # 262
posted 03. May 2003 08:12
If you look closer, I think you will find that the experimental design I am proposing for testing scientific procedures is a common ‘successive approximation’ applied science experimental design. The experimental design involves:
1) define the goal or purpose G of a task or process- in this application formulate, test and validate predictive scientific hypotheses,
2) identify and/or define a set of procedures Y to compete with the existing procedures X- in this application, procedures Y are developed to address what perceived as potential weaknesses in existing procedures. These potential weaknesses would include inadequate requirements for explicitly and precisely defining predictive hypotheses, and the use of subjective and potentially biased evaluation techniques.
3) determine if X and Y produce different results with respect to G- In this application this involves determining if one set of procedures rejects or fails to validate hypotheses validated by the other- this is readily easily determined
4) For identified differences in result- determine which result in more likely valid- In this application- if there are two analytical processes that produce conflicting conclusions regarding the validity of a predictive hypotheses- the step by step processes used to generate the two conclusions would be reviewed in detail to determine if there are errors or mistakes in either process- In general, effective techniques exist for performing such step by step analysis, for identifying the source of differences in result, and for determining if one or both types of review/validation procedures are flawed. Note that ultimately the validity of competing predictive hypotheses is determined by the ability of the hypotheses to satisfy established scientific standards and by the ability of the hypotheses to produce accurate, verifiable, and reliable predictions.
5) Having reconciled the differences in results obtained by X and Y, determine if a)one set of procedures is clearly better than the other, b)there is some set of procedures Z combining the best features of X and Y which is more effective than either, or c)neither set of procedures produces effective results.
6) The above experimental design generates a ‘current best practices set of procedures XN’. The successive approximation process of finding best practices procedures would continue with development of a new set of procedures YN. Note that new technologies and new discoveries mean that even if the set produces XN used today are best practices, the set XN would not necessarily be best practices in the future.
It is worth noting that the ‘successive approximation’ experimental design for evaluating scientific procedures, as outlined above and as suggested by the dynamic best practices topology is not new. The experimental design outlined above is a standard applied science design. It can also be argued that the dynamic best practices topology or logical structure is not new, but a topology or logical structure widely used in applied sciences.
As I suggested in an earlier post, the use of the proposed topology in scientific analysis depends on 1)its logical/mathematical soundness and 2)its usefulness. Existing applied science applications support both the soundness and the usefulness of the dynamic best practices topology.
As I also stated earlier, the dynamic best practices topology makes it possible to formulate predictive hypotheses from ‘dynamic and teleological causal relationships’. Although this type of hypotheses would appear to be in conflict with certain views in conventional science, it appears to be consistent with current applied sciences practices.
Finally, the dynamic best practices topology suggests that explicitly constructing/defining complex logical/mathematical structures may be useful in trying to analyze and construct models and hypotheses of the complex processes associated with life forms. This raises, at least in my mind, the question "Is it possible/practical to construct complex predictive models and hypotheses of complex causal relationships without first explicitly defining a topology or logical/mathematical universe with the appropriate logical structures?’. Are both ID and conventional life science limited by the lack of appropriate and explicitly defined topologies?