Data processing: measuring – calibrating – or testing – Measurement system – Statistical measurement
Reexamination Certificate
1999-12-22
2001-11-13
Hoff, Marc S. (Department: 2857)
Data processing: measuring, calibrating, or testing
Measurement system
Statistical measurement
C604S890100
Reexamination Certificate
active
06317700
ABSTRACT:
BACKGROUND OF THE INVENTION
1.1. Technical Field
This invention relates to computer-based computational methods and systems to perform empirical induction. Empirical induction involves procedures to arrive at generalized conclusions and to make predictions from data. In particular, this document addresses procedures for using repeated measures data to quantify, discover, analyze, and describe longitudinal associations between events and variables for individuals.
1.2. Description of Related Art
Statistical analysis is the prevailing computational method to perform empirical induction. Empirical induction is used to gain scientific knowledge, to develop and evaluate treatments and other interventions, and to help make predictions and decisions. This document focuses on empirical induction about patterns of association between and among variables.
Computational methods and systems of empirical induction are designed to provide high quality generalized conclusions and predictions. Generalized conclusions and predictions based on generalized conclusions are considered to be of high quality when they meet four criteria. First, the generalized conclusions and predictions are of high quality when they are based on observation and experience that is recorded as data that can be shared. Second, the generalized conclusions and predictions are of high quality when the data are properly analyzed by computational procedures that can be specified in detailed protocols, the protocols making the procedures transparent. Third, the generalized conclusions and predictions are of high quality when application of the protocols to the data yield results that can be reliably repeated by the same investigator and reproduced by other investigators. Fourth, the generalized conclusions and predictions are of high quality when they are not apt to be falsifiable by new or additional data.
The Appendix is an outline that helps reveal the logical structure of this document. Section 2.9 defines many terms used in this document.
1.2.1. Fundamental Limitations of the Statistical Method and a Derivative Nexus of Problems and Needs
This section identifies four fundamental limitations of the statistical method and illustrates a common set of conditions under which these limitations lead to a nexus of related problems and needs. This section also offers a prime example of how the nexus of problems and needs hinders progress in science, some professions, and the advancement of human welfare.
There are two major research strategies for investigating individuals. First, individuals can be investigated directly as individuals. Second, individuals can be investigated indirectly as members of groups or collective entities. The statistical method primarily is a component of the second research strategy. The statistical method includes descriptive statistics for describing groups and populations as well as inferential statistics. Inferential statistics uses statistical descriptions of statistical samples to make inferences about populations.
The first fundamental limitation is that the statistical method is not well suited to perform empirical induction for individuals. In other words, the statistical method often is not well suited to provide high quality generalized conclusions about and predictions for individuals. For example, the value of a statistical measure such as a group mean may not describe any individual member of the group.
It is possible for applications of both the direct and the indirect research strategies for investigating individuals to arrive at similar high quality generalized conclusions and predictions. However, conditions suitable for the achievement of similar high quality generalized conclusions and predictions with the two different research strategies often do not obtain.
Conditions not favorable for similar high quality generalized conclusions and predictions with the direct and indirect research strategies for investigating individuals can be illustrated in the context of medicine. Investigations of phenomena in which individuals could be investigated either directly as individuals or indirectly as members of groups could be expected to arrive at similar high quality generalized conclusions about individuals if individual patients were clones with identical histories. Problems arise in clinical research and medicine because patients are not clones with identical histories. In areas of investigation such as medicine, it is perfectly possible for applications of the statistical method to arrive at high quality generalized conclusions about groups and high quality inferences about populations but low quality predictions for individual members of the groups or populations. For example, individual patients may not respond to a treatment in the same way that most patients in a group respond to the treatment.
The first fundamental limitation of the statistical method has two parts. The first part of the first limitation is that the statistical method is not well suited to be applied during investigations of unique individuals. Individuals can be unique either because they are so particular or unique because they are so inclusive. Individual patients with particular genomes and histories are unique because they are so particular. The world economy, the worldwide investment market, and the worldwide health-related environmental system are each unique because each is so inclusive.
The second part of the first fundamental limitation of the statistical method is that the statistical method is not well suited to reveal that which may make individual group members different with respect to associations between and among variables. Without recognizing that which may make individual group members different, it is difficult to develop the classification systems that help make the statistical method useful. The classification systems at issue are, for example, classifications of medical disorders that can be applied to form more homogeneous groups of individuals for investigations and to predict responses of individual patients to treatments.
The second fundamental limitation is that the statistical method is not well suited to arrive at high quality generalized conclusions about longitudinal associations or to make high quality predictions about longitudinal associations. Longitudinal associations are quantified within individuals. The quantification of longitudinal associations would help enable investigations of dynamic functioning including the internal and external control of individuals.
The statistical method is well suited to arrive at high quality generalized conclusions about cross-sectional associations. Cross-sectional associations are quantified across individuals for particular variables effectively at particular times. But generalizations about associations do not have to be generalizations across individuals for particular variables to be generalizations. Generalizations about associations can be generalizations across variables and over time for particular individuals. For example, it is a generalization for an individual to conclude that her allergy symptoms generally get worse after she pets a cat and rubs her eyes.
Biotechnology is making rapid progress in identifying that which makes individuals different in terms of genetic characteristics that are relatively stable over time. It could also be valuable to identify that which makes individuals different in terms of dynamic functioning, functioning that can involve longitudinal associations between the products of genetic expression that fluctuate in level over time.
The limitation of the statistical method for quantifying longitudinal associations, together with the almost exclusive role of the statistical method as a computational method of empirical induction, appears to be the reason why there are so few investigations of longitudinal associations in, for example, the medical literature.
The third fundamental limitation is that the statistical method is not well suited to investigate complexity and multidimensionality. The capability
Brooks & Kushman P.C.
Hoff Marc S.
Raymond Edward
LandOfFree
Computational method and system to perform empirical induction does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with Computational method and system to perform empirical induction, we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Computational method and system to perform empirical induction will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-2617809