Tamar Kakiashvili

Pairwise Comparisons and Visual Perceptions of Equal Area Polygons

The number of studies related to visual perception has been plentiful in recent years. Participants rated the areas of five randomly generated shapes of equal area, using a reference unit area that was displayed together with the shapes. Respondents were 179 university students from Canada and Poland. The average error estimated by respondents using the unit square was 25.15%. The error was substantially decreased to 5.51% when the shapes were compared to one another in pairs. This gain of 20.24% for this two-dimensional experiment was substantially better than the 11.78% gain reported in the previous one-dimensional experiments. This is the first statistically sound two-dimensional experiment demonstrating that pairwise comparisons improve accuracy.

Improving the medical scale predictability by the pairwise comparisons method

In the clinical practice of psychiatry, presence or absence of particular symptoms is based on the subjective interpretation, by the clinician, of mental and behavioural descriptions offered by the patient. However, this subjectivity that characterizes the diagnostic decision making process may limit the reliability of diagnosis. In this current study, the pairwise comparisons (PC) method is used to investigate whether the psychometric properties of a medical screening questionnaire can be improved. The pilot data described herein did indeed demonstrate that modest improvements in diagnostic accuracy could be achieved using PC, and provides early evidence that the inconsistency produced by subjective clinical ratings can be reduced using this method, thus providing impetus for further investigation.

A note on relevance of diagnostic classification and rating scales used in psychiatry

In the clinical practice of psychology and psychiatry, presence or absence of particular disorder or syndromes is based on the subjective interpretation of mental and behavioral descriptions offered by the patient. This is often done by questionnaires (also called instruments or scales) or by interviews. This subjectivity of the diagnostic decision-making process may limit the reliability of diagnosis. In the present study, a new method of scale relevance, based on double cluster analysis, is proposed as it is important to verify what we are trying to find with the proposed scale. If two data sets cluster differently, we must consider them as different.

How to reduce the number of rating scale items without predictability loss

Rating scales are used to elicit data about qualitative entities (e.g., research collaboration). This study presents an innovative method for reducing the number of rating scale items without the predictability loss. The “area under the receiver operator curve method” (AUC ROC) is used. The presented method has reduced the number of rating scale items (variables) to 28.57% (from 21 to 6) making over 70% of collected data unnecessary. Results have been verified by two methods of analysis: Graded Response Model (GRM) and Confirmatory Factor Analysis (CFA). GRM revealed that the new method differentiates observations of high and middle scores. CFA proved that the reliability of the rating scale has not deteriorated by the scale item reduction. Both statistical analysis evidenced usefulness of the AUC ROC reduction method.

Assessing the properties of the World Health Organization’s Quality of Life Index

This methodological study demonstrates how to strengthen the commonly used world health organizationpsilas quality of life index (WHOQOL) by using the consistency-driven pairwise comparisons (CDPC) method. From a conceptual view, there is little doubt that all 26 items have exactly equal importance or contribution to assessing quality of life. Computing new weights for all individual items, however, would be a step forward since it seems reasonable to assume that all individual questions have equal contribution to the measure of quality of life. The findings indicate that incorporating differences of importance of individual questions into the model is essential enhancement of the instrument.

Approximate reasoning by pairwise comparisons: “Topodynamics of metastable brains” by Arturo Tozzi, et al.

The innovative approach in [1], “Topodynamics of Metastable Brains” by Arturo Tozzi, James Peters, Andrew Fingelkurts, Alexander Fingelkurts, and Pedro Marijuan has a high potential of becoming a paradigm shift in the brain research. It seems that this study has successfully explored the possibility of applying a celebrated Borsuk–Ulam theorem to the operational architectonics of the fundamental brain-mind processes. It has been already in use in practically all branches of dynamics in classical mechanics, quantum physics, fluid and gas dynamics. Among the most recognized names contributing to this approach are Sophus Lie and Henri Poincaré. In our opinion, [2] provides a comprehensive introduction. An exposition on holonomy is given in [3] and in the context of pairwise comparisons in [4]. Fig. 1 shows the parallel transport which is an illustration of a holonomy (one of the invariants). The first step in the discussed method is to identify quantities related to the dynamics in order to describe the internal structure of the analyzed system. These quantities can have several names which carry very explicitly the corresponding concepts: symmetries, invariants, characteristic classes, curvatures (just to name a few of them). In fact, there is a plethora of such “invariants” expressing some obstructions, or the ability to get some kind of stability or structure. Non-linear systems are not stable enough to be sufficiently analyzed by classical numerical methods. For them, invariants are of particular use. To the best of our knowledge, the most refined approach (in this spirit) appears in the mathematical quantum physics (see [3] for more on this subject). By “topodynamics”, the authors signify hidden structures of metastable brains via symmetries. The principal topological tool is the Borsuk–Ulam theorem (BUT). It is strongly related to the theory of invariants and symmetries. More precisely, it is a part of fixed-point theory. BUT is also related to symmetries on a hypersphere which are special

Attacking the Web Cancer with the Automatic Understanding Approach

The new method, based on automatic understanding is proposed for fighting spam in web information exchange (especially email correspondence). The web cancer term is used to reflect the variety and sophistication of web contaminations. The notable oncology achievements in medicine could inspire more research towards finding solutions to what can easily turn into an analogous civilization crisis. Automatic understanding is appropriate for the semantic-level content analysis and is expected to substantially reduce the wasted user time for semi-automatic analysis needed for the massive processing as most filters are too tight or too loose.