The paper tries to answer the following question: "How should we treat ordinal data in hierarchical clustering?" The question is strongly connected to the use of questionnaires with ordinal scales in the social sciences. The results could help to differentiate among answers to the questions from questionnaires that could be considered as scale variables, those it would be better to convert to ranks and those that should be treated as nominal variables. To make the results general several two-dimensional combinations of group sizes, shapes and differences between their centers were used as well as one three-dimensional combination. Each combination was simulated both with and without unessential variables. All datasets consisted of 3 groups, each with its own multivariate distribution (2 or 3 variables) with known means and covariances. From each design several datasets were simulated. Each variable was cut and recoded to achieve an ordinal scale. Different cutting schemes were used (the intervals were of equal size, either increasing/decreasing from the lowest to the highest value or decreasing from the mean to both extremes). These new variables were then treated as interval, converted to ranks and treated as nominal. Then hierarchical clustering algorithms were used. Ward's algorithm with Squared Euclidean distance was used when data were considered interval or converted to ranks, and Ward's algorithm with matching coefficient as dissimilarity measure was used when they were considered nominal. The quality of the results was assessed by comparing the gained partitions with the three original groups. We also compared results from clustering the original (uncut) data with the three original groups for comparison. The comparison was made using Corrected Rand Index. The results indicate that in most cases treating the data as interval or converting them to ranks yields better results than treating them as nominal, but the differences are sometimes diminished when cutting into a smaller number of intervals.

Floods not only induce vast economic damages but also pose a great danger to human life. In Slovenia, floods rank number one on the scale of damage magnitude. Different factors external to the hazard of flooding influence the gravity and extent of the impacts. A comprehensive collection and analysis of the information related to the understanding of causative factors of human impacts can substantially contribute to the mitigation and the minimisation of fatalities and injuries. In this work, historical analysis was performed for flood fatalities in the years between 1926 and 2014, with 10 flood events that induced 74 casualties considered. A detailed collection and review on human impacts was made. Victims in cars have seemed to increase in recent years and rural areas tend to be more vulnerable and susceptible to having victims of flooding. With regards to gender, the majority of victims are male. The focus was on demographic aspects (age and gender) of fatalities and analysis of the circumstances of loss of life. Based on a description of the activities of victims during flood events and repetitive patterns, groups were made based on the type of flood fatality. Eight interviews were performed with rescuers and people who were affected by floods in order to obtain more extensive information on preventative measures, received help, and flood risk perception.

The aim of this article is to compare methods for detecting and/or measuring regular equivalence in valued networks. The methods compared can be divided into direct and indirect methods. The indirect methods considered here are all variants of the REGE algorithm (White, 2005 White , D. R. ( 2005 ). REGGE (web page) . Retrieved May 12, 2005, from http://eclectic.ss.uci.edu/~drwhite/REGGE/ [Google Scholar] ). All direct methods have roots in Batagelj, Doreian, and Ferligoj ( 1992 Batagelj , V. , Doreian , P. , & Ferligoj , A. ( 1992 ). An optimizational approach to regular equivalence . Social Networks , 14 , 121 – 135 . [Crossref] [Google Scholar] ); however they have been adapted for valued networks (Žiberna, 2006b Žiberna , A. ( 2006b ). Generalized Blockmodeling of Valued Networks . Social Networks(in press) , Retrieved May 30, 2006, from http://dx.doi.org/10.1016/j.socnet.2006.04.002 [Google Scholar] ). All approaches are compared on an empirical valued social network. Variants of the REGE algorithm are also compared on a few artificial valued networks in order to clarify their functionality. This is not necessary for direct approaches, which have clearly defined ideal blocks. In discussion, advantages and disadvantages of both indirect and direct approaches are discussed.