The integration of fuzzy sets and statistics: toward strict falsification in the social sciences

Abstract

Whilst statistics take up a prominent place in the social science research toolkit, some old problems that have been associated there with have not been fully resolved. These problems include bias through the inclusion of irrelevant variation and the exclusion of relevant variation, which may lead to hidden and spurious correlations in more extreme—however not at all unthinkable—cases. These issues have been addressed by Ragin by building a case for the usage of fuzzy set theory in social science. In this paper, we take a complementary view, insofar as we incorporate fuzzy set theory in current statistical analyses. Apart from shedding new light on the main issues associated with (population based) statistics, this approach also offers interesting prospects for the falsification of theories - rather than single relations between variables - in the social sciences

Solving group decision-making problems in manufacturing systems by an uncertain compromise ranking method

“…In recent years, the fuzzy sets theory and their extensions have been widely used in many different fields, such as social science (Heylen and Nachtegael, 2012), economics (Vis et al, 2013), engineering (Zhou et al, 2012;Makui et al, 2010;Mousavi et al, 2013bMousavi et al, , 2013c, management (Mousavi et al, 2011;Patil and Kant, 2014) and decision making (Gitinavard et al, 2016b;Gitinavard and Zarandi, 2016;Vahdani et al, 2014aVahdani et al, , 2014bVahdani et al, , 2015Salimi et al, 2013).…”