Overconfidence Bias in Groups of Russian Students

                National Research University – Higher School of Economics

Faculty of Economics

Strategic Corporate Finance

Overconfidence in groups: the case of Russian students

Made by:

Novikova A.M.

Checked by:

Stepanova A.N

Moscow, 2015

Introduction

Confidence was called one of five basic constructive emotions at the conference organized by a group of neuroscientists and buddhists in 2000. A confident person is considered to be productive, as he is considered to be able to pay proper amount of attention to activity in which he takes part. However, there is still a broad field for investigation of the factors that influence confidence and measurable consequences of confidence in different activities. Overconfidence is included by behavioral economists in the list of biases that influence the economic and financial decisions of agents. It shows an interest for research as the deeper understanding of mechanism that influence overconfidence could be used in institutional formation.

This paper presents a survey of overconfidence – ‘above average effect’ - in skills and grading expectations of three groups of students. The following text contains literature review and the research part.

Literature review

Studied both by economists [Glaser, Noth, Weber (2004); Glaser, Weber (2007)] and psychologists [Weinstein (1980); Taylor & Brown (1988)], the bias of overconfidence could be defined as overestimation of one’s actual ability, performance, level of control, or chance of ‘success’ [Moore & Healy (2008); Xia, Wang, Li (2014)]. Most research showed that being too optimistic about the results is typical for human nature [De Long, Shleifer, Summers & Waldmann (1991); Kyle & Wang (1997); Wang, A. (1997); Weinstein, (1998); Benos (1998); Odean (1998)]. The studies of overconfidence could be divided into two groups. The works of first class are about absolute (‘stand-alone’) overconfidence [Yates, Lee, Sieck, Choi & Price (2002)] – when individuals assess only their own skills. Index of overconfidence, constructed as the difference between measures of subjective confidence and objective performance, is regressed on some other value.  For example, such an index was constructed by [Biais (2005); Brenner, Griffin & Koehler (2005); Larrick, Burson & Soll (2007); Xia, Wang, Li (2014)]. The second group of papers measures (‘referential’) confidence of participants relative to the average [Alicke, Klotz, Breitenbecher, Yurak & Vrendenburg (1995); Glaser, Weber (2007); Grieco, Hogarth (2009)].

The majority of popular research on overconfidence lie in one of three topics – education, auto and finance.  A large amount of papers on knowledge and school/university grades, for example [Fischhoff, Slovic, & Lichtenstein (1977); Sinkavich (1995); Liberman & Trope (1998); Dunning, Johnson, Ehrlinger, & Kruger (2003); De Paolo, Gioia, Scoppa (2014)], could be explained by the availability of data – sample of students’ answers is the simplest sample to gather. An interest in driving [Svenson (1981); Broberg, Willstrand (2014)] is specific for the USA: driving licenses are traditionally got there in early age and by a large part of population. Svenson showed that 82% of a sample of students placed themselves among the top 30% safest drivers.