Gender effects on bias in complex financial decisions

Abstract

Overconfidence in decision making is known to negatively affect performance (Sieck and Arkes, 2005) and task satisfaction (McGraw et al., 2004). If individuals cannot accurately judge their own performance capability in such important functions as financial decision making, however, their performance may suffer. For example, people who are overconfident may not use available information aids (e.g., Arkes et al., 1986; Sieck and Arkes, 2005; Whitecotton, 1996). In addition, overconfident individuals may risk resources based on falsely inflated expectations of future performance (e.g., Campbell et al., 2004). Furthermore, initial overconfidence can lead to a feeling of failure when the person does not perform as he/she expected (Bandura and Schunk, 1981; Stone, 1994). Consequently, future effort and self-perceived ability may suffer. Despite a great deal of research on decision making bias, however, sufficient research has not been conducted regarding gender differences in over- and underconfidence (Pallier, 2003). Moreover, those studies that address gender differences often lack control variables and statistical power (Bajtelsmit and Bernasek, 1996) and fail to compare self-perceived ability to objective measures (Ackerman et al., 2002). In fact, researchers often state that men are overconfident and women are underconfident without actually using objective performance measures (Beyer, 1990). Finally, the number of investigations that center on overconfidence far outweighs those that focus on underconfidence (Bandura and Locke, 2003). In the current study, we investigate gender differences in self-efficacy to make complex financial decisions. This study uses the robust self-efficacy construct, personal goals, objective performance feedback, and an objective performance measure of decision accuracy. First, we review the relevant literature and develop hypotheses. Next, we describe methods and present results. Last, we discuss limitations and implications for managers and researchers. LITERATURE REVIEW Gender Differences in Self-efficacy to Make Complex Financial Decisions Self-efficacy Construct. Self-efficacy is defined as one's perceived ability to perform a specific task (Bandura, 1997). Self-efficacy is formed from an individual's evaluation of the task, the context in which it is performed, and of personal ability. The result of this evaluation is a perception that may or may not reflect actual objective ability as measured by skill tests. In fact, individuals with the same level of skill may report different levels of self-efficacy because they have different types of environmental or personal supports. Four main external influences form self-efficacy (in order of most to least effective influence): personal performance experiences, vicarious performance, persuasion, and physiological arousal (e.g., excitement or fear) (Gist and Mitchell, 1992). These four external factors affect a person's perceptions of the task and how difficult it will be to perform. Prior to performance, serf-efficacy leads to intermediate outcomes such as effort and personal goals. An appropriate level of effort is applied to fit the desired performance level (Bandura, 1997). Finally, once a person does perform a task, feedback also may change perceived self-efficacy when the performance is evaluated and attributed to personal or environmental factors. Because self-efficacy is dynamic and changes over time depending on a person's perception of past performance and changes in one's self and environment, objective performance feedback is important to increase complex task performance (Earley, 1985). Self-efficacy predicts performance across a wide range of tasks and subjects (Bandura, 1997; Stajkovic and Luthans, 1998). Self-efficacy is a superior predictor of performance versus simple measures asking if a subject is confident in his/her ability to succeed in a task (Lee and Bobko, 1994). …

Cite this work

No public copy is stored for this record, so only the metadata is cited.

Stored only when both papers are in the corpus — 15 citations exist upstream.

Papers that cite the same references — the relational fallback while embeddings are still backfilling.