How cognitive biases and regulatory foci influence manager’s risk taking propensity
MSc in Business Administration
Specialization: Business Development
University of Groningen
Faculty of Economics and Business
March 2013
First supervisor: Dr. J.D. van der Bij
Second supervisor: Drs. H.P. van Peet
Sandra Reiners
j.m.reiners@student.rug.nl
Abstract
This paper strives to understand which determinants influence managerial risk taking propensity. Risk taking propensity is determinative for organizations nowadays. It influences an organization’s creativity, innovation and in the end a firm’s performance. Therefore, it is essential to understand what influences one’s risk taking propensity. This study investigates how six cognitive biases and the regulatory foci are related to risk taking propensity. Furthermore the influence of promotion and prevention focus on the six biases, emanating from the availability and representativeness heuristics, is investigated. The sample consisted of 117 students from the University of Groningen. Results show that promotion focus increases the use of the regression fallacy bias. Furthermore, promotion focus and prevention focus were both found to lower risk taking propensity.
1. Introduction
Markets are changing fast (Tien-Shang Lee and Munir Sukoco, 2011) and organizations have to respond rapidly. Strategic decisions, for example about new product development (NPD) or alliances, have to be taken quickly in order to survive. However, many of these decisions are based on the likelihood of uncertain events (Tversky and Kahneman, 1974). Lately, risk taking has been widely researched as one of the important variables influencing such decisions (i.a. Sitkin and Pablo, 1992; Tien-Shang Lee, 2011). In new product development (NPD) risks are associated with changes in beliefs or routines. Sitkin and Pablo (1992, p.10) define risk as “the extent to which there is uncertainty about whether potentially significant or disappointing outcomes of decisions will be realized”. Decisions become riskier when outcomes are uncertain or include extraordinary consequences or when goals are challenging (Bryant and Dunford, 2008; Sitkin and Pablo, 1992). Managers responsible for these decisions are faced with a lot of challenges. For example, risks are always involved in the NPD process, as the outcome of new product introductions can never be known beforehand (Tien-Shang Lee, 2011). When an R&D manager has to take a major decision about producing or adopting a technological innovation, risks are involved. Current research makes a distinction between risk perception, “a decision maker’s assessment of the risk inherent in a situation” (p. 12, Sitkin and Pablo, 1992; Petrakis, 2005) and risk taking propensity, an individual’s risk taking tendencies (p. 12, Sitkin and Pablo, 1992; Petrakis, 2005). This paper will focus on the risk taking propensity of managers. Risk taking propensity is the degree to which managers are prepared to take decisions involving risky resource commitments, which consequently could lead to big losses (Miller and Friesen, 1978). Thus risk taking propensity is the willingness to take or avoid risks (Lumpkin and Dess, 1996; Sitkin and Weingart, 1995).
Although some researchers explain that risk taking behavior does not play a dominant role in every situation, team or organization (Gibb and Haar, 2010; Grey and Gordon, 1978), many researchers highlight the importance of risk taking behaviour (i.e. Amabile, Conti, Coon, Lazenby and Herron, 1996; Cabrales, Medina, Lavado and Carbera, 2008; Nystrom, Ramamurthy and Wilson, 2002). Risk-taking propensity can be a vital factor for organizations, because it promotes creativity, which is the starting point for innovation (Amabilem et al, 1996). Risk taking behaviour is also related to radical innovations (Cabrales et al., 2008; Nystrom et al., 2002; Sethi, Smith & Park, 2001), which in turn influences new product success and a firm’s financial and market position positively (Rubera and Kirca, 2012; Tien-Shang Lee and Munir Sukoco, 2011). Furthermore Gibb and Haar (2010) discovered a positive relationship between risk taking, high development and financial performance. Thus, risk taking propensity is important for the level of success of organizations. In many situations it supports strategic decisions and eventually determines firm performance (Simsek, 2007; Walls & Dyer, 1996).
From this, it can be concluded that it is important to know what influences risk taking propensity. Previous research has identified age, gender, personality, inertia, outcome history, risk perception and self-efficacy already (Endriulaitienè and Martisius, 2010; Killgore, Grugle, Killgore, Balkin, 2005; Simsek, 2007; Sitkin and Pablo, 1992; Sitkin and Weingart, 1995). This study will extend this field of research with two additional variables: cognitive biases derived from heuristics and regulatory foci. Stewart and Roth (2004) explain that cognitive biases and risk taking propensity are interconnected and request to explain this relationship.
Heuristics help to speed-up and simplify the decision making process, which might result in efficient decision making (Kahneman, 2003; Tversky and Kahneman, 1974). However, using heuristics could also lead to errors in decisions, which are called cognitive biases (Kahneman and Tversky, 1972). Until now, the use of heuristics has been investigated in particular risk situations (Camerer and Lovallo, 1999; Simon and Houghton, 2003). As mentioned before, risk taking propensity is of critical importance to make strategic decisions and therefore it is useful to understand how biases influence risk taking propensity in general. The present study focuses on six biases emanating from the representativeness and the availability heuristic: base-rate fallacy, sample size fallacy, illusion of control, regression fallacy, illusory correlation and overconfidence (e.g. Tversky and Kahneman, 1974) and tend to investigate the effects of these biases on risk taking propensity.
Secondly, in this study a relatively new research area will be added, namely regulatory focus theory. Each person has a different way of approaching a certain outcome. The regulatory focus theory differentiates between the promotion focus and the prevention focus (Higgins, 1997). Persons with a promotion focus are eager to achieve goals through hard-working and development and focus on the presence or absence of positive outcomes (Crowe and Higgins, 1997; Higgins, 2005; Langens, 2007).
Individuals with a prevention focus act with vigilance in a responsible way and focus on the presence or absence of negative outcomes (Crowe and Higgins, 1997; Higgins, 2005; Langens, 2007). Regulatory focus theory has not been investigated as a determinant of risk taking propensity before. However, some researchers assume that the eagerness of promotion-focused individuals inevitably leads to a more risky response bias or risky choices, whereas the vigilance of prevention-focused individuals is expected to show a natural tendency towards making safe choices (Crowe and Higgins, 1997; Hamstra, Bolderdijk and Velstra 2011). Therefore, this research will add value by providing a clarification of this relationship.
Because of its novelty in the research field, practically no knowledge exists about the effect of the regulatory foci on the cognitive biases. Only Langens (2007) has found significant relationships between the illusion of control bias and regulatory foci. Trevelyan (2008) has discovered that overconfidence and risk taking propensity are not related. The current research will widen this field of research with the influence of the regulatory foci on the six biases, in a comprehensive model with risk taking propensity as dependent variable. It can be expected that promotion-focused individuals are more prone to make use of heuristics, because they will try to do anything in their power to achieve a positive outcome, regardless of any failures involved. It is likely that individuals with a prevention focus will think thoroughly before making decisions, because they would try to avoid making mistakes. Therefore, they are expected to make rely less on heuristics.
In short, I presume that the cognitive biases and promotion focus positively relate to risk taking propensity and that the prevention focus negatively relates to risk taking propensity. Furthermore, I presume that the promotion focus positively, and the prevention focus negatively, influence the cognitive biases. Therefore, this study will strive to address the following research questions. 1) How do cognitive biases influence risk taking propensity? 2) How do promotion and promotion focus influence risk taking propensity? 3) How do the promotion and prevention focus influence the cognitive biases?
In this research, students are used to provide the necessary data. There has been a persistent debate about students serving as surrogates for business men (Ashton and Kramer, 1980; Chang and Ho, 2004; Hughes and Gibson, 1991; Khera and Benson, 1970; Tangpong and Ro, 2008). Contradictory outcomes are found in these studies. Clearly, caution has to be taken when using business students as representatives for business men. To get a better understanding of this debate the results will be compared with results of similar studies involving managers. Also, it can be compared with studies about entrepreneurs to discover differences between all groups, for example with the study of Song et al. (UR).
This study will further contribute to current decision-making, strategy and management literature by providing an explanation about how managers make decisions in uncertain situations. This research will enlarge the understanding of managerial decision making by addressing new determinants of risk taking propensity of managers, which is an important factor in strategic decision making. It can be useful
for organizations to gain insights into the cognitive variables influencing risk taking propensity. Therefore, six biases emanating from the representativeness and availability heuristic will be investigated. Additionally, value will be added by providing a new theory into the risk taking propensity field. Regulatory focus theory is expected to influence risk taking propensity. Moreover, promotion and prevention focus are studied as stimulating or obstructive factors of the six biases. In this study, more insights will be gained into these relationships.
2. Literature review and conceptual model
2.1 Heuristics and biases
We all remember situations in which we had to make a tough decision and even in those circumstances we were able to make one, be it either good or bad. But how do we come to these decisions? In the seventies, Tversky and Kahneman conducted far-reaching research on the decision making process and concluded that people rely on heuristics in their decision making process. Heuristics help to simplify and speed-up decision-making (Tversky and Kahneman, 1974). This study focuses on two heuristics: the representativeness and availability heuristic (Kahneman and Tversky, 1972; Tversky and Kahneman, 1973; 1974). A judgement is influenced by heuristics when people do not use the target attribute for an assessment, but replace it with a related heuristic attribute, which is derived from mental models (Kahneman, 2003; Kahneman and Frederick, 2002). Decisions based on these heuristics could lead to errors in judgements, which are called cognitive biases (Busenitz, 1999; Tversky and Kahneman, 1974). This paper will focus on six biases derived from the representativeness and availability heuristic.
Representativeness heuristic
The representativeness heuristic is the tendency of people to associate one situation with another and conclude that those situations are similar to each other. Thus people evaluate probabilities by the degree to which they believe A represents or resembles B (Tversky & Kahneman, 1974). Decision makers, using the representativeness heuristic, order situations by their subjective probabilities. Thus, they base the outcome on one’s own beliefs or experiences, instead of statistical information (Kahneman and Tversky, 1972). In many situations, this simplified idea of reasoning is to one’s satisfaction and a proper decision can be made. Nevertheless, in other situations it could lead to errors in decisions making. It could lead to, for instance, people neglecting important factors such as prior probabilities when making a judgement (Kahneman and Tversky, 1973). Consequently, it could cause people to make wrong predictions. This study will focus on the following four biases originating from the representativeness
heuristic: base rate fallacy, sample size fallacy, illusion of control and regression fallacy (Langer, 1975; Tversky & Kahneman, 1974).
People applying the base rate fallacy bias use irrelevant information to make a probability judgement and statistical information regarding prior probabilities (base rates) is ignored (Tversky and Kahneman, 1974). Thus, people arrange information by its perceived importance and ignore relevant information (Bar-Hillel, 1980). They rely on personal, subjective information, because they believe it
represents the actual situation. The following experiment of Tversky and Kahneman (1974) illustrates
this. They use an example of a population which consists of 70% lawyers and 30% engineers. Subjects are asked to read a description of Max which was applicable to both occupations. The subjects judged the probability of Max being a lawyer to be 50% and thus equal to Max being an engineer. The subjects ignored statistical information about the population and based their decision on subjective probabilities (Tversky & Kahneman, 1974). Judging a situation on the basis of e.g. personal information rather than prior probabilities implies that statistical information is considered less relevant (Bar-Hillel, 1980; Greening and Chandler, 1997; Juslin, Nilsson and Winman, 2009). However, base rates are very precise indicators, since they measure the relative frequency of events (Plous, 1993).
Before continuing, it is important to shed light on managers and their characteristics. Staw and Ross (1980) conducted research regarding managerial decision making. They found that organizational decision makers received the greatest esteem when making consistent decisions. Also, managers prefer to be seen as effective administrators of organizational resources, which makes them inclined to hold on to their initial choices (Sleesman, Conlon, McNamara and Miles, 2012). Additionally, Rabin and Schrag (1999) state that people are inclined to interpret conflicting information about their initial ideas as supporting information that satisfies these ideas. Furthermore, managers are inclined to focus on positive information and neglect or distort negative information (Boulding Morgan and Staelin, 1997; Kahneman and Lovallo, 1993; Simon et al., 1999). They believe their skills and abilities are strong enough to overcome any situation (Boulding et al., 1997; Duhaime and Schwenk, 1985). Even though risk assessments may warn them for failures, managers are inclined to believe they will not have to bear the risks (March and Shapira, 1987). Because of their great confidence in their own capabilities and their desire to be a great manager, they are inclined to continue along the same line of action, and thus stay consistent to their preferences (Simon et al., 1999).
Managers with a high level of base rate fallacy are inclined to believe that their subjective information is sufficient to make an accurate decision. As stated before, they feel they have to stay consistent to their earlier choices, and therefore they would overestimate information which suits their preferred decision. They tend to ignore or underestimates base rates (Bar-Hillel, 1980; Juslin et al., 2009; Tversky and Kahneman, 1974) in favour of their initial ideas. They believe this biased information
represents the actual situation. Because decisions will be based on incomplete or biased information (Busenitz, 1999), managers will fail to recognize the risks involved which in turn could lead to hazardous situations.
Sample size fallacy is also known as the belief in the law of small numbers. It will occur, for
example, when decisions regarding a whole population are based on a small sample (Tversky and Kahneman, 1971). Conclusions are drawn from a limited number of information sources and people with a high level of sample size fallacy consider the small sample size as representative for the whole population (Rabin, 2002; Tversky & Kahneman, 1974). The law of large numbers guarantees that a large sample size will resembles a population (Busenitz, 1999; Kahneman, Slovic and Tversky, 1982; Rabin, 2002; Tversky and Kahneman, 1971). A large sample size is less variable and its mean is closer to the population mean. Consequently, conclusions are more significant and reliable. Persons with a high level of sample size fallacy neglect this law and overestimate the reliability of the data from the small sample (Rabin, 2002; Simon, Houghton and Aquino, 1999). Managers try to stay consistent to earlier choices or promises made (Simon et al., 1999; Sleesman et al., 2012). When making a decision, they tend to distort negative information into positive information and shed additional light on information that suits their preferences (Boulding et al., 1997; Kahneman and Lovallo, 1993; Simon et al., 1999). Since managers are inclined to overestimate such preference consistent information (Sleesman et al., 2012), managers with a high level of sample size fallacy will consider a small sample size as representative if it matches their preferences. They will then base their decisions on such a small sample. These decisions, for example, could be based on a few success stories about technological innovations. Managers treat this information as representative and base their decisions on this information. This could involve a lot of risks, since decisions are based on incomplete, biased or irrelevant information (Busenitz, 1999).
Illusion of control is the tendency of people to believe they can control a certain outcome while,
in fact, they are unable to exert any influence on it (Langer, 1975; Zuckerman, Knee, Kieffer and Rawsthornel, 1996). Thompson, Armstrong, & Thomas (1998) explain that people with a high level of illusion of control overestimate their personal control and exaggerate the probability of success. Langer (1975) sets out a distinction between “skill and luck situations”. In the case of the former, people can control the outcome, while in the case of the latter, the situation cannot be controlled. People with a high level of illusion of control believe they can also influence the outcome in situations in which luck is determinative, as these persons overestimate their own abilities (Simon et al., 1999). In principle, they judge according to their own perceived skills instead of statistical data about chance or luck, i.e. they believe chance determined situations and skills determined situations are similar to each other. Besides, individuals that have experienced prior successes before, have a greater tendency to experience this bias (Duhaime and Schwenk, 1985). Consider a situation in which people are playing a game with dice. Some
people may believe that throwing slow will lead to low numbers on the dice and throwing hard will lead to high number on the dice. These people believe they can influence the situation, while in reality the number on the dice cannot be influenced (Langer, 1975). Generally spoken, managers believe themselves to be above the median (Duhaime and Schwenk, 1985; Kahneman and Lovallo, 1993) and therefore they tend to overestimate the controllability of events. Managers with a great tendency to the illusion of control bias tend to overestimate their own skills and are therefore inclined to believe they can influence a certain outcome (Duhaime and Schwenk, 1985; Kahneman and Tversky, 1974; Schwenk, 1984). Even though managers perceive some risks, they believe they can overcome these risks because of their capabilities (March and Shapira, 1987). It can be expected that managers with a high level of illusion of control strongly believe their managerial abilities can surmount challenges (Kahneman and Lovallo, 1993; Simon et al., 1999). Since they rely on their skills to control an outcome, they will perceive fewer risks (Busenitz, 1999). For those reasons, it can be expected that managers with a high level of illusion of control are less perceptive to possible dangers and are therefore more willing to take risks.
Regression fallacy suggests that people erroneously neglect the regression to the mean principle
(Tversky and Kahneman, 1974). This statistical phenomenon implies that when the outcome of one situation is extremely high or extremely low, it is very likely that in the next situation, the outcome will be shifted towards the mean (Tversky & Kahneman, 1974). Simply stated: when someone/thing performed very bad/well the first time, people will believe the next time it will perform the same. They believe the extreme outcome represents the mean. Because the first result was extraordinary, people took some sort of action. People with a high level of regression fallacy will believe to have influenced the situation with these actions. However, the regression to the mean phenomenon shows that if the outcome was extreme the first time, it will naturally tend to move closer towards the mean the second time (Tversky & Kahneman, 1974) and thus these persons overestimated their own actions. As can be recalled, managers are very confident about their own skills and abilities and they will believe their skills are able to influence certain situations (Duhaime and Schwenk, 1985; Kahneman and Tversky, 1974; Schwenk, 1984). Managers with a high level of regression fallacy erroneously overvalue their own actions and believe they are able to influence extreme outcomes. As a consequence, they are inclined to make erroneous links between their actions and a certain performance and will pay no attention to the natural cycle of evolution. They are now even more confident about their abilities to influence extreme outcomes, which make them perceive fewer risks and more inclined to make decisions containing a lot of risks.
People with a high level of base rate fallacy or sample size fallacy will be more likely to base decisions on their ideas or experiences instead of on statistical information, such as prior probabilities and a large sample size. These managers will rather use information consistent with their ideas. Thus, to match their preferences they will rather use information that is positively biased than using statistical
information. People with a high level of illusion of control or regression fallacy will overestimate their skills and the controllability of events. They believe they can control situations determined by luck or believe they are able to influence extreme outcomes. These managers will not be discouraged by the risks involved, because they believe their skills and abilities will enable them to overcome these risks. They will therefore also perceive fewer risks. Hence, it can be expected that managers with a high level of base rate fallacy, sample size fallacy, illusion of control or regression fallacy tend to take more risks than managers with a low level of one of these biases. Therefore, the following hypothesis can be derived:
Hypothesis 1: The higher the degree of a manager’s base rate fallacy, sample size fallacy, illusion of
control or regression fallacy – biases resulting from the representativeness heuristic, the higher the risk taking propensity.
Availability heuristic
People relying on the availability heuristic assess information on the basis of events that spring easily to mind. People associate events with memories that can be easily recalled. Thus, only limited information is used to make decisions, which in turn can lead to systematic biases (Tversky and Kahneman, 1974). An illustration; parents are more worried about their child being kidnapped, than involved in a fatal car accident, because the former comes to mind more easily. This study will focus on the following two biases originating from the availability heuristic: illusory correlation and overconfidence (Russo and Schoemaker, 1992; Tversky & Kahneman, 1974).
The illusory correlation bias occurs when people erroneously believe that a correlation exists between two events. Due to the conspicuousness of events, behaviours or people (Sherman, Kruschke, Sherman, Percy, Petrocelli, Conrey, 2009; Tverksy and Kahneman, 1974), it is easy to recall those
situations. People feel like these events happen frequently and believe these events correlate with each
other (Tversky and Kahneman, 1973; 1974; Fiedler, Hemmeter and Hofmann, 1984). Examples of the illusory correlation bias include the belief that it will rain after a rain dance (Chapman, 1967), as well as the belief that suspicious persons are more likely to draw idiosyncratic eyes (Chapman and Chapman, 1967). People inclined towards a high degree of illusory correlation would judge a certain situation according to these erroneous relationships, which could lead to biased decisions and risky situations (MacDonald, 2000). As stated before, managers seek out information that is consistent with their preferences. Managers prefer to assess situations on the basis of information that satisfies their own ideas. (Boulding et al, 1997; Simon et al., 1999; Sleesman et al., 2012). To illustrate: a manager would merely remember positive (whether or not distorted) stories about investing in high technology innovations and the successes these innovation brought to different firms. Those few success stories can be remember
more easily and may cause the manager to believe a correlation exists. Subsequently, managers are more willing to believe that investing will lead to successes and they will take the risk of investing a lot of money in a project. However, this erroneously reasoning could lead to systematic errors (Kahneman et al., 1982). Managers with a high level of illusory correlation tend to believe events are correlated, especially if they cohere to their initial ideas. They take decisions with regard to these incorrect relationships, because it satisfies their arguments. These managers will be inclined to take more risks, since decisions are made on the basis of wrong background information.
Overconfidence occurs when people believe they are in the right, while in fact they are not
(Fischhoff, Slovic and Lichtenstein, 1977). They “fail to know the limits of one’s knowledge” and will treat their assumptions as facts (Russo and Schoemaker, 1992; Simon et al, 1999). Because these persons are unaware of their overconfidence and so they do not know other relevant information is available, they truly believe their assumptions are correct. They base their confidence in decisions made on the “ease
with which they can recall reasons for confidence” (p.117, Simon et al., 1999). Simon and Houghton
(2003) argue that overconfidence is more likely to appear in uncertain situations, such as product introductions. Camerer and Lovallo (1999) dedicate overconfidence to the optimism about ones abilities. To illustrate, an example about R&D managers can prove to be useful; R&D managers are responsible for go/no go decisions and eventually may feel personally responsible for negative consequences, because of their commitment to a certain product (Staw, 1976). Managers with a high level of overconfidence may believe they know everything, while, in fact, they do not. Because of manager’s optimism (Boulding et al., 1997) and their willingness to act in a manner consistent with their initial ideas (Simon et al., 1999l; Sleesman et al., 2012), they will trust information that supports their preferences (Rabin and Schrag, 1999). They will therefore not look for additional information to make more accurate decisions, as they do not know other information exists and find the information at hand satisfactory. As a consequence, manager do not take into consideration all information during the decision making process, which in turn could lead to a strategy involving a lot of risks.
Rabin and Schrag (1999) explain that people tend to distort information that contradicts their ideas into information that supports their ideas. Also, managers search for information that is in line with their initial decisions (Sleesman et al., 2012). Managers with a high level of illusory correlation or overconfidence will believe certain relationships exist, or will assume they are aware of all essential information. Thus they are unaware that their information is incorrect or that other relevant information exists. They will make decisions based on data that can be remembered easily, but that information tends to be positively biased towards their own ideas. Consequently, managers are likely to choose strategies that come with a lot of risks. Because of these reason, it can be expected that managers with a high level
of illusory correlation or overconfidence, are more likely to take risks. Therefore, one can come to the following hypothesis:
Hypothesis 2: The higher the degree of a manager’s illusory correlation or overconfidence – biases
resulting from the availability heuristic, the higher the risk taking propensity.
2.2 Regulatory foci
In recent research, regulatory focus theory has received much attention. According to Haws, Dholakia and Bearden (2010) it is of great importance because psychological processes and behaviours can be predicted by this theory. Regulatory focus theory distinguishes between two different ways of approaching a desired outcome or achieving a goal (Higgins, 1997). In short, a distinction is made between those who approach pleasure, and those who avoid pain, respectively either through eagerness or vigilance (Higgins, 1997; 2005). Goals can be pursued most effectively when one’s way of achieving this goal, fits someone’s regulatory orientation (Higgins, 2000; 2005; Spiegel, Grant-Pillow and Higgins, 2004).
Individuals with a promotion focus orientation try to achieve their goals by hard working, self-improvement and -development (Crowe and Higgins, 1997; Higgins, 2005). They are in search of gains and non-gains, i.e. their focus is on the absence or presence of positive outcomes (Aaker and Lee, 2006; Crowe and Higgins, 1997; Haws et al., 2010; Langens, 2007). These outcomes are characterized by hopes, wishes and aspirations made by either themselves, or by significant others (Higgins, 1997). Promotion-focused individuals are more inclined to generate new ideas and are more open to change (Brockner, Higgins and Low, 2004; Trevelyan, 2008). They typically worry less about negative outcomes, or the possibility of failure, since their focus is on positive results. Due to the eagerness of these persons, they are naturally more inclined to take risks than prevention-oriented individuals (Boldero and Higgins, 2011; Crowe and Higgins, 1997; Gino and Margolis, 2011; Hamstra et al., 2011). The focus of managers with a high level of promotion focus is on positive outcomes, and they will be less worried about making mistakes. These managers will try to stay consistent to their initial ideas and act so as to not deviate from them, even though their actions can come with failures. Because they are not afraid of making mistakes, it can be expected that they are more willing to take risks. For example, managers with a promotion focus striving for a new product investment will act more quickly when new opportunities arise concerning this investment (Spanjol, Tam, Qualls and Bohlmann, 2011). Even though there is a real chance of failure, the chance to hit the desired end-state is greater as well (Higgins, 1997; Spanjol et al., 2011). Therefore, the following hypothesis can be derived:
Hypothesis 3: The higher the degree of a manager’s promotion focus, the higher the risk taking
propensity.
Individuals with a prevention focus orientation will typically try to reach their targets with vigilance. Their goals are based on their, or significant others’, duties, responsibilities and obligations (Higgins, 1997). They approach goals by acting responsible and with a certain degree of confidence (Crowe and Higgins, 1997; Higgins, 2005). Prevention-focused individuals focus on negative outcomes (Langens, 2007) and they are sensitive to losses and non-losses (Aaker and Lee, 2006). Since they are afraid of making mistakes (Higgins, 1997), they will try to avoid making them (Crowe and Higgins, 1997; Spanjol et al., 2011), causing them to make safer choices (Hamstra et al., 2011). In turn, they will try to avoid risky or new situations, because they cannot assure the absence of a negative outcome (Brockner et al., 2004; Treveleyan, 2008). Crowe and Higgins (1997) stated that these persons have a greater conservative response bias, and that they will try to avoid making mistakes (Higgins, 1997). Managers with a prevention focus also try to stay consistent to their initial ideas, but their approach is different to that of promotion-focused managers. Prevention-focused managers will feel very responsible for their decisions and outcomes and will try to make safe decisions. They want to make the correct decisions, ideally without any involved failures. Therefore, it can be expected that they will be less willing to take risks. Spanjol et al. (2011) provide the example of prevention-focused managers who try to avoid situations involving possible failure when decisions have to be made about a new product investment. Therefore, they will explore all information and deliberate long before making a decision. Therefore, one can come to the following hypothesis:
Hypothesis 4: The higher the degree of a manager’s prevention focus, the lower the risk taking
propensity.
2.3 Influence on heuristics
Research on the influence of the regulatory foci on the six biases is quite new. However some of them have been examined in previous literature. Langens (2007) found support for the relationship between the regulatory foci and illusion of control. After three studies, he concluded that promotion focus has a positive influence on illusion of control. Promotion-focused persons are very set on matching their desired end states and tend to remember positive events better than negative events. Promotion-focused individuals with a high level of illusion of control believe they are able to control those positive outcomes. Prevention-focused persons would more easily remember the moments the desired outcome did not match the actual outcome and will believe they are not able to exert control over a specific
outcome. Furthermore, Trevelyan (2008) argues that overconfidence holds no relation to the prevention focus, which, as she explains, is caused by the fact that individuals with a prevention focus engage in low risk activities. She also explains that overconfident people tend to engage in riskier situations, because they believe they are knowledgeable or skilled enough to get involved. Further data about the relationships between the six biases and the two regulatory foci does not exist. This study enhances this field by adding more variables.
One can presume that the prevention focus relates negatively to the six biases, because prevention-focused individuals are more careful when taking any decisions and feel highly responsible for their actions. They will try to make safe decisions. Therefore, they will think twice before making decisions and will rely less on heuristics. It can be expected that promotion focus positively influences the biases. As previously mentioned, individuals with a promotion focus act in a positive manner and are ambitious when pursuing their goals and aspirations. They will worry less about making mistakes and will therefore be less afraid to make decisions quickly. Decisions will be taken fast and less careful (Spanjol et al., 2011), which in turn could lead to biases. Promotion-focused managers are very eager to achieve their goals quickly, even though they will have to face failures first. They will tend to make decisions based on information that can be recalled easily. Persons with a prevention focus will, on the contrary, seek for all evidence needed to make decisions and will consider more details concerning the situation. They will thus over-think situations to ensure the absence of negative outcomes (Brockner et al., 2004; Treveleyan, 2008) and will be less inclined to experience biases. Therefore, the following hypotheses can be derived:
Hypothesis 5: The higher the degree of a manager’s prevention focus, the lower the degree of base rate
fallacy, sample size fallacy, illusion of control, regression fallacy, illusory correlation and overconfidence.
Hypothesis 6: The higher the degree of a manager’s the promotion focus, the higher the degree of base
rate fallacy, sample size fallacy, illusion of control, regression fallacy, illusory correlation and overconfidence.
2.4 Conceptual model
Figure 2.1: Conceptual model
3. Methodology
3.1 Sample and data collection
Business students from the Rijksuniversiteit Groningen were approached to complete the questionnaire regarding risk taking propensity. A long debate is going on about using students as surrogates for business men (Ashton and Kramer, 1980; Chang and Ho, 2004; Hughes and Gibson, 1991; Khera and Benson, 1970; Tangpong and Ro, 2008). Students can be seen as representative, because they are experienced in the same field and will, in the future, take over the jobs of managers. Asthon and Kramer (1980) and Chang and Ho (2004) have reviewed literature on this debate and both concluded with contradictory outcomes found in previous research. However, Asthon and Kramer (1980) did conclude that students can be used as surrogates for managers in the decision-making process. Therefore, a choice was made to use students in this research. If business students are indeed representative, it will decrease time, costs and effort involved in this sort of research in the future. Therefore, this research should be compared with results of future, similar studies concerning managers to make a proper conclusion concerning this debate. The questionnaire was distributed in the spring of 2012 both digitally and in the form of hard copy. 117 students completed the questionnaire, 108 from the faculty of Business and Economics, and 9 students from other faculties, such as the Faculty of Law and the faculty of Behavioural and Social Sciences. Of the business students, 32.41% is enrolled in an innovation master. The respondents are aged between the ages of 19 and 29, the average age being 22.9. The business faculty consists of relatively
more males, which translates to the distribution of respondents as well; 63.2% male and 36.8% female. It took students about 25 minutes to complete the questionnaire. To increase the response rate, follow-up e-mails were sent and the students were approached at different times and over the course of several days.
3.2 Measurements
Cases and scales from previous research were used in this study. Due to limitations concerning the length of the questionnaire, items with high factor loadings were chosen. Some cases were adapted to fit the student’s situation. A pre-test was necessary because of adjustments made in the questionnaire, as well as to observe if students were able to understand the questions asked. Three business students were asked to complete the questionnaire by thinking out loud. The analysis of the pre-test resulted in the reformulation of items and cases and changes in sentence structure. Appendix A presents the constructs and questions of the questionnaire.
For the dependent variable risk taking propensity three different means of measurement were used. The first method was obtained from Weber, Blais and Betz (2002) and involves ten statements about decision-making concerning five different domains: ethical, financial, health/safety, recreational and social. Respondents were asked to fill in how likely they are to get involved in a certain situation on a 5 point Likert scale. For example, students were asked how likely it is that they would cheat on an exam. Secondly, a few new cases were provided and respondents were asked to choose between an uncertain, risky decision, and a certain, risk averse decision. See appendix A for examples of the cases. The last question is based on Kahneman and Tversky (1979) and taken from Song et al. (UR). Students were asked to choose between 10 options. The first option mentioned is the most risk-averse option and involves receiving €900 for sure. The last option consists of 10% chance of winning €9,000, and 90% chance of winning nothing, and is thus the option that involves the most risks. Even though all options have the same expected value, choosing the first options is optically choosing for more certainty and make respondents feel more confident about their possibility of winning (Song et al., UR). Hence, choosing these options implies that a respondent has a low level of risk taking propensity.
Base rate fallacy occurs when people base their decisions on irrelevant information instead of
statistical information; they rather make assessments based on personal knowledge and ignore base rates (Tversky and Kahneman, 1974; Bar-Hillel, 1980). Base rate fallacy was measured by using four cases, derived from Lynch and Ofir (1989) and Song et al. (UR). Respondents were asked to give the estimated probability of a certain situation. In each case, relevant statistical information was given, supplemented with irrelevant information. For example: “Your friend Emma is looking for a new laptop. You have done
some research and found that sixty per cent of the laptops of the model she prefers need repair after three till four years. After you have left your lecture, she called you to let you know that she looked at it once
more and that she really likes the design and colour of the laptop. What is your estimated probability that Emma will need some major repairs after three till four years?(0-100%)” Respondents with a high level
of base rate fallacy would ignore the statistical information and deviate from the correct answer. The more they deviate, the higher the base rate fallacy.
Sample size fallacy occurs when people make assessments with reference to a small sample size,
even though a bigger sample size would lead to more accurate decisions (Tversky and Kahneman, 1971). This bias is measured on the base of five questions on a 5 point Likert scale. Questions are based on Song et al. (UR) and adapted to fit the student’s situation. For example: “I usually just check the questions from
the last two exams to imagine what questions might be asked when I have to prepare for the exams”. A
higher answer indicates a greater level of small sample size fallacy. Additionally, respondents filled in four cases derived from Tversky and Kahneman (1971; 1974). To illustrate: “A certain town is served by two hospitals. In the larger hospital circa 45 babies are born each day, and in the smaller hospital circa 15 babies are born each day. As you know, about 50% of all babies born are boys. However, the exact percentage varies from day to day. Sometimes it might be higher than 50%, sometimes lower. For a period of 1 year, each hospital recorded the days on which more than 60% of the babies born were boys. Which hospital do you think recorded more such days”? Most respondents answered with “about the
same” while actually the correct answer is the smaller hospital, because a larger sample size is more likely to straight out the distribution between boys and girls. All cases consisted of three answers, in which the answer that indicates the highest level of sample size fallacy was measured with a 5.
Illusion of control is the tendency to believe one can control a certain outcome, while in fact one
cannot exert any influence on it (Langer, 1975; Zuckerman et al, 1996). Questions are derived from Zuckerman et al. (1996) and measured on a 5 point Likert scale.Respondents indicated to what degree a certain situation applied to them; 1 indicating true and 5 indicating not true. An example of a statement is:
“There is no such thing as misfortune; everything that happens to us is the result of our own doing”. Regression fallacy suggests that people mistakenly neglect the regression to the mean principle
and believe that their actions guide them in a certain way (Tversky & Kahneman, 1974). Respondents indicated to what degree they agreed with the statements on a 5 point Likers scale. New items were based on Song et al. (UR) and Tversky and Kahneman (1974). For example: “Punishing students is very
effective in improving students’ grades”. Complementary, three cases were added. To illustrate: “Assuming that your firm operates in a stable economic environment. Two years ago, the sales of your products increased by 15%. You made the decision to increase your advertising budget by 25% last year. However, you just got a report showing that the sales decreased by 5% last year. How likely would you conclude that the advertising was not effective?” Due to the regression to the mean phenomenon, the
et al., UR). Therefore, concluding that advertising was ineffective indicates a high level of regression fallacy.
Illusory correlation occurs when two events happen simultaneously and people erroneously
believe that a correlation exists between those two events (Tversky and Kahneman, 1974; Fiedler et al., 1984). Items are based on Song et al. (UR) and are measured on a 5 point Likert scale. Examples are “sugar makes children hyperactive” or “Guys are more likely to fail exams than girls are”.
People with a high level of overconfidence will fail to know the limits of one’s knowledge and will treat their assumptions as facts (Russo and Schoemaker, 1992; Simon et al, 1999). Questions were made following Klayman, Soil, González-Vallejo and Barlas (1999) and Bernstein, Erdfelder, Meltzoff, Peria and Loftus (2011). Respondents were asked to fill in five 2-answer questions and specify how confident they were that their answer was correct. The more confident a respondent was about answering correctly, even though in reality the answer was incorrect, the higher the level of overconfidence.
The regulatory focus theory describes how individuals pursue their goals. The theory distinguishes between promotion focus and prevention focus (Higgins, 1997). For both foci, respondents ranked five different statements on a 5 point Likert scale following the procedure of Haws et al. (2010). Promotion-focused individuals view their goals as ideals (Higgins, 1997) and are eager to strive for positive outcomes (Langens, 2007). Prevention-focused individuals pursue their goals by acting cautiously, responsible and safe (Crowe and Higgins, 1997; Higgins, 2005). The statements are a composition of Higgins et al. (2001), Carver and White, (1994) and Lockwood, Jordan and Kunda (2002). Respondents were asked to rank statements about themselves and their actions and feelings in a decision making process. For example; “I frequently think about how I can prevent failures in my life” represents the prevention focus. Ranking this statement with 1 or 2 would indicate a high level of prevention focus. The statement “When I see an opportunity for something I like, I get excited right away” represents promotion focus. Ranking this statement with 1 or 2 would imply a high level of promotion focus. In the database these questions were reversed, so that 4 or 5 indicate a high level of promotion or prevention focus.
4. Analysis
4.1 Common method bias
Common method bias may occur because of different reasons (Podsakoff, MacKenzie, Jeong-Yeon Lee and Podsakoff, 2003). Since the only method used in this research is the questionnaire, measures were taken to diminish common method bias (MacKenzie and Podsakoff, 2012). Besides the Likert-scale questions, cases were added, reversed items were used and related items were dispersed
throughout the questionnaire. Furthermore, common method bias was tested statistically by a common latent factor in LISREL (Podsakoff et al, 2003). Items were loaded on a latent variable called “factor”. To ascertain no common method bias, the model has to be less significant than the original model. The model fit of the latent variable model is: 2 /df is 1.57, RMSEA is 0.067, standardized RMR is 0.082, CFI is 0.81, NFI is 0.65 and GFI is 0.95. The model fit of the latent model is not totally sufficient, but the differences compared to the original model are not very big, although worse. Therefore, a mild form of common method bias might exist in this study. However, multicollinearity of the variables is measured as well to preclude multicollinearity.
The multicollinearity of the variables is calculated in SPSS. Table 4.1 shows the tolerance and VIF value for each variable, which are respectively more than .01 and less than .10 (Field, 2009; Myers, 1990). Following the line of reasoning of Myers (1990) multicollinearity does not exist. Table 4.2 presents the correlations, means, Cronbach’s Alpha’s and standard deviations of the variables. The Cronbach’s Alpha’s of the multiple-item constructs range between .607 and .774, which suggests reasonable reliabilities (Field, 2009; Nunnaly, 1978).
TABLE 4.1 Multicollinearity
Variable Tolerance VIF
Base rate fallacy .896 1.117 Sample size fallacy .949 1.054 Illusion of control .895 1.118 Regression fallacy .786 1.272 Illusory correlation .928 1.078 Overconfidence .964 1.037 Promotion focus .885 1.130 Prevention focus .942 1.061
4.2 Confirmatory factor analysis
After the exploratory factor analysis in SPSS, a confirmatory factor analysis was conducted via Maximum Likelihood estimation in LISREL 8.80 on all variables with a metric scale (Hair et al, 1998). Each construct was reviewed and items loading on multiple constructs or with low item-to-construct loadings were deleted. The models showed a good fit (Hair et al., 1998; Byrne, 1989). The model fit for risk-taking propensity is 2 /df is 1.018, RMSEA is 0.012, standardized RMR is 0.041, CFI is 1.00, NFI is 0.93 and GFI is 0.98. For the biases: 2 /df is 1.47, RMSEA is 0.064, standardized RMR is 0.043, CFI is
0.95, NFI is 0.90 and GFI is 0.99. For the regulatory foci: 2 /df is 2.39 RMSEA is 0.11, standardized RMR is 0.06, CFI is 0.95, NFI is 0.92 and GFI is 0.96. The measurement models of the CFA’s are depicted in Table 4.3. Note that the factor loading of BR2 is greater than 1, which is acceptable but might suggest a certain degree of multicollinearity (Perryer, 2005). However, the previous paragraph demonstrated no multicollinearity.
Because of the mediating character of the study, structural equation modelling (SEM) was used to test the hypotheses. The model is built on the measurement model of the CFA’s. The full model presented a good fit (Hair et al., 1998; Byrne, 1989): 2 /df is 1.36, RMSEA is 0.055, standardized RMR is 0.08, CFI is 0.88, NFI is 0.84 and GFI is 0.87.
TABLE 4.3
Confirmatory Factor Analysis
Risk-taking propensity: 2 /df is 1.018, RMSEA is 0.012, standardized RMR is 0.041, CFI is 1.00, NFI is 0.93 and GFI is 0.98. Biases: 2 /df is 1.47, RMSEA is 0.064, standardized RMR is 0.043, CFI is 0.95, NFI is 0.90 and GFI is 0.99. Regulatory foci: 2 /df is 2.39 RMSEA is 0.11, standardized RMR is 0.06, CFI is 0.95, NFI is 0.92 and GFI is 0.96
See the appendix for the meaning of each variable
5. Results
Table 5.1 shows the results of the study. This research focuses on the relationship between regulatory foci and biases on risk-taking propensity, and biases mediating the relationship between the foci and risk-taking propensity. Results turned out quite different than expected. They will be discussed below.
Hypothesis 1 and 2 deal with the relationship between the cognitive biases and risk-taking propensity. Unfortunately, all hypotheses had to be rejected because the relationships between the variables proved to be insignificant. Hypothesis 3 and 4 respectively suggest a positive relationship between promotion focus
Construct Item
Factor Loading T-value
Base rate fallacy BR2 BR3 1.52 0.41 2.37 2.05 Promotion focus PROF1 PROF3 PROF4 PROF5 0.72 0.54 0.62 0.65 8.83 6.21 6.94 7.57 Risk-taking propensity R2 R5 R7 R8 R9 0.51 0.59 0.46 0.58 0.59 4.36 4.79 3.63 5.14 3.76
and risk-taking propensity and a negative relationship between prevention focus and risk-taking propensity. Both foci have a significant relationship to risk taking propensity. However, in contrast to the expectations, promotion focus seems to have a negative relationship to risk taking propensity (β = -0.51, p = <0.01). In line with earlier reasoning, prevention focus also has a negative relationship to risk taking propensity (β = -0.48, p = <0.01). Hypothesis 5 and 6 concern the relationships between the regulatory foci and the biases. Regrettably, support for only one relationship could be found. As expected, promotion focus has a positive effect on regression fallacy (β = 0.23, p = <0.05). Accordingly, hypothesis 5 partly found support. Other relationships provide no support for hypothesis 6.
6. Discussion
In current research, managerial risk taking behavior has gained much attention because of its importance for organizational success (Rubera and Kirca, 2012; Tien-Shang Lee and Munir Sukoco, 2011). This study examined factors that were expected to influence risk taking propensity. Unfortunately, only a few significant relationships were found, which is quite surprising, since other studies have been able to find relationships about the six biases, the regulatory foci and risk taking propensity (e.g. Langens, 2007; Simon et al., 1999; Trevelyan 2008).
None of the biases mentioned had any significant effect on risk taking propensity, which is contradictory to initial expectations, as well as to those of other researchers (Simon et al., 1999; Simon and Houghton, 2003). However, some concerns exist about the biases and/or their relationship to risk taking propensity (read: Cosmides and Tooby, 1995; Dorn and Huberman, 2005; Gigerenzer, Hell and Blank, 1988; Keh, Foo & Lim, 2002; Koehler, 1996; Nosic and Weber, 2010). Furthermore, only one relationship between the regulatory foci and the six biases has proved to be significant. Despite the fact that only some of these relationships have been investigated before (Langens, 2007; Trevelyan, 2008), and thus only limited evidence exists regarding them, it is still quite surprising that these results appeared to be insignificant. Promotion-focused individuals tend to worry less about mistakes, whereas prevention- focused persons think more carefully before taking decisions, which will respectively lead to a higher level of the use of heuristics and its biases, and a lower level of its use. However, relationships were not found and I believe in this study the most detrimental reason for these findings is the sample consisting of students, which will be discussed later on.
The prevention and promotion focus were both negatively related to risk taking propensity. Prevention focus was expected to be negative, since risks might cause the presence or absence of negative outcomes, and prevention oriented persons will try to avoid that. This relationship is also in line with the reasoning of Boldero and Higgins (2011), Crowe and Higgins (1997), Gino and Margolis (2011) and
TABLE 4.2
Means, Standard deviations, Correlations and Reliabilities
Figures on the diagonal line represent Cronbach’s α *p<0.05; **p<0.01
Construct: Mean St.Dev. 1 2 3 4 5 6 7 8 9
1. Risk-taking propensity 2.09 0.70 .61
2. Base rate fallacy 0.80 0.98 .067 .65
3. Sample size fallacy 3.56 1.28 .083 -.070 -
4. Illusion of control 2.63 1.16 -.062 -.061 .042 - 5. Regression fallacy 2.32 0.95 -.126 -.226* .168 .274** - 6. Illusory correlation 2.73 0.90 .033 .176 -.083 .052 .065 - 7. Overconfidence 0.65 1.16 -.061 -.005 .008 .043 -.056 .067 - 8. Promotion focus 3.73 0.71 -.202* -.162 -.067 -.005 .240** .072 .116 .77 9. Prevention focus 2.95 1.12 -.315** .094 .011 -.118 -.044 -.115 .089 -.108 -
TABLE 5.1 Results
Significance levels are based on unstandardized coefficients *p<0.05; **p<0.01
Model fit:df = 141.2, 2 = 104, 2 /df is 1.36, RMSEA is 0.055, standardized RMR is 0.08, CFI is 0.88, NFI is 0.84 and GFI is 0.87
Independent variables: Risk-taking propensity Base rate fallacy (H5/6) Sample size fallacy (H5/6) Illusion of control (H5/6) Regression fallacy (H5/6) Illusory correlation (H5/6) Over- Confidence (H5/6)
Base rate fallacy (H1a) 0.17 Sample size fallacy (H1b) 0.06
Illusion of control (H1c) -0.09 Regression fallacy (H1d) -0.83
Illusory correlation (H2a) 0.43
Overconfidence (H2b) -0.23
Promotion focus (H3) -0.48** -0.07 0.01 -0.01 0.23* 0.07 0.17
Hamstra et al., (2011). However, it is surprising that promotion focus is related in the same direction, which is clearly not in line with the reasoning of the previous researchers. Questions about the prevention and promotion focus were provided at the end of the questionnaire and it is possible that students had trouble finding the motivation to read accurately and might have filled in these questions randomly. Another possibility is that they had an idea about the subject of the questionnaire by then, and tried to give a “socially acceptable response” (MacKenzie and Podsakoff, 2012).
It is surprising that the six biases emanating from the representativeness and availability heuristic, had almost no effect on the other variables. For future research, it is therefore important to evaluate the method used. One of the purposes of this study was to use students as representatives of managers. However, even without considering comparing the results of students with these business men in the future, I think this questionnaire was not appropriate for students. Despite all measures taken to provide an interesting, motivating and relatively easy questionnaire for students, the results still turned out to be disappointing. Therefore, it is important to shed light on the motivation and interests of students. MacKenzie and Podsakoff (2012) conducted research on to the motivation and ability of respondents to fill in a questionnaire accurately. This questionnaire had possibly demotivated students in three ways. Firstly, the questionnaire could have been too long for students. Students were motivated because the subject seemed interesting; however, it took students 20 minutes or more to complete it, which proved to be too long for them. Secondly, afterwards, many students complained about the difficulty of the cases, meaning the time they needed to answer each question correctly. MacKenzie and Podsakoff (2012) explain that a seemingly long or difficult questionnaire reduces one’s motivation and ability to fill it in correctly. Lastly, students may not see the advantages of completing the questionnaire, even though they could win prices. Possibly, students lost their concentration, which could have led to biases in the results. Even though the questions were changed into easily understandable questions, some questions remained tough, because this was needed to measure the variables. In a next study, the number of (difficult) questions could be diminished, by providing fewer questions, or by researching fewer variables. Another cause of the disappointing results could be the numerous one item constructs used for the measurements, which might have reduced the reliability of this research
This study strives to give more insights into the literature concerning heuristics, regulatory foci and risk taking propensity. This research extended the existing field by adding regulatory foci to the research about biases and risk taking propensity. It can be expected that the sample consisting of students was the most harmful reason for the surprisingly insignificant relationships. Even though no relationships were found, it can still be considered as a new starting point for future research. Putting this research in a new form and using other occupations such as managers or entrepreneurs as sample, might provide interesting results in the future.
7. Managerial implications
In this research both promotion focus and prevention focus seem to be negatively related to the risk taking propensity of managers, however both foci are quite different. As stated before goals can be pursued most effectively when one’s way of achieving this goal fits someone’s regulatory orientation (Higgins, 2000; 2005; Spiegel, Grant-Pillow and Higgins, 2004). Which would also imply that a person would perform at its best when his/her regulatory focus matches the focus of the organization or department/team. Managers should take this in considerations while creating teams. Furthermore, even though one’s own regulatory focus will always predominate, an organization or its (top)managers could be able to influence one’s regulatory focus temporary. Through an organization’s mission statement, policy, incentive scheme and culture, an individual’s regulatory foci can be shaped in the desired way (Gino and Margolis, 2011; Haws et al., 2004). Emphasis on gains would lead to an increase in the promotion focus level, whereas emphasis on losses would increase the level of prevention focus (Gino and Margolis, 2011). However, with regard to risk taking propensity this research has found no big differences between both foci, which makes it useless to make a difference between both foci. However, promotion focus appeared to have a positive relationship to regression fallacy. In this research we have not found the importance or usefulness of regression fallacy for risk taking propensity. However, if it is of importance to have a higher level of regression fallacy, managers could be temporarily directed to be more prone to the promotion focus in the same way as mentioned before. But it is important to remember that one’s regulatory focus will always predominate (Haws et al., 2004), and therefore it is important to take this in consideration during the hiring process.
8. Limitations
This study consisted of various limitations; partly due to the limitation posed by the fact that the sample consisted of students, and partly due to the many one item constructs used. The first limitation has been analyzed in the discussion. The questionnaire was quite long and difficult which could have reduced the student’s motivation to complete it accurately. Therefore I would propose to include fewer variables and/or to choose between statements or cases in the future, instead of using both. I tried to increase motivation by offering the students a chance to win prices, but this did not work out. For these reasons I would advise, for future research including students, to look for more positive incentives to increase their motivation. Moreover, for further similar research I would recommend to use managers instead of students to discover psychological characteristics of managers. In the end students do differ from managers; they are in a different phase of their life and they do not have the same experiences and skills as managers (Ashton and Kramer, 1980), which would logically results in biased answers. I further
believe that in general students are less motivated to complete a questionnaire seriously. Moreover, respondents filled in the questionnaire at the faculty or on the internet, which gave them time think carefully about the questions. They possibly answered differently than in a situation where time is short, the situations where heuristics are mostly used (Kahneman, 2003; Tversky and Kahneman, 1974).
Also, the confirmatory factor analysis turned out to be quite disappointing as well. The analysis consisted of many one item constructs, which might have influenced the reliability of the data and subsequently the results. Most constructs consisted of one question, because after the CFA and EFA no consistency was found between the questions.
This study could be extended in different ways. Eventually, most organizations work in teams and therefore it would be interesting to measure the variables in management teams to discover the influence of other team members. Besides it would be interesting to found out the best way of managers to encourage or, in some situations discourage, risk taking behaviour when needed. Furthermore, it might be interesting to discover other characteristics of managers and their relationship to the biases, regulatory foci and risk taking propensity. For example, how manager who tend to escalate to commit to a project (e.g. Staw, 1976) are influenced by biases or regulatory foci.
References
Aaker, J. L., & Lee, A. Y. (2006). Understanding regulatory fit. Journal of Marketing Research (JMR),
43(1), 15-19.
Amabile, T. M., Conti, R., Coon, H., Lazenby, J., & Herron, M. (1996). Assessing the work environment for creativity. Academy of Management Journal, 39(5), 1154-1184.
Ashton, R. H., & Kramer, S. S. (1980). Students as surrogates in behavioral accounting research: Some evidence. Journal of Accounting Research, 18(1), 1-15.
Bar-Hillel, M. (1980). The base-rate fallacy in probability judgments. Acta Psychologica, 44(3), 211-233. Bernstein, D. M., Erdfelder, E., Meltzoff, A. N., Peria, W., & Loftus, G. R. (2011). Hindsight bias from 3 to 95 years of age. Journal of Experimental Psychology: Learning, Memory, and Cognition, 37(2), 378-391.
Boldero, J. M., & Higgins, E. T. (2011). Regulatory focus and political decision making: When people favor reform over the status quo. Political Psychology, 32(3), 399-418.
Boulding, W., Morgan, R., & Staelin, R. (1997). Pulling the plug to stop the new product drain. Journal
of Marketing Research (JMR), 34(1), 164-176.
Brockner, J., Higgins, E. T., & Low, M. B. (2004). Regulatory focus theory and the entrepreneurial process. Journal of Business Venturing, 19(2), 203.
Bryant, P., & Dunford, R. (2008). The influence of regulatory focus on risky decision-making. Applied
Psychology: An International Review, 57(2), 335-359.
Busenitz, L. W. (1999). Entrepreneurial risk and strategic decision making: It's a matter of perspective.
Journal of Applied Behavioral Science, 35(3), 325-340.
Byrne, B.M. (1989). A primer or LISREL: Basic applications and programming for confirmatory factor
analytic models. NEW York: Springer.
Cabrales, Á. L., Medina, C. C., Lavado, A. C., & Cabrera, R. V. (2008). Managing functional diversity, risk taking and incentives for teams to achieve radical innovations. R&D Management, 38(1), 35-50. Camerer, C., & Lovallo, D. (1999). Overconfidence and excess entry: An experimental approach.
American Economic Review, 89(1), 306-318.
Carver, C. S., & White, T. L. (1994). Behavioral inhibition, behavioral activation, and affective responses to impending reward and punishment: The BIS/BAS scales. Journal of Personality & Social
Psychology, 67(2), 319-333.
Chang, C. J., & Ho, J. L. Y. (2004). Judgment and decision making in project continuation: A study of students as surrogates for experienced managers. Abacus, 40(1), 94-116.
Chapman, L. J. (1967). Illusory correlation in observational report. Journal of Verbal Learning & Verbal
