A study of how the better-than-average effect can be minimised in low performers through individuation and personal contact

(1)

A study of how the better-than-average effect can be minimised in low

performers through individuation and personal contact.

Research has shown that the better-than-average effect can be minimised through the use of individuation and personal contact. However, research has not shown whether these factors also influence individuals situated in the bottom percentile the same way. This study was conducted in order to demonstrate that the magnitude of the better-than-average effect depends on the level of abstraction for both the general sample and the bottom percentile. While this study has proven that personal contact reduces the better-than-average effect in both the general sample and the bottom percentile, the same cannot be said about individuation. This study showed that individuation did not reduce the better-than-average effect in its group when it was compared to the average college student group. This implies that the level of individuation was not significant enough for the participants to be able to consider the level of abstraction less ambiguous.

Sayed Ahmad S.A.S.H. Sayed Hasan 10630473

Rob van Hemert 17.08.2016

(2)

Page

I. Introduction 3

II. Literature review 5

III. Methodology 12

IV. Results and Discussion 13

V. Conclusion 23

Bibliography 26

(3)

I. Introduction

“I cannot say this too strongly: Do not compare yourselves to others. Be true to who you are, and continue to learn with all your might.” (Daisaku Ikeda, 2013)

According to Daisaku Ikeda, people should focus on cultivating their own abilities, without taking into consideration how their peers are performing compared to themselves. However, during this age, in many societies, success in activities, such as school, job, or sports, is seen as a function of how one’s ability and performance compare to those of others (Burson, Klayman & Larrick, 2006). For this reason, estimating one’s standing relative to others can have a significant impact on both the life choices they make and on the satisfaction that comes from those choices.

Nevertheless, when people tend to estimate themselves, they are susceptible to biases. One such bias is the self-serving attribution bias which, according to Miller and Ross (1975), refers to attributing success to personal factors while associating external factors to failures. A clearer explanation of this phenomena is given by Libby and Rennekamp (2012), who state that those people who do well attribute their performance to their own skill and effort, while those who perform poorly attribute their failure to either to the difficulty of the task or to bad luck. However, Miller and Ross (1975) argue that there is not enough research developed in the attribution of causality for success and failure at skill tasks in social contexts. They explain that there is no clear evidence that questions whether a person attributes more causality to themselves when performing a successful performance than they would for an equally well performance of someone else. For this reason, Alicke, Klotz, Breitenbecher, Yurak & Vredenburg (1995) researched and discovered that individuals who estimate how their performance is in comparison with others tend to evaluate themselves more positively than others.. This phenomena is called the better-than-average effect, which is considered by Alicke et al. (1995) as a type of self-serving attribution bias. For this reason, the better-than-average effect will be considered the dependent variable of the present paper.

Furthermore, when individuals engage in estimating their performance in relation to others, both top and bottom performers suffer from an inaccuracy in their self-assessment (Ehrlinger, Johnson, Banner, Dunning & Kruger, 2008). While bottom performers have been proven to overestimate their ranking, which is in accordance with the better-than-average effect, the top performers have showed that they suffer from underestimation. These displays of erroneous self-assessment between the two types of performers have shown to have different causes. The top performers consider the task to be easy and, thus, consider that their peers have done the same. This leads to them judging their own performance as unexceptional. On the other hand, Kruger and

(4)

Dunning (1999) consider that bottom performers are less likely to distinguish correct from incorrect answers than their more competent peers. Thus, they overestimate their performances.

The focus of this paper will be directed towards the poor performers’ capacity to erroneously estimate their performance. An explanation for this can also be seen in Kruger and Dunning (1999), who consider that a metacognitive deficit can explain the tendency to inflate self-appraisals.

Intriguingly, current literature only investigates how bottom performers overestimate their results when comparing themselves with their peers (e.g., Kruger & Dunning, 1999; Schlosser et. al, 2013; Dunning et al., 2003). While these studies prove that there is a difference in estimation, they do not provide solutions on how to minimise this inaccuracy. This paper will attempt to demonstrate how to minimise the miscalibration between actual and perceived standing. This assertion will be developed below.

For the purpose of better understanding the subject matter, some terms will be briefly introduced, with a more detailed analysis of them to be present in Section II. The first term, overconfidence, is divided into three definitions by Moore and Healy (2008): overestimation, over-placement, and over-precision. These three definitions form the theory of confidence, which states that people often have imperfect information about their own performances but even worse information about the performance of others. The second term, performance, is described by Kruger & Dunning (1999) as being the skills necessary to evaluate the competence in a certain domain. Lastly, the third term, individuation, is referred by Alicke et al. (1995) as the recognition of a person’s identity.

With the introduction of these terms, the purpose of this paper will be to investigate how the usage of the moderators, individuation and personal contact, reduces the better-than-average effect when low performers are engaged in intellectual tasks. Thus, the aim of the paper will be to what extent does individuation and personal contact minimise the better-than-average effect in low performers when engaging in intellectual tasks.

In order to accomplish the aim of the paper, a survey study will be used. This survey will be answered by students attending the introduction course of industrial engineering at the Kuwait University. The survey that will be conducted will be a between subjects design, with 2 experimental groups and a control group. While all three groups will be given the same 15 logical reasoning questions to answer, the difference will be in the targets with whom the students will compare themselves. The control group will compare themselves with the average college student, while the first experimental group will compare themselves with a specific target, which in this case will be a professor with whom they have no personal contact. The second experimental group will

(5)

compare themselves again with a specific target, this time being their own professor. In order to simplify the identification of these groups, the following notations will be used. The group that compares with the average college student will be noted as Group 1, while the group that compares with an individuated person without contact will be named Group 2. Lastly, the group that compares with an individuated person with whom it also had personal contact will be termed Group 3.

The remainder of this paper will be divided into sections. In section II, a literature review will be conducted that will enlist all the present relevant theory, along with current findings on the subject matter. With the help of these, hypotheses will be constructed, which will help in providing an answer to the research question. Furthermore, the terms presented above will be analysed in more detail. Following the literature review, section III will be the Methodology. This section will contain a detailed description on the participants and the content of the survey. Afterwards, in section IV, Results and Discussion, the results will be analysed and discussed. Finally, an answer to the question of the paper will be provided in Section V, Conclusion, which will also contain the limitations of the present study and future considerations.

II. Literature review

In this section, all the present relevant theory and current findings will be identified and subjected to reinforcement or criticism. Additionally, the discussed literature will contribute to the formulation of hypotheses that this current paper will use in order to answer the central research question. The literature review will first discuss the role that overconfidence has on the estimation of performance. Afterwards, the self-serving attribution bias will be presented. This will be followed by a discussion regarding a type of self-serving attribution bias, the better-than-average effect bias. Finally, the reason why low performing individuals are more prone to overestimating their performance will be analysed.

The role of overconfidence

When discussing the reason people inflate their perceived performance, Moore and Healy (2008) consider that overconfidence may be the source of this phenomena. Furthermore, Plous (1993, p.217) considers that, “No problem in judgment and decision making is more prevalent and more potentially catastrophic than overconfidence”. For this reason, Moore and Healy (2008) concluded that overconfidence can be used to explain wars, entrepreneurial failures and even stock market bubbles. In order to better understand the role that overconfidence plays in the gap between actual and perceived performance, Moore & Healy (2008) give three definitions of overconfidence.

(6)

The first definition of overconfidence is “the overestimation of one’s actual ability, performance, level of control, or chance of success” (p. 502). In order to easily distinguish this definition of overconfidence from the rest, this definition was named “overestimation” by Moore & Healy (2008). One example of overestimation is when a student is asked, after an examination, how well does he think he did. If the actual value is lower than the perceived value then it is considered to be an overestimation.

The second definition of overconfidence is when “people believe to be better than others, such as when a majority of people rate themselves better than the median” (p.502). Moore & Healy (2008) call this “over-placement”. An example of this is when a student is asked where she considers that she ranked. If the actual rank is lower than the perceived rank then it is an over-placement.

The third definition of overconfidence is “excessive certainty regarding the accuracy of one’s beliefs” (p.502), or “over-precision”. The example provided my Moore & Healy (2008) best describes over-precision. They state that if 100 students are 90% certain that they answered at least 5 out of 10 quiz questions but, in fact, only 20% of them have scored 5 or more, then their erroneous certainty would indicate over-precision.

Intriguingly, Moore & Healy (2008) consider that there are three notable problems with the research that is available on overconfidence. Firstly, the most popular research paradigm confounds overestimation and over-precision. In other words, it is impossible to distinguish between the relative influence of each. A second problem with overconfidence is that the literature has notable instances of under-confidence. Similar to overconfidence, under-confidence is formed of underestimation, under-placement and under-precision. The third problem of overconfidence is the inconsistency between overestimation and over-placement. Moore & Healy (2008) state that easy tasks, which produce the most underestimation, also produce the most over-placement. On the other hand, hard tasks, which produce the most overestimation, produce the most under-placement. One reason for this inconsistency is the fact that overestimation research has focused on difficult domains, such as challenging trivia questions (Campbell, Goodie, & Foster, 2004), while research about over-placement has tended to focus on easier tasks, such as driving a car or getting along with others (Svenson, 1981).

In order to account for all the problems, Moore & Healy (2008) have developed a theory, called the Theory of Confidence. This theory states that, after experiencing a task, people often have imperfect information about their own performances but even worse information about the performances of others. This results in people’s post-task estimates of themselves to be regressive,

(7)

while the estimates of others to be even more regressive. Therefore, when performance is especially high, people will underestimate their own performance, underestimate others even more, which will make them think that they are better than others. When performance is low, people will overestimate themselves, overestimate others more, and thus consider themselves worse than others (Moore & Healy, 2008). This means that the theory states that there is a negative correlation between estimation and placement. For this reason, this paper will focus on overestimation and over-placement.

The present paper will use the Theory of Confidence's assumption that people have imperfect information in order to test how people estimate themselves when comparing to different targets (in the case of this paper, with the Average College Student, a College Professor, and a College Professor with whom students had personal contact). Therefore, when performance is low, people will tend to overestimate their performance. Thus, the following hypothesis can be suggested:

H1: When performance is low, people will overestimate their performances.

Self-serving attribution bias

While the previous sub-section discussed the effect that overconfidence has on people, this part will look at one possible cause of overconfidence. Libby and Rennekamp (2012) state that one reason for managers to engage in forecasts is managerial overconfidence. This managerial overconfidence results from past positive performances, which lead the manager to enlist self-serving attributes in order to explain these positive performances. Additionally, the self-self-serving attribution can lead managers to be too confident that future performances will improve. While this has proven to be one of the factors that cause overconfidence, more insights in the origins and development of the self-serving attribution bias is necessary to understand the full scope of it. In support of these claims the current paper will present the following.

The origin of the self-attribution bias can be traced back to Heider’s (1958) “naive analysis of action” model. Heider (1958) states that in order to select an acceptable causal attribution depends on two factors: “(1) The reasons have to fit the wishes of the person and (2) the datum has to be plausibly derived from the reason” (p.172). Thus, Ross & Miller (1975) consider that, if it plausible to do so, people try to explain their own behaviour in terms that will flatter them and also put them in a good light. Additionally, this model is consistent with the notion of “effective control” developed by Kelley (1971). He considers that it is important for individuals to be able to exercise control over their environment. He describes the relationship between the need for effective control and the attribution process as follows: “The attribution to self of success and the attribution to

(8)

external factors of failure provides for the continuation of control attempts” (p.23). For this reason, Ross & Miller (1975) describe the self-serving attribution bias as the association of successes to inner factors, while failures are associated to exterior ones.

While most discussions acknowledge the existence of the self-serving biases in perception of causality, some authors have questioned the empirical and theoretical basis of the assumption (Bem, 1972; Kelley, 1971). This is why Ross & Miller (1975) conducted a critical examination of the principle’s empirical base. They have shown that there are some evidence that support the basis of the self-serving bias, however they argue that there is little experimental evidence that shows the fact that people in general attempt to avoid blame for unfavourable outcomes by associating responsibility to external factors. One possible explanation for this can be seen in the research of Lerner (1970), who states that people are in fact willing to accept responsibility for nontrivial negative consequences. However, Ross & Miller (1975) consider that a possible reason for why people do not blame external factors for failures is that the studies failed to touch the necessary level of ego involvement, which is important in extracting self-protective reactions.

Intriguingly, recent research conducted by Campbell & Sedikides (1999) have managed to test whether the presumption made by Ross & Miller (1975) was indeed a possible reason for the little evidence in support of the self-serving bias. They showed that self-threat can lead to displays of self-serving attribution bias. In order to better understand how self-threat can lead to an increase in the self-serving attribution bias, the concept itself will be discussed. According to Baumeister (1997), the self is constructed as having at least two aspects. The first one, the representational structure, views the self as an object of perception, self-concept, or me. The second one, the executive function, views the self as the subject of perception, ego, or I. The aspect of the self that is most vulnerable to threat is considered to be the “me” rather than the “I” by Campbell & Sedikides (1999). The reason for this is that the “I” has no inherent qualities that can be threatened. Therefore, the authors refer to the self-threat as the threat to the self-concept.

While there are people who posses negative global or specific self-views, most people have a positive self-concept (Kendall, Howard, & Hayes, 1989; Schwartz, 1986). Therefore, Campbell & Sedikides (1999) state that, when people are faced with feedback that threatens their self-concept, they will experience a temporary drop in self-esteem. The authors further consider that people will make self-serving attributions in order to get through the uncomfortable situation. Thus, the higher the threat to self-concept is, the greater is the self-serving attribution bias. After Campbell & Sedikides (1999) conducted the meta-analysis, they have proven that the self-serving attribution bias has both empirical and theoretical basis.

(9)

Better than-average effect

According to Alicke et al. (1995) the better-than-average effect provides evidence that people maintain unrealistically positive images about themselves in relation to others. One such example is present in the study conducted by Cross (1977) in which 94% of college instructors consider themselves better than the average. For this reason, according to some, the self-serving attribution bias can be considered the cause of the better-than-average effect, where people evaluate their own characteristics more favourably than that of others. Additionally, the authors state that the better-than-average effect lies within the scope of influence of the social comparison theory. This theory entails three fundamental elements: a motivation for self-evaluation or self-protection, a target with whom comparisons are made, and the particular dimension that is being evaluated, such as the behaviour, trait or attitude.

While the better-than-average effect can be widespread, there have been identified a number of moderating variables that reduce this bias. It has been seen that people exhibit this bias less on traits that are perceived as being uncontrollable than on traits that are perceived as being controllable (Alicke, 1985). Additionally, Dunning, Meyerowitz, and Holzberg (1989) state that the bias is greater when people use their own definition of ambiguous trait dimensions. Finally, it has been shown that the bias is reduced when people are given specific information that indicates that the person to whom the comparison is made is not worse-off on the comparison dimension (Weinstein, 1980, 1983). Thus, as observed by Alicke et al. (1995), the main finding that can be seen throughout these themes is that the better-than-average effect is diminished as the social comparison becomes less ambiguous, or is defined more objectively.

Intriguingly, Alicke et al. (1995) have found evidence that most papers that study the better-than-average effect give participants, as a comparison target, the average peer. This is, in most cases, the average college student. In doing so, this gives a high degree of ambiguity to the comparison process. Under such circumstances, participants tend to take the “I am better than the average” heuristic. For this reason, Alicke et al. (1995) considered two factors in order to minimise the effect of the better-than-average effect.

The first factor, individuation, refers to the “recognition of a person’s distinct identity” (p. 806, Alicke et al., 1995). By using individuation, the authors consider that they are accessing the first level of concreteness, or objectivity, which detaches people from the abstract perception that they are better than others. By providing individuation, it has been observed that the better-than-average effect has been reduced. One reason for this is that, when specific information regarding a

(10)

real person is absent, people tend to assume at least moderately positive characteristics about these individuals (Matlin & Stang, 1978).

The second factor that separates real people from hypothetical social comparison targets is the personal contact. With the establishment of personal contact, Alicke et al. (1995) consider that the better-than-average effect will be diminished even further than through individuation. One such evidence of how personal contact can have a significant impact on people can be seen in the works on obedience by Milgram (1965, 1974). During these experiments, the participants delivered the most shocks when the ‘learner’ was not visible and audible and less shocks when initial contact was made between them.

When assessing how individuation can reduce the better-than-average effect, Alicke et al. (1995) have used in their study the probability of being victimised by 24 unfortunate life events, which were taken from Weinstein (1980). The reason for choosing this experimental design is because by using it, people demonstrate the effects of optimistic bias (Weinstein, 1980) and the

illusion of invulnerability (Pedoff & Fetzer, 1986). This occurs when individuals consider that they

are less susceptible to misfortune than the average college student.

In order to test whether the better-than-average effect is reduced through individuation, Alicke et al. (1995) conducted interviews. For this, the researchers divided the participants into five groups, which they have called conditions. In the first two conditions, the interviewer and live-observer condition, personal contact was established with the target of comparison. The next two, the videotape condition and the transcript condition, only established individuation, without any personal contact. Lastly, the fifth condition had the participants compare themselves with the average college student (average-peer condition).

The results of Alicke et al. (1995) have shown that there was no difference in the

comparison between the first two conditions, the interviewer and live-observer one, nor between the third and fourth one (videotape and transcript conditions). When comparing the better-than-average effect, it demonstrated that the bias was greater in the videotape and transcript conditions, which had only individuation, than the first two conditions in which personal contact was established. On the other hand, the individuation only conditions had a lesser better-than-average effect than the group that compared themselves with the average college student.

The current paper will try to build on the experiment conducted by Alicke et al. (1995) in order to provide the necessary foundation on which the research question will ultimately be answered. For this reason, the participants will be divided into three groups, each with their own

(11)

target of comparison. Group 1 will compare themselves with the average college student. Group 2 will compare with an individual with whom they had no personal contact. Lastly, the third group, Group 3, will compare themselves with an individual with whom they have had personal contact. With the notation of the groups being established, the following hypothesis is constructed:

H2: The higher the individuation effect, the lower the better-than-average effect will be.

Lack of self-insight among low performers

In order to better understand why people make errors in self-assessment, Ehrlinger et al. (2008) suggest that the individuals who make the most mistakes in self-judgement must be able to be identified. They argue that by examining how those individuals who are more prone to error differ from their more accurate peers can lead to the identification of sources of error in general.

Following the same method, Kruger & Dunning (1999) suggest that the poorest performers are those who hold the least accurate assessments of their skills and performance, greatly overestimating how well their performance is compared to their peers. One example that the authors give, is that the bottom 25% of the students that have been tested through the use of grammar, logical reasoning and humour tests considered that they were performing above the 60th percentile.

Kruger & Dunning (1999) consider that this great overconfidence occurs because people who lack skill are not in a position to accurately detect their own shortcomings. This creates a double curse. First, because of the lack of skill, it is hard for them to produce correct answers, thus making mistakes. Second, this lack of skill also impairs them from recognising whether a particular decision is correct or incorrect. To illustrate a case of the double curse, Ehrlinger et al. takes the example of producing a grammatically correct sentence. In order to produce such a sentence, one must first know something about the rules of grammar. However, one must also have adequate knowledge about these rules in order to recognise when a sentence is grammatically correct. Thus, those who lack the knowledge about the rules of grammar are also unable to recognise whether a sentence is grammatically correct or not.

In the logical reasoning study that was conducted by Kruger & Dunning (1999), the results have shown that the participants who were in the bottom quartile (25 per cent) were the ones who had the greatest miscalibration when estimating their own performance. Taking into account that this paper will use a similar logical reasoning test, the following hypotheses are considered:

H3: The bottom 25 per cent of the participants in each group will overestimate their performance more than the top 25 per cent.

(12)

H4: The better-than-average effect will be greater in the bottom 25 per cent of the participants when compared to the top 25 per cent.

H5: The better-than-average effect of the bottom 25 per cent of Group 1 will be greater than the better-than-average effect in the bottom 25 per cent of Group 2.

In conclusion, with the help of the hypotheses, the research question “to what extent does individuation and personal contact minimise the better-than-average effect in low performers when engaging in intellectual tasks” will be answered.

III. Methodology

The study consists in giving a group of university students a quiz that will test their logical reasoning, similar to the study performed by Kruger & Dunning (1999). By testing an individual's intellectual ability means that the quiz will only have one correct answer. This will enable an accurate calculation of the difference between actual and perceived performance.

Participants

The participants were 73 second year students from Kuwait University attending the industrial engineering course. The 20 participants that were part of Group 3 (individuation with personal contact) also got extra credits for their participation in the experiment.

Materials

Participants were given a quiz that contained 15 logical reasoning questions that were created using questions from an online Graduate Management Admission Test (GMAT) preparation test. For Group 1 (the average college student) and Group 2 (individuation) the quiz was given through the usage of an online survey provided by the open-source website Lime Survey. After the conclusion of the 15 questions, the students were asked two questions regarding their perceived performance. The first question asked the participants to estimate how many questions do they think they have gotten correctly. The second question asked them to compare themselves with their target, whether it be the average college student, an individual with whom they have no personal contact, or a person with whom they have personal contact, using a 9-point scale (0 = much less than the

(13)

average college student/the professor/my professor; 4 = about the same as the average college student/the professor/my professor; 8 = much more than the average college student/the professor/ my professor).

Procedure

The participants in Group 1 (n = 30) were volunteered students who were given a QR code to scan and solved the quiz in their final day at the university.They have received ambiguous information about their target of comparison, the average college student. The second group, Group 2 (n = 23), were contacted through e-mail and the university website in order to respond to the survey. The target of comparison for Group 2 was a professor who managed to solve 9 out of the 15 available questions. The use of the word professor in Group 2 represents a knowledgeable person that is not personally known by the survey participants against whom they may psychologically compare themselves. Lastly, the third group, Group 3 (n = 20), were given the quiz on paper by their professor. They were asked to answer the questions keeping in mind that the target of comparison was their own professor who managed to solve 9 out of the 15 questions. All three versions of the quiz will be available in the Appendix.

IV. Results and Discussion

As it was expected from Hypothesis 1, participants have overestimate how many logical reasoning questions they have answered correctly. On average, across all three groups, participants estimated that they have answered 9.38 questions, which was significantly higher than the actual number of correct answers, 5.53, one sample t(73) = 12.44, p < .0001. By taking a closer look to each individual group, it can be noticed that Group 3 has the lowest average estimated correct

(14)

answers with 7.95, however it also has the lowest average correct answers, with 5.1, one sample

t(20) = 6.69, p < .0001. On the other hand, Group 2 showed the highest average number of correct

answers, with 5.7391, though it also contains the highest average estimate of correct answers, 10.70, one sample t(23) = 9.39, p < .0001.

In order to answer the second hypothesis, the mean scores of each Group was determined in all the comparison conditions. The number of participants in each group, along with the mean and standard deviation are present in Table 2. To determine the overall effect of different comparison targets on the degree of the better-than-average effect, an ANOVA was performed on the mean scores, with the three comparison conditions as the independent variable.

(15)

The analysis on the mean scores have shown that there is a significant overall effect of the comparison conditions, F(2, 72) = 15.42, p < .0001 (Table 3). However, the analysis of the first two groups have demonstrated that the better-than-average effect did not minimise. The mean score for the group that compared themselves with the average college student, M = 5.6333, was lower than for the group that compared themselves with an individuated person, M = 5.6522, F(1, 52) = .00, p < .9607 (Table 4). A possible reason for this is that the survey failed to actually individuate the target of comparison. By giving them a target that has managed to score 9 out of 15 correct answers, the conditions did not provide the survey participants the necessary information to view the target in a less ambiguous way than they would view the average college student. On the other hand, the data have shown that the better-than-average effect has been reduced when comparing Group 2 with Group 3. The mean score of Group 2, M = 5.6522, was greater than the mean score of Group 3, M = 3.75, F(1, 42) = 20,62, p < .0001 (Table 5). Lastly, the comparison between Group 1 and Group 3 have shown that the better-than-average effect has been minimised. The mean score of Group 1, M = 5.6333, was greater than the mean score of Group 3, M = 3.75, F(1, 49) = 32.58, p < .0001 (Table 6). As it was expected from the literature of Alicke et al. (1995), the group with individuation and personal contact has the lowest better-than-average bias. However, the data did not support the claim that individuation reduces the better-than-average effect when comparing to the average college student. This finding has refuted the claim made by Matlin & Stang (1978) that, when specific information is absent about an individual, people tend to assume at least moderately positive characteristics in this target of comparison. The participants considered that by comparing themselves with a professor with whom they had no personal contact was similar to comparing to an ambiguous target. In other words, the comparison condition of Group 2 was similar to the comparison condition of Group 1 due to the fact that Group 2 failed to make the target of comparison less ambiguous.

(16)

(17)

In order to find an answer to the central research question, the bottom performers will be analysed. Following the definition of Kruger & Dunning (1999), the bottom performers are those

(18)

individuals who have scored in the bottom percentile (bottom 25%) of the data. When comparing the low performers to the top performers, the data shows that there is a significant difference between the estimations of the two percentiles and the actual results. For the low performers, the average difference between estimated and actual results is significant, with M = 5.1667, paired t(18) = 9.61, p < .0001 (Table 7), while the average difference between the estimated and actual results for high performers is not significant with M = 1.6111, paired t(18) = 1.93, p < .0711 (Table 8). When comparing the differences of the bottom percentile with the upper percentile the results show that there is a significant difference, paired t(18) = 3.37, with p < .0036 (Table 9) . These results supports the claims of Hypothesis 3 and reinforces the idea that Kruger & Dunning (1999) developed, stating that the bottom performers are overestimating their abilities more than the top performers.

(19)

As the analysis above has shown, the bottom performers have overestimated their performance when compared to the top performers. In order to test whether the bottom 25% also over-placed their performance compared to the top 25%, the better-than-average effect will be compared between the two percentiles in each group. In Group 1, the bottom 25%, with M = 5.625, when compared to the top 25%, M = 5.875, did not produce a significant difference in the than-average effect, paired t(8) = -.61, p < .563 (Table 10). When looking at Group 2, the better-than-average effect between the two types of performers, the bottom and the top ones, have also proven to be insignificant, with the mean of the bottom performers, M = 5.5, and the mean of the top performers, M = 4.6667, paired t(6) = 1.54, p < .1852 (Table 11). The same observation can be deduced from the analysis of Group 3 as well. The better-than-average effect does not change significantly between the bottom percentile, M = 3.2, and the top percentile, M = 3.6, paired t(5) = -.49, p < .6483 (Table 12). While these findings are contrary to what Kruger & Dunning (1999) found, which is that low performers over-place their performance when compared to the top performers, it shows that the impact of the better-than-average effect is similar, whether it be in the people situated in the bottom percentile or the top one. The following part will analyse how the better-than-average effect can be minimised when looking at the bottom performers.

(20)

In order to see whether the better-than-average effect in the bottom 25% follow the same pattern that were shown in the comparison of the entire group, the bottom percentile of all the groups will be extracted from the data and analysed against each other. As it was done in the

(21)

previous part, the means along with the standard deviation of all three groups will be presented in Table 13. This will be followed by an ANOVA table, which will first determine the overall impact of the better-than-average effect on the average scores of the bottom percentile, with the three conditions representing the independent variables.

As it was expected from the previous calculations of the better-than-average effect, a significant overall impact is present on the comparison conditions, F(2, 18) = 8.24, p <.0035 (Table 14). The comparison between the bottom 25% of Group 1 and Group 2 has followed the same trend as the comparison of the entire data set of the two groups. The better-than-average effect of Group 1, with the mean score of M = 5.625, was not significantly different from Group 2, mean score M = 5.5, F(1, 13) = .04, p < .8507 (Table 15). As mentioned above, this finding has been consistent with the findings found when comparing the entire data sets of Group 1 and Group 2. Consequently, the average score of Group 2, M = 5.5, was greater than the average score of Group 3, M = 3.2, with

F(1, 10) = 12.01, p < .0062 (Table 16). In order to see the complete effect that individuation with

personal contact has on minimising the better-than-average effect, the comparison between Group 1, M = 5.625, and Group 3, M = 3.2, has shown that the difference is significant, F(1, 12) = 15.70, p < .0022 (Table 17).

(22)

(23)

V. Conclusion

The data collected have managed to prove that the low performers minimise the influence of the better-than-average effect when individuation and personal contact is present. In addition to this, the data have given enough evidence that people possess an unrealistically positive image about

(24)

themselves. This is consistent with the work of Ross & Miller (1975) who consider that people maintain self-enhancing and self-protective behaviours, which are known as the self-serving attribution bias. For this reason, the better-than-average effect, which can be considered a cause of the serving attribution bias, is viewed as representing a more fundamental type of self-enhancement and self-protective behaviour. Alicke et al. (1995) best defines the better-than-average effect by stating that it is “the tendency to view one’s behaviours, opinions, characteristics, and prospects more favourably than those of others” (p. 822). This construct was used as the independent variable of this study in order to see how the effect can be minimised by individuation and personal contact.

Although the data have shown that individuals who are in the bottom percentile overestimate their results more than the top percentile, when comparing to the experimental comparison targets, they do not differ significantly. This demonstrates that the conclusions that Alicke et al. (1995) have reached regarding the fact that the better-than-average effect can be minimised by individuation and personal contact also applies to the less competent individuals. However, the data have shown that there is no significant difference between the two levels of abstraction, involving the average college student and individuation without personal contact. While Alicke et al. (1995) consider that individuation means the recognition of a person’s identity, just by giving the target a title (that of a professor) and a score (9 out of 15 correct answers to the survey) has proven to not influence the way participants compare themselves to them. These findings also refuted the claim of Matlin & Stang (1978) that people tend to give moderately positive characteristics to other individuals of whom they are lacking specific information. This was also proven by Alicke et al. (1995), which consider that providing concrete case information through means of still video images, videotapes, audiotapes and transcripts have a greater influence on the social judgement than abstract statistical information.

While the discussion has focused more about how individuation and personal contact may have an impact on the better-than-average effect, less was discussed about the overestimation. In all the groups, overestimation was present. As it was shown from the data, the number of correct answers for the three groups was less than the median. This is consistent with the view of Moore & Healy (2008), who consider that when the performance is low, people will tend to overestimate their abilities. The participants who are most responsible for the miscalibration in estimations are the bottom 25%. When looking at the experimental data set, the bottom percentile were significantly overestimating their performance when compared to the top percentile. A reason for this can be the

(25)

dual burden that Krueger & Dunning (1999) have identified, and which states that people not only reach mistaken conclusions that produce errors, but they also lack the ability to realise it.

The results of this study can be an addition to the analysis of the people who are in the bottom percentile of performance. In addition, it shows a possible solution that can be used in limiting the miscalibration of their estimates. While it has shown that personal contact reduces the better-than-average effect, the sample size is too small to generalise the current findings. Rather, this paper should be seen as a first step into the research of whether individuation and personal contact can have a greater impact in lowering the estimation errors of the bottom 25% of the study participants than the remaining participants.

For future work, researchers should repeat the current study with a larger sample size in a controlled environment. Additionally, several surveys should be made, which have different difficulty levels, from easy, to medium, to hard.

(26)

Bibliography

Alicke, M. D. (1985). Global self-evaluation as determined by the desirability and controllability of trait adjectives. Journal of Personality and Social Psychology, 49, 1621-1630.

Alicke, M. D., Klotz, M. L., Breitenbecher, D. L., Yurak, T. J., & Vredenburg, D. S. (1995). Personal contact, individuation, and the better-than-average effect. Journal of personality

and social psychology, 68(5), 804.

Baumeister, R. F. (1997). The self. In D. T. Gilbert, S. T. Fiske, & G. Lindzey (Eds.), Handbook of

social psychology (4th ed., pp. 680-740). New York: McGraw-Hill

Bern, D. J. Self-perception theory. In L. Berkowitz (Ed.), Advances in experimental social

psychology (Vol. 6). New York: Academic Press, 1972.

Burson, K. A., Larrick, R. P., & Klayman, J. (2006). Skilled or unskilled, but still unaware of it: how perceptions of difficulty drive miscalibration in relative comparisons. Journal of

personality and social psychology, 90(1), 60.

Campbell, W. K., & Sedikides, C. (1999). Self-threat magnifies the self-serving bias: A meta- analytic integration. Review of general Psychology, 3(1), 23.

Campbell, W. K., Goodie, A. S., & Foster, J. D. (2004). Narcissism, confidence, and risk attitude.

Journal of Behavioral Decision Making, 17(4), 297-311.

Cross, P. (1977). Not can but will college teaching be improved? New Directions for Higher

Education, 17, 1–15.

Dunning, D., Johnson, K., Ehrlinger, J., & Kruger, J. (2003). Why people fail to recognize their own incompetence. Current directions in psychological science, 12(3), 83-87.

Dunning, D., Meyerowitz, J. A., & Holzberg, A. D. (1989). Ambiguity and self-evaluation: The role of idiosyncratic trait definitions in self-serving assessments of ability. Journal of personality

and social psychology, 57(6), 1082.

Ehrlinger, J., Johnson, K., Banner, M., Dunning, D., & Kruger, J. (2008). Why the unskilled are unaware: Further explorations of (absent) self-insight among the incompetent.

Organizational behavior and human decision processes, 105(1), 98-121.

Features. (n.d.). Retrieved May 27, 2016, from https://www.limesurvey.org/ Heider, F. The psychology of interpersonal relations. New York: Wiley, 1958. Ikeda, D. (2013). Discussions on Youth. Chicago: Middleway Press.

(27)

Kendall, P. C., Howard, B. L., & Hays, R. C. (1989). Self-referent speech and psychopathology: The balance of positive and negative thinking. Cognitive Therapy and Research, 13(6), 583-598.

Kruger, J., & Dunning, D. (1999). Unskilled and unaware of it: how difficulties in recognizing one's own incompetence lead to inflated self-assessments. Journal of personality and social

psychology, 77(6), 1121.

Logical Reasoning GMAT Practice Tests. (n.d.). Retrieved May 27, 2016, from http:// www.testpreppractice.net/GMAT/Logical-Reasoning-GMAT-Practice-Tests.aspx Matlin, M., & Stang, D. (1978). The Polyanna principle. Cambridge, MA: Schenkman.

Milgram, S. (1965). Some conditions of obedience and disobedience to authority. Human relations,

18(1), 57-76.

Milgram, S. (1974). Obedience to authority. New York: Harper & Row.

Miller, D. T., & Ross, M. (1975). Self-serving biases in the attribution of causality: Fact or fiction?.

Psychological bulletin, 82(2), 213. 

Moore, D. A., & Healy, P. J. (2008). The trouble with overconfidence. Psychological review, 115(2), 502.

Perloff, L. S., & Fetzer, B. K. (1986). Self–other judgments and perceived vulnerability to victimization. Journal of Personality and social Psychology, 50(3), 502.

Plous, S. (1993). The psychology of judgment and decision making. New York: McGraw-Hill. Schlösser, T., Dunning, D., Johnson, K. L., & Kruger, J. (2013). How unaware are the unskilled?

Empirical tests of the “signal extraction” counterexplanation for the Dunning–Kruger effect in self-evaluation of performance. Journal of Economic Psychology, 39, 85-100.

Schwartz, R. M. (1986). The internal dialogue: On the asymmetry between positive and negative coping thoughts. Cognitive Therapy and Research, 10(6), 591-605.

Svenson, O. (1981). Are we less risky and more skillful than our fellow drivers? Acta Psychologica,

47(2), 143–148.

Weinstein, N. D. (1980). Unrealistic optimism about future life events. Journal of personality and

social psychology, 39(5), 806.

Weinstein, N. D. (1983). Reducing unrealistic optimism about illness susceptibility. Health

(28)

Appendix

Group 1 Questionnaire:

Thank you for taking the time to answer this survey! The purpose of this research is to see how students answer these 15 logical reasoning questions and then estimate how well they did.

Question 1: A financial genius is someone who manages to earn more than his family can spend.

Which of the following is the closest in meaning to the above statement? a) One should earn as much as one needs to spend.

b) A financial genius is one who can manage his own personal finances well. c) A financial genius is one who can save a little of his earnings. (x)

Question 2: Coming late for class, to a party, a movie or anything else is not just fashionable, but has become a habit these days.

Which of the following statements does the above give maximum support to? a) People are not so bothered about coming late these days. (x)

b) Being punctual is out of fashion.

c) People come late for every important meeting now-a-days.

Question 3: I could no longer see the corruption going on in the company and finally I decided to speak up against it, though I knew it would be risky.

Which of the following about the speaker is shown from the sentence? a) Responsible (x)

b) Honest c) Truthful

Question 4: You must be the change you wish to see in the world. What does the speaker of the above sentence mean to imply?

a) People will not change until you do.

b) If you set an example, others will follow. (x)

c) For making a difference in others, make a difference in yourself first.

Question 5: “It totally depends on the situation – if it's a professional occasion or if you're meeting someone for the first time, then being late is totally unacceptable.”

What is the best suited inference one can draw from the above sentence? a) The speaker is punctual according to the occasion.

b) The speaker is conscious about being punctual.

c) The speaker does not give much importance to being punctual when meeting friends. (x)

Question 6: In an attempt to deliver as much information as possible about themselves, applicants sometimes answer questions at length and stray from the topic.

Why is the above behaviour unacceptable? a) The applicants sound self-centered. (x) b) The applicant seems to be bragging. c) he applicants end up seeming nervous.

Question 7: Indulge yourself, eat well; after all, you will live only once.

Which of the following statements does the above paragraph give maximum support to? a) People should be carefree about eating.

(29)

c) People should not bother about eating a proper diet.

Question 8: The colours used in the house of Rita Ramit are very subdued. The look of the house is very minimal. The walls are mostly plain with earthy colours like beige and crème. Even the furniture is simple. What will attract your attention is that the windows are larger than usual and she claims to keep them mostly wide open.

What can one best interpret about Rita Ramit on the basis of the above paragraph? a) She is simple and sober

b) She is simple and loves nature (x) c) She does not like dark colours on walls

Question 9: Which one of the following is the weakest claim in the advertisement of Morning Walker, an electronic device which can be used as a substitute of walking?

a) Morning Walker is one of the best ways to maintain physical and mental health. b) Morning Walker has been tested and certified by leading medical institutions. (x) c) It can be used safely by people of all age groups.

Question 10: Our idea is to make aged care homes a better place to live in. These facilities should not be viewed as a dumping ground. Senior citizens should live in dignity and people should be proud of being a part of it.

The above paragraph best supports which of the following? a) Old age homes need a modification (x)

b) Senior citizens need to live in dignity

c) People need to change their attitude towards aged care homes

Question 11: At this time of the year, the transition period, the skin gets affected by the changes in the weather. The skin becomes very dry as it has to face the cold harsh winds and the scorching sun at the same time. It's hence very important to take extra care of your skin, so that it does not get damaged.

Which of the following is most likely the advice given by the author of the above paragraph? a) Moisturize the skin on a daily basis.

b) Massage the hands, legs and other parts of the body with a good skin lotion. (x) c) Scrubbing the skin to free it of dead cells in the dry areas is also very important.

Question 12: While the life-expectancy gap between men and women has shrunk to five years, the narrowest since 1946, it is no secret that men still need to pay more attention to their bodies.

Which of the following is the most suitable reason for the above? a) Men do not seek medical help as often as women. (x)

b) Men tend to join in fearless, risky, dangerous behaviour more often than women. c) Men do not accept that they are sick and need help.

Question 13: A leading child psychologist and education consultant gives the following advice on how special children should be dealt with.

Which of the following is the most unimportant advice given by her? a) Slow learners should be exempted from some subjects. (x) b) Make them answerable for themselves.

c) Do not frustrate the child.

Question 14: Ann had a bad headache. She tried to take medicine on her own but that did not help. She felt miserable. Finally she had to ask her sister Amy, with whom she was not talking since two days, to take her to a doctor.

Which of the following situations represents a similar case as is mentioned above?

a) Tim and Titan stop talking to each other as a result of a small fight. Titan sees Tim in a financial trouble and gives him help.

(30)

b) Sim knows that his mother is very angry but still he dares to ask her if he can go with his class for a picnic. His mother lets him go. (x)

c) Sam stops talking to Mat when he returns his bike with a flat tyre. Later Mat apologises and they patch up.

Question 15: Some stories are for the mere purpose of making the listener feel good. Like say the fairy tales which always have a happy ending. On the other hand C. S. Lewis's Narnia tales are Christian allegories.

Which of the following is the most likely meaning of allegory? a) A folklore

b) A romantic story c) A symbolic story (x)

Thank you for answering the questions! Now there are only 3 more small questions left! Hang in there!

Question 1: How many questions do you think you have answered correctly? Answer:

Question 2: How do you think you have done compared to the average college student? Please give a rating from 0 to 8. The following illustrate how the ranking works:

0 = much less than the average college student 4 = about the same as the average college student 8 = much more than the average college student Answer:

(31)

Group 2 Questionnaire:

Before starting, it should be noted that a university professor has also taken this test and has scored 9 out of 15 without any prior preparation.

b) A financial genius is one who can manage his own personal finances well. c) A financial genius is one who can save a little of his earnings.

Which of the following statements does the above give maximum support to? a) People are not so bothered about coming late these days.

Which of the following about the speaker is shown from the sentence? a) Responsible

a) People will not change until you do. b) If you set an example, others will follow.

Why is the above behaviour unacceptable? a) The applicants sound self-centered. b) The applicant seems to be bragging. c) he applicants end up seeming nervous.

(32)

b) She is simple and loves nature

c) She does not like dark colours on walls

a) Morning Walker is one of the best ways to maintain physical and mental health. b) Morning Walker has been tested and certified by leading medical institutions. c) It can be used safely by people of all age groups.

The above paragraph best supports which of the following? a) Old age homes need a modification

b) Massage the hands, legs and other parts of the body with a good skin lotion. c) Scrubbing the skin to free it of dead cells in the dry areas is also very important.

Which of the following is the most suitable reason for the above? a) Men do not seek medical help as often as women.

Which of the following is the most unimportant advice given by her? a) Slow learners should be exempted from some subjects. b) Make them answerable for themselves.

(33)

b) Sim knows that his mother is very angry but still he dares to ask her if he can go with his class for a picnic. His mother lets him go.