Does Joy Influence Overconfidence?

Different factors are bound to cause a person’s to be overconfident. Various research studies have been carried out to identify and determine the extent to which these factors influence overconfidence. This review seeks to analyze one such factor, joy and its effect on confidence. The background of joy and overconfidence, the purpose of the study, the research objective, research question, and hypothesis statement will be reviewed in this section. Decision making is a vital part of our everyday lives. Overconfidence is one heuristic that affects decision making and a well-documented psychological trap that afflicts people in almost every aspect of life, often with severe and costly consequences. It can arise from the desire to see oneself as capable and proficient (Blanton, et al., 2001) and past studies have examined overconfidence as a context and/or task dependent effect (Moore & Small, 2007).

In the fields of business and finance, investors are often overconfident about their stock picks and forecasts, usually believing that their knowledge is more accurate than others, which causes them to trade more often while taking on stocks with greater risks. They end up paying more in taxes, commissions and transaction fees (Baker & Nofsinger, 2002). Forbes (2005) suggested entrepreneurs display more overconfidence than non-entrepreneurial managers, although that is usually the result of individual and contextual factors such as age affecting one’s cognitive profile. The overconfidence bias has also been suggested to play a part in the global financial crisis of 2008 (Abbes, 2013). Berner & Graber (2008) noted that even physicians and those working in the medical field tend to overestimate the likelihood of their diagnoses being accurate while Oskamp (1965) discovered that increasing feelings of confidence did not indicate an increased accuracy of psychologists’ clinical decisions.

Koellinger and Treffers (2015) measured the effects of joy on both relative and absolute overconfidence as well as a countermeasure for said overconfidence. In their experiment, researchers gave half the participants (who were in the joy and joy awareness groups) an unexpected gift in the form of a bag of small gummy bears, while the control group (no gift and overconfidence awareness groups) did not receive a gift. They were distributed discreetly so the other half of the participants were unaware of the occurrence of the gift. The purposes of the gifts were to amplify the participants’ positive baseline mood, as well as provide participants in the joy awareness group a salient reason for their positive mood. Participants in the joy awareness group had to answer a quiz that made mood manipulation from the gift obvious. All participants then had to complete a general knowledge quiz with ten multiple-choice questions of medium difficulty and their results were recorded. Participants were given monetary payoffs depending on their performances. They received 4 Euros for attending the experiment and could earn up to 12 Euros. Results suggested participants in the joy only group who received the gummy bears were more overconfident of their answers in the quiz compared to their peers in the control group who did not receive a gift, hence suggesting that joy affects and induces overconfidence.

These authors, however, suggested an alternative explanation to their findings. They opined that it was possible that the participants in the joy group consumed the gummy bears before the experiment and thus sugar consumption may have caused overconfidence as a result of improved positive moods. Various other studies point to the possibility of the effect of sugar consumption on moods (Christensen & Redig, 1993, Benton, 2002).

This paper seeks to bridge this gap by replicating the study by this experiment without the effects of sugar consumption by substituting gummy bears with a non-sugar related gift in the form of a pen. We will be testing the following hypotheses:

There is a difference in overconfidence judgments between the joy and control group.
Joy mood ratings are associated with overconfidence scores.
Receiving a gift unrelated to the judgment task will amplify positive mood states and increase overconfidence.

A noteworthy aspect is that the research by Koellinger and Treffers (2015) involved monetary incentives for the participants. Given that these projects involve a replication of the study, a different incentive is provided and a new approach adopted to measure overconfidence following the techniques embodied in a paper by Dunning, Griffin, Milojkovic, and Ross (1990).

The key purpose of this review is to determine whether joy affects overconfidence. The research seeks to shed more light on the issue given the fact that different scholars share varied opinions on the subject. Ideally, the study seeks to confirm if joy causes overconfidence that is whether a joyful person is more likely to embody higher levels of confidence than an unhappy person. Furthermore, it seeks to establish and confirm that other factors such as age and gender are not likely to influence overconfidence.

The primary objective of this research is to examine whether joy affects overconfidence. In particular, it seeks to determine if confidence precedes joy or vice versa. Therefore, data on joy before confidence and joy after confidence will be examined.

Does joy affect overconfidence?

Ho: A linear relationship exists between overconfidence and joy.

Ha: No linear relationship exists between joy and overconfidence.

The sample size for this research was one hundred categorized into two groups representing the joy and control groups. These respondents were randomly selected between individuals aged 18-26 years from students taking Psychology degrees at the University of Reading Psychology Department. All participants were asked to sign the consent form indicating that their participation is voluntary and that they can at any time pull out of the experiment. A person’s joy before (Joy_Before), their joy after (Joy_After), percentage of confidence levels (Confidence), and personal identifiers that is age and gender were the data metrics used in the study. Using content analysis techniques, the data was coded, groped based on theme categories and counted based on the number of occurrence. Quantitative and qualitative data were linked to facilitate confirmation and detailed analysis of the study variables. A descriptive statistical analysis was carried out on all the metric variables of the dataset and outcomes corresponding to the mean, median, mode, standard deviation, and skewness were relayed. It is worth noting that both quantitative and qualitative procedures were used to collect non- discrete and discrete data. Descriptive research methods were used to collect information from the samples individuals who were presented with questionnaires. The Oxford questionnaires entailed ten questions corresponding to confidence, ten for joy before and an addition ten for joy after. An average of these responses was obtained and used as the data for analysis.

From the survey data, this section presents an overview of how tests were carried out to investigate whether joy influences overconfidence. The data was measured and analyzed using different statistic tools for varying constructs and variables in the study. Consequently, the results were summarized and discussed in this section. Data on the confidence level was collected and an average of all the responses obtained from the ten different questions obtained. Given that gender was presented as a string variable, the data was coded such that 1 represented Males while 0 indicated Females. The information was then displayed in the column denoted as Gender1. The Statistical Package for Social Sciences (SPSS) was used to organize, code, analyze, and convert the data into quantitative summary reports. The analysis performed helped the researcher obtain descriptive statistics on the study variable in form of frequencies. Also, visual representation of the study metric was obtained.

Respondents were recruited via SONA, a research panel used by the University of Reading’s Psychology Department. Based on the current study’s effect size of (d».5) and practical considerations, a priori power calculation using G*Power 3.1.7 (Erdfelder, Faul & Bucher, 1996) suggests that with N=50 and d».5,  one gets 80% power to find the between subjects differences in means. Since there are 2 different groups (Joy and Control), a total of at least 100 participants were required for this study.

Participants were given an Oxford Happiness Questionnaire (OHQ) (Hills & Argyle, 2002) alongside a General Knowledge Quiz. The General Knowledge Quiz contained 10 multiple choice questions. The Oxford Happiness Questionnaire contained 29 questions and participants were required to rank their responses on a likert scale from 1-6, with 1 being strongly disagree and 6 being strongly agree. 12 questions were reverse coded and the overall average score denoted the participant’s happiness score. Scores of 1-2 indicate unhappiness; 2-3 indicate slight unhappiness. Scores of 3-4 indicate a form of indifference, while 4-5 indicate slight happiness. Scores of 5-6 indicate happiness while a score of 6 indicates the participant is too happy. They were equally presented with a pen and a “Thank you” note for their participation.

 

A between subjects independent T-test was performed with all participants given the same materials. The independent variables were mood manipulation (gift vs no gift) and the dependent variables were levels of overconfidence judgement measures. Half the participants were gifted a pen at the start of the experiment and thus assigned to the joy condition, while the other half in the control condition were only gifted the pen at the end of the experiment.

 

Participants were briefed on what the experiment is about and what was required of them. They were each given a quiet individual booth. Participants in the “joy” group were gifted a pen before the experiment, while those in the control group only received it after the experiment was completed. After filling up the consent form, they were presented with the Oxford Happiness Questionnaire (OHQ) to test their baseline mood followed by the general knowledge quiz. They were then tasked to repeat the OHQ at the end of the experiment before a final debrief. The entire experiment took about 15 minutes per participant.

All participants were asked to sign the consent form indicating that their participation is voluntary and that they can at any time pull out of the experiment. Participants were informed that all physical consent forms and raw data will be withheld by the University of Reading’s Psychology department for 5 years. Participants’ information and details were also stored anonymously with no way for the participants’ identities to be revealed. Participants were given access to their experiment data upon request. Researchers were present at all times to answer any queries that participants had during the experiment. Ethical approval from University of Reading School Research Ethics Committee (SREC) was obtained before the experiment proceeded.

This experiment was based on a research paper presented by Koellinger and Treffers (2015) about how a positive mood state (joy) can affect overconfidence. Given that a replication of the same was done, a positive mood was induced by presenting half the participants with a gift specifically a pen. Given that the gift was a surprise the participants were only made aware of the same (the mood manipulation). The other half from the control group were unaware of this addition in the experiment as the gifts were distributed discreetly. Nonetheless, upon completion of the experiment, individuals in the control group equally received pens to ensure fairness was upheld during the experiment.

Following a descriptive analysis done on the metric variables in the study, Table 1 presents a summary of the same. On average, the respondents were twenty years old, and most were female. The respondents also had a mean of 4.109 and 4.14 corresponding to joy before and joy after variables. The participants had an average of 57% corresponding to Overconfidence with a mode of 62%. Also the maximum level of confidence registered was 96% while the minimum level was 10%. Visual representations of the data are relayed in the Appendix.

Table 1.1. Descriptive Statistics of All Study Variables

Descriptive Statistics

 
Mean
Std. Deviation
N

Joy_Before
4.109
0.60098
100

Age
20.2
2.202
100

Joy_After
4.1414
0.64979
100

Confidence
57.262
15.9676
100

Gender1
0.32
0.469
100

 

Table 1.2. Descriptive Statistics Summary of Joy Before, Joy After, and Percentage Confidence

 

Joy (Before)

Joy(After)
Confidence (%)

Mean
4.109
Mean
4.1414
Mean
57.262

Standard Error
0.060098
Standard Error
0.064979
Standard Error
1.59676

Median
4.03
Median
4.03
Median
59.5

Mode
3.83
Mode
4.72
Mode
62

Standard Deviation
0.600985
Standard Deviation
0.649791
Standard Deviation
15.9676

Sample Variance
0.361183
Sample Variance
0.422228
Sample Variance
254.9642

Kurtosis
0.0528
Kurtosis
-0.06149
Kurtosis
0.727679

Skewness
0.017881
Skewness
0.037549
Skewness
-0.6614

Range
2.93
Range
3.21
Range
86

Minimum
2.59
Minimum
2.41
Minimum
10

Maximum
5.52
Maximum
5.62
Maximum
96

Sum
410.9
Sum
414.14
Sum
5726.2

Count
100
Count
100
Count
100

Confidence Level(95.0%)
0.119248
Confidence Level(95.0%)
0.128933
Confidence Level(95.0%)
3.168318

 

Background of respondents

To obtain the profile backgrounds of the respondents, the collected data was analyzed to ascertain that the respondents were sourced from the target population.

The study sought to determine the gender distribution of the respondents. From the responses, majority of the respondents (68%) are female while 32% of the respondents were male as shown in Figure 1 below. The gender imbalance is likely to affect the results of the study as the nature of the research and the research questions. Actually, studies indicate that females who are joyful are likely to depict high levels of confidence. Given that most of the questions relied on the opinions and perceptions of the respondents, the gender distribution is likely to accommodate the opinions of different gender. Moreso, the study relied on facts and data and focused primarily on whether joy influences confidence thus the gender imbalance ids not likely to affect the outcome of the study.

Figure 1. Distribution of respondents by gender.

The study sought to establish the ages of the respondents. From the findings, majority (40%) of the respondents were 18 years, 31% of the respondents were 21years, 11% of the respondents were 25 years,7% were 19 years, and 5% were 22 years. Also, 2% of the respondents were aged 20 years and 23 years while 1% of the respondents were 24years. The age imbalance is not likely to affect the study as the nature of the research and questions asked were not age sensitive and any unlikely error as a result may be tolerated.

 

Figure 2. Distribution of respondents by Age

Cross tabulations of Gender of the respondent against Age

Table 2 below presents the findings of a cross tabulation of gender by age. Since the majority of the respondents were 18 years’ old, 57% are female while 43% were male respondents. From the respondents aged nineteen years, 67% are female while 33% of the respondents were male. For those aged twenty and twenty-one, 87% and 74% of the respondents are female while 13% and 26% are male respectively. Also, for respondents aged 25 years 44% were female while 56% were male. The remaining age categories had 100% of the respondents being of one gender type either male or female.

 

 

 

 

Table 2. Cross Tabulation of Gender Against Age

 
Gender1
Total

 
 
Female
Male

Age
18
20
15
35

19
4
2
6

20
13
2
15

21
20
7
27

22
4
0
4

23
2
0
2

24
0
1
1

25
4
5
9

26
1
0
1

Total
68
32
100

 

 

The hypothesis of the study sought to determine whether joy influenced overconfidence. A regression analysis was carried out to examine whether a relationship exists between overconfidence (denoted as confidence) and all the other study variables. From the analysis results it is evident that the model is not a good fit for the data given the adjusted R square value 0.041 approaching zero. On the contrary, an ANOVA test done on the data was statistically significant given the p value 0.093 is less than 0.05.  The beta coefficients of all the variables save for that corresponding to age were statistically insignificant. Their p values were greater than 0.05. Since age was the only significant variable with an unstandardized beta coefficient is 1.796, and a positive correlation coefficient, it implies that as one advances in age, they tend to become more overconfident and vice versa. Table 3, 4 and 5 present an overview of the data.

Table 3. Model Summary

 

Model
R
R Square
Adjusted R Square
Std. Error of the Estimate

1
.282a
.080
.041
15.63714

 

Table 4. ANOVA Table of the Regression Analysis

ANOVAb

Model
Sum of Squares
Df
Mean Square
F
Sig.

1
Regression
2012.041
4
503.010
2.057
.093a

Residual
23229.415
95
244.520
 
 

Total
25241.456
99
 
 
 

 

 

Table 5. Regression analysis of overconfidence against Joy_Before, Joy_After, Gender, and Age.

 

Model
Unstandardized Coefficients
Standardized Coefficients
t
Sig.

B
Std. Error
Beta

1
(Constant)
16.970
17.251
 
.984
.328

Joy_Before
-3.293
6.077
-.124
-.542
.589

Joy_After
4.293
5.607
.175
.766
.446

Gender1
4.337
3.368
.127
1.288
.201

Age
1.716
.718
.237
2.390
.019

a.                Dependent Variable: overconfidence
 
 
 

 

Table 6. Correlation Analysis of All the Regression Variables

 

Correlations

 
 
Joy_Before
Age
Joy_After
Confidence
Gender1

Joy_Before
Pearson Correlation
1
.104
.902**
.070
.089

Sig. (2-tailed)
 
.303
.000
.491
.377

N
100
100
100
100
100

Age
Pearson Correlation
.104
1
.094
.237*
-.023

Sig. (2-tailed)
.303
 
.352
.018
.817

N
100
100
100
100
100

Joy_After
Pearson Correlation
.902**
.094
1
.094
.073

Sig. (2-tailed)
.000
.352
 
.350
.470

N
100
100
100
100
100

Confidence
Pearson Correlation
.070
.237*
.094
1
.123

Sig. (2-tailed)
.491
.018
.350
 
.221

N
100
100
100
100
100

Gender1
Pearson Correlation
.089
-.023
.073
.123
1

Sig. (2-tailed)
.377
.817
.470
.221
 

N
100
100
100
100
100

**. Correlation is significant at the 0.01 level (2-tailed).
 
 
 

*. Correlation is significant at the 0.05 level (2-tailed).
 
 
 

 

A comparison of the means between the gender and overconfidence of the participants revealed that females have lower overconfidence levels than males, thus the mean overconfidence levels are not equal. Table 6 below presents a summary of the mean comparisons. A study carried out to examine the confidence gap in gender for top economist revealed that women were less overconfident in answering questions raised outside their field of expertise (Sarsons, & Xu, 2015). In fact, the overconfidence levels in females were believed to be lower than that of males from a tender age. Although it is often perceived that confidence is a pre- determined state, certain innate emotions are bound to influence confidence such as joy.

Table 7. Comparison of means on overconfidence against gender

 
 
 

Gender1
Mean
N
Std. Deviation

Female
55.9162
68
16.41824

Male
60.1219
32
14.80642

Total
57.2620
100
15.96760

 

 

Table 7.1. T Test Summary On the Mean Comparisons

t-Test: Two-Sample Assuming Equal Variances

 
Gender
overconfidence (%)

Mean
0.32
57.262

Variance
0.219798
254.964198

Observations
100
100

Pooled Variance
127.592

Hypothesized Mean Difference
0

df
198

t Stat
-35.6456

P(T<=t) one-tail 2.13E-88 t Critical one-tail 1.652586 P(T<=t) two-tail 4.27E-88 t Critical two-tail 1.972017  

 

The comparison of means of overconfidence of the respondents against ages revealed that respondents aged twenty-five years’ have the highest mean overconfidence percentage that is 69.8%. From the findings it was uncovered that respondents who were eighteen years old had 53% overconfidence levels, age 19 had 50% overconfidence levels, age 20 had 60% overconfidence levels, respondents aged 22, 23, and 24 years old had overconfidence percentages of 38%, 56%, and 65% respectively. Respondents aged twenty-six years had 61% confidence. From the data the lowest mean overconfidence percentage levels was 50% relating to twenty-two-year-old participants.

Table 8. Comparisons of Means of overconfidence Against Age of Respondents

 

overconfidence
 
 

Age
Mean
N
Std. Deviation

18
53.2543
35
16.93712

19
49.6667
6
15.01555

20
60.2800
15
13.37034

21
61.2259
27
11.18727

22
37.8750
4
25.63323

23
55.7500
2
7.42462

24
65.0000
1
.

25
69.7778
9
16.16924

26
51.0000
1
.

Total
57.2620
100
15.96760

 

The result of the experiment seeking to examine the effect of joy on overconfidence that is whether there was a difference in the overconfidence judgements between the joy and control groups are relayed in Table 9 below. From the findings, it was observed that the confidence means of respondents in the joy group were higher than that of the control group. The results were found to be normally distributed and the overconfidence score positively skewed (n = 100).

 

Figure A. Mean overconfidence scores for mood manipulation.

 

 

Table 1.

Mean Overconfidence Score and Standard Deviation of mood manipulation.

Group
Mean
SD

Joy
15
1

Control
10
1

Note. n = 100

 

By using mood manipulation as the independent variable and levels of overconfidence judgement as the dependent variable, results show that participants in the joy group (M= 15, SD=1) had higher levels of confidence versus participants in the control group (M=10, SD=1) as evident from Table 9 above. Figure 3 and Table 10 below show that participants in the joy group (Before; M= 4.5, SD= 0.5; After; M=4.2, SD=0.3) displayed higher mood rating scores than those of the control group (Before; M= 3.8, SD= 0.5; After; M=3.0, SD=0.2). A one-way independent T test was used to examine the subjects in the data.

Figure 3. Mean mood rating scores (Before & After)

 

Table 10. Mean Mood Rating Score and Standard deviation for Mood Manipulation

Mean mood ratings score and Standard Deviation (SD) for mood manipulation.

Group
Mean
SD

Joy (Before)
4.5
0.5

Joy (After)
4.2
0.3

Control (Before)
3.8
0.5

Control (After)
3.0
0.2

Note. n = 100

 

The scatterplot of standardized residuals shows the data met all assumptions of homogeneity of variance and linearity.

Figure 4. Scatterplot for correlation between joy mood rating scores and overconfidence scores (measured in %).

 

Furthermore, the Cronbach’s Alpha was also computed for the Oxford Happiness Questionnaire (OHQ) for consistency reasons. The OHQ was found to be highly reliable (29 items; α=.93).

 

Cronbach’s Alpha for Oxford Happiness Questionnaire.

Self-report questionnaire
N
Alpha

OHQ
29
0.931

 

A person’s confidence is often influenced by factors such as age and emotions like joy. Actually, joy comes out as an affecting factor on overconfidence levels as shown by vaious studies carried out by different researchers. Diener, King and Libemirsky (2005) point that happiness leads to success in a person’s various life domains. Consequently, excellence in the most significant areas of a person’s life is bound to increase their confidence. Thus, though not directly joy influences overconfidence. However, the same outcome was not depicted in the regression analysis carried out on the variables against overconfidence. The variables corresponding to Joy that is Joy_Before and Joy_After are not statistically significant; hence, they do not explain the variations registered by overconfidence. In the likely event that these variables were statistically significant, then an increase in Joy before and/or after would result in a decrease and increase in overconfidence respectively. In fact, Age was the only factor that was significant in predicting confidence. This assertion is supported by a study that sought to examine how age and gender influence self-improvement.

The research findings pointed out that as a person ages they tend to become more self-aware thereby increasing a person’s level of self-confidence. The study pointed out that women who are in their late 20s seem to have higher confidence levels than males who felt the need of improving their levels of self-confidence in their 40s. Although the above findings have shown that joy is a key player in shaping an individual’s confidence, it is significance to know that increased joy beyond the acceptable bandwidth may result in overconfidence, an aspect that often has fatal implications. Overconfident individuals are less likely to pay attention to detail when they are tasked to accomplish certain duties. Their continued belief in their past success creates the impression that they are immune to failure. In the event they fail, such individuals are often depressed and it takes time to build their confidence. To them, failure is one of the aspects they have never thought of. This sometime leads to low self-esteem.

It is significant to note that low confidence levels do not always translate to reduced performance levels among individuals. An individual can be joyous about his or her current state. However, this does not imply that the person is satisfied with this position. Instead this joy acts as a foundation towards the attainment of greater goals. With this, one’s joy gradually increases and eventually reaches its optimum. At this level, an individual’s confidence can be clearly depicted in his or her hard work, diligence, and determination. For an individual to exhibit high levels of self-control, his or her joy levels should be within the acceptable limits. Joy affects the normal functioning of the brain making it easy to miss key detail.  People who are averagely joyous are more likely to make the right decisions compared to those who are overly joyous. As an entrepreneur high levels of joy can have detrimental effects on the success and progress of the firm. It may also lead to reduced authority and control over the employees.

An individual’s mood is bound to affect their judgment and their decision making a process explained through an affect as information premise. Basically, this implies that a person can confuse pre-existing feelings as a consequence of a certain objective. A noteworthy aspect on the subject is that suppose an insignificant aspect of confidence is made prominent, the likelihood of overconfidence occurring is minimized. From the study performed, the respondents were offered pens as a gift and because of that they made informed judgements than those who did not receive these incentives. Also, a mean comparison of gender against confidence yields a t value of -35.6456 and a p value of 4.27E-88, which is less than 0.05. The values were collected from a t statistic test done on the mean comparisons of confidence against gender. Research supports the premise that gender influences overconfidence with male respondents depicting higher overconfidence levels than female respondents. As a matter of fact, a study examining the effect of adverts directed to access female respondents indicated the ad was not effective in fostering confidently held beliefs as was the case for males. Ultimately, from the mood manipulations, joy affects overconfidence. Despite the fact that other factors such as gender and age influence confidence, increased levels of the same are a consequence of joyous emotions.

 

In conclusion, it is evident that joy influences overconfidence. The mood manipulation techniques adopted in the study proved that to be true. Respondents were also seen to depict increased confidence levels based on their ages with older respondents portrayed to be making more confident decisions than their younger colleagues. Unfortunately, a regression analysis on the variables indicated the non-existence of a significant relationship between joy and confidence for both provisions of joy. It is important to note that joy does not translate to satisfaction or high confidence levels this is because a person may be joyful but still have depict low levels of overconfidence. Also, joy induces one’s mood causing them make unexpected remarks and decisions and judgements on aspects that may affect their overconfidence. While the insights derived concerning joy and its influence on overconfidence are true, further experiments are still necessary to prevent confounding of study variables. Numerous researchers have concluded that joy affects overconfidence thus the outcome presented in this particular study is valid. Nonetheless, further research on the subject is still necessary.

Major recommendations of the study

Since the study focused on how joy affects overconfidence, more research is necessary on factors interrelated with joy that influence overconfidence.
Additional analysis seeking to explain why mood manipulations were effective in determining if joy influenced overconfidence while regression analyses done on the variables did not register the same is essential.
More information on how age influences overconfidence would also be important.
Increasing the sample size and using questionnaire with questions relating to overconfidence and joy would enhance the research results.

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

References

Abbes, M. B. (2013). Does overconfidence bias explain volatility during the global financial crisis?. Transition Studies Review, 19(3), 291-312.

Baillon, A., Koellinger, P. D., & Treffers, T. (2016). Sadder but wiser: The effects of emotional states on ambiguity attitudes. Journal of Economic Psychology, 53, 67-82.

Baker, H. K., & Nofsinger, J. R. (2002). Psychological biases of investors. Financial services review, 11(2), 97.

Benton, D. (2002). Carbohydrate ingestion, blood glucose and mood. Neuroscience & Biobehavioral Reviews, 26(3), 293-308.

Berner, E. S., & Graber, M. L. (2008). Overconfidence as a cause of diagnostic error in medicine. The American journal of medicine, 121(5), S2-S23.

Blanton, H., Pelham, B. W., DeHart, T., & Carvallo, M. (2001). Overconfidence as dissonance reduction. Journal of Experimental Social Psychology, 37(5), 373-385.

Christensen, L., & Redig, C. (1993). Effect of meal composition on mood. Behavioral Neuroscience, 107, 346-346.

Dunning, D., Griffin, D. W., Milojkovic, J. D., & Ross, L. (1990). The overconfidence effect in social prediction. Journal of personality and social psychology, 58(4), 568.

Erdfelder, E., Faul, F., & Buchner, A. (1996). GPOWER: A general power analysis program. Behavior research methods, instruments, & computers, 28(1), 1-11.

Forbes, D. P. (2005). Are some entrepreneurs more overconfident than others?. Journal of business venturing, 20(5), 623-640.

Koellinger, P., & Treffers, T. (2012). Joy Leads to Overconfidence–and A Simple Remedy.

Koellinger, P., & Treffers, T. (2015). Joy leads to overconfidence, and a simple countermeasure. PloS one, 10(12), e0143263

Lundeberg, M. A., Fox, P. W., & Punćcohaŕ, J. (1994). Highly confident but wrong: Gender differences and similarities in confidence judgments. Journal of educational psychology, 86(1), 114.

Moore, D. A., & Small, D. A. (2007). Error and bias in comparative judgment: on being both better and worse than we think we are. Journal of personality and social psychology, 92(6), 972.

Oskamp, S. (1965). Overconfidence in case-study judgments. Journal of consulting psychology, 29(3), 261.