Self-esteem, risk-taking and social
information seeking in adolescence
M. L. Scheurwater, 11219726,19-06-2020, University of Amsterdam Bachelor thesis
Supervisors: W. van den Bos, S.K. Slagter
Key words: Risk-taking; popularity; adolescence; social network searching; role models
Abstract
During adolescence, peer relations become very important and influential. Young adolescents show more conformity to social influence from peers than people from other age groups do. Several factors have been found to affect the susceptibility to influencing, like low self-esteem and the popularity of the influencer. According to the social learning theory, whether the peer is a role-model or not may also modify the effect of influencing. Peer influence is often studied by giving participants information about peers, like their opinions.
However, no study has investigated which peers adolescents actively observe to inform their decisions. Therefore, this study will look into adolescents’ interest in peer popularity and role models when making risky decisions, and how this is affected by self-esteem. Furthermore, self-esteem is also associated with risk behaviour in adolescent, but the evidence is mixed, finding both negative and positive associations. So as a second goal, current research studied this
relationship. Participants completed a gambling task, able to select peers who’s answers they would like to see. Social relations were captured using a peer
nomination questionnaire. Results show that peer-popularity had a positive effect on selection. The interaction effect of self-esteem and popularity on search
behaviour was only significant when using a categorical measure of self-esteem. There was a significant effect of role models on selection, but no interaction with self-esteem. The data did not show an effect for self-esteem on risk taking.
Introduction
That we, as humans, are social creatures may not come as a surprise. Although some more than others, we need other people. And because there are a lot of us around, social situations are an everyday job. According to the ‘need to belong’- hypothesis we are fundamentally motivated to form interpersonal relationships. These relationships have positive effects on both our physical and mental health (Baumeister & Leary, 1995; Ju & Lee, 2018). Although social interaction is
important for all ages, it seems particularly significant during adolescence. Where first the most important relationship was formed with parents, this now shifts to peers. There also is an increase in the salience of social and emotional
information. A peak of activity is seen in the salience network (anterior cingulate and anterior insula) in early adolescence, compared to childhood and late
adolescence (Rosen et al., 2018). Emotional faces also induce more activity in the adolescent brain than in the adult brain (Monk et al., 2003). Furthermore, an
increase in oxytocin receptors is found within the social-emotional brain system of this age-group, which enhances attention and memory for social information (Steinberg, 2008). All of these findings indicate that young adolescents are very sensitive to emotional and social information, which highlights the importance of social functioning at this age.
In our social interactions we influence and are influenced by others.
Susceptibility to social influence varies between people. We know that age plays an important role in this level of sensitivity. From early adolescence to young adulthood the resistance to peer pressure grows (Steinberg & Monahan, 2009) and young adolescents show more conformity to social influence than people from different age groups (Hoving, Hamm & Gavin, 1969; Knoll et al., 2015). Other factors, besides age, also seem to play a role in susceptibility, like self-esteem and peer status. Self-self-esteem is the positive or negative evaluation of the self (Smith, 2015). Individuals with lower self-esteem make more use of social comparison than people with higher self-esteem (Wayment & Taylor, 1995), which suggests that they might be more sensitive to the influence of others. Indeed, low self-esteem adolescents’ behaviour is influenced more by peers than that of adolescents with high self-esteem (Tian et al., 2019). When studying the effect of self-esteem on the level of risk-taking in a stoplight game, this study found that when a peer was present, adolescents with low self-esteem took significantly more risk than those with high self-esteem. This difference was not seen when the task was conducted alone. That lower self-esteem is a vulnerability factor for peer influencing is supported by the sociometer theory of Leary (Leary &
Baumeister, 2000). According to this theory, our self-esteem acts somewhat like a thermometer, measuring the amount of valuing or devaluing we get from others. A lower self-esteem indicates lower levels of valuing and acceptance, making that person more vulnerable for peer contagion in order to increase acceptance. A similar reasoning is followed in Cohen and Prinstein (2006), who found that social anxiety moderates the effect of peer influencing, making socially anxious
adolescents more vulnerable. Adolescents scoring higher on a social anxiety scale conformed more to both high and low status peers, while those with low social anxiety scores were influenced only by high status peers. The article suggests that fear of negative evaluations makes adolescents more vulnerable to peer conformity. These results also show that, besides anxiety for rejection, the status of the influencer also plays a role in peer contagion. The social learning theory of Bandura supports this. This theory states that social interactions are important to us, because we learn from others how to behave in the world (Bandura, 1971; McLeod, 2016). We learn from our social environment through classical and operant conditioning and modelling. Modelling means imitating someone's behaviour. If the behaviour of the model leads to valued rewards, more attention will be payed to this model’s behaviour and it will be imitated more likely. Compared to someone with a high self-esteem, someone with a low self-esteem could see more qualities in more peers that lead to social rewards, and he might like to possess these qualities to enhance his self-evaluation. Lower self-esteem can therefore increase the observing of others for information, advice or certain types of behaviour. It may also broaden the range of peer status that is observed, because when self-evaluation is low, not only very popular, but also less popular peers may have qualities or behaviours that can increase
acceptance and social rewards. These factors, peer status, role-models and self-esteem, have yet to be studied together to look at the combined effect they have on peer influencing.
Most of the studies that are mentioned, looked at passive information processing in adolescents, for instance reading about someone’s opinion. However, humans are not only at the passive, receiving end of the information
chain. They also actively look for it, even more so for social information, like advice from others (Puskaric, von Helversen & Rieskamp, 2018). The already mentioned sociometer theory also supports this (Leary & Baumeister, 2000). Looking to others for information and conforming to this information can make us more valued and accepted, which is why it can be expected that those with lower self-esteem will do this more often and to more people, than those with higher self-esteem. But this active searching by adolescents in their social network is understudied and remains mostly unknown.
As we mentioned self-esteem in the context of peer influence, it is
important to know that its impact extends much further. It shapes our behaviour and our lives to a great extent and is an important risk/protective factor for our psychological well-being and overall life-satisfaction (Forrester et al., 2017; Montes-Hidalgo & Tomás-Sábado, 2017; Du, King & Chi, 2017; Newark, Elsässer & Stieglitz, 2016; Ju et al., 2018; Arslan, 2019). Moreover, self-esteem is also
associated with risk behaviour. Tian et al., (2019), as discussed, found that adolescents lower on self-esteem showed more risk behaviour when a peer was present. Lower self-esteem is also linked to other types of risky behaviours like smoking or drug abuse (Khajehdaluee et al., 2013; Murphy et al., 2019). Different subsets of low self-esteem were linked to different risk behaviours in adolescents by Wild et al., (2004). Low global self-esteem in girls, for instance, was linked to more risky sexual behaviour. Low family self-esteem increased the chance of alcohol abuse in both girls and boys. The sociometer theory, as mentioned, may explain this association. Low self-esteem increases motivation to change
behaviour in order to get accepted. Sometimes, people can resort to more extreme and dangerous methods to get accepted (Leary & Baumeister, 2000). However, evidence on the link between self-esteem and risk-behaviour is mixed. In Wild et al. (2004) results also show a link between higher self-esteem and more risk behaviour. Eleventh grade boys had higher self-esteem than girls from the same grade on almost all self-esteem scales. These boys also showed more risk behaviour (alcohol, smoking, bullying and risky sexual behaviour) than the girls. In line with this is the study of Johansen et al. (2000) where participants could choose between a safe option of two dollars or a fifty percent chance on four dollars or nothing. Those with higher self-esteem chose more for the risky option, than those with low self-esteem, confirming a positive correlation. More studies found similar evidence (Josephs et al., 1992; Cole et al., 1997; Cavallo et al., 2012). All in all, the evidence about this relationship is not clear. This
contradiction in research makes it important to study the relationship between low self-esteem and risk behaviour further.
If we put all of this together two questions arise that need to be answered. Firstly, adolescents’ interest in peer popularity and role models when making risky decisions, and how this is affected by self-esteem1. It is hypothesized that higher levels of self-esteem will lead to observing more popular peers, while lower levels of self-esteem will observe both popular and nonpopular peers (Bandura, 1971; Leary & Baumeister, 2000; Cohen et al., 2006; McLeod,
2016). Secondly, the contradicting evidence on the effect of self-esteem on taking behaviour leads to the question: Does self-esteem affect the level of risk-taking in adolescents? It is hypothesised that it does affect the level of risk taking, and we expect to see a positive relationship between risk-taking and self-esteem (Josephs et al., 1992; Johanson et al., 2000).
Adolescents in first and third year of high school will take part in a peer nomination questionnaire to assess adolescents’ perception of their classmates. 1Because of the spread of the coronavirus, we were unable to obtain all of the data. Therefore, we were not able to study conformity of adolescents towards their peers.
A gambling task will be conducted, at the end of which participants can select peers who they would like to see the answers from, to gain insight in their search behaviour. A self-esteem questionnaire will also be completed. For the way
adolescents use their social network it is anticipated that higher self-esteem adolescents will choose mainly popular peers for their answers, while those with lower self-esteem scores will pick both popular and unpopular people (Bandura, 1971; Leary & Baumeister, 2000; Cohen et al., 2006). According to the social learning theory it is expected that adolescents will look for information in people they look up to (role-models), which will be more so for adolescents with a lower self-esteem. For the gambling task it is expected self-esteem scores will be positively associated with the number of risky choices. This study will help us to detect those with a higher vulnerability for peer influence and risk behaviour.
Method
Participants
Participants were recruited from 2 high schools in the Netherlands. Different educational levels were tested (MAVO/HAVO/VWO). There were 28 classes, resulting in a total of 409 participants who took part in the study. Two different groups of data were gathered due to a misunderstanding concerning the self-esteem questionnaires. The first group completed the Self-Esteem Questionnaire and consisted, after subtraction of incomplete responses, of 201 participants. Mean age was 14 (SD = 1.17). The sample consisted of 89 boys and 112
girls. The Adolescent Self-esteem Questionnaire was filled in by 98 adolescents, from which 57 remained, 35 boys and 22 girls, after removing incomplete data. Age range was 13-16 years (M = 14). The high number of incomplete responses in both questionnaires was caused by some technical problems and lack of time. The exact samples differ between different research questions, because not all participants completed all of the tasks.
Informed consent was given online by parents via Qualtrics. Students older than 16 did not need their parent’s consent. Participants gave their own consent at the beginning of the experiment. This study was approved by the ethics committee of the University of Amsterdam.
Measurements
Peer nomination questionnaire
A Peer Nomination Questionnaire (PNQ) was filled in to provide a broad view of the social relations in each class. It consisted of questions like: ‘Which classmates are your friends?’, ‘Which classmates would you tell a secret?’, ‘Which classmate do you not like?’ (see appendix 1 for the complete questionnaire). This
questionnaire was designed by putting together different questionnaires used in two studies (Franken et al., 2016; Van den Bos et al., 2018) and adding some for different projects. To answer the research question about peer popularity and role models, two questions were used for the current analysis: ‘Which classmates are the most popular?’ and ‘Which classmates do you look up to?’.
Self-esteem Questionnaire
As mentioned, two separate questionnaires were used. The intended
questionnaire composed different subscales of the Self-Esteem Questionnaire (SEQ) made by Dubois (Dubois, et al., 1996). The global scale and the peer-, family- and body-image subscales were used. A global scale was necessary and
the subscales were of interest to different researchers joining the project. The questionnaire counted an 8-item global scale and three different subscales. Only the global scale scores were used in current study. Participants rated the
statements on a 5-point Likert scale ranging from ‘completely disagree’ to
‘completely agree’. A neutral option was added to the original rating scale: ‘agree nor disagree’. Maximum possible score was 40 and minimum score was 8. Higher scores indicated higher self-esteem. In the questionnaire in its original form (Dubois et al., 1996) all scales had a high internal consistency (Cronbach’s Alpha = 0.81-0.91) and a good test-retest reliability with correlations of r = 0.70-0.87. Average validity was 0.55.
The other questionnaire used, was the Adolescence Self-esteem
Questionnaire (ASQ) (Hafekost et al., 2017). The same rating scale was used as in the SEQ and it consisted of 12 items. No subscales were present in the ASQ, the questionnaire was similar to the global scale of the SEQ. Maximum possible score was 60, minimum score was 12. Higher scores indicated higher self-esteem. A good internal consistency was found for the ASQ, with a Cronbach’s Alpha of 0.91 and a test-retest reliability correlation of r = 0.86. A negative correlation of r = -0.78 was found in comparison with a depression questionnaire, showing good construct validity (Hafekost et al., 2017). Statements of both questionnaires were translated into Dutch and translated back to English by a third party to check for any inconsistencies.
Gambling task
The vase game was used to create risk dilemmas. It was inspired by the wheel-of-fortune task from the study of Blankenstein et al., (2016) and the monetary risk task from Van den Bos et al., (2017). In the solo condition of the task students could choose between a safe option or a risky option to earn points. At the end of the game, points from three randomly picked trials would be added up and
translated into lottery tickets. These tickets increased participants’ chance of winning a 40 euro gift-card in a lottery held in each class at the end of the study. Choosing the safe option always earned participants five points. In the risky option a ball would be randomly pulled from a vase containing black and green balls, a hundred in total. A black ball resulted in no points, and a green ball always resulted in more points than the safe option. The amount that could be earned with a green ball varied between trials and was either 8, 14, 20, 32 or 50 points.
The task consisted of three levels of uncertainty, each containing 16 trials. In the risk-level, the distribution of coloured balls was shown on the vase (f.e. 15 green and 85 black). The ambiguity low-level showed a random pull of 15 balls from the vase, to give participants information about the vase content. The exact distribution was unknown to participants. In the last level, ambiguity high, three randomly pulled balls were shown (see Figure 1). All pulls from the vase were included in the total of hundred balls, they would be put back in the vase. The game was played in six blocks, consisting of eight trials each. The order was as follows: - risk; - ambiguity high or low; - ambiguity high or low; - risk; - ambiguity high or low; - ambiguity high or low. Whether the ambiguity trials were high or low was determined randomly, but all participants got two ambiguity high blocks and two ambiguity low blocks. The position (left or right) of the options changed between trials to prevent habitual choosing.
The total number of risky choices was calculated over all the levels. The maximum number of risk choices was 48. After completing the task, participants were asked which classmates they would like to see the answers from in the next
session. The selected peers were used together with the peer nomination data to study the effect of peer popularity and role models on peer selection for information.
Figure 1. The lay-out of the task. From top to bottom: a risk trial, a low ambiguous trial and a high ambiguous trial. In the risk levels the distribution of balls was on the vase. Ambiguity low levels had a random pull of 15 balls, ambiguity high levels had a random pull of three balls from the vase. On the left is the safe option of five points, which was the same in every trial. On the right is the risky option, which varied between trials in the content of the vase and the number of points that could be earned.
The vase game originally had a social search condition, which would be
completed a week later, in a second session. Students would have been able to observe answers from their classmates before they made the decision.
Unfortunately, this second part could not be conducted2. Testing procedure
Testing took place inside the classroom with as many classmates present as possible to create a natural social context. Tasks and questionnaires were
completed on tablets from the Connected Minds Lab (University of Amsterdam). Plastic screens were put on the tables to make sure participants worked alone. Firstly, participants answered questions about the informed consent letter, to check whether they understood the information. Next, the PNQ was filled in, followed by the vase game. In the last part of testing students filled in a self-esteem questionnaire.
Statistical analysis
Excel and RStudio were used for cleaning, ordering and analysing the
data. Selection of classmates for information and selection on the PNQ items was coded binary, selected classmates getting 1, not selected classmates getting 0. A model was built using the classmate selection for information as the outcome 2 Because of the spread of the coronavirus, we were unable to obtain all of the data. Therefore, we were not able to study conformity of adolescents towards their peers.
variable. Self-esteem data was centralised using the grand mean. The PNQ item ‘Who do you think is popular?’ and the self-esteem score (global scale SEQ or ASQ) were used as predictor variables. An interaction between these two was included in the model. The interaction between the PNQ item ‘Who do you look up to?’ and the centralised self-esteem score was also added to the model, so that both hypotheses would be tested simultaneously. Class was included as level two unit, because data could be nested. A multilevel logistic regression was carried out. As an extra analysis a multilevel logistic model was built with a high and low self-esteem group, splitting the scores in two, using the mean.
A multilevel linear regression analysis with class as a level two unit was conducted to look into the effect of self-esteem on dealing with risky decision making. From the vase game, the total number of risky decisions was calculated as dependent variable. Two different models were built, one using the ASQ-score and the other the global scale SEQ-score as the predictor variable. The different risk levels from the task were not used in this analysis, because it extended beyond the scope of the current research question.
Results
Self-esteem and search behaviour
The total sample of 201 participants was used to look into self-esteem and search behaviour. The mean SEQ score was 32.55 (SD = 6.51). The data was slightly skewed towards the more higher scores. For this question only SEQ scores were used in the analysis, because a lot of participants that filled in the ASQ did not complete the PNQ. The variables peer popularity and looking up had a correlation of r = - 0.11 and a VIF-value of 1.02, so there was no multicollinearity in the model. The variance explained by class was negligible (Variance < .00001). Therefore, the model was run without class as level 2 unit. There was a
significant effect of popularity on selection (Z = 3.14 , p =.002, b = 0.47). With an odds ratio (OR) of 1.6, there was a bigger chance of picking a peer for
information that one also nominated as popular. The interaction between
popularity and self-esteem was not significant (Z = 1.88, p = 0.06, b = 0.05). In the model from the extra analysis with self-esteem as categorical variable, the interaction between popularity and self-esteem was significant (Z = 2.14, p = . 033, b = 0.62, OR = 1.86). According to the expectations, the chance to select popular vs. nonpopular peers was higher for adolescents in the high self-esteem group, compared to those in the low self-esteem group.
The effect of looking up to peers on selecting them was significant (Z = 10.97, p < .0001, b = 2.5, OR = 12.57). The chance of selecting someone for information increased highly, when the peer was also looked up to. The interaction between looking up and self-esteem was also not significant (Z = -0.30, p = 0.76, b = -0.01, OR = 0.99).
Self-esteem and risky decisions
Two analyses were done using either the SEQ score or the ASQ score. The SEQ-score analysis will be discussed first, followed by the ASQ analysis.
A subpart of the total sample was used to study this research question. For the analysis with the SEQ 198 participants were included, with mean age 14 (SD = 1.18). The sample consisted of 89 boys and 109 girls. Mean SEQ score was 32.52 (SD = 6.55). Mean number of risky choices was 28.18 (SD = 6.10). Class as level two unit explained part of the variance (var= 0.50). No significant effect of
self-esteem on the number of risky decisions was found (T = 0.86, p = 0.39, b = 0.06) (see figure 2). The results were more or less the same for most classes (see appendix 2).
Figure 2.
Y-axis displays the number of risky decisions. On the X-axis the SEQ self-esteem score are plotted. The effect of self-esteem on the number of risky decisions is not significant (T = 0.86, p = 0.39, b = 0.06, OR = 1.06).The sample that filled in the ASQ had a mean score of 46.96 (SD = 6.09). The model showed that class did not explain any of the variance, so it was removed from the model to be able to converge; a normal linear regression was conducted. In this analysis there was no significant effect of self-esteem on the number of risky decisions ( T = 1.45, p = 0.15, b =0.19). See figure 3.
Number risky decisions
Figure 3.
The relationship between risky decisions(Y-axis) and ASQ self-esteem score(X-axis). There was no significant effect of self-esteem on risky decisions (T = 1.45, p = 0.15, b =0.19, OR = 1.21).Discussion
Current research studied the effect of peer popularity, role-models and self-esteem on selecting classmates for information using peer nominations and a gambling task. At the end of this task participants were able to select a peer who they would want to see the answer from. It was hypothesized that self-esteem would affect the way adolescents use their peers’ status and role-models for information. As a second goal of this study, the association between risk behaviour and self-esteem was studied. It was hypothesised that there is a positive relationship between self-esteem and risky behaviour. Results showed a positive effect of popularity on being chosen for information, but no significant interaction effect for self-esteem and peer popularity when using a continuous variable. When self-esteem was divided in a high and a low group, there was a significant positive interaction effect of popularity and self-esteem on selection. Being a role-model also increased the chance of being selected for information, but the interaction effect with self-esteem was not significant. There was no difference in risk-taking found between different levels of self-esteem. In the following section we will discuss the research questions separately.
Popularity, role-models, self-esteem and search behaviour
The positive effect of popularity on being chosen for information is in line with the literature (Cohen et al., 2006; Teunissen et al., 2012; Choukas-Bradley et al., 2015). This result also provides us with new information, namely that
adolescents are not only passively influenced by popular peers, but also actively look for information in these peers. Against expectation, the effect of peer
popularity on selection for information did not depend on self-esteem. This corresponds to the study of Prinstein (2007), who concluded that there is an effect of social anxiety, but not of self-esteem on how adolescents are influenced by peer status, although he did not look into active information seeking.
Number risky decisions
Contradictory, current study did find a significant effect for peer popularity and self-esteem on selection when comparing a high and low self-esteem group, which suggests that there is a difference in how adolescents search for information in their social network. The effect found was according to our
expectation that adolescents low on self-esteem would be influenced by popular and non-popular peers alike, while those with high self-esteem would only be influenced by popular peers. However, the result of the first analysis was not significant. This difference may be explained by the change in data presentation. When using a continuous variable there is more variation in the data that needs to be explained by the model. Because the categorical variable uses just two groups, the variation becomes smaller, which makes it easier for the model to explain. A downside of this data transformation is that a categorical variable provides us with less information than a continues one, which makes the latter more reliable to represent the reality of the relationship. However, there are also some advantages to dichotomisation as it makes an association more
understandable. Above that, it can be helpful to handle skewed data and is beneficial when the predictor is not linearly related to the outcome (Farrington & Loeber, 2000). This is an ongoing discussion in the statistical field. For now, it can be said that when the difference in self-esteem is big (as is automatically the case when using a categorical variable), there would be an effect on how
participants selected others for information. So when one person scores eight and another scores 39 (out of 40), for example, the effect might be evident and we will see the person with high self-esteem choose more popular peers for
information, than the one with low self-esteem.
The other studied factor was looking up to peers in relation to selection. The strong positive effect of looking up to someone on selecting them for information that was found, is in line with the social learning theory (Bandura 1971). Adolescent may indeed use role-models behaviours to guide their own. If they also conform their decisions to those role-models should be looked into in future studies. This could also incorporate more specific role-models, to see if all role-models are used in all situations. There was no interaction with self-esteem was, against expectations. It might be that there is no difference between levels of self-esteem, because learning from peers or modelling those you look up to is such a strong mechanism in adolescence.
Several limitations in current study should be mentioned, to interpret these results correctly. The first limitation is the lack of a social component in the task. The task focussed on earning points by choosing either a risky option or a safe option, which did not have anything to do with participants’ social network. The study of Cohen et al., 2006, had more of a social component in his design, as he created a virtual chat room in which computer-simulated peers talked to
participants about aggressive behaviour and/or health risk behaviour. In a non-social task there is no opportunity for adolescents with lower self-esteem to enhance their acceptance or self-evaluation, so they do not have a reason to act differently than others. This may also be the reason that there was no interaction effect of looking up to someone and self-esteem on selecting someone for
information. Looking for information using one’s social network may not differ between different levels of self-esteem when the situation one is faced with is non-social. Future research could differ between social and non-social designs and study the same variables in separate paradigms. Secondly, the level of risk might have been too low for adolescents to worry about information from others, even for people with a low self-esteem. Johansen et al., (2000) did not
incorporate very high risk either, but the choice directly translated into a monetary reward, which made it more important to make a good decision and prompted people low on self-esteem to avoid the negative consequence of
leaving the experiment empty-handed. In current research, the points
transferring into lottery tickets would be gathered from three randomly picked trials. These lottery tickets gave a chance to win a gift-card in the lottery two weeks later. The reason no effect was found might be due to the indirectness of the risk, which may have made it not important enough for participants to make a thought-out decision about who’s answers they wanted to see.
Risky decisions and self-esteem
The other goal of this research was, to unravel the mixed evidence on the link between self-esteem and risk behaviour in adolescents. On the one hand studies found a positive link between self-esteem and risk behaviour (Josephs et al., 1992; Cole et al., 1997; Johansen et al. (2000); Wild et al. (2004); Cavallo et al., 2012). On the other hand there was evidence for a negative link between self-esteem and risky behaviour (Wild et al., 2004; Khajehdaluee et al., 2013; Tian et al., 2019; Murphy et al., 2019). Current result is not in line with this literature and places itself in the midst of the ongoing discussion, favouring neither side. Only one study we found did not find an effect of self-esteem on risky decisions, but this may have been due to small sample size (n =28), (Yang, Dedovic & Zhang, 2010).
There are a lot of perspectives to take when explaining these results. Maybe the most logical explanation is that the current gambling task, did not incorporate a social component, as is said previously. Most studies finding a link between self-esteem and risk-taking (Wild et al., 2004; Cavallo et al., 2012; Khajehdaluee et al., 2013
;
Tian et al., 2019; Murphy et al., 2019) had some sort of social component in their study. For example, Tian et al., (2019) looked at the effect of peer presence. Other studies looked at behaviours like smoking and drug abuse, which are behaviours often carried out in a social context. For this question, he absence of a social component meant that there was no actual threat of rejection, nor an opportunity to gain more social acceptance. This may have made it unnecessary for those with lower self-esteem to show riskybehaviour in order to get accepted (Leary & Baumeister, 2000). It may be concluded from this perspective in combination with the current results that people with different levels of self-esteem only exhibit differences in risk taking when a social component is in play, but this should be further investigated. However, Johansen et al. (2000) did not have a social component incorporated in his design. His task was quite similar to the current study, letting participants choose between two dollars or a fifty-fifty chance on 4 dollars. He found that people with lower self-esteem did show fewer risky choices, which suggests that a social context does not necessarily have to be present to find a difference in risk-taking.
The second perspective might be the low perceived level of risk in the gambling task, which could have influenced the results. This point is also explained in the discussion of the previous question. That the importance of a task can play a role in decision making is supported by the risk sensitivity theory, which says that, based on the height of goals and desires someone has, he prefers a different level of risk-taking (Mishra, Barclay & Sparks, 2016). In our study, adolescents with lower self-esteem would have a higher social goal to achieve than those with higher self-esteem, namely social acceptance(Leary & Baumeister, 2000). However, the task had very little to do with the goal that was studied, which might be the reason no effect of self-esteem was found. The task simply did not give an opportunity to reach those goals or desires.
A third perspective might explain the mixed evidence that is found in general, including current results. It could be the result of the measured
constructs, self-esteem and decision-making, which are broad and influenced by a lot of different factors. Self-esteem is a very popular and well-studied concept, but it encompasses a lot of different subsets, like body-image or academic esteem. Decision-making is influenced by a lot more factors than just self-esteem, like sensation seeking, impulsivity and short- or long term perspective (Zimbardo, Keough & Boyd, 1997; Donohew et al., 2000). When studying two broad concepts, sometimes in combination with fairly specific tasks, it is inevitable to come across different and contradicting results between studies. Future research should focus on narrowing down both of these concepts and match the tasks, to really understand the relationship between risk-taking and esteem. For instance a study looking into risky sexual choices, physical self-esteem and level of information seeking in pretty or handsome classmates. This narrows the concepts down, and creates a more specific relationship measure.
A couple of future study directions have already been mentioned. Studying the complete, original question about searching for information, peer conformity and self-esteem is another interesting direction. Because studies into social information searching in adolescents are still few in number, it would be an important contribution, for which current results have paved the way.
Current research tried to find characteristics that make adolescents more vulnerable for peer contagion and risky behaviours. We found that for risky behaviour, self-esteem is not such a characteristic. Low self-esteem might make adolescents more vulnerable for peer influence, because they might look to both popular and unpopular peers. With all this said, a couple of conclusions can be drawn. Firstly, peer popularity increases the chance of observing them for information and there may be a difference in the way high and low self-esteem adolescents use their peers’ popularity to inform their decisions. However, the latter is said with caution, as more studies are needed for a clear answer. Secondly, looking up to someone will increase the chance that adolescents choose them for information, but there is no difference between self-esteem levels. And lastly, there is no difference in risk-taking between levels of self-esteem in a non-social task. Current conclusions can be evaluated and used for the future of the field of social psychology to study a most turbulent and
fascinating period in all of our lives: adolescence.
Literature
Arslan, G., (2019). Mediating role of the self-esteem and resilience in the association between social exclusion and life satisfaction among adolescents. Personality and individual differences, 151.
https://doi.org/10.1016/j.paid.2019.109514
Baumeister, R.F., Leary, M.R. (1995). The need to belong: desire for interpersonal attachments as a fundamental human motivation. Psychological Bulletin, 117(3), 497-529.
Blankenstein, N. E., Crone, E. A., van den Bos, W., van Duijvenvoorde, A. C. (2016).
Dealing with uncertainty: Testing risk-and ambiguity-attitude across adolescence. Developmental neuropsychology, 41(1-2), 77-92. https://doi.org/10.1080/87565641.2016.1158265
Van Den Bos, W., Hertwig, R. (2017). Adolescents display distinctive tolerance to ambiguity and to uncertainty during risky decision making. Scientific reports, 7, 40962.
Van den Bos, W., Crone, E. A., Meuwese, R., Güroğlu, B. (2018). Social network cohesion
in school classes promotes prosocial behavior. PLoS one, 13(4), e0194656. doi: 10.1371/journal.pone.0194656
Cavallo, J.V., Holmes, J.G., Fitzsimons, G.M., Murray, S.L., Wood, J.V. (2012). Managing Motivational Conflict: How Self-Esteem and Executive Resources Influence Self-Regulatory Responses to Risk. Journal of Personality and Social Psychology, 103(3), 430-451. doi: 10.1037/a0028821
Cohen, G. L., & Prinstein, M. J. (2006). Peer contagion of aggression and health risk behaviour among adolescent males: An experimental investigation of effects on public conduct and private attitudes. Child development, 77(4), 967-983. doi: 10.1111/j.1467-8624.2006.00913.x
Cole, F.L. (1997). The role of self-esteem in safer sexual practices. Journal of the association of Nurses in AIDS-care, 8(6), 64-70. https://doi.org/10.1016/S1055-3290(97)80059-3
Choukas-Bradley, S., Giletta, M., Cohen, G.L., Prinstein, M.J. (2015) Peer Influence, Peer Status, and Prosocial Behavior: An Experimental Investigation of Peer
Socialization of Adolescents' Intentions to Volunteer. Journal of Youth and Adolescence, 44(12), 2197-2210. doi: 10.1007/s10964-015-0373-2
Donohew, L., Zimmerman, R., Cuppa, P.S., Novak, S., Colon, S., Abella, R. (2000) Sensation seeking, impulsive decision-making, and risky sex: implications for risk-taking and design of interventions. Personality and Individual Differences, 28(6), 1079-1091. https://doi.org/10.1016/S0191-8869(99)00158-0
Du, H., King, R.B., Chi, P., (2017). Self-esteem and subjective well-being revisited: the roles of personal, relational and collective self-esteem. PLos one, 12(8). doi: 10.1371/journal.pone.0183958
Dubois, D.L., Felner, D.R., Brand, S., Phillips, R.S.C., Lease, A.M., (1996) Early Adolescent Self-Esteem: A developmental-ecological framework and assessment strategy. Journal of research on adolescence, 6(4), 543-579.
Farrington, D. P., & Loeber, R. (2000). Some benefits of dichotomization in psychiatric and criminological research. Criminal Behaviour and Mental Health, 10, 100-122. doi:10.1002/cbm.349
Forrester, R.L., Slater, H., Jomar, K., Mitzman, S., Taylor, P.J.,(2017). Self-esteem and non-suicidal self-injury in adulthood: A systematic review. Journal of Affective disorders, 221, 172-183. https://doi.org/10.1016/j.jad.2017.06.027
Franken, A., Prinstein, M. J., Dijkstra, J. K., Steglich, C. E., Harakeh, Z., Vollebergh, W.
A. (2016). Early Adolescent Friendship Selection Based on Externalizing Behavior: the
Moderating Role of Pubertal Development. The SNARE Study. Journal of abnormal child
psychology, 44(8), 1647–1657.
Fu, X., Padilla-Walker, L.M., Brown, M.N., (2017).Longitudinal relations between adolescents' self-esteem and prosocial behavior toward strangers, friends and family. Journal of Adolescents, 57, 90-98.
Hafekost K., Boterhoven de Haan K., Lawrence, D., Sawyer M.G., Zubrick S.R. (2017) Validation of the Adolescent Self-Esteem Questionnaire: Technical Report: Telethon Kids Institute and the Graduate School of Education, The University of Western Australia, Perth, Australia.
Hoving, K. L., Hamm, N., & Galvin, P. (1969). Social influence as a function of stimulus ambiguity at three age levels. Developmental Psychology, 1(6), 631– 636. https://doi.org/10.1037/h0028268
Johanson, J.C., (2000) Correlations of self-esteem and intolerance of ambiguity with risk aversion. Psychological Reports, 87(543).
Josephs, R. A., Larrick, R. P., Steele, C. M., Nisbett, R. E. (1992). Protecting the self from the negative consequences of risky decisions. Journal of Personality and Social Psychology, 62(1), 26–37. https://doi.org/10.1037/0022-3514.62.1.26 Ju, S., Lee, Y., (2018). Developmental trajectories and longitudinal mediation effects of self-esteem, peer attachment, child maltreatment and depression on early adolescents. Child Abuse and Neglect, 76, 353-363.
https://doi.org/10.1016/j.chiabu.2017.11.015
Khajehdaluee, M., Zavar, A., Alidoust, M., Pourandi, R., (2013). The relation of self-esteem and illegal drug usage in high school students. Iran red crescent medical journal, 15(11). doi: 10.5812/ircmj.7682
Knoll, L.J., Magis-Weinberg, L., Speekenbrink, M., Blakemore, S. (2015). Social Influence on Risk Perception During Adolescence. Psychological science, 26(5), 583-592. doi:10.1177/0956797615569578
Leary, M.R., Baumeister, R.F., (2000). The nature and function of self-esteem: Sociometer theory. Advances in experimental social psychology, 32, 1-62. https://doi.org/10.1016/S0065-2601(00)80003-9
McLeod, S. A., (2016). Bandura - social learning theory. Simply psychology:
https://www.simplypsychology.org/bandura.html
(
Viewed at 10-03-2020). Mishra, S., Barclay, P, Sparks, A. (2016) The Relative State Model: Integrating Need-Based and Ability-Based Pathways to Risk-Taking. Personality and sociology Review, 21(2), 1-23. doi: 10.1177/1088868316644094Monk, C. S., McClure, E.B., Nelson, E.E., Zarahn, E., Bilder, R.M., Leibenluft, E., Charney, D.S., Ernst, M., Pine, D.S., (2003). Adolescent immaturity in attention-related brain engagement to emotional facial expressions. Neuroimage, 20, 420-428. doi:10.1016/S1053-8119(03)00355-0
Montes-Hidalgo, J., Tomás-Sábado, J., (2016). Self-esteem, resilience, locus of control and suicide risk in nursing students. Enfermería Clínica, 26(3), 188-193. doi: 10.1016/j.enfcli.2016.03.002
Murphy, C.M., Janssen, T., Colby, S.M., Jackson, K.M., (2019). Low self-esteem for physical appearance mediates the effect of Body Mass Index on smoking
initiation among adolescents. Journal of Pediatric Psychology, 44(2), 197-207. doi: 10.1093/jpepsy/jsy070
Newark, P.E., Elsässer, M., Stieglitz, R., (2016) Self-Esteem, Self-Efficacy, and Resources in Adults With ADHD. Journal of Attention Disorders, 20(3):279-90. doi: 10.1177/1087054712459561
Prinstein, M.J., (2007). Moderators of Peer Contagion: A Longitudinal Examination of Depression Socialization Between Adolescents and Their Best Friends. Journal of Clinical Child & Adolescent Psychology, 36(2), 159-170.
https://doi.org/10.1080/15374410701274934
Puskaric, M., von Helversen, B., Rieskamp, J., (2018). How social information affects information search and choice in probabilistic inferences. Acta
Psychologica, 182, 166-176. doi:10.1016/j.actpsy.2017.08.004
Rosen, M.L., Sheridan, M.A., Sambrook, K.A., Dennison, M.J., Jenness, J.L., Askren, M.K., Meltzoff, A.N., McLaughlin, K.A. (2017). Salience network response to
changes in emotional expressions of others is heightened during early
adolescence: relevance for social functioning. Developmental science, 21(3). https://doi.org/10.1111/desc.12571
Smith, E.R., Mackie, D.M., Claypool, H.M., (2015). Social psychology (4th ed.). New York, USA: Psychology Press.
Steinberg, L. (2008). A Social Neuroscience Perspective on Adolescent Risk-Taking. Developmental Review, 28, 78-106. doi: doi.org/10.1016/j.dr.2007.08.002 Steinberg L., Monahan, K.C., (2009). Age differences in resistance to peer
influence. Developmental Psychology, 43(6), 1531-1543. doi: 10.1037/0012-1649.43.6.1531
Teunissen, H.A., Spijkerman, R., Prinstein, M.J., Cohen, G.L., Engels, R. C.M.E., Scholte, R.H.J. (2012) Adolescents’ conformity to their peers’ pro-alcohol and anti-alcohol norms: the power op popularity. Alcoholism, clinical and experimental research, 36(7), 1257-1267. doi: 10.1111/j.1530-0277.2011.01728.x
Tian, L., Dong, X., Xia, D., Liu, L., Wang, D., (2019). Effect of peer presence on adolescents’ risk-taking is moderated by individual self-esteem: An experimental study. International Journal of Psychology, 55(3), 373-379
.
https://doi.org/10.1002/ijop.12611
Wayment, H.A., Taylor S.E., (1995). Self-evaluation processes: motives, information use, and self-esteem. Journal of personality, 63(4), 729-757. doi: 10.1111/j.1467-6494.1995.tb00315.x
Wild, L.G., Flisher, A.J., Bhana, A., Lombard, C. (2008). Associations among adolescent risk behaviours and self-esteem in six domains. The journal of child psychology and psychiatry, 45(8), 1454-1467. https://doi.org/10.1111/j.1469-7610.2004.00330.x
Yang, J., Dedovic, K., Zhang, Q. (2010) Self-esteem and risky decision-making: an ERP study. Neurocase, 16(6), 512-519. doi: 10.1080/13554791003785893
Zimbardo, P.G., Keough, K.A., Boyd, J.N. (1997) Present time perspective as a predictor of risky driving. Personality and individual differences, 23(6), 1007-1023. https://doi.org/10.1016/S0191-8869(97)00113-X
Appendix 1 Complete peer nomination questionnaire
1. Which classmates do you find nice? 2. Which classmates do you not find nice? 3. Which classmates are the most popular? 4. Which classmates are your friends?
5. Which of these classmates is your best friend or are your best friends? 6. Who are the friends of your best friend(s)?
7. With which classmates would you share a secret or your feelings? (f.e. that you have a crush)
8. Who do you look up to in your class? 9. Which classmates are cool?
10.Which classmates are smart?
11.Who makes other classmates do as he/she wants? 12.Who often takes charge within your class?
13.Which classmates are handsome?
14.Which classmates are mean to other classmates? (f.e. bullying, excluding others, humiliating others)
Appendix 2
Figure 3.
This figure shows the regression effect for the different classes. On the Y-axis are the number of risks taken, the X-axis displays the self-esteem score on the Dubois Self-esteem Questionnaire. Classes show different effects. Some show a positive relationship and some a negative one.Some of the classes showed a divergent effect (see figure 3). In one class there was a clear positive effect of self-esteem on risk-taking. This was seen in some other classes, but less strong. Other classes that the opposite effect: a negative relationship between self-esteem and risky decisions. Some classes, however, contained very few participants (120, 122, 123, 126, 127, 128) and the separate effects for the classes were not tested on significance.
