1. Introduction

Risk‐taking refers to behavior that results in uncertain outcomes (Mohr et al. 2010). According to models of decision theory (largely based on von Neumann and Morgenstern 1944), risk‐taking behaviors are often conceptualized by two opponent patterns. “Risk aversion” is marked by the avoidance of potential gain to maintain stability. In contrast, “risk‐seeking” is marked by avoidance of stability to increase potential gains. Traditionally, risk‐seeking behaviors are associated with maladaptive outcomes including, but not limited to, impulsivity and poor decision making (Cho et al. 1999; Del Popolo et al. 2024; Gong et al. 2022) as well as psychopathology (e.g., attention‐deficit/hyperactivity disorder, addictive behaviors) (Boyer 2006; Mishra et al. 2010; Petry 2001; Secades‐Villa et al. 2016). Consequently, risk‐aversion is often considered to be adaptive (note, however, that clinically elevated levels of risk aversion may also be indicative of anxiety disorders, e.g., Eisenberg et al. 1998; Hartley and Phelps 2012; Maner et al. 2007). However, accumulating evidence suggests the outcomes of risk‐seeking behaviors are context‐dependent and that, in some cases, may support learning.

Recent research using the Balloon Analog Risk Task (BART) provides evidence of risk‐related learning. The BART is an objective assessment of risk‐taking tendencies where risk‐seeking behaviors are rewarded up to a certain threshold, beyond which additional risk leads to negative outcomes (Lejuez et al. 2002). Using a unique adaptation of the BART, Meshi et al. (2020) reported that excessive social media use was associated with heightened risk‐aversion, but only after participants experienced loss. These results demonstrate probabilistic reversal learning, as risk‐taking behavior changed depending on a learned reward contingency (high‐risk vs. low‐risk scenarios). Likewise, Sebri et al. (2023) performed trial‐level analysis of BART data and reported changes in risk‐taking behaviors based on prior outcomes (specifically, gains and losses), further illustrating risk‐related learning.

Adapted from the BART, the Balloon Emotional Learning Task (BELT) is a computerized task that measures individual differences in learning through risk‐taking. In contrast to traditional risk‐taking tasks (e.g., the BART and Iowa Gambling Task), the BELT includes variable and stable conditions that allows the tracking of condition differentiation that implicitly facilitates risk‐related learning. Using the BELT, Humphreys et al. (2013) reported that individuals who engaged in high rates of sensation‐seeking behaviors and showed high levels of sensitivity to learning from those risks, performed significantly better on the task than participants who engaged in sensation‐seeking behaviors but did not learn, and participants who were otherwise risk‐averse. Thus, risk‐seeking behaviors led to improved task performance only when participants learned from the risks previously taken.

To build on the work of Humphreys et al. (2013), further research is needed to bridge the gap between objective measures of risk‐taking and their associations with adaptive behaviors in adult populations. To date, a limited number of studies have empirically evaluated the associations between risk‐taking tendencies and adaptive behaviors in adult samples (Fischer and Smith 2004; Hansen and Breivik 2001; Wood et al. 2013; Blair et al. 2018).

Further, most of the existing literature is limited to subjective assessments such as participant self‐report (Fischer and Smith 2004; Hansen and Breivik 2001; Wood et al. 2013). In this literature, relations between risk‐taking and adaptive outcomes were often examined very narrowly (e.g., in the context of certain personality traits or behavioral constructs; Fischer and Smith 2004; Wood et al. 2013; Blair et al. 2018; Kwon et al. 2022) or not at all (Hansen and Breivik 2001). As such, additional research assessing objective measures of risk‐taking behavior in relation to specific adaptive behaviors in adults is sorely needed.

Work in adolescents suggests that risks which (1) benefit an individual's well‐being, (2) carry potential costs that do not harm an individual's health or well‐being, and (3) are socially acceptable, can positively influence developmental outcomes (see Duell and Steinberg 2019 for review). Consider for example, the choice to initiate a new friendship, try a new dish at a restaurant, or enroll in a challenging course. All these scenarios are inherently risky as the consequences are unknown—however, those consequences come at a minimal cost and may result in advantageous outcomes.

Although much of the existing literature in adult populations highlights the potential negative outcomes associated with risk‐taking behaviors, nuanced theoretical models account for the complexities of risk‐taking and emphasize the importance of examining contexts where such behaviors might yield adaptive benefits. For example, Ernst's Triadic Model offers valuable insights into the neural and motivational processes underlying these behaviors. The Triadic Model (Ernst et al. 2006; Ernst 2014) outlines three key neural systems—approach (reward‐driven, primarily involving the nucleus accumbens), avoidance (harm‐avoidant, centered around the amygdala), and regulation (mediated by the prefrontal cortex)—and their interactions in motivated behaviors. This model can be adapted to examine how these systems influence relations between risk‐taking behaviors and motivation and emotional processes in adults as the neural systems involved are presumed more mature during adulthood.

Relatedly, research indicates that the outcomes derived from different risk‐taking measures vary with age. For example, Braams et al. (2015) reported that performance on the BART varied in an adolescent sample, whereas a self‐reported measure of risk‐taking tendencies—namely, the Behavioral Inhibition System/Behavior Activation System (BIS/BAS) Scales—was stable across adolescence and early adulthood (Braams et al. 2015; Takahashi et al. 2007). The BIS/BAS Scales (Carver and White 1994) measure generalizable risk‐taking tendencies related to an individual's inherent motivational or emotional responses. Interestingly, however, Braams et al. (2015) reported a low correlation between the BART and BIS/BAS, further demonstrating a difference in measurement in this age group. In contrast, Li et al. 2019 reported that approach motivation was associated with increased risk‐taking behaviors in adolescents. Taken together, these findings suggest that objective and subjective measures may estimate different aspects of individual differences in risk‐taking behavior in adolescents. Whether relations between objective and self‐reported risk‐taking outcomes are evident in adults, when the neural systems underlying approach and avoidance are matured, warrants further research.

1.1. Current Study

The goals of the current study were two‐fold. First, we aimed to evaluate risk‐related learning in a sample of neurotypical young adults. It was hypothesized that participants who engaged in early, exploratory risk‐taking behaviors would demonstrate improved performance over the course of the task. Second, we were interested in how risk‐taking outcomes correlate to individual differences in motivation and emotional processing (as measured by BIS/BAS Scales) in this sample. Work in adolescents indicates that approach motivation positively correlates with risk‐taking behavior (Li et al. 2019). Moreover, motivation and positive emotional state are reported to positively relate to risk‐taking behaviors in young adults (Cooper et al. 2000; Leikas et al. 2009). As such, we hypothesized that risk‐taking behavior would positively correlate with motivation and positive emotional state in our young adult sample.

2. Materials and Methods1

2.1. Participants

Participants were recruited through the student research participation platform, SONA, facilitated by the Department of Psychology at Assumption University. All SONA participants completed a prescreening questionnaire to evaluate study eligibility. Participants with a history or current diagnosis of a self‐reported addictive behavior disorder and/or color blindness were ineligible to participate in this study. Data were obtained from 58 participants (M _age = 19.66 years, SD = 1.43 years; 74% female). In this sample, 74.1% of participants were White/Caucasian, 8.6% of participants were Black/African American, 5.2% of participants were Chinese, and 6.9% of participants identified as “other” race.2

2.2. Measures

2.2.1. Balloon Emotional Learning Task (BELT)

The BELT is an exploration‐based, implicit learning task that measures differences in risk‐taking behavior (Humphreys et al. 2013; Figure 1). During the task, participants pressed a button on a keyboard to inflate images of balloons displayed on the computer screen. Each time the participant pressed a button to inflate the balloon, they earned one point (i.e., one pump = one point). The goal of the task was to earn as many points as possible. Importantly, however, participants were informed that balloons would pop if over‐inflated and that not all balloons popped at the same point. Participants were able to “bank” (i.e., save) their points after at least one pump. When points were saved, participants heard a tone that signified success, and the number of pumps were converted to “points” on a visible prize meter which displayed the total number of points accumulated throughout the task. The next trial then began. If a participant over‐inflated a balloon, participants heard a tone indicative of an explosion and the word “Pop!” appeared on the screen. No points were added to the prize meter and the next trial began.

Balloons were shown in three different colors that, unbeknownst to the participant, determined the explosion rate of that color balloon. Two of the colors represented stable conditions in which the explosion rate is fixed at a certain short number of pumps (e.g., 6 pumps before explosion) or a certain long number of pumps (e.g., 18 pumps before explosion). The third color represented a variable condition in which the explosion rate was variable (e.g., explosion occurred variably after either 6, 12, or 18 pumps). The different colored balloons were presented equally across 54 trials (18 trials per balloon type) and blocks of the task (18 trials per block).

The two stable conditions allowed for direct examination of learning across the task, as participants should learn which colored balloons to avoid pumping (low explosion rate) and continue pumping (high explosion rate) using feedback from previous trials. Likewise, the variable condition provided an opportunity to evaluate risk‐taking propensity in the context of uncertainty. Consistent with Humphreys et al. (2013), several outcome measures were derived from the BELT including

1. Pumps as a measure of general risk‐taking (more pumps = more risk‐taking)

2. Adjusted pumps as a measure of successful risk‐taking (number of pumps excluding balloons that exploded; adapted from the BART (Lejuez et al. 2002))

3. Post‐explosion pumps as a measure of sensitivity to negative feedback

4. Optimal pumps as a measure of implicit learning (difference between number of pumps and balloon durability).

2.2.2. Behavioral Inhibition System/Behavioral Activation System (BIS/BAS) Scale

The BIS/BAS Scales measure individual differences in two opponent processes related to motivation and emotional processing (Carver and White 1994). Participants use a four‐point response scale to rate the degree to which they agree or disagree with 24 items, where responses of “1” represent strong agreement (“very true for me”) and responses of “4” represent low agreement (“very false for me”). All but two items are reverse scored (e.g., responses of “4” become “1”). Responses across subsets of items are then averaged to create scales of two dimensions of motivation and emotional processing.

The behavioral inhibition system (BIS) scale measures negative emotional responses to punishment (e.g., “Criticism or scolding hurts me quite a bit.”) whereas the behavioral activation or approach system (BAS) scale measures positive emotional responses to reward. BAS items are further categorized by behaviors related to Reward Responsiveness (e.g., “When I'm doing well at something I love to keep at it.”), Drive (e.g., “I go out of my way to get things I want.”), and Fun‐Seeking (e.g., “I am always willing to try something new if I think it will be fun.”). Individual differences in these scales map to clinical outcomes in the realm of anxiety and impulsivity, respectively. Items within these scales have high internal reliability (all α ≥ 0.66). The BIS/BAS also demonstrates strong convergent and divergent validity with other measures of motivation, emotion processing, extraversion/introversion, and personality (see Table 2 in Carver and White 1994). In the current study, BIS/BAS Scale outcomes were used to explore individual differences in our sample as they relate to performance on the BELT.

2.3. Procedures

All procedures were approved by the Institutional Review Boards (IRBs) at Assumption University and Merrimack College. Undergraduate research participants were recruited through SONA. All participants completed a prescreening survey to determine eligibility (described above).

Research assistants first collected written informed consent. Participants were reminded that they could ask questions or withdraw from the study at any time. After consent was obtained, participants completed the BIS/BAS Scale on an iPad (untimed). Participants then completed the BELT. The BELT took approximately 20 minutes to complete. After completion of the BELT, participants were dismissed from the study protocol.

2.4. Data Analysis

Wilcoxon signed‐rank tests were used to test for gender differences in total adjusted pumps and total number of explosions across the entire BELT task to determine whether gender needed to be included as a covariate in the planned statistical models. Pearson correlations were computed to explore (i) the association between early exploratory behavior, as measured by the total adjusted pumps during the first task block (i.e., first ⅓ of trials), and total points accumulated across the task and (ii) relations between total adjusted pumps across the entire task, as a measure of successful risk‐taking, and scales from the BIS/BAS.

Generalized linear models (GLM, lme4 package in R) were used to evaluate relations between (i) the number of explosions across task blocks (block 1: first ⅓ of trials, block 2: middle ⅓ of trials, and block 3: last ⅓ of trials), (ii) the number of adjusted pumps by balloon type (variable, stable long, stable short) and task block, (iii) the performance (i.e., total pumps) on the next balloon by balloon type and outcome on the previous balloon (i.e., explosion or reward collected), and (iv) the difference between the number of adjusted pumps and the optimal number of pumps by balloon type and task block. Balloon type and task block were entered in the GLM as fixed effects and subject was entered as a random effect. Models for “number of balloon pumps” and “number of explosions” as the outcome variable were modeled using the glmer function and family = poisson. Effect sizes for significant effects were reported as the rate ratio with 95% confidence interval (CI). The significance level was set at p = 0.05 for all analyses. The Holm–Bonferroni method was applied to adjust for multiple comparisons in the correlational analyses.

3. Results

There were no significant differences by gender for either total adjusted pumps (W = 257.5, p = 0.68) or total explosions (W = 268, p = 0.83) across the task; therefore, gender was not included as a covariate in any of the subsequent analyses.

3.1. Risk‐Taking Supports Implicit Learning

There was a significant positive relationship between the total adjusted pumps during block 1 (first ⅓ of trials) and the total points accumulated across the task (r = 0.71, p < 0.001; Figure 2A). Furthermore, in the model evaluating the number of explosions by task block (block 1: first ⅓ of trials, block 2: middle ⅓ of trials, and block 3: last ⅓ of trials), there was a significant main effect of block (𝜒² = 10.10, p = 0.006) such that the number of explosions was significantly lower during both block 2 (rate ratio = 0.85, 95% CI = 0.72‐1.00, p = 0.05) and block 3 (rate ratio = 0.77, 95% CI = 0.66‐0.91, p = 0.002) compared to block 1 (Figure 2B); there was no difference in the number of explosions between block 2 and block 3 (rate ratio = 0.9, 95% CI = 0.76‐1.08). Together, these results indicate that exploratory risk‐taking behavior early in the task supports performance later in the task.

In the model evaluating changes in adjusted pumps by balloon type (variable, stable long, stable short) and task block, there was a significant main effect of balloon type on the adjusted pumps per balloon (𝜒² = 159.02, p < 0.001) and a significant interaction between balloon type and block (𝜒² = 34.06, p < 0.001) but no significant main effect of block (𝜒² = 1.08, p = 0.58). Compared to balloons with stable, short explosion rates, participants made significantly more pumps on balloons with variable explosion rates (rate ratio = 1.48, 95% CI = 1.35‐1.61, p < 0.001) and stable, long explosion rates (rate ratio = 1.68, 95% CI = 1.55‐1.83, p < 0.001; Figure 3). Based on post‐hoc comparisons, the significant interaction effect was due to an increase in the number of pumps on balloons with long, stable explosion rates over blocks of the task (block 1 to block 2: rate ratio = 1.11, 95% CI = 1.00‐1.24, p = 0.05; block 1 to block 3: rate ratio = 1.18, 95% CI = 1.05‐1.31, p = 0.002).

To evaluate sensitivity to negative feedback, another model was generated to evaluate the relation between outcome on the previous balloon (explosion or saved points) and performance (total pumps) on the next balloon by balloon type. There was a significant main effect of the previous outcome (𝜒² = 20.29, p < 0.001) and the current balloon type (𝜒² = 310.01, p < 0.001) but no significant interaction effect (𝜒² = 0.88, p = 0.64) on the number of pumps on the current balloon (Figure 4). There was a significant increase in the number of pumps following an outcome of saved points compared to an explosion (rate ratio = 1.14, 95% CI = 1.08‐1.21, p < 0.001) and a significant increase in pumps if the current balloon type was variable (rate ratio = 1.28, 95% CI = 1.20‐1.37, p < 0.001) or stable, long (rate ratio = 1.72, 95% CI = 1.62‐1.83, p < 0.001) compared to stable, short.

A final model was generated to evaluate the difference between the number of adjusted pumps and the optimal number of pumps (i.e., balloon durability) by balloon type across task block. There was a significant main effect of block (𝜒² = 19.47, p < 0.001) and balloon type (𝜒² = 726.71, p < 0.001), and a significant interaction effect between block and balloon type (𝜒² = 34.46, p < 0.001; Figure 5). The difference between the number of adjusted pumps and the optimal number of pumps decreased significantly from block 1 to block 2 (rate ratio = 0.65, 95% CI = 0.50‐0.84, p < 0.001) and from block 1 to block 3 (rate ratio = 0.59, 95% CI = 0.46‐0.76, p < 0.001), providing evidence that participants learned the optimal number of pumps across the task. The difference between the number of adjusted pumps and the optimal number of pumps was significantly higher on balloons with variable explosion rates (rate ratio = 7.87, 95 % CI = 6.58‐9.42, p < 0.001) and on balloons with stable, long explosion rates (rate ratio = 10.56, 95% CI = 8.85‐12.59, p < 0.001) compared to balloons with stable, short explosion rates, reflecting the increased number of pumps possible on these balloon types.

3.2. Risk‐Taking Positively Relates to Persistent Pursuit of Goals

To evaluate relations between risk‐taking and individual differences in motivation and emotional processing, bivariate correlations were computed between the total adjusted pumps on the BELT and the BIS/BAS Scales. The total number of adjusted pumps on the BELT was significantly, positively correlated with the BAS Drive Scale (n = 57, r = 0.277, p = 0.037; Figure 6) but was not correlated with the BAS Fun Seeking Scale (r = 0.098, p = 0.470), BAS Reward Response Scale (r = ‐0.053, p = 0.70), or BIS Scale (r = 0.135, p = 0.316). Notably, however, when p values were adjusted for multiple comparisons, the correlation between total number of adjusted pumps and the BAS Drive Scale was no longer significant (p = 0.149, Holm–Bonferroni method).

4. Discussion

The primary aim of this study was to evaluate risk‐related learning in neurotypical young adults. Consistent with our hypothesis and the work of others (Humphreys et al. 2013), our results indicate that early, exploratory risk‐taking supported learning in our sample. Early exploration, measured by the total adjusted number of pumps in the first third of the task, was associated with accumulating more total points across the task (Figure 2A). Moreover, the number of explosions decreased across the task (Figure 2B) and participants increased the number of times they pumped the safer balloons with stable, long explosion rates over the course of the task (Figure 3). Given that there was no explicit feedback on task performance (other than points earned and balloon explosions), this suggests that risk‐taking supported implicit learning in our sample as participants learned which balloons were safer and, consequently, more advantageous for task success. Relatedly, the difference between the number of pumps and the optimal number of pumps (i.e., balloon durability) decreased across the task, suggesting that participants not only learned which balloons were safer but started to learn how many pumps were possible before the balloon would explode (Figure 5).

Humphreys et al. (2013)’s sample demonstrated a consistent number of pumps across blocks for both stable conditions (short and long) and a decrease in the pumps for the variable condition. In contrast, our young adult sample significantly increased the number of pumps on long, stable balloons across blocks. These differences may be due to the number of trials administered in each study, as we doubled the number of trials (54 trials) that Humphreys’ study used (27 trials). A larger number of trials may have provided a greater opportunity for learning. Nonetheless, both instances provide evidence of learning, as these changes in behavior ultimately improved overall task performance.

Our data also indicate that participants utilized feedback from previous trials to improve future task performance. This is evidenced by the increased number of pumps observed following trials when a balloon did not explode (e.g., points were saved, indicating successful risk‐taking) versus trials when a balloon exploded (e.g., balloon was overinflated, unsuccessful risk‐taking; Figure 4). Reduced pumps after explosions represent sensitivity to negative feedback. Consistent with Humphreys et al. (2013, 2011), this finding suggests that the knowledge of outcomes from earlier risks shaped future behaviors which, ultimately, improved task outcomes.

Together, these results suggest that context‐dependent risk‐taking supports learning—an adaptive outcome—in young adults. This work extends a growing body of literature outlining negative implications for risk‐taking (Boyer 2006; Gong et al. 2022; Mishra et al. 2010; Petry 2001; Secades‐Villa et al. 2016) and supports theoretical perspectives of adaptive risk‐taking in adolescence (Duell and Steinberg 2019; Ellis et al. 2012). Ellis et al. 2012 argue that risk‐taking behavior may serve evolutionary advantages during critical developmental transitions such as adolescence. For example, aggressive behaviors may signal adaptive traits such as bravery or social status. Following this framework, Ellis et al. (2012) propose that interventions should promote social pathways by cultivating environmental conditions that reduce stress, increase predictability, and remove harsh dynamics (e.g., “zero tolerance” policies) to match adolescents’ goals and motivations rather than working against their instincts. Whether or not such suggestions improve risk‐related outcomes in adulthood—a critical developmental transition when additional social challenges arise—warrants additional research.

The second aim of our study was to determine how risk‐taking outcomes would correlate to individual differences in motivation and emotional processing (as measured by BIS/BAS Scales). Our hypothesis was supported, and was consistent with prior research (Cooper et al. 2000; Leikas et al. 2009; Li et al. 2019), as risk‐taking (BELT total adjusted pumps) was positively correlated to persistent pursuit of goals (as measured by the BAS Drive Scale; Figure 6). The BAS Drive Scale is based on participant self‐report on 4‐items that measure agreement with statements related to motivation for goal‐directed behavior including “I go out of my way to get things I want,” “When I want something I usually go all‐out to get it,” “If I see a chance to get something I want I move on it right away,” and “When I go after something I use a ‘no holds barred’ approach.”

This finding has compelling implications as goal‐directed behaviors strongly predict academic success (Eppler and Harju 1997; Steinmayr et al. 2011), particularly in non‐traditional students who take time off during undergraduate studies (Eppler and Harju 1997). Importantly, however, these results should be interpreted with caution, as the effect was no longer significant after controlling for multiple comparisons. A post hoc power analysis (G*Power 3.1.9.7) indicated that we only had 74% and 73% power to detect a significant correlation at an alpha level of 0.05 between for BELT total pumps and total adjusted pumps and the BAS Drive Scale. As such, research with a larger, more diverse sample is needed to further explore relations between risk‐taking and motivation and emotional processing in adults. Additionally, future research is needed to determine how risk‐related learning may support academic outcomes in diverse bodies of college students, as academic outcomes were not evaluated in the current study. Moreover, neuroimaging work is needed to understand how the development and activity of neural systems—such as those posed in Ernst's Triadic Model (Ernst et al. 2006; Ernst 2014)—influences risk‐taking behaviors and relations with motivation and emotional processes in adults.

Although underpowered, our correlational results may be of interest to individuals in educational or professional settings where risk‐related operations affect performance, physical safety (e.g., military, healthcare settings), financial gain or stability (e.g., investment banking, entrepreneurship), and creativity/innovation (e.g., artists, writers, filmmakers, technology, research and design) (Bechara et al. 1997; Breivik et al. 2019; Giaccone and Magnusson 2022; Macko and Tyszka 2009; McGowan 2007). In these careers and others, successful assessment and management of risk relates to improved outcomes.

5. Conclusions

The current study was innovative as we used an objective measure of risk‐taking behavior to demonstrate risk‐related learning in adults. These findings corroborate prior research which indicates that risk‐taking supports learning (Humphreys et al. 2013). Additionally, risk‐taking tendencies were positively related to goal‐driven behavior in our sample. Together, these findings suggest that context‐dependent risk‐taking is adaptive and may improve learning‐based educational and professional outcomes. Results obtained from this study are expected to inform interventions that shape adaptive outcomes during this critical developmental period.

Introduction

Risk-taking involves behaviors that have uncertain results. In decision-making, these behaviors often show two main patterns. "Risk aversion" describes avoiding possible gains to keep things stable. In contrast, "risk-seeking" involves avoiding stability to get potentially larger gains. Historically, risk-seeking behaviors have been linked to negative outcomes such as impulsive actions, poor decision-making, and mental health issues like ADHD or addictive behaviors. Because of this, risk aversion has often been seen as a positive, adaptive trait, though very high levels of risk aversion can also point to anxiety disorders. However, a growing amount of evidence suggests that the results of risk-seeking behaviors depend on the situation and can, in certain cases, help with learning.

Newer studies, especially those using tasks like the Balloon Analog Risk Task (BART), show how learning can happen through taking risks. The BART is an objective test that measures how much a person is willing to take risks. In this task, risk-seeking behaviors are rewarded up to a certain point; going beyond that point leads to negative outcomes. One study found that people who used social media excessively showed more risk aversion, but only after they had experienced a loss in the task. This suggests a type of learning where risk-taking changed based on what they learned about rewards. Similarly, another study analyzed BART data and reported that risk-taking behaviors changed based on past results, such as gains or losses, further showing that learning occurs from taking risks.

The Balloon Emotional Learning Task (BELT) is a computer-based task adapted from the BART. It measures how individuals learn through risk-taking. Unlike other risk-taking tasks, the BELT includes both changing and stable conditions. This allows researchers to track how people learn to distinguish between these conditions, which helps with risk-related learning. Using the BELT, a study found that individuals who were both highly sensation-seeking and good at learning from risks performed better on the task. This performance was better than those who sought sensations but did not learn, or those who avoided risks. This highlights that risk-seeking behaviors led to better task performance only when participants learned from their previous risks.

More research is needed to connect objective ways of measuring risk-taking with positive behaviors in adults. Currently, only a few studies have looked at the links between risk-taking tendencies and adaptive behaviors in adult populations.

Furthermore, most of the existing research relies on self-reported information from participants. In these studies, the relationship between risk-taking and positive outcomes was often looked at in a very specific way, such as in relation to certain personality traits or behaviors, or not at all. Therefore, more research is greatly needed to assess objective measures of risk-taking behavior in connection with specific adaptive behaviors in adults.

Research in teenagers suggests that certain risks can positively influence their development. These include risks that benefit a person's well-being, have potential costs that do not harm their health, and are socially acceptable. Examples include starting a new friendship, trying a new food, or signing up for a challenging class. These situations involve risk because the outcomes are unknown, but the costs are minimal, and they could lead to good results.

While much of the existing research on adults emphasizes the potential negative outcomes of risk-taking, complex theoretical models acknowledge the many aspects of risk-taking. These models stress the importance of looking at situations where such behaviors might offer benefits. For example, Ernst's Triadic Model offers insights into the brain and motivational processes behind these behaviors. This model describes three key brain systems—one for seeking rewards, one for avoiding harm, and one for regulating these behaviors—and how they interact. This model can be used to study how these systems affect the links between risk-taking and motivation and emotional processes in adults, as these brain systems are expected to be more developed in adulthood.

Relatedly, studies show that the outcomes from different risk-taking measures change with age. For instance, one study found that performance on the BART varied in teenagers. However, a self-reported measure of risk-taking tendencies, called the Behavioral Inhibition System/Behavior Activation System (BIS/BAS) Scales, remained stable from adolescence into early adulthood. The BIS/BAS Scales measure general risk-taking tendencies related to a person's natural motivational or emotional responses. Interestingly, that study also reported a weak connection between the BART and BIS/BAS, showing that these measures might assess different aspects of risk-taking in this age group. In contrast, another study reported that a motivation to approach rewards was linked to increased risk-taking in teenagers. These findings together suggest that objective and self-reported measures may capture different aspects of individual differences in risk-taking behavior in teenagers. Whether similar connections between objective and self-reported risk-taking outcomes exist in adults, when the underlying brain systems are fully developed, requires further investigation.

Current Study

The current study had two main goals. First, the aim was to evaluate how adults learn from risks. Researchers predicted that participants who took more risks early in the task would perform better overall. Second, the study sought to understand how risk-taking outcomes relate to individual differences in motivation and emotional processing, as measured by the BIS/BAS Scales. Based on previous research showing that approach motivation relates positively to risk-taking behavior and that motivation and positive emotional states are linked to risk-taking in young adults, researchers predicted that risk-taking behavior would positively relate to motivation and positive emotional states in the young adult participants.

Materials and Methods

Participants

Participants were recruited from a university student research platform. All participants completed an initial survey to check if they were eligible for the study. Those with a history or current diagnosis of an addictive behavior disorder or color blindness were not allowed to participate. Data were collected from 58 participants. The average age was 19.66 years. Most participants were female (74%). The majority (74.1%) identified as White/Caucasian, with smaller percentages identifying as Black/African American (8.6%), Chinese (5.2%), or "other" race (6.9%).

Measures

The Balloon Emotional Learning Task (BELT) is a computer-based task that explores implicit learning and measures differences in risk-taking behavior. In the task, participants inflated balloon images on a computer screen by pressing a button. Each press earned one point. The goal was to earn as many points as possible. Participants knew that balloons would pop if over-inflated and that not all balloons had the same breaking point. They could "bank" or save their points after at least one pump, which caused a success tone and added points to a visible prize meter. If a balloon popped, an explosion tone sounded, "Pop!" appeared on screen, and no points were added. Balloons came in three colors, which secretly determined their explosion rate. Two colors had stable explosion rates (e.g., 6 pumps or 18 pumps), while the third had a variable rate (e.g., 6, 12, or 18 pumps). Each color was presented equally across 54 trials. The stable conditions allowed for tracking learning, as participants were expected to learn which balloons to avoid and which to continue pumping based on feedback. The variable condition assessed risk-taking when outcomes were uncertain. Outcomes measured from the BELT included total pumps (general risk-taking), adjusted pumps (successful risk-taking), post-explosion pumps (sensitivity to negative feedback), and optimal pumps (implicit learning).

The Behavioral Inhibition System/Behavioral Activation System (BIS/BAS) Scales assess individual differences in two opposing processes related to motivation and emotional responses. Participants rated their agreement with 24 statements using a four-point scale. The BIS scale measures negative emotional responses to punishment (e.g., "Criticism or scolding hurts me quite a bit."). The BAS scale measures positive emotional responses to reward, with sub-scales for Reward Responsiveness, Drive, and Fun-Seeking. These scales show individual differences linked to anxiety and impulsivity. In this study, BIS/BAS Scale outcomes were used to explore individual differences related to performance on the BELT.

Procedures

All study procedures were approved by the ethics committees at the participating universities. Undergraduate students were recruited. First, research assistants obtained written informed consent. Participants were reminded they could ask questions or withdraw at any time. After consent, participants completed the BIS/BAS Scale on an iPad. Following this, they completed the BELT, which took about 20 minutes. After finishing the BELT, participants were dismissed.

Data Analysis

Statistical tests were used to check for gender differences in adjusted pumps and total explosions. This helped determine if gender needed to be considered in later analyses. Correlations were calculated to explore the link between early risk-taking behavior (adjusted pumps in the first block) and total points earned, and between overall successful risk-taking (total adjusted pumps) and the BIS/BAS Scales. Generalized linear models were used to examine how the number of explosions changed across task blocks, how adjusted pumps changed by balloon type and block, how performance on a balloon was affected by the outcome of the previous balloon, and how the difference between adjusted pumps and optimal pumps changed by balloon type and block. Balloon type and task block were set as fixed factors, and individual participants as a random factor. The significance level for all analyses was set at p = 0.05. A method called Holm–Bonferroni was used to adjust for multiple comparisons in the correlation analyses.

Results

No significant gender differences were found for either total adjusted pumps or total explosions across the task. Therefore, gender was not included in any further analyses.

Risk‐Taking Supports Implicit Learning

A strong positive relationship was found between successful risk-taking during the first third of the task and the total points earned throughout the task. Also, when examining the number of explosions across the three task blocks, there was a significant decrease in explosions during the middle and last blocks compared to the first block. This suggests that the number of explosions became lower over time. These results indicate that taking exploratory risks early in the task helps improve performance later in the task.

In the analysis of changes in adjusted pumps based on balloon type (variable, stable long, stable short) and task block, there was a significant effect of balloon type and a significant interaction between balloon type and block. Participants made significantly more pumps on balloons with variable and stable long explosion rates compared to those with stable, short explosion rates. This significant interaction was due to an increase in the number of pumps on balloons with long, stable explosion rates as the task progressed.

To understand how participants reacted to negative feedback, another analysis looked at the relationship between the outcome of the previous balloon (exploded or points saved) and the number of pumps on the next balloon for each balloon type. There was a significant effect of the previous outcome and the current balloon type. Participants significantly increased the number of pumps after a previous balloon's points were saved (indicating success) compared to when a balloon exploded. Also, there was a significant increase in pumps if the current balloon was of the variable or stable long type compared to the stable short type.

A final analysis examined the difference between the number of adjusted pumps and the optimal number of pumps for each balloon type across task blocks. There were significant effects for both block and balloon type, as well as a significant interaction between them. The difference between adjusted pumps and optimal pumps significantly decreased from the first to the second block, and from the first to the third block. This provides evidence that participants learned the best number of pumps over the course of the task. This difference was significantly higher for variable and stable long balloons compared to stable short balloons, reflecting that more pumps were possible on these balloon types.

Risk‐Taking Positively Relates to Persistent Pursuit of Goals

To examine how risk-taking relates to individual differences in motivation and emotional processing, correlations were calculated between successful risk-taking on the BELT and the BIS/BAS Scales. The total number of adjusted pumps on the BELT showed a significant positive correlation with the BAS Drive Scale, which measures the persistent pursuit of goals. However, this correlation was no longer significant after adjusting for multiple comparisons. No significant correlations were found with the BAS Fun Seeking Scale, BAS Reward Response Scale, or BIS Scale.

Discussion

The main goal of this study was to examine learning through risk-taking in healthy young adults. In line with predictions and previous research, the results show that early, exploratory risk-taking helped learning in the participants. Taking risks early in the task, measured by the number of successful pumps in the first third of the task, was linked to earning more total points throughout the task. Additionally, the number of balloon explosions decreased as the task progressed, and participants increased the number of times they pumped the safer, stable long-explosion balloons. Since there was no direct feedback other than points earned and explosions, these findings suggest that risk-taking supported implicit learning, as participants learned which balloons were safer and more beneficial for success. Similarly, the difference between the number of pumps and the optimal number of pumps decreased, indicating that participants learned not only which balloons were safer but also how many pumps were generally possible before an explosion.

Another study reported that participants maintained a consistent number of pumps for both stable conditions and decreased pumps for the variable condition across blocks. In contrast, this study's young adult participants significantly increased the number of pumps on long, stable balloons over time. These differences might be due to the larger number of trials used in the current study, which may have offered more learning opportunities. Nonetheless, both studies provide evidence of learning, as these behavioral changes ultimately led to improved task performance.

The data also show that participants used feedback from previous trials to improve their performance. This was evident as they increased the number of pumps after a balloon did not explode (meaning points were saved and risk-taking was successful) compared to after a balloon exploded (meaning it was overinflated and risk-taking was unsuccessful). Pumping less after an explosion indicates sensitivity to negative feedback. Consistent with other research, this finding suggests that knowing the outcomes of earlier risks influenced future behaviors, ultimately improving task results.

Taken together, these findings suggest that risk-taking, when appropriate for the context, helps with learning—a positive outcome—in young adults. This work adds to a growing body of research that typically highlights negative consequences of risk-taking and supports ideas that risk-taking can be adaptive, particularly during developmental periods like adolescence. Some theories propose that risk-taking behaviors might offer evolutionary benefits during important developmental transitions. Following this idea, certain approaches suggest that interventions should encourage positive social pathways by creating environments that reduce stress and increase predictability, rather than working against natural tendencies. Whether such suggestions improve risk-related outcomes in adulthood, another key developmental stage with new social challenges, requires more investigation.

The second aim of the study was to see how risk-taking outcomes correlated with individual differences in motivation and emotional processing. The results were consistent with prior research, as risk-taking (successful pumps on the BELT) was positively related to the persistent pursuit of goals (as measured by the BAS Drive Scale). The BAS Drive Scale is based on self-reported agreement with statements about motivation for goal-directed behavior.

This finding has interesting implications, as goal-directed behaviors strongly predict academic success, especially for certain student populations. However, these results should be interpreted cautiously because the effect was no longer significant after adjusting for multiple comparisons. A power analysis indicated that the study had insufficient power to confidently detect a significant correlation for this specific measure. Therefore, larger and more diverse samples are needed to further explore the relationship between risk-taking and motivation and emotional processing in adults. Future research should also investigate how learning from risks might support academic outcomes in various college student populations, as academic outcomes were not assessed here. Furthermore, brain imaging studies are needed to understand how the development and activity of brain systems, such as those proposed in Ernst's Triadic Model, influence risk-taking behaviors and their links to motivation and emotional processes in adults.

Even with limited statistical power, the correlational results might be relevant for individuals in educational or professional environments where managing risk affects performance, safety (e.g., military, healthcare), financial success (e.g., finance, entrepreneurship), and creativity/innovation (e.g., arts, technology). In these and other fields, successful risk assessment and management are linked to better outcomes.

Conclusions

This study uniquely used an objective measure of risk-taking behavior to demonstrate learning from risks in adults. The findings confirm previous research showing that risk-taking supports learning. Additionally, risk-taking tendencies in the participants were positively related to goal-driven behavior. Together, these findings suggest that risk-taking, when appropriate for the situation, is an adaptive behavior that can improve learning-based educational and professional outcomes. The results from this study are expected to help inform interventions aimed at fostering positive outcomes during this important adult developmental period.

Introduction

Risk-taking involves behaviors where outcomes are not certain. In decision-making, risk-taking often shows two opposing patterns. "Risk aversion" means avoiding possible gains to keep things stable. In contrast, "risk-seeking" involves avoiding stability to get potentially larger gains. Traditionally, risk-seeking behaviors are linked to negative outcomes like acting on impulse, making poor choices, and mental health issues such as ADHD or addiction. Because of this, risk aversion is usually seen as a positive trait, though extreme risk aversion can indicate anxiety. However, new evidence suggests that the results of risk-seeking behaviors depend on the situation, and sometimes, they can even help with learning.

Recent studies using tasks like the Balloon Analog Risk Task (BART) show how people learn from risk. The BART measures risk-taking where higher risks can lead to rewards up to a point, after which more risk leads to losses. One study found that heavy social media users became more risk-averse after experiencing losses on an adapted BART. This shows that risk-taking behaviors can change based on what is learned about potential rewards and losses. Another study also found that risk-taking changed depending on previous outcomes, highlighting risk-related learning.

The Balloon Emotional Learning Task (BELT), adapted from the BART, is a computer task that measures how individuals learn through risk-taking. Unlike other risk tasks, the BELT includes conditions that are both stable and variable, which helps track how people learn from different risk situations. One study using the BELT found that people who actively sought out new experiences and learned from those risks performed better than those who took risks but did not learn, or those who avoided risks altogether. This means risk-seeking improved performance only when individuals learned from their past risky actions.

More research is needed to connect objective measures of risk-taking with positive behaviors in adults. Currently, only a few studies have looked at the relationship between risk-taking and adaptive behaviors in adult populations. Most existing research uses self-reports, which are subjective. These studies often look at risk-taking in narrow contexts, such as specific personality traits, or do not examine adaptive outcomes at all. Therefore, there is a clear need for more research that assesses objective risk-taking behaviors in relation to specific positive behaviors in adults.

Some research in adolescents suggests that certain risks can be beneficial. These include risks that improve well-being, have minimal costs that do not harm health, and are socially acceptable. Examples include starting a new friendship, trying a new food, or taking a challenging course. These situations involve uncertainty but have small potential downsides and could lead to good results.

While much of the research on adults focuses on the negative outcomes of risk-taking, more complex models acknowledge the intricate nature of risk-taking and emphasize examining situations where these behaviors might have benefits. For example, Ernst's Triadic Model offers insights into the brain and motivational processes behind these behaviors. This model describes three main brain systems: an approach system for rewards, an avoidance system for harm, and a regulation system that helps control these. This model can be used to understand how these systems influence the connections between risk-taking, motivation, and emotions in adults, as these brain systems are considered more developed in adulthood.

Research also shows that the outcomes from different risk-taking measures change with age. For example, performance on the BART varied in adolescents, while a self-report measure of general risk-taking, the BIS/BAS Scales, remained stable across adolescence and early adulthood. The BIS/BAS Scales measure general tendencies related to motivation or emotional responses. Interestingly, studies have shown a low connection between BART and BIS/BAS results in adolescents, indicating they might measure different aspects of risk-taking. More research is needed to see if these relationships between objective and self-reported risk-taking are also seen in adults, when the brain systems for approach and avoidance are fully developed.

Current Study

The current study had two main goals. First, it aimed to examine learning related to risk in healthy young adults. It was expected that participants who engaged in early, exploratory risk-taking would show better performance as the task progressed. Second, the study investigated how risk-taking outcomes relate to individual differences in motivation and emotional processing, as measured by the BIS/BAS Scales. Based on previous research showing that motivation and positive emotional states are linked to increased risk-taking in young adults, it was hypothesized that risk-taking behavior would have a positive relationship with motivation and positive emotional states in the young adult participants.

Materials and Methods

Participants

Participants were recruited from a university psychology department's research platform. To be eligible for the study, participants completed a pre-screening questionnaire. Individuals who reported a history or current diagnosis of an addictive disorder or color blindness were not allowed to participate. Data were collected from 58 participants. The average age was 19.66 years, with a standard deviation of 1.43 years; 74% of participants were female. Most participants (74.1%) identified as White/Caucasian, 8.6% as Black/African American, 5.2% as Chinese, and 6.9% as "other" races.

Measures

The study used two main measures: the Balloon Emotional Learning Task (BELT) and the Behavioral Inhibition System/Behavioral Activation System (BIS/BAS) Scales.

The Balloon Emotional Learning Task (BELT) is a computer-based task designed to measure differences in risk-taking behavior and implicit learning, meaning learning that happens without conscious effort. In the task, participants inflated balloon images by pressing a button, earning one point for each pump. The goal was to earn as many points as possible. Participants were told that balloons would pop if over-inflated and that balloons did not all pop at the same number of pumps. They could "bank" their points at any time after at least one pump, which saved the points and added them to a prize meter. If a balloon popped, no points were added. Balloons appeared in three colors, which secretly determined their explosion rate: two stable conditions (one popping at a low number of pumps, one at a high number of pumps) and one variable condition (popping at different pump counts). These different balloon types allowed researchers to see how participants learned which balloons to avoid or continue pumping. Key measures from the BELT included the total number of pumps (general risk-taking), adjusted pumps (successful risk-taking, excluding exploded balloons), post-explosion pumps (sensitivity to negative feedback), and optimal pumps (a measure of implicit learning based on how close pumps were to the balloon's maximum durability).

The Behavioral Inhibition System/Behavioral Activation System (BIS/BAS) Scales assess individual differences in two opposite motivational and emotional processes. Participants rated how much they agreed or disagreed with 24 statements using a four-point scale. The Behavioral Inhibition System (BIS) scale measures negative emotional responses to punishment, such as feeling hurt by criticism. The Behavioral Activation System (BAS) scale measures positive emotional responses to rewards. The BAS items are further divided into sub-scales for Reward Responsiveness (enjoying success), Drive (pursuing goals actively), and Fun-Seeking (desire for new and enjoyable experiences). Differences in these scales are linked to clinical outcomes like anxiety and impulsivity. These scales are reliable and have been shown to be valid in relation to other measures of motivation, emotion, and personality. In this study, the BIS/BAS Scales were used to explore how individual differences relate to performance on the BELT.

Procedures

All study procedures received approval from the Institutional Review Boards (IRBs) at Assumption University and Merrimack College. Undergraduate research participants were recruited through the university's research platform. Before participation, all individuals completed a pre-screening survey to confirm their eligibility. Research assistants obtained written informed consent from participants, who were reminded that they could ask questions or withdraw at any time. After consenting, participants first completed the BIS/BAS Scale on an iPad, which was untimed. Following this, participants completed the BELT, a task that took approximately 20 minutes. After finishing the BELT, participants were dismissed from the study.

Data Analysis

To determine if gender needed to be considered in the statistical models, Wilcoxon signed-rank tests were used to check for gender differences in total adjusted pumps and total explosions during the BELT task. Pearson correlations were calculated to examine two main relationships: first, the link between early exploratory behavior (measured by total adjusted pumps in the first third of the task) and the total points accumulated across the entire task; and second, the connections between successful risk-taking (total adjusted pumps across the task) and the scales from the BIS/BAS. Generalized linear models (GLMs) were employed to evaluate relationships between several factors, including: the number of explosions across different task blocks (first, middle, and last thirds of trials), the number of adjusted pumps based on balloon type (variable, stable long, stable short) and task block, how performance on the next balloon (total pumps) was affected by the outcome of the previous balloon (explosion or reward), and the difference between adjusted pumps and the optimal number of pumps for each balloon type across task blocks. For these models, balloon type and task block were considered fixed effects, and each participant was treated as a random effect. Models for "number of balloon pumps" and "number of explosions" used a specific statistical distribution suitable for count data. Effect sizes for significant results were reported as rate ratios with 95% confidence intervals. The significance level for all analyses was set at p = 0.05, and the Holm–Bonferroni method was applied to adjust for multiple comparisons in the correlation analyses to control for false positives.

Results

No significant differences were found between genders for either total adjusted pumps or total explosions during the task. Therefore, gender was not included in any of the subsequent analyses.

The study found that risk-taking behaviors supported implicit learning. A strong positive relationship existed between early exploratory risk-taking, measured by adjusted pumps in the first third of the task, and the total points earned throughout the task. Additionally, the number of explosions significantly decreased from the first block to the second and third blocks, indicating that early exploration helped improve performance later on. Participants also increased the number of pumps on balloons with long, stable explosion rates as the task progressed. When evaluating sensitivity to negative feedback, there was a significant increase in pumps after points were successfully saved (meaning the balloon did not explode) compared to after an explosion. This suggests participants used feedback from previous trials to adjust their future behavior. Furthermore, the difference between the number of adjusted pumps and the optimal number of pumps decreased across the task, demonstrating that participants learned not only which balloons were safer but also how many pumps were generally possible before a balloon would pop. These results collectively indicate that exploratory risk-taking early in the task helped participants learn implicitly and improve their overall performance.

Regarding the relationship between risk-taking and individual differences, the total number of adjusted pumps on the BELT showed a significant positive correlation with the BAS Drive Scale, which measures persistent pursuit of goals. The BAS Drive Scale assesses motivation for goal-directed behavior, reflecting statements like "I go out of my way to get things I want." However, it is important to note that this correlation was no longer statistically significant after adjusting for multiple comparisons, suggesting that more research with larger samples is needed to confirm this relationship. No significant correlations were found between total adjusted pumps and the BAS Fun Seeking, BAS Reward Response, or BIS Scales.

Discussion

This study aimed to understand learning related to risk in healthy young adults. The findings supported the hypothesis that early, exploratory risk-taking aided learning. Early exploration, measured by successful pumps in the first third of the task, was linked to earning more total points throughout the task. Moreover, the number of explosions decreased over time, and participants learned to pump the safer balloons (those with stable, long explosion rates) more frequently. Since there was no explicit instruction or feedback beyond points earned or explosions, these results suggest that risk-taking helped participants learn implicitly which balloons were safer and more beneficial for task success. The decreasing difference between actual pumps and optimal pumps further indicates that participants learned about balloon durability.

Our findings differ slightly from a previous study using the BELT, which showed consistent pumping across stable conditions but a decrease in the variable condition. In contrast, our young adult participants significantly increased pumping on the long, stable balloons over time. This difference might be due to the increased number of trials in our study, which could have provided more opportunities for learning. Nevertheless, both studies provide evidence of learning, as behavioral changes ultimately led to improved task performance.

The data also showed that participants used feedback from earlier trials to improve their later performance. They pumped more after successfully saving points (meaning a balloon did not explode) than after a balloon exploded. This sensitivity to negative feedback, consistent with prior research, suggests that knowing the outcomes of earlier risks influenced future behaviors, ultimately improving task results. Together, these findings suggest that risk-taking, when appropriate for the context, supports learning—a beneficial outcome—in young adults. This expands on existing literature that often highlights the negative aspects of risk-taking and aligns with theories suggesting adaptive risk-taking in developmental stages like adolescence. These theories propose that certain risky behaviors might have evolutionary advantages, encouraging interventions that foster positive environmental conditions to support beneficial risk-taking.

The second goal of the study was to explore how risk-taking correlates with individual differences in motivation and emotional processing. As hypothesized and consistent with previous research, risk-taking (total adjusted pumps on the BELT) was positively related to the persistent pursuit of goals, as measured by the BAS Drive Scale. This scale assesses an individual's self-reported motivation for achieving desired outcomes. While this finding is compelling, it is important to interpret it cautiously, as the correlation lost its statistical significance after adjusting for multiple comparisons. A power analysis indicated that the study might have been underpowered to detect a significant correlation, suggesting the need for research with a larger, more diverse sample to further investigate this relationship. Additionally, future studies should explore how risk-related learning might affect academic outcomes in college students, as this was not examined here. More neuroimaging research is also needed to understand how brain systems, such as those described in Ernst's Triadic Model, influence risk-taking behaviors and their connections with motivation and emotions in adults.

Despite being potentially underpowered, these correlational results could be relevant for individuals in educational or professional fields where managing risk impacts performance, safety (e.g., military, healthcare), financial success (e.g., investment banking, entrepreneurship), or creativity and innovation (e.g., arts, technology, research). In such careers, effectively assessing and managing risk is often linked to better outcomes.

Conclusions

This study uniquely used an objective measure of risk-taking behavior to demonstrate that adults engage in learning related to risk. These findings support previous research showing that risk-taking helps with learning. Additionally, risk-taking tendencies were positively connected to goal-driven behavior in the participants. Overall, these results suggest that risk-taking, when approached thoughtfully and in relevant contexts, is beneficial and can improve learning-based outcomes in both educational and professional settings. The insights gained from this study are expected to help inform interventions aimed at shaping positive outcomes during adulthood.

Introduction

Risk-taking involves actions that lead to uncertain results. People often view risk-taking in two main ways. "Risk aversion" means avoiding possible rewards to keep things stable. On the other hand, "risk-seeking" means giving up stability to get bigger rewards. Traditionally, risk-seeking behaviors have been linked to negative outcomes like acting without thinking and making bad choices. They are also connected to certain mental health conditions, such as ADHD or addictive behaviors. Because of this, avoiding risk is often seen as a good thing, though too much risk aversion can also point to anxiety disorders. However, a growing amount of evidence shows that the results of risk-seeking behaviors depend on the situation, and sometimes these behaviors can actually help with learning.

Recent studies using tasks like the Balloon Analog Risk Task (BART) show how people learn through risk. In the BART, taking risks can lead to rewards up to a certain point, after which more risk leads to losses. Researchers have found that how people react to risk can change based on what they have learned from past rewards or losses. For example, some studies found that people who used social media excessively became more risk-averse after they experienced losses. Other research has also shown that risk-taking behaviors change based on previous outcomes, like gains and losses, which further proves that people learn from taking risks.

The Balloon Emotional Learning Task (BELT) is another computer task, adapted from the BART, that measures how individuals learn through risk-taking. Unlike other common risk tasks, the BELT has both changing and stable conditions. This allows researchers to track how people differentiate between conditions, which helps them learn about risks without being directly told. One study using the BELT found that people who enjoyed seeking new sensations and were good at learning from those risks performed better on the task. This suggests that risk-seeking behaviors led to better performance only when individuals learned from the risks they had previously taken.

There is a need for more research to connect objective ways of measuring risk-taking with positive behaviors in adults. Currently, only a few studies have looked at how risk-taking relates to adaptive behaviors in adult populations. Much of the existing research relies on what participants report about themselves. In these studies, the links between risk-taking and positive outcomes were often explored in a very narrow way, such as only looking at certain personality traits or behaviors. More research is needed that uses objective ways to measure risk-taking behavior and examines its links to specific adaptive behaviors in adults.

In adolescents, certain types of risks can have positive effects on development. These are risks that benefit a person's well-being, have potential costs that do not harm health, and are socially acceptable. For instance, choosing to make a new friend, trying a new food, or signing up for a challenging class all involve some risk because the results are unknown. However, these risks often have minimal downsides and can lead to good results.

Current Study

This study had two main goals. First, it aimed to study how young adults who do not have specific health conditions learn from taking risks. Researchers believed that participants who explored and took risks early in the task would perform better as the task continued. Second, the study wanted to see how the results of risk-taking related to differences in a person's motivation and emotional processing. Based on earlier research in adolescents and young adults, it was expected that risk-taking behavior would be positively related to a person's motivation and positive emotional state in the young adult group studied.

Materials and Methods

Participants

Participants were recruited from a university's psychology department using a student research platform. Everyone first filled out a survey to check if they were eligible for the study. Those with a past or current self-reported addiction or color blindness were not allowed to participate. Data was collected from 58 participants, with an average age of about 19 and a half years old. Most participants were female (74%) and identified as White/Caucasian (74.1%).

Measures

Balloon Emotional Learning Task (BELT)

The BELT is a computer task that measures differences in risk-taking behavior through exploration and learning. During the task, participants inflated images of balloons on a screen by pressing a button. Each button press earned one point, and the goal was to earn as many points as possible. Participants were told that balloons would pop if inflated too much and that not all balloons popped at the same point. They could "bank," or save, their points after at least one pump. Saving points created a success sound, and the points were added to a visible prize meter. If a balloon popped, participants heard an explosion sound, "Pop!" appeared on the screen, and no points were added.

Balloons came in three different colors. Unknown to the participants, each color had a different explosion rate. Two colors had stable explosion rates—one popped quickly (e.g., after 6 pumps) and one popped slowly (e.g., after 18 pumps). The third color had a variable explosion rate, meaning it could pop after 6, 12, or 18 pumps. All three balloon colors were shown an equal number of times across 54 trials.

The two stable conditions allowed researchers to see how participants learned over time, expecting them to learn which balloons to avoid (low explosion rate) and which to keep pumping (high explosion rate) based on past experiences. The variable condition allowed for testing risk-taking when the outcome was uncertain. Several measurements were taken from the BELT, including the number of pumps (general risk-taking), adjusted pumps (successful risk-taking, excluding popped balloons), post-explosion pumps (sensitivity to negative feedback), and optimal pumps (how well participants learned the best number of pumps).

Behavioral Inhibition System/Behavioral Activation System (BIS/BAS) Scale

The BIS/BAS Scales measure individual differences in two opposing processes related to motivation and how emotions are handled. Participants rated how much they agreed or disagreed with 24 statements using a four-point scale. Most items were scored in reverse. The average responses for groups of items created scores for two main areas of motivation and emotion.

The Behavioral Inhibition System (BIS) scale measures negative emotional reactions to punishment, such as feeling hurt by criticism. The Behavioral Activation System (BAS) scale measures positive emotional reactions to rewards. BAS items are further divided into categories like Reward Responsiveness (enjoying continuing something when doing well), Drive (working hard to get what is wanted), and Fun-Seeking (always willing to try new, fun things). Scores on these scales are linked to clinical issues like anxiety and impulsivity. The items in these scales are reliable and consistent, and the scales have been shown to be valid when compared to other measures of motivation, emotion, and personality. In this study, the BIS/BAS Scale results were used to examine how individual differences related to performance on the BELT.

Procedures

All study procedures were approved by the review boards at the participating universities. Undergraduate students were recruited, and they first completed a pre-screening survey to confirm their eligibility.

Research assistants obtained written informed consent from all participants, reminding them that they could ask questions or leave the study at any time. After consent, participants completed the BIS/BAS Scale on an iPad, without a time limit. Next, they performed the BELT task, which took about 20 minutes. Once the BELT was finished, participants were excused from the study.

Data Analysis

Statistical tests were used to check for gender differences in total successful pumps and total explosions during the BELT task. This was to see if gender needed to be considered in later statistical models. Pearson correlations were used to explore two things: first, the link between early exploratory behavior (successful pumps in the first third of the task) and the total points earned throughout the task; second, the relationships between total successful pumps across the whole task and the scores from the BIS/BAS Scales.

Generalized linear models were used to examine several relationships. These included: how the number of explosions changed across different parts of the task; how the number of successful pumps changed based on the balloon color (variable, stable long, stable short) and task block; how performance on the next balloon was affected by the outcome of the previous balloon (explosion or points saved); and how the difference between successful pumps and the ideal number of pumps changed based on balloon color and task block. Balloon color and task block were treated as fixed factors, and each participant was treated as a random factor. A significance level of 0.05 was set for all analyses, and a method was used to adjust for multiple comparisons in the correlation analyses.

Results

No significant differences were found between genders for either total successful pumps or total explosions during the task. Therefore, gender was not included in any of the following analyses.

Risk-Taking Supports Implicit Learning

A strong positive link was found between the number of successful pumps during the first third of the task and the total points earned throughout the task. Also, when looking at the number of explosions across the task, there was a clear effect of the task block. The number of explosions was significantly lower in the middle and last thirds of the task compared to the first third, with no difference between the middle and last thirds. These findings suggest that exploring and taking risks early in the task helped improve performance later on.

When looking at changes in successful pumps based on balloon type and task block, there was a significant main effect of balloon type on the number of successful pumps per balloon, and a significant interaction between balloon type and block. Participants pumped balloons with variable explosion rates and stable, long explosion rates significantly more times than balloons with stable, short explosion rates. The significant interaction effect showed an increase in pumps for balloons with long, stable explosion rates over the course of the task.

To see how sensitive participants were to negative feedback, another analysis looked at the relationship between the outcome of the previous balloon (explosion or saved points) and the number of pumps on the next balloon. There was a significant main effect of the previous outcome and the current balloon type. Participants pumped significantly more times after they had saved points (meaning a successful pump) compared to when a balloon exploded. They also pumped significantly more if the current balloon type had variable or stable, long explosion rates compared to a stable, short explosion rate.

A final analysis explored the difference between the number of successful pumps and the ideal number of pumps (how many pumps a balloon could take before exploding) based on balloon type across task blocks. There was a significant effect of both block and balloon type, as well as a significant interaction between them. The difference between successful and ideal pumps decreased significantly from the first to the second block, and from the first to the third block. This provides evidence that participants learned the optimal number of pumps as the task progressed. The difference was significantly higher for balloons with variable and stable, long explosion rates compared to stable, short explosion rates, which reflected the greater number of pumps possible for these balloon types.

Risk-Taking Positively Relates to Persistent Pursuit of Goals

To examine how risk-taking related to individual differences in motivation and emotional processing, correlations were calculated between the total successful pumps on the BELT and the BIS/BAS Scales. The total number of successful pumps on the BELT showed a significant positive correlation with the BAS Drive Scale, which measures persistent pursuit of goals. However, this correlation was no longer significant after adjusting for multiple comparisons. No significant correlations were found with the BAS Fun Seeking Scale, BAS Reward Response Scale, or the BIS Scale.

Discussion

A main goal of this study was to evaluate how young adults learn from taking risks. In line with the study's predictions and previous research, the results show that early, exploratory risk-taking helped participants learn. Exploring early in the task, measured by the number of successful pumps in the first third of the task, was linked to earning more total points throughout the task. Also, the number of explosions decreased as the task went on, and participants increased how many times they pumped the safer balloons with long, stable explosion rates. Since there was no direct feedback other than points and explosions, these findings suggest that risk-taking supported learning, as participants figured out which balloons were safer and more beneficial for success. Additionally, the difference between the number of pumps and the optimal number of pumps decreased, indicating that participants learned not only which balloons were safer but also how many pumps they could take before a balloon exploded.

Compared to an earlier study, the young adult participants in this research significantly increased the number of pumps on long, stable balloons across blocks, while the previous study showed consistent pumping. These differences might be because this study used twice as many trials, giving participants more chances to learn. Regardless, both studies show evidence of learning, as these changes in behavior ultimately improved task performance.

The data also show that participants used feedback from previous trials to improve their performance. This was seen in the increased number of pumps after a balloon did not explode (meaning points were saved and risk-taking was successful) compared to when a balloon did explode (meaning the balloon was overinflated and risk-taking was unsuccessful). Fewer pumps after explosions indicate sensitivity to negative feedback. This finding, consistent with previous work, suggests that knowing the outcomes of earlier risks influenced later behaviors, which ultimately improved task results.

Overall, these results suggest that taking risks in specific situations supports learning, which is a positive outcome, in young adults. This work adds to existing research that often highlights the negative aspects of risk-taking, and it supports ideas that suggest risk-taking can be adaptive, especially during important developmental stages like adolescence. For example, some theories propose that risk-taking behaviors might have evolutionary benefits during adolescence.

The study's second goal was to see how risk-taking outcomes correlated with individual differences in motivation and emotional processing. Consistent with previous research, risk-taking, measured by total successful pumps on the BELT, was positively linked to a persistent pursuit of goals, as measured by the BAS Drive Scale. This scale uses self-reported items that gauge motivation for goal-directed behavior, such as "I go out of my way to get things I want."

These findings have important implications, as goal-directed behaviors are strong predictors of academic success. However, these results should be viewed carefully because the effect was no longer significant after accounting for multiple comparisons. More research with a larger, more diverse group of adults is needed to further explore these connections. Future studies should also investigate how learning through risk might support academic outcomes in college students, as this was not evaluated in the current study. Additionally, brain imaging research is needed to understand how brain systems influence risk-taking and its relationship with motivation and emotions in adults.

Even with the limitations, these correlation results could be valuable for professionals in fields where risk management impacts performance, safety (e.g., military, healthcare), financial success (e.g., investing, entrepreneurship), and creativity (e.g., arts, technology). In these and other careers, successfully assessing and managing risks is linked to better outcomes.

Conclusions

This study used an objective measure of risk-taking behavior to demonstrate that adults learn from taking risks. These findings confirm earlier research showing that risk-taking helps with learning. Additionally, risk-taking tendencies were positively related to goal-driven behavior in the study participants. Together, these results suggest that taking risks in the right situations can be beneficial and may improve learning-based outcomes in education and professional life. The findings from this study are expected to help create programs that encourage positive outcomes during this important stage of adult development.

Introduction

Taking risks means doing things where the result is not certain. There are two main ways people act when faced with risk. One is "risk aversion," which means avoiding a possible gain to keep things steady and safe. The other is "risk-seeking," which means avoiding steady situations to try for bigger gains.

People have often linked risk-seeking to bad outcomes, like acting without thinking or making poor choices. It has also been connected to mental health issues such as ADHD or addiction. Because of this, avoiding risk is often seen as a good thing. (However, very high levels of avoiding risk might also be a sign of anxiety problems.) But more and more facts show that the results of taking risks depend on the situation. Sometimes, taking risks can even help people learn.

New studies using a task called the Balloon Analog Risk Task (BART) show how learning happens with risk. In the BART, taking risks can earn rewards up to a certain point. After that point, more risk leads to bad results. One study found that people who used social media a lot became more careful about risks, but only after they had lost points in the game. This showed that they learned to change their risk-taking behavior based on what happened. Another study also found that people changed how much risk they took based on past wins or losses, further showing learning related to risk.

A similar task, the Balloon Emotional Learning Task (BELT), helps measure how individuals learn through taking risks. Unlike other risk tasks, the BELT has both changing and steady situations. This helps track how people learn from different risks. One study using the BELT found that people who liked seeking thrills and were also good at learning from those risks did much better on the task. This means that taking risks helped them do better, but only if they learned from the risks they had taken.

More studies are needed to connect real ways of measuring risk-taking with good behaviors in adults. So far, not many studies have truly looked at how risk-taking connects to good behaviors in adults. Also, most past studies relied on what people said about themselves. In these studies, the links between risk-taking and good results were often looked at in a very narrow way, or not at all. Because of this, more research is greatly needed to use real measures of risk-taking and see how they relate to specific good behaviors in adults.

Studies with teenagers show that risks can have good effects if they: 1) help a person's health and happiness, 2) have possible costs that won't harm a person's health or happiness, and 3) are generally accepted by others. For example, deciding to make a new friend, trying a new food, or taking a challenging class all involve risk because the outcomes are unknown. However, these risks usually don't cost much and can lead to good results.

While much of the research on adults points to the possible bad results of risk-taking, some detailed ideas explain the complex nature of risk-taking. These ideas highlight that it's important to look at situations where such actions might lead to good outcomes. For example, Ernst's Three-Part Idea offers helpful thoughts on the brain and what drives certain behaviors. This idea describes three main brain systems: the "go for it" system (for rewards), the "stay away" system (for avoiding harm), and the "control" system (for making choices). This idea can be used to see how these systems affect how risk-taking connects to a person's drive and feelings in adults, as these brain systems are thought to be more developed in adulthood.

Related studies also show that results from different ways of measuring risk-taking change with age. For instance, a study found that how teenagers did on the BART varied. However, what teenagers and young adults said about their general risk-taking (using a test called BIS/BAS Scales) stayed the same. Interestingly, that study also found little connection between the BART and the BIS/BAS Scales for this age group. This suggests that these two ways of measuring risk-taking look at different things in teenagers. More research is needed to see if connections between real and self-reported risk-taking are found in adults, when the brain systems for "going for it" and "stay away" are fully grown.

Current Study

This study had two main goals. First, the researchers wanted to see how typical young adults learn from risk. They thought that people who took risks early on would do better as the task went on. Second, they wanted to see how risk-taking results connected to how people differ in what drives them and how they feel (measured by the BIS/BAS Scales). Studies with teenagers show that wanting to go for things connects positively with taking risks. Also, being driven and having a good mood are linked to taking risks in young adults. So, the researchers thought that taking risks would connect positively with a person's drive and good mood in this group of young adults.

Materials and Methods

Participants

People for the study were found using a student study website at Assumption University's Psychology Department. Everyone filled out a short survey first to see if they could be in the study. People who had or said they had an addiction problem or could not see colors were not allowed to join. The study collected information from 58 people (average age was about 19.66 years old, with most being female). Most of the people in the study were White/Caucasian, with smaller numbers from other racial backgrounds.

Measures

The Balloon Emotional Learning Task (BELT) is a computer game that measures how people learn by exploring and taking risks. In the game, people pressed a button to pump up pictures of balloons on a screen. Each pump earned one point. The goal was to get as many points as possible. People were told that balloons would burst if pumped too much, and that not all balloons burst at the same point. People could "bank" (save) their points after at least one pump. When points were saved, they heard a sound and the points were added to a score on the screen. Then a new balloon appeared. If a balloon was pumped too much, people heard a burst sound and the word "Pop!" appeared. No points were added, and a new balloon appeared.

The balloons came in three different colors. People did not know that the color decided how quickly the balloon would burst. Two colors represented steady situations: one always burst after a small number of pumps (e.g., 6 pumps), and the other always burst after a large number of pumps (e.g., 18 pumps). The third color represented a changing situation where the balloon burst at different times (e.g., after 6, 12, or 18 pumps). All three colors of balloons were shown an equal number of times across 54 tries in the game.

The two steady situations helped the researchers see how people learned. People should learn which colored balloons to be careful with (low burst rate) and which to keep pumping (high burst rate) by using what happened in past tries. The changing situation also allowed them to see how likely someone was to take risks when things were unclear. The BELT measured several things, including:

1. Pumps: how much risk people took (more pumps meant more risk).

2. Adjusted pumps: how many successful risks people took (pumps on balloons that did not burst).

3. Post-explosion pumps: how much bad results (bursts) affected their next pumps.

4. Optimal pumps: how well people learned the best number of pumps for each balloon.

The Behavioral Inhibition System/Behavioral Activation System (BIS/BAS) Scales measure how people differ in two opposite ways related to what drives them and how they feel. People answered 24 questions on a scale of 1 to 4, saying how much they agreed or disagreed. Most answers were scored so that 1 meant strong agreement and 4 meant strong disagreement. The scores from groups of questions were then added up and averaged to create different scores.

The BIS scale measures bad feelings about being punished (e.g., "Criticism or scolding hurts me quite a bit."). The BAS scale measures good feelings about getting a reward. BAS items are further broken down into things like how much someone reacts to rewards, how much "drive" they have (e.g., "I go out of my way to get things I want."), and how much they like seeking fun (e.g., "I am always willing to try something new if I think it will be fun."). Differences in these scores can relate to issues like anxiety or acting without thinking. These scales are very consistent and measure what they are supposed to. In this study, the BIS/BAS scores were used to see how individual differences connected to how people did on the BELT.

Procedures

All study steps were approved by groups that make sure studies are safe and ethical at Assumption University and Merrimack College. College students were found for the study through SONA. Everyone filled out a short survey first to see if they could take part.

Study helpers first got written permission from each person to join the study. People were told they could ask questions or quit the study at any time. After they gave permission, people filled out the BIS/BAS Scale on an iPad, with no time limit. Then, they completed the BELT, which took about 20 minutes. After finishing the BELT, people were allowed to leave the study.

Data Analysis

The researchers used certain math tests to look for differences between genders in total successful pumps and total bursts. This helped them decide if gender needed to be considered in the results. They also used other math tests to see how early risk-taking connected to total points earned. And they looked at how total successful pumps connected to the scores from the BIS/BAS test.

They used specific computer models to study: 1) how many balloons burst in different parts of the task, 2) how many successful pumps were made for each balloon type and in each part of the task, 3) how many pumps were made on the next balloon based on its type and what happened with the last balloon, and 4) how far off the successful pumps were from the best number of pumps for each balloon type and task part. A result was considered important if the chance of it happening by accident was less than 5%. A special method was used to check for many connections at once.

Results

There were no big differences between genders for either total successful pumps or total bursts in the task. So, gender was not included in any of the later analyses.

Risk-Taking Supports Learning

There was a clear good connection between successful pumps in the first part of the task and the total points earned throughout the task. Also, when looking at the number of bursts in different parts of the task, bursts were much lower in the middle and last parts compared to the first part. This means that taking risks and exploring early in the task helped people do better later on.

When looking at changes in successful pumps for different balloon types and task parts, there was a clear effect for balloon type and a clear connection between balloon type and task part. People made many more pumps on balloons that burst at changing times and on balloons that burst after many pumps, compared to balloons that burst after few pumps. This connection was because people increased their pumps on balloons that burst after many pumps as the task went on.

To see how people reacted to bad results, another analysis looked at what happened with the previous balloon (burst or points saved) and how many pumps were made on the next balloon. There was a clear effect for what happened before and for the current balloon type. People pumped more after saving points compared to after a balloon burst. They also pumped more if the current balloon type was changing or burst after many pumps, compared to balloons that burst after few pumps.

A final analysis looked at how far off the successful pumps were from the best number of pumps for each balloon type and task part. This difference went down a lot from the first part of the task to the later parts. This shows that people learned the best number of pumps as the task went on. This difference was much higher for balloons that burst at changing times and for balloons that burst after many pumps, which makes sense because more pumps were possible on those balloons.

Risk-Taking Connects to Working Hard for Goals

To see how risk-taking connected to how people differ in what drives them and how they feel, the researchers looked at the links between total successful pumps on the BELT and the BIS/BAS Scales. The total number of successful pumps on the BELT connected positively with the BAS Drive Scale. This scale is about working hard to reach goals. However, this connection was not considered clear anymore after the researchers carefully checked for many connections at once. It did not connect with the BAS Fun Seeking Scale, BAS Reward Response Scale, or BIS Scale.

Discussion

The main goal of this study was to see how typical young adults learn from risk. As thought, and like other studies found, the results show that taking risks and exploring early on helped people in this study learn. Taking risks early in the task was connected to getting more points in total. Also, the number of balloons that burst went down as the task went on, and people pumped the safer balloons (those that burst after many pumps) more over time. Since there was no clear information given on how well people were doing (other than points and bursts), this suggests that risk-taking helped people learn without even trying. They learned which balloons were safer and therefore better for success. Also, the difference between the number of pumps and the best number of pumps went down, meaning people learned not only which balloons were safer, but also how many pumps they could do before a balloon burst.

A past study found that people kept pumping the same amount in the steady situations and pumped less in the changing situation. In contrast, this study's young adult group pumped the long, stable balloons much more as the task went on. These differences might be because this study had more tries (54 tries compared to 27 in the other study), giving more chances to learn. Still, both studies show that learning happened, as these changes in behavior eventually led to better overall performance.

The results also show that people used what happened in past tries to do better in the future. This is seen by the increased number of pumps after a balloon did not burst (meaning successful risk-taking) compared to after a balloon burst (meaning unsuccessful risk-taking). Pumping less after bursts shows that people were affected by bad results. This finding suggests that knowing what happened with earlier risks changed future actions, which ultimately improved how people did on the task.

Together, these results suggest that taking risks that fit the situation helps learning, which is a good outcome, in young adults. This work adds to many studies that show bad effects of risk-taking, and it supports ideas that risk-taking can be helpful in young people. Some ideas say that risk-taking can have benefits for survival during important growth stages like the teenage years. For example, aggressive actions might show good traits like bravery or social standing. Following this idea, some experts suggest that programs should encourage good social paths by creating environments that reduce stress, make things more clear, and remove harsh rules. This would better match what young people want and feel, instead of going against their natural urges. More research is needed to see if such ideas help with risk-related outcomes in adulthood, which is an important stage of growth with its own social challenges.

The second goal of this study was to see how risk-taking results connected to how people differ in what drives them and how they feel. This idea was supported and matched earlier research: risk-taking (total successful pumps on the BELT) connected positively to working hard for goals (measured by the BAS Drive Scale). The BAS Drive Scale is based on what people say about themselves in 4 questions about their drive to reach goals, such as "I go out of my way to get things I want."

This finding has important meanings, as working hard for goals strongly predicts success in school, especially for older students who take breaks during college. However, these results should be looked at carefully, as the effect was not considered clear after checking for many connections. A later check of the study's strength showed that the study did not have enough power to find a clear connection between BELT successful pumps and the BAS Drive Scale. So, more studies with a larger and more varied group of people are needed to learn more about how risk-taking connects to what drives adults and how they feel. Also, future research needs to see how learning from risk might help with school success for different college students, as this study did not look at school results. Moreover, brain scan studies are needed to understand how brain systems, like those in Ernst's Three-Part Idea, grow and work, and how they affect risk-taking and connections to what drives people and how they feel in adults.

Even though the study was small, the results about connections might be helpful for people in schools or jobs where risky actions affect how well someone does, physical safety (like in the military or hospitals), money gain or steadiness (like in banking or starting a business), and new ideas (like for artists, writers, tech workers, or researchers). In these and other jobs, good checking and handling of risk leads to better results.

Conclusions

This study was new because it used a real way to measure risk-taking to show how adults learn from risk. These findings confirm earlier research that shows risk-taking helps people learn. Also, how likely someone was to take risks connected positively to working hard for goals in this study. Together, these findings suggest that taking risks that fit the situation is helpful and can lead to better results in school and work that depend on learning. The results from this study are expected to help create programs that lead to good outcomes during this important time of life.