Boredom is typically defined as “the aversive experience of wanting, but being unable, to engage in satisfying activity” (Eastwood et al., 2012, p. 483). Prior research has demonstrated that boredom can lead to both positive and negative outcomes, such as increased prosocial intentions (van Tilburg & Igou, 2017a), lower performance in work and study (O'hanlon, 1981; Tze et al., 2016), and binge eating, binge drinking, gambling, and drug use (Biolcati et al., 2016; Moynihan et al., 2015). Additionally, experimental studies in college samples demonstrated that boredom leads to increased self-administration of (painful) electric stimulation (Havermans et al., 2015; Nederkoorn et al., 2016; Wilson et al., 2014; Yusoufzai et al., 2022). In these studies, the voluntary self-administration of electric schocks was conceived as an experimentally controlled proxy for non-suicidal self-injury (NSSI). NSSI is defined as “the deliberate destruction of one’s own body tissue in the absence of intent to die” (Nock et al., 2009, p. 78). Self-shocking out of boredom could be explained through mismatched task requirements and attentional capacity, suggesting that boredom results from suboptimal levels of arousal (Gerritsen et al., 2014; Pattyn et al., 2008). While the literature concerning the dimensions of arousal associated with boredom is mixed (Elpidorou, 2021), and different instances of boredom are accompanied by both low and high arousal (Goetz et al., 2014), the current paper is specifically concerned with the stereotypical low arousal and monotonous boredom experience. Insufficient arousal could lead to seeking stimuli that increase arousal, resulting in e.g., drug use, gambling, or NSSI. In line with this hypothesis, NSSI prevalence is high in situations where people cannot escape boredom, such as correctional settings, especially in solitary confinement (Casiano et al., 2013; Dixon-Gordon et al., 2012; Kaba et al., 2014). Furthermore, among other emotions, boredom commonly preceded NSSI behavior in a self-report study of female inmates (Chapman & Dixon-Gordon, 2007). Alternatively, another theory states that boredom results from opportunity costs being high (Kurzban et al., 2013), which would imply that NSSI as an alternative option may be appraised as an activity with a high potential value. Altogether, boredom might trigger NSSI as an attempt to escape this negative emotion, though further research is warranted to examine contextual (environmental and person-based) factors modulating this effect (Yusoufzai et al., 2022).

One important contextual factor needing further research is the presence of an alternative behavioral response while being bored. The aforementioned experimental studies provided participants only with the option to shock themselves or do nothing at all (Havermans et al., 2015; Nederkoorn et al., 2016; Wilson et al., 2014; Yusoufzai et al., 2022). In daily life however, behavioral options are rarely limited to negative ones. It is therefore unclear whether boredom also leads to NSSI when a positive behavioral alternative is available, since the latter could also alleviate the monotony of boredom through stimulation. In other contexts, the effect of boredom under circumstances of multiple choice has been explored tentatively: one study (Pfattheicher et al., 2020) demonstrated that boredom increased negative behaviors towards others, though after including the option to behave prosocially, this effect only persisted in participants with high dispositional sadism. Another study (Bench & Lench, 2019) found that bored participants chose to view negative images despite having the option of positive images. The current study will therefore examine the robustness of the effect of boredom on self-administered aversive stimuli as a proxy for NSSI behavior, by also offering a positive behavioral option during boredom. Self-administration of aversive and positive sounds was used as a proxy for NSSI and a positive response, respectively. These two options are procedurally similar, differing only in valence. Noxious sounds are widely used in e.g., conditioning studies (e.g., Waters et al., 2017; You et al., 2021), showing their effectiveness as aversive stimuli that participants want to avoid (e.g., Klein et al., 2021). In addition, positive natural sounds can facilitate mood recovery (Benfield et al., 2014) and thereby diminish monotony (i.e., increase arousal).

Boredom is not the only negative emotion that has been linked to NSSI, as anger is often reported to elicit NSSI behavior as well. Anger, like boredom, is experienced as frustrating and negative (Kuppens et al., 2007). It has previously been reported to be a common NSSI antecedent in both female inmates (Chapman & Dixon-Gordon, 2007) and patients with eating disorders (Claes et al., 2010; Paul et al., 2002). Furthermore, self-reports in male inmates also demonstrated that anger was associated with high levels of concurrent self-harm ideation (Humber et al., 2013). In earlier work (Yusoufzai et al., 2022) we examined for the first time whether boredom and anger have a similar effect on self-administered aversive stimuli as a proxy for NSSI. While the results showed that the boredom induction led to more self-administered shocks than the anger induction, shock intensity (i.e., painfulness) did not differ between both conditions. Additionally, the anger experience induced via video was presumed to have been of an indirect nature out of sympathy for the characters in the video, rather than being a direct first-hand anger experience. The current study will use an anger induction method requiring participants to reflect on personal anger-provoking experiences through a writing task. Similar guided writing tasks and verbal autobiographical recall have repeatedly been used and found to be effective anger inductions (e.g., Lobbestael et al., 2008; Moons & Mackie, 2007; Moons & Shields, 2015; Tiedens & Linton, 2001).

Research already demonstrated the relevance of personality characteristics in the effect of boredom on self-administering aversive stimuli. Self-shocking out of boredom, for example, was shown to be especially prominent in participants with a history of NSSI (i.e., engaged in NSSI behavior at least once in their life; Nederkoorn et al., 2016; Yusoufzai et al., 2022). A possible explanation for this could be that prior NSSI experiences increased participants’ pain tolerance, and/or taught them the ‘benefits’ of this behavior, as NSSI can improve emotional state (Hooley & Franklin, 2018). A second personality factor that has been linked to NSSI is negative urgency, a sub-construct of impulsivity that refers to the tendency to engage in impulsive action in response to negative emotions like depression, boredom, and stress (Lynam et al., 2006). Previous studies in university student samples demonstrated that negative urgency predicted the onset and presence of NSSI symptoms (Allen & Hooley, 2018; Riley et al., 2015). Furthermore, a meta-analysis (Hamza et al., 2015) found that out of all impulsivity sub-constructs, only negative urgency showed consistent effects in relation to NSSI.

Taken together, the current experiment replicates and extends prior research on the effect of boredom on self-administered aversive stimuli as a proxy for NSSI behavior, by examining its robustness in the presence of a positive behavioral alternative, and its specificity by comparing the effect of boredom with that of anger. Furthermore, personal characteristics are examined as potential moderators of the link between boredom and self-administered aversive stimuli. Specifically, we will test the moderating impact of NSSI history and negative urgency.

Methods

Participants

The current study was approved by the Ethical Research Committee of the Faculty of Psychology and Neuroscience, Maastricht University (approval code 184_01_10_2017). Sample size was based on previous studies examining the effect of boredom on self-administered aversive stimuli, which found large and moderate-to-large effects for the main effect of boredom and the interaction effect between boredom and history of NSSI (η2partial = 0.13 to 0.41; Havermans et al., 2015; Nederkoorn et al., 2016; Yusoufzai et al., 2022). G*Power 3.1 (Faul et al., 2009) was used for power analysis, selecting “Linear multiple regression: Fixed model, R2 increase”, a moderate effect size of f2 = 0.15, an alpha rejection criterion of 0.05, power of 0.95, 4 tested predictors and 8 total predictors (due to the use of 2 separate dummy variables for condition), which indicated 129 participants were needed. We aimed to test between 129 and 150 participants to account for potential drop out. Participants were recruited via study advertisements at Maastricht University. We did not mention NSSI in our advertisements, meaning that we did not target participants specifically familiar with NSSI. A total of 130 participants were recruited. One participant was excluded from the dataset due to suspicion of copy-pasting instead of typing during the writing task, as their character length in the writing task far exceeded the sample mean (11,552 as opposed to 1961, respectively). Thus, a total of N = 129 participants were included in the current analyses. Mean age of participants was 20.70 years, SD = 3.02. The majority was female (77.52%), and a university student (98.5%). More than half (61.24%) of participants reported a history of performing NSSI at least once in their life.

Materials

Self-administered stimuli

Participants were presented with two different sound clips of four seconds each. The aversive sound clip was of a screaming pig, while the positive clip was of chirping birds, both of which could be played by pressing a designated button on the keyboard in front of the participants. Before starting the task, participants were asked to listen to the positive and aversive sound once and rate their pleasantness on a scale from 0 (not at all) to 100 (extremely). This initial introduction of the sounds was also meant to familiarize participants with the stimuli and reduce novelty effects. The number of times participants played each of the sound clips (i.e., frequency) during the writing task was recorded. In the current study, a paired samples t-test showed that the screaming pig sound (M = 11.95; SD = 12.66) was rated as significantly less pleasant than the chirping bird sound (M = 81.19; SD = 18.64); t(128) = -31.30, p < .001, Cohen’s d = 4.38.

Emotional state

Participants rated their emotional state on visual analogue scales ranging from 0 (does not describe my feelings) to 100 (clearly describes my feelings). They were asked how negative, relaxed, angry, bored, sad, aroused, happy, positive, and frustrated they felt in the present moment.

Negative urgency

Negative urgency was measured using a subscale of the Urgency, Premeditation (lack of), Sensation Seeking, Positive Urgency (UPPS-P) Impulsive Behavior Scale (Cyders, 2013). This self-report scale measures five aspects of impulsivity, including negative urgency. Participants rated 12 statements on a 4-point Likert-scale, ranging from 1 (agree strongly) to 4 (disagree strongly). Internal consistency of the Negative Urgency subscale was good in previous studies, with α = .77 to .89 (Carlson et al., 2013; Cyders, 2013; Sanchez-Roige et al., 2019). In the current study, internal consistency was also good, with McDonald’s ω = .81.

History of NSSI

History of NSSI was measured with the Self-Injury Questionnaire (SIQ; Claes & Vandereycken, 2007). This scale contains questions about five types of non-suicidal self-injurious behavior (scratching, cutting, burning, bruising, and biting) and gave participants the opportunity to add a sixth. Of each behavior, the incidence was asked on a 5-point scale, ranging from 1 (last week) to 5 (never). Following the instructions of the SIQ (Claes & Vandereycken, 2007), only when participants reported the occurrence of a behavior within the last month, more questions about the behavior were asked. These follow-up questions included what body parts were injured, how many days and how many times a day during the last month participants self-injured, how often they felt pain during the act, to what degree they felt pain, whether it was planned beforehand or not, whether participants hid the act from others, whether they took care of the wounds, emotional state before and after the act, and reasons why participants engaged in the act. In the present study, participants were dichotomously identified as having a history of NSSI or not (0–1). Participants with at least one type of NSSI, performed once or more in their life, were identified as having a history of NSSI. In the freely described sixth NSSI behavior, hair pulling (mentioned three times), soft head beating (mentioned once) and falling on purpose (mentioned once) were classified as self-injurious behavior by the researchers. Pinching (mentioned three times), cracking knuckles (mentioned once), thoughts of self-hatred (mentioned once), hit something really hard (mentioned once), breaking one’s arm (mentioned once), unconscious scratching of dry skin (mentioned once), swallowing pills (mentioned once), smoking (mentioned once), and watching pornography were not considered as self-injurious behavior.

Procedure

After entering the lab, the procedure and exclusion criteria were explained, and participants were asked to sign a consent form. Participants were randomly assigned to the boredom, neutral, or anger condition.

Then, participants rated the positive and aversive sound, after which they were asked to fill out the questionnaire measuring their current emotional state. Next, they spent 15 min completing a writing assignment serving as the emotion induction, while they had the opportunity to listen to either or both of the two sound clips, however many times they wanted. Participants were not informed about reasons why they could select the sound clips, and were only told about the option to do so during the writing task. To induce emotional state, in the boredom condition, participants were asked to type the word “Abramson” over and over again. In the neutral condition, participants were instructed to describe in detail how they travel from home to university. In the anger condition, participants were asked to describe in detail an autobiographical memory in which they felt very angry. Participants were instructed to write at a moderate pace, and to keep writing until they were given the signal that the task was over. Similar writing tasks have been used before, and were shown to be effective emotion inductions, as participants found boredom inducing writing tasks significantly more boring than an interest inducing task where participants wrote stories in response to ambiguous pictures of people (Abramson & Stinson, 1977), and autobiographical anger inducing writing tasks led to increased anger (Jallais & Gilet, 2010). After the writing task, they rated their emotional state again. To ensure that the emotion induction had no lingering effect on subsequent trait measures and emotions, participants watched a short cheerful filler video of a flash mob (De Boeck, 2009). Next, participants rated their emotional state again and filled out the UPPS-P and SIQ. At the end of the study, participants were fully debriefed and could choose course credits or a 5-euro gift certificate as reward for participation.

Statistical analyses

As a manipulation check, a MANOVA was used, with the emotional state changes as outcomes (post- minus pre-induction emotional state scores on the nine VAS scales), and condition as between-subjects variable. For significant results, post hoc tests were performed, Bonferroni corrected. Changes in all measured emotional states were analyzed, to test manipulation effectiveness as well as specificity: i.e., testing all emotional states allows us to examine whether our manipulations targeted the specific emotional states they were designed to target.

To examine the differences in stimulus frequency due to boredom compared to anger and neutral states, an ANOVA was used, with condition as independent variable and stimulus frequency as dependent variable, followed up by post-hoc tests, Bonferroni corrected.

To test whether NSSI history and negative urgency positively moderate the effect of boredom on self-administered aversive stimuli, linear regression analyses were performed using the SPSS macro PROCESS (Hayes, 2017). Condition (boredom, neutral, anger) was entered as the multicategorical predictor (x) variable through indicator (i.e., dummy) coding, and frequency of aversive sounds as outcome (y) variable. History of NSSI and negative urgency were entered as moderators (w and z) in a single model (model number 2 in the PROCESS dialog box; Hayes, 2017). See Fig. 1 for a conceptual overview of the tested model. Scores for negative urgency were mean centered to reduce multicollinearity. Indicator coding was used for the multicategorical × variable (condition). Condition labels were as follows: 1 = boredom, 2 = neutral, 3 = anger, making the boredom condition the reference level. Then, analyses were run again, with the labels re-arranged to 1 = neutral, 2 = boredom, 3 = anger, to allow for the comparison between the neutral and anger condition.

Fig. 1
figure 1

Conceptual diagram of the tested model

Full size image

Results

Sample characteristics

The majority of the sample was female, and studied psychology. The three experimental groups did not differ in any of the demographic variables (see Table 1). Mean stimulus frequency for the total sample was 4.53 (SD = 7.88) for the aversive sound, and 12.23 (SD = 14.36) for the positive sound. Table 1 also shows the mean levels of negative urgency and numbers of participants with and without a history of NSSI across experimental groups: no significant differences were found between groups for these trait variables. Out of the total sample, 61.2% of participants reported having a history of NSSI, which is higher than other studies using college samples, ranging from 7.3 to 17.7% (Kiekens et al., 2021; Taliaferro & Muehlenkamp, 2015).

Table 1 Demographic information of study sample
Full size table

Manipulation check

Overall, the MANOVA showed a significant effect of condition on emotional state (F(18,238) = 8.95; p < .001, η2partial = .41). Significant differences were found for all 9 emotional states (see also Table 2 and Fig. 2). Specifically, boredom change was significantly higher in the boredom condition compared to both the neutral and anger condition. Anger change was significantly higher for the anger condition compared to the neutral condition, although no significant difference was found when compared to the boredom condition. Additionally, sad mood change was higher for the anger condition compared to both the neutral and boredom condition. Lastly, the current boredom and anger induction both led to increased negativity and frustration, and to decreased happy, relaxed, and positive emotions, compared to the neutral condition.

Table 2 Differences in emotional states and writing tasks in the neutral, anger and boredom conditions
Full size table
Fig. 2
figure 2

Changes in emotional states in the neutral, anger and boredom conditions. *p < .05; **p < .01; ***p < .001

Full size image

Additionally, significant differences were found between the boredom and anger condition, and between the boredom and neutral condition, in character length written during the writing task.

Lastly, emotional state score changes due to the filler video can be found in Table A1. State boredom post-video was significantly higher in the boredom condition compared to the neutral and anger condition, and state frustration was significantly higher in the boredom condition compared to the neutral condition.

Main analyses

Table 1 displays an overview of mean scores of stimulus frequency per condition. Frequency of positive sounds was higher than aversive sounds in the overall sample (t(128) = -8.06, p < .001) as well as in each separate condition (t(42) = -7.09, p < .001; t(40) = -5.31, p < .001; t(44) = -2.58, p = .013, respectively for the boredom, neutral and anger condition).

An ANOVA was used to test whether boredom leads to more self-administration of aversive sounds than the other emotional state inductions even in the presence of a positive behavioral alternative, and showed a significant effect of condition, F(2,126) = 5.41, p = .006, η2partial = .08. Post-hoc tests, Bonferroni corrected, showed that participants self-administered more aversive sounds in the boredom condition compared to both the neutral (M difference = 5.22, SE = 1.66, p = .01) and anger condition (M difference = 3.94, SE = 1.63, p = .05). No significant difference was found in frequency of aversive sounds between the neutral and anger condition (M difference = -1.28, SE = 1.65, p = 1). Additionally, an ANOVA was used to compare self-administered positive stimuli, and also showed a significant effect of condition, F(2,126) = 21.08, p < .001, η2partial = .25. Post-hoc tests, Bonferroni corrected, showed that participants self-administered more positive sounds in the boredom condition compared to both the neutral (M difference = 13.38, SE = 2.74, p < .001) and anger condition (M difference = 16.41, SE = 2.67, p < .001). No significant difference was found in frequency of positive sounds between the neutral and anger condition (M difference = 3.03, SE = 2.71, p = .795).

Moderation regression analyses were used to test whether the effect of boredom on self-administered aversive stimuli was positively moderated by NSSI history and negative urgency, (see Table 3 and Fig. 3 and 4 in Appendix B). No significant moderation effects of NSSI history or negative urgency were found for the effect of boredom on self-administered aversive stimuli, with B-values ranging from -6.72 to 4.29, t-values ranging from -1.90 to 1.10, and p-values ranging from .06 to .60. The moderation analysis was run again, with re-arranged dummy coding to ensure that the anger condition was also compared to the neutral condition (as this is not done automatically when a multicategorical variable is included in PROCESS). Again, no significant moderators were found. The dataset generated in the current study is accessible via https://osf.io/.

Table 3 Results of moderation analyses with self-administration of aversive sound as outcome variable, using the PROCESS SPSS macro
Full size table

Discussion

The current study aimed to replicate and extend prior findings on the effect of boredom on self-administered aversive stimuli as a proxy for NSSI behavior, by examining whether this effect persists when a positive behavioral alternative is present, and whether the effect of boredom on this behavior is greater than the effect of anger. Additionally, the moderating effects of negative urgency and history of NSSI on the link between boredom and self-administered aversive stimuli were examined. Boredom and anger were induced through a writing task, during which frequency of self-administered aversive versus positive sounds was measured.

The boredom manipulation through the writing task was successful: state boredom increased more in the boredom condition compared to the neutral and anger condition. This finding is in line with earlier research (Abramson & Stinson, 1977) showing that the monotony of repeatedly writing the same word indeed is an effective boredom induction method. The anger manipulation was also successful, as state anger increased more in the anger condition compared to the neutral condition, as did sadness and arousal. In our previous study (Yusoufzai et al., 2022), we also found that the video anger induction increased both anger and sadness. These findings are in line with an earlier study (Jallais & Gilet, 2010) using a similar autobiographical anger induction task, where this induction increased both anger and sadness. Such a mixed emotional state seems to be typical for anger, as a large daily self-report study (Trampe et al., 2015) demonstrated that anger most frequently co-occurs with other negative emotions, including sadness. Interestingly, anger increase did not differ significantly between the boredom and anger condition. While on the one hand boredom and anger are conceptually different emotional states (Van Tilburg & Igou, 2017a, 2017b), on the other hand the literature demonstrates a positive association between anger and boredom at a trait level (Isacescu et al., 2017; Rupp & Vodanovich, 1997). Further research is needed to clarify the possible overlap between these two emotional states, and their effect on self-administered aversive stimuli. Other boredom induction methods may be considered in examining this overlap, for example the writing task used by Van Tilburg & Igou (2012), which demonstrated a significant increase in boredom but not anger for a high-boredom as compared to a low-boredom writing task. Additionally, character length written during the writing task differed significantly between the boredom and anger condition, and between the boredom and neutral condition. Specifically, participants typed more characters during the boredom condition compared to the anger and neutral condition. This could be explained through the contents of the writing task, as in the boredom condition, participants were simply required to write the same word over and over again, whereas in the other two conditions they were instructed to write about topics that presumably required more time and cognitive resources to type up.

Our data demonstrated that boredom led to increased selection of aversive sounds, compared to both the neutral and anger condition, even though a positive behavioral alternative was present. The current findings support prior studies demonstrating that boredom leads to self-administration of electric shocks (Havermans et al., 2015; Nederkoorn et al., 2016; Yusoufzai et al., 2022), and establish that this effect extends to aversive experiences other than painful electric shocks. More importantly, the current study is the first to demonstrate that boredom can lead one to exert negative behaviors towards oneself even when a positive alternative is available. Interestingly, anger did not have this effect: no differences were found between the neutral and anger condition in self-administered aversive sounds. Considering that self-administered aversive sounds was used as a proxy for NSSI in the current study, the lack of an anger effect is not in line with self-report studies demonstrating that (other-directed and self-directed) anger often precedes NSSI, in both clinical (Briere & Gil, 1998; Claes et al., 2010) and nonclinical samples (Laye-Gindhu & Schonert-Reichl, 2005; Muehlenkamp et al., 2013; Nock et al., 2009) This lack of significant findings may be attributed to the difference between the current lab paradigm using sound and actual NSSI as it is experienced in daily life. Altogether, the current findings demonstrated that in the context of a college sample, there was no evidence for a causal link between anger and self-administered aversive sound stimuli.

History of NSSI did not moderate the link between boredom and self-administered aversive stimuli. This might be due to the fact that there was no physical pain involved in the current proxy for NSSI behavior. Instead, we used aversive sounds here, while typical NSSI in daily life involves physical pain. Such physical pain involved in NSSI has been shown to serve the function of emotion-improvement through e.g., gratification of self-punishment desires, distraction from negative thoughts, and offset relief (Hooley & Franklin, 2018). The current study could thus imply that this learned function of earlier NSSI experiences involving physical pain (Hooley & Franklin, 2018) did not carry-over to the current proxy for NSSI, i.e., self-administering aversive sounds. Whereas people without a history of NSSI might experience a barrier to self-inflicting pain, this barrier could be weaker or absent in the case of self-administering non-painful aversive stimuli, blurring the distinction between people with and without a history of NSSI in the current context. Negative urgency also did not moderate the boredom-NSSI behavior link. This might be explained by our choice of induction, as negative urgency refers to the tendency to impulsively act in response to negative emotions (Lynam et al., 2006). However, our participants were already typing, which means they were already performing an activity. Thus, negative urgency might only be a relevant characteristic in the boredom-NSSI relation when the individual is not already engaged in some form of action, no matter how boring it may be. Alternatively, negative urgency may simply not be a relevant moderator in the effect of boredom on NSSI, regardless of the boredom setting.

Importantly, positive sound frequency was higher than aversive sound frequency across conditions overall, as well as within each condition separately. Additionally, the effect size of the pleasant stimulus frequency across the experimental conditions was higher than the effect size of the aversive stimulus frequency. A possible explanation for these findings is that boredom promotes the need for novel experiences that increase arousal, to escape the monotonous boredom experience. In other words, boredom leads to more activation in general, regardless of whether the behavior is positive or negative. This is in line with studies demonstrating that boredom leads to more self-administered electric shocks (Havermans et al., 2015; Wilson et al., 2014), but also to increased prosocial intentions (Van Tilburg et al., 2017a) and eating behavior (Havermans et al., 2015). Additionally, the current findings suggest that the negative effect of boredom is not confined to situations with only negative behavioral options. This implies that boredom may lead to NSSI behavior in settings beyond penitentiary and clinical institutions, where lack of positive behavioral options were previously thought to explain the increased NSSI (Casiano et al., 2013; Chapman & Dixon-Gordon, 2007; Kaba et al., 2014). Nonetheless, it remains to be elucidated whether the current boredom effect persists when a broad range of alternative options are present, in addition to other contextual and individual factors, as is the case in daily life.

The current study has a number of limitations. Firstly, the current use of self-administered aversive sounds requires caution when extrapolating our findings to NSSI behavior in real life, considering that the sound clips used in the current study were merely aversive and not painful, and there was no “deliberate destruction of one’s own body tissue in the absence of intent to die” (Nock et al., 2009, p. 78). Ethical considerations evidently pose a challenge in experimental research in NSSI, though the current design provides a tentative starting point for future endeavors. Secondly, emotional state was measured before and after, but not during, the emotion induction and NSSI behavior. This limits our insight in the functions and emotional consequences of the self-administered stimuli, e.g., whether listening to aversive sounds decreased negative and increased positive emotional state. Additionally, pleasantness of the sound stimuli was measured only once, at the beginning of the experimental procedure, thus we cannot determine whether habituation took place for either the aversive or positive stimuli, nor whether reversed effects took place by way of the pleasant sounds possibly becoming unpleasant after repeated exposure. Furthermore, emotional state between the experimental conditions differed after the filler video before the last questionnaires: specifically, state boredom post-video was significantly higher in the boredom condition compared to the neutral and anger condition, and state frustration was significantly higher in the boredom condition compared to the neutral condition. This may have affected responses in the subsequent trait questionnaire, thus caution is needed when interpreting the results. Lastly, the current sample’s generalizability to the general population is limited, as the majority of participants were female and university students. Additionally, 61.24% of participants reported having a history of NSSI, which is much higher than other studies using college samples (e.g., 7.3–17.7%; Kiekens et al., 2021; Taliaferro & Muehlenkamp, 2015). The higher proportion of participants with an NSSI history in the current sample may be attributed to our use of a longer questionnaire than these other studies, as NSSI history prevalence was shown to be higher when measured with questionnaires rather than shorter (e.g., single-item) measures (Muehlenkamp et al., 2012). Earlier studies in our lab using the same NSSI measure in similar samples (Nederkoorn et al., 2016; Yusoufzai et al., 2022) also found higher proportions of these samples reporting an NSSI history (39.13% and 54.79%, respectively) that was more comparable to the NSSI history level we found in the current study. The current study should be replicated in samples with lower NSSI history rates, as found by Kiekens et al. (2021) and Taliaferro and Muehlenkamp (2015), to allow for generalizations to the general population.

The current study supports the causal role of boredom on NSSI behavior, and suggests that while boredom seems to lead to more activation in general, the effect of boredom on self-administered aversive stimuli persists even when a positive alternative is available. The latter findings support and extend prior studies (Havermans et al., 2015; Nederkoorn et al., 2016; Wilson et al., 2014; Yusoufzai et al., 2022), and, importantly, further lend plausibility to the hypothesized boredom-NSSI link. If boredom is a relevant causal factor in NSSI behavior in both clinical and non-clinical populations, it can provide an effective target in prevention and intervention methods through e.g., environmental alterations to lower boredom or training of individual boredom coping strategies. Additionally, perhaps boredom also leads to self-defeating behaviors other than NSSI. Further experimental research could help determine the boundaries of this effect, and may identify less conspicuous but equally maladaptive behaviors that were not considered thus far in the negative consequences of boredom. Considering that social media is often used to alleviate boredom (Stockdale & Coyne, 2020), a possible maladaptive behavior to examine would be “doomscrolling”, which refers to excessively reading large quantities of negative online news.

In conclusion, our findings suggest that people are so motivated to escape boredom that they will resort to choosing aversive experiences, even when a positive alternative option is available. Additionally, this effect was restricted to boredom and not found for anger in the current study, despite both emotional states being identified as precursors to NSSI in earlier self-report studies (e.g., Chapman & Dixon-Gordon, 2007; Claes et al., 2010). Considering that boredom is a universal emotion, the pervasiveness of its negative consequences requires scrutiny, and provides a promising avenue for future studies.