The Role of Information Processing as an Underlying Mechanism in the Experience of Anxiety Reactivity and Anxiety Perseveration, Two Dissociable Dimensions of Trait Anxiety

Anxiety is the most prevalent mental health concern worldwide, with approximately 301 million people experiencing an anxiety related condition in 2019 alone (World Health Organisation, 2023). Not all experiences of anxiety are maladaptive, often serving the adaptive purpose of prompting individuals to avoid potential environmental threats (Gutiérrez-García & Contreras, 2013). However, when the intensity and frequency of anxiety symptoms is out of proportion to the current situation, this can cause significant emotional and cognitive disturbances, limiting the quality of life for the affected individual (Gutiérrez-García & Contreras, 2013). Due to the significant individual variability in the experience of anxiety, research has focused on identifying the mechanisms underpinning the frequency with which people experience anxiety symptoms, with the goal of limiting the impairment associated with this prevalent mental health concern (Ouimet at el., 2009). A mechanism of particular interest has been the manner in which people process information, with people often choosing to process information in a way that fits with their worldview and perceptions of self (Ouimet et al., 2009). Prominent cognitive theories of anxiety describe how people with anxiety often seek out and select information which matches their perceptions of the world as threatening or dangerous (Beck & Clark, 1997). Inherent in this process, is a matched perception that they do not have adequate resources to cope with this threat, prompting actions such as impaired disengagement with the threat or avoidance (Ouimet et al., 2009). As a result, individuals with anxiety-related dysfunction or anxiety disorders, experience hypervigilance for cues that indicate a potential threat and are at an increased likelihood of misinterpreting benign cues as dangerous, prompting unnecessary evasive actions, limiting their engagement with the outside world (Beck & Clark, 1997).

The frequency with which people experience anxiety symptoms is termed trait anxiety and is a well-established predictor of anxiety-related dysfunction (Endler & Kocovski, 2001). Anxiety is distinct from fear, in that anxiety is a more diffuse, future-oriented emotional state focusing on potential or anticipated threats (Grogans et al., 2023), whereas fear is tied to an immediate stimulus and therefore present-focused (Beck & Emery, 2005). In addition, anxiety involves more elaborate cognitive elaboration and anticipatory thinking (Grogans et al., 2023). Trait anxiety has been identified as a factor contributing to the aetiology of anxiety-related disorders (Knowles & Olatunji, 2020), making research into the construct itself, and the contributing mechanisms, an important element in extending our understanding of anxiety.

Until recently, trait anxiety has commonly been considered a unitary dimension that is frequently assessed using questionnaires such as the trait version of the Spielberger State-Trait Anxiety Inventory (STAI-T; Spielberger et al., 1983), which requires individuals to report on the frequency with which they generally experience specific anxiety symptoms (Julian, 2011) (see method section and Knowles & Olatunji (2020) for review of STAI-T discriminant validity concerns with depressive symptomatology). Theoretically, however, a number of different types of anxious disposition could underlie variation in trait anxiety scores. Research by Rudaizky and colleagues (2012, 2014) has revealed that two dimensions of anxious disposition each account for unique variance in trait anxiety scores (Rudaizky et al., 2012). One dimension, referred to as anxiety reactivity, reflects individual differences in the magnitude of state anxiety responses elicited by a stressor. For example, in response to an impending examination, most people would experience some elevation in state anxiety. Individuals who experience greater reactivity would exhibit more frequent and intense state anxiety reactions. The other dimension of trait anxiety, referred to as anxiety perseveration, reflects individual differences in the persistence of a state anxiety elevation, once induced by a stressor (Rudaizky et al., 2012). For example, following an examination, most people would experience anxiety reducing emotions such as relief. Individuals who experience greater perseveration would remain anxious for longer after leaving the examination room.

Over the last decade, the suggestion has been made within the literature that different cognitive mechanisms may independently drive variation in anxiety reactivity and in anxiety perseveration (Gole et al., 2012). However, to the authors’ knowledge, the present study is the first to systematically assess candidate cognitive mechanisms, specifically information processing, underlying these two dimensions of anxiety vulnerability. Cognitive theorists have suggested that individual differences in anxiety reactivity and anxiety perseveration reflect the differing impact of alternative types of information which people process (Rudaizky et al., 2014). Building on this foundation, the present study’s hypotheses were derived from literature focusing on similar emotional processes with well-established cognitive underpinnings (Kross et al., 2005; Watkins et al., 2008). Relevant to understanding individual differences in anxiety reactivity is literature focusing on emotional regulation (Davis et al., 2011). Research by Kross and colleagues (2005, 2014) posits that thinking about a situation in a manner that involves focusing on situational details, leads to heightened emotional arousal. More specifically, by using a self-immersed perspective where one prioritises processing and visualising the emotionally arousing and specific elements of a situation (e.g., the feelings elicited and challenging aspects), individuals run the risk of experiencing enhanced physiological stress and greater emotional reactivity (Kross & Ayduk, 2017). This model suggests that individuals who are prone to taking a self-immersed perspective, focusing on tangible and emotional aspects of a situation (i.e., situational information), are more likely to experience increased emotional reactivity (Kross & Ayduk, 2017), a process not dissimilar to anxiety reactivity. There is empirical evidence to support this account. For example, a study by Bornstein and colleagues (2020) identified that when participants tended to process situational details concerning a stressful event (e.g., writing how and in what order the event happened), they experienced greater elevation of negative emotions than did those who tended to process more self-distanced and abstract details. Though Kross and colleagues did not examine anxiety reactivity specifically, their findings suggest that elevated anxiety reactivity could be driven by the processing of situational information concerning the stressor. Therefore, the first hypothesis under test in the current study is that processing situational information about a stressful event leads to greater anxiety reactivity (but not to greater perseveration) in response to this event.

Relevant to understanding anxiety perseveration is literature focusing on repetitive thoughts, such as depressive rumination (a process akin to anxiety perseveration). Within this literature base, there is evidence to suggest that preferential processing of information, which is decontextualised and abstract in nature, is one of the crucial processes which differentiates dysfunctional from functional repetitive thoughts (Watkins & Teasdale, 2004). More specifically, Watkins’ (2008) Processing Mode Theory posits that utilising an implicational processing mode inhibits adaptive emotional processing of negative/stressful events (White & Wild, 2016). By disrupting the formation of a cohesive and specific narrative concerning a situation or event, negative interpretations are strengthened, giving rise to repetitive negative, or intrusive thoughts (White & Wild, 2016). Thus, Watkins and colleague propose that processing implication information which focuses on the reasons for, and the consequences of an event, increases repetitive negative thinking and so elevates negative affect (Watkins et al., 2008).

There is evidence to support Watkin’s theory, with increased processing of implicational information processing found to elevate negative affect (Kambara et al., 2019). Most relevant to the present research, one study identified that individuals who were presented with statements that contained abstract thoughts about the reasons for and the consequences of a past event experienced longer duration of negative mood and arousal following a stressful event (Ehring et al., 2009). Another study by Altan-Atalay and colleagues (2022) exposed participants to a stressful situation and employed a manipulation that required participants to either process implication information (by answering questions such as “why is this happening to me?”), or non-implicational information (by answering questions such as “how can I solve the problem at hand?”). They found that participants who were induced to process implicational information experienced higher levels of subsequent state rumination (Altan-Atalay et al., 2022). While these studies provide support for Watkin’s theory, their methodology does not distinguish between the impact that processing implicational information has on anxiety perseveration. Therefore, the current literature base is unable to determine whether the processing of implicational information contributes to elevated anxiety perseveration specifically. However, given Watkins’ theory and this preliminary empirical support, it is plausible to suggest that elevated anxiety perseveration could be increased by the processing of implicational information. The second hypothesis tested in the current study is therefore that processing information which is implicational in nature leads to greater anxiety perseveration (but not anxiety reactivity) in the wake of a stressful event.

Distinguishing the cognitive mechanisms that differentially contribute to variation in anxiety reactivity and anxiety perseveration, respectively, has important theoretical implications for advancing the understanding of individual differences in trait anxiety. Cognitive models of anxiety, while recognising that focusing on negative content elevates trait anxiety, do not differentiate between the distinctive ways in which information can be processed (i.e. the nature of this negative focus) that may differentially give rise to elevated anxiety reactivity and anxiety perseveration (Rudaizky et al., 2012). By exploring the influence of such information processing styles, it becomes possible for us to better understand how varied information and/or content can influence individuals’ responses to stressful events, in both the short and longer term. This opens the possibility of examining how manipulating the information people process prior to a stressful event, impacts the level of anxiety reactivity and anxiety perseveration they experience both throughout and after the event (Rudaizky et al., 2012).

The current study sought to test these two hypotheses by examining whether participants who process either situational or implication information about an upcoming stressor respectively exhibit heighted anxiety reactivity and anxiety perseveration in response to this event. Participants were told that would be exposed to a lab-based stressor, which took the form of a simulated interview. At several junctures across the experimental session, both as they approached this stressor and also after they had completed it, participants’ state anxiety was assessed. This made it possible to compute separate measures of their anxiety reactivity (i.e., reflecting state anxiety elevation elicited by the approaching interview) and anxiety perseveration (i.e., reflecting persistence of anxiety following the interview). To examine the impact of processing either situational or implicational information on each dimension of anxiety response, every participant was required to process either one or the other type of information concerning the stressor. This information was presented in the form of video clips, in which supposed former participants shared either situational information about the interview (i.e. what they experienced in the moment) or implicational information about the interview (i.e. what the consequences of the interview were for them).

While information type was the main experimental manipulation, consideration was given to the valence of the information, given the wealth of research that implicates negative information in the experience of anxiety (Mogg & Bradley, 2016) and rumination (Watkins & Roberts, 2020). For example, there is extensive evidence which implicates biased attentional processing of negative information in the experience of trait anxiety (Bar-Haim et al., 2007). Negative information, in this context, is threat-related and therefore focuses on threatening aspects of the Interview Challenge. Conversely, positive information is threat-incongruent, focusing on neutral or pleasant aspects of the Interview Challenge. This gave rise to an exploratory manipulation whereby each information type category was broken into two separate conditions: positive and negative. Given the well-established link between negatively valenced information and elevated anxiety-linked responses (Ouimet et al., 2009; Mogg & Bradley, 2016), it is plausible to suggest that only the processing of negative situational and implicational information would give rise to anxiety reactivity and anxiety perseveration, respectively. To ensure that any effects observed could be attributed to the processing of negative situational or implicational information (and not to the processing of situational or implicational information more generally) this between-subjects manipulation led to participants being presented either with exclusively negative or exclusive positive (implicational or situational) information about the upcoming interview.

The hypotheses predict that participants who are required to process negative information concerning the upcoming Interview Challenge Stressor (see experimental tasks p. 14) that is situational in nature, as compared to those who are required to process negative information that is implicational in nature or that is positive, will experience greater anxiety reactivity (but not perseveration) during the period when they are introduced to the Interview Challenge. Additionally, it is predicted that participants who are required to process negative information concerning the upcoming Interview Challenge that is implicational in nature, as compared to those who are required to process negative information that is situational in nature or that is positive, will experience greater anxiety perseveration (but not reactivity) in the period following participation in the Interview Challenge.

The present study served as an initial investigation into the cognitive mechanisms underlying anxiety reactivity and anxiety perseveration, two dissociable dimensions of trait anxiety. Drawing from literature focusing on rumination (Watkins et al., 2008) and emotional regulation (Kross et al., 2005, 2014), we hypothesised that processing of negative information which is situational in nature leads to greater anxiety reactivity (but not perseveration) in the lead up to a stressful event, and that processing negative information which is implicational in nature leads to greater anxiety perseveration (but not anxiety reactivity) in the wake of a stressful event.

Results revealed that processing the negative situational information contained in the video clips did not significantly influence participants’ experience of state anxiety during the anxiety reactivity assessment period. Thus, these findings do not support the first hypothesis under test. This result is inconsistent with the support for this theoretical account found in the emotional regulation literature, where it was identified enhanced processing of situational details concerning a stressful event led to greater intensity of negative emotions (Bornstein et al., 2020). This discrepancy could be explained by methodological variations between the present study and Bornstein and colleagues (2020) study. For one, the present study utilised pre-recorded videos to induce a processing mode, which constrained participants to viewing information that might not have been in line with how they would naturally process a stressful situation. In contrast, Bornstein and colleagues (2020) asked participants to draw from their own past experiences, to provide a retelling of a stressful interpersonal experience, which was then coded as either situational or implicational in construal (Bornstein et al., 2020). This difference may have impacted the level of psychological distance perceived by the participants, i.e., listening to others’ experiences, rather than drawing from their own, may have attenuated the intensity of the emotional experience (Kross, 2009). There is evidence to support this, with Verduyn and colleagues (2012) finding that the duration of negative emotions was decreased when individuals processed an event from a self-distanced perspective (seeing themselves as an observer), compared to individuals who processed the event from a self-immersed perspective (first-person).

The present results additionally revealed that the processing of negative implicational information, presented in the video clips, did not significantly influence participants’ experience of state anxiety during the anxiety perseveration assessment period. Therefore, the second hypothesis under test was not supported. This result contrasts with those found in the depressive rumination literature, which highlight the contribution of the implicational processing mode to depressed mood and ruminative cognitions (Kambara et al., 2019; Watkins & Teasdale, 2004). These prior works similarly utilised convenience samples of unselected undergraduate students (i.e., students were not selected for high trait anxiety), so it is not unreasonable to assume that enhanced processing of implicational information could influence the experience of anxiety perseveration in a subclinical sample. Therefore, authors examined potential methodological explanations that may have influenced the present study’s ability to detect significant effects. One candidate explanation for our non-significant effect could be that individuals did not experience substantial anxiety perseveration in the wake of the Interview Challenge stressor. This explanation is given weight by the finding that state anxiety significantly reduced across the anxiety perseveration assessment phase, irrespective of valence and information type contained in the videos. Perhaps this significant decrease in state anxiety during the perseveration assessment phase is due to a mismatch between expectancies formed during the reactivity assessment phase and the experience of the stressor itself. It is plausible that while participants’ state anxiety was elevated by both the introduction to the Interview Challenge stressor, the stressor itself did not match their expectations in terms of difficulty or arousal. This could have elicited feelings of surprise and relief (Gaschler et al., 2014), and a subsequent drop in state anxiety. Relief has been described a “tension reducing emotion” (Pekrun et al., 2007 p. 25), which is linked to a reduction in subjective negative states, such as state anxiety. Research examining expectation mismatch and prediction error lend weight to this suggestion, as it has been identified that a mismatch between predictions and/or expectations of outcomes, and the actual outcome itself can influence performance, and the experience of anxiety (White et al., 2017; Gaschler et al., 2014).

A condition requiring processing of positive situational information, and a condition requiring processing of positive implicational information, were also included to disentangle the impact of processing situational vs. implicational information from the impact of processing negative vs. positive information. The valence of the video clips emerged as a factor which influenced participants’ anxiety reactivity. More specifically, while participants who processed positive information (situational or implicational) experienced a reduction in their level of state anxiety following their viewing of the video clips, individuals who were exposed to negative information remained elevated in their state anxiety ratings after viewing the video clips. This finding is consistent with a significant body of research that implicates the processing of negative information in the experience of heightened state anxiety and anxiety related dysfunction (Bar-Haim et al., 2007; Mathews & MacLeod, 2005). Of note, valence of processed information influenced only anxiety reactivity, without having any impact on anxiety perseveration. Prior research into anxiety reactivity and perseveration has not considered the influence of valence on the experience of these independent dimensions (Rudaizky et al., 2012, 2014), making this a novel finding. This finding will be important to consider when aiming to target elevated anxiety, as exposure to positive information may only attenuate anxiety reactivity, and not anxiety perseveration.

As an additional endeavour, this research sought to conduct analyses akin to those presented in Rudaizky and colleagues’ (2014) study. These analyses aimed to explore the ability of in vivo measures of anxiety reactivity and anxiety perseveration to explain independent variance in trait anxiety scores. The study found that only anxiety perseveration accounted for independent variance in trait anxiety scores, contrasting with Rudaizky et al., (2014), but aligning with Mann (2020), who also identified anxiety perseveration as the sole independent predictor. Differences in experimental design may explain the discrepancy; Rudaizky et al.‘s (2014) study used a stressor where participants anticipated a speech without delivering it, while the current study and Mann’s methodology involved a simulated interview or speech with preparation time. Furthermore, Rudaizky and colleagues (2014) allowed more time for participants to experience anxiety reactivity, by assessing state anxiety across a five-minute window. In contrast, the present study assessed participants’ state anxiety immediately after exposing them to the video clips, providing less opportunity for anxiety reactivity to develop.

Although the present design did elicit elevations in state anxiety during the anxiety reactivity and anxiety perseveration assessment phases, future studies could modify the study protocol, with the goal of increasing the degree to which participants experience anxiety reactivity and anxiety perseveration. For example, such refinements may involve amending the questions involved in the interview, to reflect topics of relevance to undergraduate samples (e.g., housing, climate change, and education). Increasing the personal relevance of experimental stimuli to undergraduate participants could be a pathway to increased engagement with the stressor (Grafton et al., 2012), and greater negative affectivity, characteristic of anxiety perseveration (Grafton et al., 2012). Additionally, it could be the case that the lack of anxiety reactivity effects in the present study were at least in part driven by a ceiling effect, occurring as a result of the stressor introduction. It is plausible that participants anticipation of the Interview Challenge stressor elevated anxiety to a point that eclipsed any further elevation that could have been elicited by the information contained in the videos. To avoid this, future experimental designs should endeavour to assess anxiety in a manner that acknowledges the contribution of the stressor introduction and the subsequent exposure to information concerning the interview, to provide a holistic measure of reactivity. By including both in the measure of anxiety reactivity, any ceiling effect will be accounted for.

Despite its strengths, the present study has several limitations of note. First, the current study did not recruit participants based on their level of trait anxiety, instead electing to recruit a convenience sample of undergraduates. It could be the case that impact of processing situational information and implicational information, on anxiety reactivity and anxiety perseveration respectively, may be more evident in people with high levels of trait anxiety. The design of the present study precluded the detection of effects that may only exist for participants with high levels of trait anxiety. Future research could overcome this limitation by recruiting participants using a measure of trait anxiety, such as the STAI-T (Spielberger, 1983), to enable comparing the impact of processing situational information and implicational information on anxiety reactivity and anxiety perseveration experienced by participants high in trait anxiety, compared to participants low in trait anxiety.

An additional limitation is that the present study was not able to determine the natural processing style of participants, which may vary across individuals in terms of whether this involves a preference for, or tendency to selectively process, either situational information or implicational information. Instead, individuals in the current study were presented with pre-selected information, which was either situational or implicational in nature, and so the information they were required to process may not have aligned with their own information processing preferences. In line with this proposal, works by Watkins and colleagues (2008, 2020) have identified that individuals who experience greater depressive rumination have an inherent tendency to process information relating to the causes, consequences, and implications of negative mood and when provided with incongruent, often concrete information, they experience reduced duration of negative mood and ruminative cognitions (Watkins & Roberts, 2020). For the present investigation, it could be the case that some individuals who have a tendency to preferentially process situational information were given implicational information, which may have led to diminished anxiety reactivity and/or perseveration (and vice versa) with the implicational processing style. Future extensions of this work could employ methodology that permits identification of participants’ information processing biases (e.g., by asking participants to report on the contents of their thoughts). This could help uncover individual differences in information processing biases which may drive preferential processing of either situational or implicational information, potentially resulting in differing levels of anxiety reactivity and anxiety perseveration. Hence, we suggest that this is an important future direction for research in this field.

Finally, mention must be made to sample characteristics, most notably, the current sample consisting of mostly young adults from a western, educated, industrialised, rich, and democratic (WEIRD) society (Wild et al., 2022). Credible concerns have been raised about the external validity of research conducted using undergraduate samples, due to the fact these individuals are rarely representative of the general population (Wild et al., 2022). Such validity concerns relate to the narrow age range, sophisticated cognitive abilities and/or skills, and higher socioeconomic backgrounds of these samples, compared to the general population (Wild et al., 2022). With these consideration in mind, the authors recommend exercising caution when considering generalising the current findings to populations outside of university undergraduate samples.

To summarise, in the present study the processing of negative situational and negative implicational information about an upcoming stressor did not serve to differentially influence the degree to which participants exhibited anxiety reactivity and anxiety perseveration in response to this stressor. Nonetheless, these findings build on previous work by Watkins (2008) and Kross (2005, 2014), as this is the first study to examine the influence of exposure to information which is either abstract and decontextualised in nature (implicational) or distinct and situation specific in nature (situational), on the two distinct facets of trait anxiety. From this foundational study, we hope further research will be conducted to illuminate the information processing biases implicated in the experience of anxiety reactivity and perseveration, furthering our theoretical, and applied understanding of these dimensions of trait anxiety. Once these have been identified, it opens the possibility for not only determining whether this relationship is causal, but for developing paradigms capable of targeting information processing biases (e.g., attention bias) with the aim of reducing individuals’ experience of anxiety reactivity and/or anxiety perseveration. One such paradigm could incorporate findings from the literature base, which highlight the benefits of training individuals to attend to and process positive information instead of negative information, in the reduction of general distress and/or anxiety symptoms (Blanco et al., 2023; Sîrbu et al., 2025). This has the potential to extend our novel valence finding, by encouraging the development of a positive information processing bias, through the use of an online bias modification paradigm, with the goal of attenuating the experience of anxiety reactivity.

These modification-based interventions are often developed to supplement treatment as usual, or to reach individuals who are currently on waitlists for psychological intervention (Vrijsen et al., 2024). While there have been queries raised as to whether these interventions are effective in modifying cognitive biases (Jones & Sharpe, 2017), a novel intervention could be designed which incorporates proposed amendments and addresses common criticisms, as the benefits of these often online based interventions are hard to overlook. Firstly, they would allow researchers to modify relatively automatic processes, which is hard to achieve through other cognitive interventions, and the digital format would allow engagement for populations who are often difficult to reach (e.g. those living rurally) (Vrijsen et al., 2024). Developing and piloting a novel intervention could prove beneficial in targeting the two dimensions of trait anxiety separately, making this proposal particularly relevant for individuals with mental health concerns which likely map onto one dimension more than the other (e.g., panic disorder and anxiety reactivity; Rudaizky et al., 2014).