Metacognition is the ability of an individual to monitor and control their own cognitive processes, which is “thinking about thinking [1].” Metacognition is divided into three functional components: metacognitive knowledge, an individual’s awareness and understanding of their own cognitive processes and abilities; metacognitive strategies, deliberate actions and techniques that individuals use to monitor, control, and regulate their cognition; and metacognitive experiences, feelings, and judgments that occur during cognitive activities [2]. While most metacognitive processes occur automatically and involuntarily, the capacity to consciously reflect on one’s thoughts provides the advantage of self-regulating cognitive functions [3]. After the introduction of the concept, metacognition became a significant topic of interest in various fields due to its crucial role in psychological processes such as insight, learning, self-regulation, and problem-solving abilities [3,4].

In the context of mental health, metacognition plays two critical roles. First, dysfunctional beliefs about one’s thoughts—such as worry, rumination, and self-blame—can serve as pathological factors that maintain cognitive distortions. These maladaptive metacognitive patterns contribute to the persistence and exacerbation of symptoms by reinforcing negative thought processes. Metacognition is deeply involved in the symptoms of numerous psychiatric disorders, including depression, bi-polar disorder, and obsessive-compulsive disorder [5-7]. Second, developing metacognitive skills offers a therapeutic intervention. By learning to regulate their cognitive processes and strategically intervene in their own thinking patterns, patients can break the vicious cycle of symptoms and negative cognition [8]. This leads to the adoption of more functional coping strategies and improves overall mental health outcomes [9].

Metacognition has been particularly emphasized in relation to pathological worry in anxiety disorders. The metacognitive model of anxiety proposed by Wells [10] has highlighted the significant impact of metacognitive beliefs on the development and maintenance of generalized anxiety disorder (GAD). According to Wells’s [10] metacognitive model, metacognitions about worry are key factors that influence both the initiation and perpetuation of GAD. This model posits that individuals hold two types of beliefs about worry: positive metacognitive beliefs (Type 1) and negative metacognitive beliefs (Type 2) [11]. These beliefs shape how individuals respond to their own thoughts and worries, thereby affecting their emotional and behavioral reactions. Type 1 worry involves positive beliefs about the utility and effectiveness of worry. Individuals with these beliefs view worry as a helpful coping strategy for dealing with potential threats or problematic situations. For example, they might think, “Worrying will help me prepare for the worst,” or “By worrying, I can prevent bad things from happening.” This type of worry is typically situation-specific and ceases once the perceived threat is resolved. It is considered a normal cognitive process that can be adaptive in managing stress and is common in both clinical and non-clinical populations [11].

However, problems arise when worry leads to negative consequences, and individuals begin to perceive worry itself as harmful or uncontrollable [11]. This shift gives rise to negative metacognitive beliefs about worry, constituting Type 2 worry, also known as meta-worry [12]. In this state, individuals start to worry about worry. Predominant thoughts might include, “I cannot control my worry,” “Worrying is dangerous and could harm me,” or “My inability to stop worrying means I’m losing control.” These negative beliefs about worry intensify anxiety and contribute to a vicious cycle where attempts to control or suppress worry fail, further reinforcing the negative metacognition. As individuals engage in ineffective behavioral strategies or thought-control techniques to alleviate their worries, such as avoidance, distraction, or suppression, they may experience temporary relief but ultimately exacerbate their anxiety symptoms. The failure of these strategies reinforces the belief that worry is uncontrollable and dangerous, heightening anxiety and perpetuating the cycle of meta-worry [11]. This process is central to the onset and maintenance of GAD symptoms, as individuals become trapped in a self-perpetuating cycle of worrying about the fact that they are worrying [13]. Empirical evidence supports the significance of negative metacognitive beliefs in GAD. Research indicates that patients with GAD experience higher levels of meta-worry compared to those without anxiety disorders [14]. Moreover, meta-worry is significantly more prevalent in GAD than in other anxiety disorders, such as panic disorder, social phobia, or somatic anxiety [15]. Additionally, meta-worry has been shown to distinguish between individuals with high and low levels of GAD symptoms [16] and plays a significant role in predicting worry severity [17]. These findings suggest that meta-worry is a key feature in GAD and is closely associated with the severity of anxiety symptoms.

While meta-worry can exacerbate psychiatric symptoms, recent research has increasingly focused on the therapeutic potential of metacognition to mitigate these effects. Cultivating an awareness and acceptance of one’s own thoughts and emotions can help interrupt the negative spirals that reinforce maladaptive cognitive patterns [18]. In this context, metacognitive therapy (MCT), developed by Wells [19], has emerged as a key intervention that targets how individuals relate to their internal experiences rather than attempting to eliminate or suppress them. MCT focuses on modifying dysfunctional metacognitions— particularly those related to worry and rumination— rather than directly challenging the specific content of these thoughts. MCT aims to break the cycle of meta-worry by altering negative metacognitive beliefs, thereby reducing anxiety symptoms and improving overall functioning. For example, MCT primarily aims to alter how individuals perceive and manage their own thoughts (e.g., questioning the purpose of worrying about accidents), and even when MCT does involve evaluating the accuracy of certain thoughts, it concentrates on beliefs about cognition (e.g., “I cannot control my worries”) rather than conventional cognitive content (e.g., “the world is dangerous”) [18]. Empirical studies have demonstrated that MCT significantly reduces worry and anxiety symptoms among individuals with conditions such as anxiety and depression, underscoring the importance of addressing metacognitive factors in treatment [8]. Consequently, a validated measure of metaworry is essential to better understand how meta-worry contributes to symptom exacerbation and to assess the effectiveness of interventions that target it.

The Meta-Worry Questionnaire (MWQ) is a self-report rating scale designed to assess negative metacognitions about worry within the framework of Wells’s [11] metacognitive model. Similar tools, such as the Meta-Cognition Questionnaire, assess metacognitive beliefs and processes [20], while the Anxious Thoughts Inventory evaluates social, health, and metaworry [21]. However, these tools do not measure beliefs about the dangerousness of worry, which is a primary component of Type 2 worry. The MWQ measures the beliefs regarding the uncontrollability and danger of worry, and provides a more precise evaluation of meta-worry by quantifying these negative beliefs. Previous studies have validated the reliability and construct validity of MWQ, and high MWQ scores were associated with a diagnosis of GAD [11,22]. Moreover, the MWQ has been used to identify pathological worry status in various contexts, such as posttraumatic stress disorder, childhood development, and during the COVID-19 pandemic [23-25]. Furthermore, the MWQ can be used to evaluate the effectiveness of treatment targeting metacognition and meta-worry [12]. Thus, the MWQ increases our understanding of the metacognitive processes underlying GAD and contributes to the development of more effective, individualized treatment strategies. A Korean version of the MWQ has not been developed, thereby limiting the application of the meta-worry construct within clinical settings, including the assessment of GAD and evaluations of treatment efficacy. In this study, we aimed to explore the reliability and validity of a Korean version of the MWQ among the general population in Korea.

Demographic information of the study participants is shown in Table 1. The total sample was randomly split into two independent subsamples: Group I (n=300) for EFA and Group II (n=300) for CFA and IRT analysis. Statistical comparisons revealed no significant differences between the two groups regarding age, sex distribution, marital status, or psychiatric history (p>0.05). Furthermore, baseline scores on the MWQ, GAD-7, ISI, MCQI-14, and the DBST index were comparable between the groups, confirming the successful randomization and equivalence of the subsamples. Regarding the data distribution of the MWQ items, mean scores ranged from 1.53 to 1.76 (SD=0.79-0.88). Skewness values ranged from 0.83 to 1.48, and kurtosis values ranged from -0.09 to 1.50, indicating data normality within the acceptable range of ±2.

This study aimed to examine the factor structure and validate the MWQ in the Korean context. Unlike previous validation studies that focused only on the classical test theory (CTT) approach [11,22], the current study used CTT and IRT approaches to assess the psychometric properties of the scale. Based on two independent subsamples of 300 participants each, the Korean version of the MWQ demonstrated excellent internal consistency (Cronbach’s α=0.946, McDonald’s ω=0.946) and strong model indices in CFA (SRMR=0.027, RMSEA=0.084, CFI=0.995, TLI=0.992). The item-level analysis further revealed strong correlations between individual item scores and the total scale score (CITC range=0.793-0.861) and strong interitem consistency (CID range=0.680-0.751), suggesting that the items contribute meaningfully to the overall measurement of meta-worry. These results indicate that the Korean MWQ is a valid, well-integrated measure for assessing negative metacognitive beliefs about worry.

Our reliability coefficients were highly consistent with those reported for the original English version and the Turkish adaptation [11,22], suggesting that each item in the Korean version retains its internal coherence after translation and cultural adaptation. Consistent with the original conceptualization of meta-worry as a unitary construct, our preliminary analysis indicated that the frequency items alone were sufficient to form a coherent dimension, even without the belief-rating component. This was further supported by a high composite reliability (0.971). The unidimensional structure revealed from EFA accounted for 71.5% of the variance, and CFA supported the single-factor solution with excellent model fit indices, which is in line with the original development and the Turkish adaptation studies [11,22]. All seven items of the MWQ showed strong factor loadings in EFA and CFA, further reinforcing the structural coherence of the scale. Results regarding the measurement invariance across sexes revealed that the Korean MWQ can be applied similarly across males and females.

IRT assumptions of the GRM model suggested the suitability of the data to perform GRM, and accordingly, GRM results revealed that the scale items provided sufficient information and could efficiently differentiate between individuals across the latent trait. The threshold parameters showed that the participants responded to each response category of the scale items at varying ability ranges. Moreover, the IICs and TIF showed that the scale items can provide information across a moderate spectrum of the latent trait, suggesting that the scale is useful in discriminating between individuals with higher and lower levels of meta-worry.

While considering the convergent validity of the scale, we observed that excessive worry about worrying was strongly associated with heightened GAD symptoms, more pronounced metacognitive beliefs related to insomnia and intrusive thoughts, and greater insomnia severity. These associations support the convergent validity of the scale. Interestingly, the correlation with the DBST index, while statistically significant, was relatively modest (r=0.12). This distinction likely arises because the DBST index quantifies a specific discrepancy between desired time in bed and total sleep time [33], which may function more as a trait-like sleep parameter or a behavioral mismatch rather than a direct measure of cognitive distress. In contrast, the ISI and MCQI-14 tap directly into the severity of symptoms and dysfunctional beliefs, constructs that share a closer conceptual overlap with the negative metacognitions assessed by the MWQ. These findings are in line with previous research showing that individuals who meet the criteria for GAD report significantly higher levels of meta-worry compared to non-anxious individuals or those experiencing other anxiety symptoms [15]. In addition, negative metacognitive beliefs such as cognitive self-consciousness, thought suppression, and intrusive thoughts tend to increase in certain situations, particularly at bedtime [49]. Intrusive thoughts before sleep have been linked to the development of primary insomnia, describing the pre-sleep thoughts as intrusive, uncontrollable, and negative [50]. Furthermore, negative metacognitive beliefs during the pre-sleep stage were found to delay the sleep onset and exacerbate the GAD symptoms and other psychopathologies [51]. Altogether, the frequency of meta-worry appears to have a direct influence on the occurrence and recurrence of GAD along with the pre-existing sleep problems, exacerbating the cognitive intrusiveness during pre-sleep conditions.

In this context, the Korean version of the MWQ offers a culturally relevant tool to assess meta-worry, which is an important metacognitive aspect of GAD. By focusing on the negative beliefs that can turn ordinary worry into a persistent and maladaptive process, the scale contributes to expanding the use of the established concept of meta-worry in diverse populations [10]. Clinically, strong item discrimination across the mid-to-upper range of meta-worry may be useful for identifying individuals who are at risk for GAD due to catastrophic beliefs about worry. Previous studies have indicated that a total score of approximately 11 on the original version may help distinguish individuals with GAD from those with other conditions [22]; future research in Korean populations could help confirm an appropriate threshold. In addition to its potential as a screening tool, the scale may aid in treatment planning by identifying individuals who are likely to benefit from interventions such as acceptance-based approaches. Moreover, as the test information curve shows good sensitivity within clinically relevant ranges, the MWQ may be useful for tracking therapeutic progress, serving as a complementary tool alongside broader symptom checklists like the GAD-7.

This study has certain limitations that should be acknowledged. First, participants were recruited through an online panel and self-selected into the study, which may limit the representativeness of the sample and introduce biases such as social desirability. Second, while the Korean version of the MWQ showed good psychometric properties in a non-clinical general population, its applicability to clinical populations remains unclear. Including participants with clinically diagnosed GAD in future studies would help determine the clinical utility of the scale, including optimal cut-off scores. Third, the study used a cross-sectional design, which does not allow for the assessment of test-retest reliability or temporal stability. Longitudinal studies are needed to evaluate the stability of the scale over time. Fourth, while the MCQI-14 was used to assess metacognitive beliefs related to insomnia, insomnia diagnoses were not verified through clinical interviews or treatment history, limiting the interpretation of findings related to insomnia severity. Despite these limitations, the Korean version of the MWQ appears to be a psychometrically sound tool with clinical implications in the field of mental health, addressing appropriate therapeutic interventions for the target individuals.

In conclusion, this study examined the psychometric properties of the Korean version of the MWQ and observed it as a reliable and valid measure of meta-worry in a general adult sample. Analyses based on CTT and IRT confirmed that the MWQ reliably measures a single construct and shows strong internal consistency and validity. These findings indicate the utility of the Korean MWQ as an effective tool for assessing meta-worry.