We conducted an online survey of adults to examine their experiences of mindfulness. Before participating in the survey, the participants were asked to read the research information and complete an electronic consent form. Only consenting participants were allowed to continue with the study. They provided sociodemographic, clinical, and individual difference information relevant to mindfulness. The questionnaire items were translated from English to Korean and modified to fit the context of the study’s topic.

Participants were recruited in December 2021 using an online panel developed by a survey company, dataSpring Korea, Inc. Adults older than 19 years and who consented to participate in the survey were recruited. Based on their responses to the survey’s screening question, they were categorized into 4 groups: (1) mindfulness mobile app users—those who used mindfulness mobile apps within a month of either downloading an app or subscribing to a paid premium service; (2) mindfulness mobile app churners—those who had experience of using mindfulness mobile apps in the past but either deleted the apps or did not use them within a month of downloading them; (3) other mindfulness behavior performers—those who did not have any experience of using mindfulness mobile apps but had experience of practicing mindfulness behaviors through other means; and (4) no mindfulness behaviors—those with no experience in performing any type of mindfulness behavior.

After the survey, participants were recategorized into 4 groups: (1) people using mindfulness mobile apps and practicing other types of mindfulness behaviors; (2) people using only mindfulness mobile apps; (3) people who only practice other types of mindfulness behavior; and (4) people not practicing any type of mindfulness behavior, with a focus on current (within 1 month) mindfulness activities. Participants only practicing other types of mindfulness behaviors were excluded from the data analysis because the study focuses on the use of mindfulness mobile apps. Consequently, the responses of 297 participants were used for the data analysis. At the time of the survey, 142/297 (47.8%) participants were using mindfulness mobile apps (either in conjunction with other types of mindfulness behavior or using mindfulness mobile apps only) and 155/297 (52.2%) were not practicing any type of mindfulness behavior. Of the 142 participants using mindfulness mobile apps, 46 (32.4%) were using Mabo, 46 (32.4%) were using Kokkiri, 17 (12%) were using Calm, 13 (9.2%) were using Harumeditation, and 20 (14.1%) were using other mindfulness mobile apps.

Of the total 297 participants, 47 (15.8%) were 20-29 years old, 68 (22.9%) were 30-39 years old, 108 (36.4%) were 40-49 years old, 72 (24.2%) were 50-59 years old, and 2 (0.7%) were 60 years old or older. In addition, 153 (51.5%) participants were male and 144 (48.5%) were female; 180 (60.6%) participants were married and 102 (34.3%) were single; and 170 (57.2%) participants had no religious affiliation. More than half of the participants were university graduates (180/297, 60.6%). Regarding the participants’ employment status, 192 (64.6%) were permanently employed. More than half of the participants (188/297, 63.3%) responded that they were in the middle-income bracket. Table 1 presents more information about participants’ demographic characteristics.

Structural equation modeling (SEM) was conducted to verify the proposed research model established to predict mindfulness behavior by integrating the Big Five personality traits and the TPB. First, a confirmatory factor analysis was conducted to examine the measurement model and assess the model fit using several goodness-of-fit indices. Next, SEM was conducted to evaluate the structural model and test the study hypotheses. Lastly, bootstrapping mediation analyses were conducted to test the significance of the mediation pathways more precisely. SPSS 22.0 (IBM, Inc.), Amos 22.0 (IBM, Inc.), and PROCESS macro for SPSS 4.1 [44] software were used for the data analysis.

This study was approved by the Institutional Review Board of Yongin Severance Hospital in Yonsei University Health System (IRB No. 9-2021-0167). We made a data collection request to an online panel research company. Only the online panel members older than 19 years had access to this study’s online survey. The members interested in the survey could thoroughly review the study explanation and voluntarily decide to participate. They could participate in the survey only if they provided informed voluntary consent. However, participants could withdraw their consent or stop participating in the study at any time according to their free will. This study was conducted through a survey; no special side effects or physical damage was expected. We received deidentified raw data from the online panel research company after the survey completion. The data were password protected, and only the research team had access.

A confirmatory factor analysis was conducted to verify the factor structure of the proposed model. First, after examining the factor loading of the observed variables constituting the latent variables, all the factor loading values were found to be statistically significant (P<.05 in all cases). The factor loading values of all the observed variables were more than 0.50, except for an observed variable in subjective norm about mindfulness (ie, the “'I feel pressure to perform mindfulness from people around me” item), which was below 0.50. Therefore, we concluded that the observed variables make up the latent variables based on an appropriate theoretical conceptualization [45,46]. In other words, content validity was established.

Next, convergent validity—the explanatory power and validity of the latent variables itself—was examined through composite/construct reliability (CR), representing the internal consistency of the observed variables, and the average variance extracted (AVE) value, representing the size of the variance that the observed variables can explain. The convergent validity was secured for all the variables because the CR exceeded 0.70 and the AVE exceeded the threshold of 0.50 [46,47].

The details of each observed and latent variable are summarized in Table 2.

After examining convergent validity, discriminant validity was examined to determine whether there was an overlap or similarity between each latent variable and whether there was differentiation. Each AVE value had to be greater than the square value of the correlation coefficient between certain variables to secure discriminant validity [47]. In this measurement model, discriminant validity was secured because the minimum value of the AVE (0.502) was greater than the largest square value of the correlation coefficient (maximum=0.696×0.696=0.484).

In terms of reliability, the Cronbach α value of each variable measurement was calculated to examine the internal consistency of the measurement tool. As the Cronbach α values of all the variables were above .70, all the measurements showed good reliability [45,48].

After confirming that the study’s measurement model met the validity and reliability requirements, an evaluation to test overall model fit was conducted. As a result, the values χ²₂₁₅=431.1 (P<.001), χ²/df=2.005, incremental fit index (IFI)=0.942, comparative fit index (CFI)=0.941, Tucker-Lewis index (TLI)=0.930, and root mean square error of approximation (RMSEA)=0.058 met all the criteria: χ² (P<.001), χ²/df≤3, IFI≥0.90, CFI≥0.90, TLI≥0.90, RMSEA≤0.08 [48]. Therefore, the overall model showed a good fit.

The details of the measurement model verified in this study are summarized in Table 3.

This research theorized and examined the indirect effect of the Big Five personality traits on behavioral intention for mindfulness based on the TPB. The study’s main finding was that conscientiousness and neuroticism could potentially impact behavioral intention for mindfulness through attitude toward mindfulness and perceived behavior control over mindfulness. The results showed that among the Big Five personality traits, only conscientiousness and neuroticism were determinants influencing people’s mindfulness behavior.

This study found no direct association between conscientiousness and behavioral intention for mindfulness; however, conscientiousness was positively associated with behavioral intention for mindfulness, via attitude toward mindfulness and perceived behavior control over mindfulness. This finding indicates that the higher the level of conscientiousness, the more positive the attitude toward mindfulness, and the stronger the perceived behavior control over mindfulness. In addition, attitude toward mindfulness and perceived behavior control over mindfulness were positively associated with behavioral intention for mindfulness, indicating that the more positive the attitude toward mindfulness and higher perceived behavior control over mindfulness, the stronger the behavioral intention for mindfulness. A strong sense of responsibility, good discipline, and effective self-regulation are therefore closely related to the characteristics of high conscientiousness [34]. People with a high level of conscientiousness are more likely to think positively about mindfulness and perceive that they are confident in performing mindfulness activities. Consequently, conscientiousness increases the likelihood of inducing behavioral intention for mindfulness.

Most importantly, neuroticism was negatively associated with behavioral intention for mindfulness via perceived behavior control over mindfulness. Inconsistent with previous studies [39,40], neuroticism was directly and positively associated with behavioral intention for mindfulness. Unlike conscientiousness, neuroticism showed a negative association with perceived behavior control over mindfulness. In other words, the higher the level of neuroticism, the weaker the perceived behavior control over mindfulness. In addition, perceived behavior control over mindfulness was positively associated with behavioral intention for mindfulness. Thus, the lower the level of perceived behavior control over mindfulness, the weaker the behavioral intention for mindfulness. As vulnerability to negative emotions is one of the characteristics of high levels of neuroticism [34], individuals with high neuroticism are less likely to perceive that they are confident in performing mindfulness. Hence, neuroticism may play a key role in reducing behavioral intention for mindfulness.

The results of this study suggest that people with high levels of neuroticism are more likely to have strong behavioral intentions for mindfulness regardless of perceived behavior control over mindfulness. One explanation is that people with high neuroticism are more likely to have poor mental health [49] but may seek treatment rather than avoidance under certain conditions. Those who perceived a high need for treatment for depression or social support are also more likely to seek professional help for depression [50]. Their strong motivation for mental health recovery may lead to stronger behavioral intention for mindfulness. In this regard, more studies are needed to verify this possible explanation and find a condition that may in turn produce results that are inconsistent with the findings of previous studies.

In the context of mindfulness, this study suggests an integrated model to explain the relationship between personality and behavior by combining the Big Five personality traits and the TPB. The results show that personality factors (ie, conscientiousness and neuroticism) and behavioral factors (ie, attitude toward mindfulness, subjective norm about mindfulness, perceived behavior control over mindfulness, and behavioral intention for mindfulness) are closely associated with changes in mindfulness behavior.

This study shows that the relationships or patterns between the Big Five personality traits and TPB variables in the context of mindfulness differ depending on personality traits. Attitude toward mindfulness and perceived behavior control over mindfulness mediate the association between conscientiousness and behavioral intention for mindfulness. However, only perceived behavior control over mindfulness mediates the association between neuroticism and behavioral intention for mindfulness. Furthermore, 3 other Big Five personality traits (ie, openness, extraversion, and agreeableness) are known to be less associated with mindfulness [39]. We posit that such inconsistencies might stem from the different characteristics of the Big Five personality traits. For example, the characteristics of high openness and liking new and diverse experiences are quite far removed from those attributes of mindfulness that require relaxation and peace. The conflict between personality traits and mindfulness characteristics may impact their associations. Examining the underlying mechanism of the effect of personality traits’ different characteristics on mindfulness behavior could require a more advanced theoretical model with higher explanatory power on mindfulness behavior. Scrutinizing the multifaceted nature of mindfulness [51,52] may also provide a useful lens to interpret the relationship between personality traits and mindfulness. Future studies could examine the reasons why different relationships exist between personality traits and mindfulness behavior.

Our findings could also help to establish successful persuasion strategies that can encourage mindfulness behavior contingent on each individual’s personality. If health professionals are able to identify a conscientious person, they could provide an intervention that may improve their attitude toward mindfulness and perceived behavior control over mindfulness to induce behavioral intentions for mindfulness and thereby promote mindfulness behavior. If health professionals are able to identify a person with high neuroticism, they could concentrate on improving the person’s perceived behavior control over mindfulness and provide appropriate intervention effectively.

When considering mindfulness mobile apps as health intervention tools, applying immersive technology in developing mindfulness mobile apps to provide users with virtual experiences of mindfulness could be useful to improve users’ perceived behavior control over mindfulness [53]. Embedding information on personality traits when designing the content of mindfulness mobile apps could be helpful in allowing users to perceive the content as personally relevant. As personalized health programs are known to have more persuasive effects than regular health messaging [54], tailoring a health intervention based on individual characteristics may improve mindfulness behavior. If a person with high neuroticism uses a mindfulness mobile app service to practice mindfulness behavior virtually and receives mindfulness tips for people with high neuroticism, the user may be more successful in overcoming the characteristics of neuroticism and improve their perceived behavior control over mindfulness. Taken together, scholars and health practitioners need to consider the indirect effects of personality traits to influence change or adherence to mindfulness behavior.

In this study, we only explored the potential of personality and behavioral factors to impact mindfulness behavior likely to drive the use of mindfulness mobile apps. We acknowledge that demonstrating other factors impacting mHealth adoption, not limited to personality traits, is important to build on the growing body of research examining mindfulness mobile apps usage. In their systematic review, Jacob and colleagues [55] introduced social and personal factors, technical and material factors, and health-related factors affecting mHealth adoption. Demographic factors (eg, age, gender, education, ethnicity, and socioeconomic factors), personal characteristics (eg, attitudes, motivation, and psychological factors), and cultural and social elements (eg, social influence, language, and culture) were representative social and personal factors. Usefulness (eg, perceived benefit, communication, and self-management), ease of use, technical factors (eg, access to technology, training, and tech support), monetary factors, data related (eg, privacy, credibility, and relevance), and user experience (eg, usability and personalization) composed technical and material factors. Disease or health condition, care team’s role, health consciousness and literacy, relation to other therapies, health behavior, and insurance status were labeled as health-related factors [55]. Jacob and colleagues [55] emphasized taking a more holistic view of these factors and developing more patient-centered approaches (eg, fit into patient journey, inclusive design, and patient education) to facilitate mHealth adoption. They also highlighted that these factors are not mutually exclusive and showed mixed results on mHealth adoption depending on the context. It means that focusing on 1 or only a few factors alone on mHealth adoption and implementation is not likely to achieve success.

The complexity of the factors affecting mHealth adoption that Jacob and colleagues [55] stressed provides an implication to this study. The different potentials of personality and behavioral factors to impact mindfulness behavior and the use of mindfulness mobile apps can be explained by not only the characteristics of personality traits themselves but also other factors and contexts beyond personality traits. For instance, even people with high conscientiousness may only perform mindfulness behavior and use mindfulness mobile apps if they have strong motivation. They may not use mindfulness mobile apps if they feel such apps are uncomfortable to use or cannot protect personal data well. By contrast, people with low conscientiousness may use mindfulness mobile apps if they were educated about the importance of mindfulness and how to use mHealth services. This complexity may result in different and inconsistent outcomes of the associations between personality traits and mindfulness behavior. More comprehensive approaches to knowing the mediating and moderating roles of various factors are crucial to predict the use of mindfulness mobile apps. Future research is required to discover successful strategies for using the characteristics of the factors affecting mHealth adoption and their interconnectivity to facilitate the use of mindfulness mobile apps.

The study has a few limitations. First, this research only looked at the Big Five personality traits and TPB variables and the mediating relationships among them. The study did not consider other possible variables that could impact mediations or other possible relationships among the Big Five personality traits and TPB variables. If this deficiency can be addressed and the proposed model can be expanded through follow-up studies, our understanding of the relationship between personality traits and behavior in the context of mindfulness will be improved. Second, although this study examined mediations supported by theoretical discussions and previous evidence, a careful interpretation of the causal assumptions is merited because this study used cross-sectional data. Future studies using experimental research methods are needed to verify causal relationships. Third, considering the development of this research’s proposed model, future studies should be conducted with other types of health behaviors to refine the model or increase the model’s degree of generalizability. Fourth, this study’s participants recruited through an online panel were already likely to favor digital technology use itself. Therefore, it is encouraged to adopt a more randomized participant selection method to deal with this inherent bias and recruit participants regardless of technology preference or knowledge. Last, the data were self-reported and could be strengthened by physiological response measures.

The COVID-19 pandemic has highlighted the importance of mental health and mindfulness and various types of eHealth and mHealth mindfulness services have appeared [56]. The question is how to promote mindfulness behavior and encourage people to adhere to mindfulness behavior. Identifying the determinants of mindfulness behavior is the first to answer this question. This paper focused on personality traits and verified that the combination of the Big Five personality traits and the TPB provides a useful theoretical framework for predicting mindfulness behavior. The research results provide a foundation to develop an advanced model that is able to illustrate the relationships between personality and behavioral factors by adding other potential variables and moderation paths in the context of mindfulness. The findings also underscore the potential of integrating other personality and behavior change theories to build a new model to explain mindfulness behavior. Furthermore, this research provides evidence to extend the research field and explore how diverse characteristics of personality traits affect mindfulness behavior. The role of personality may well explain mindfulness behavior more than people realize.