In recent years, expressive arts interventions, such as artwork, music, drama, narrative and storytelling, have been found to be effective in enhancing individuals’ psychological well-being (Phillips and Becker, 2019). According to a recent systematic review of expressive art interventions for health workers (Phillips and Becker, 2019), artwork and music interventions have had greater positive impacts on mental health than storytelling or narrative. Moreover, artwork intervention can work as a mindfulness-based therapy model to bring positive effects such as stress reduction, emotional regulation, self-esteem, self-awareness and resilience (Coholic, 2011). Specifically, the evidence about face-to-face artwork interventions has shown improved psychological well-being in different settings, including counsellors in the United States (Ifrach and Miller, 2016), nurses and nursing assistants in Lithuania (Karpavičiūtė and Macijauskienė, 2016; Hsu et al., 2021) and end-of-life workers in Hong Kong (Potash et al., 2014).

Among undergraduates, a few studies have examined the relationship between face-to-face art-based workshops and psychological state (Curry and Kasser, 2005; Walsh et al., 2005; Sandmire et al., 2012; Campenni and Hartman, 2020; Sonnone and Rochford, 2020). Anxiety could be reduced by creative art (Walsh et al., 2005), free-form painting (Sandmire et al., 2012), or coloring mandalas (Curry and Kasser, 2005; Walsh et al., 2005; Sandmire et al., 2012; Campenni and Hartman, 2020). Walsh et al. (2005) found that creative art reduced stress and increased positive emotion. Sonnone and Rochford (2020) found that group art therapy could improve undergraduates’ ability to express and disclose themselves, enhance their social connections, and help them to develop new insights. The benefits of artwork in improving psychological status among undergraduates via face-to-face modes have been found. However, considering the “new normal’ of online teaching during the COVID-19 pandemic and a post-pandemic future, the effectiveness of online artwork workshops on improving undergraduates’ psychological well-being worth exploring.

The popularity of the Zentangle method and Pastel Nagomi Art is growing. These two art forms share similar concepts. For example, they promote relaxation by emphasizing no comparisons, critiques, or judgments of the finished piece with others, and appreciating the opportunity to make choices [Japan Pastel Hope Art Association (JPHAA), 2003; Krahula, 2012]. Furthermore, both only require simple materials and little space.

A 3-arm randomized control trial study design was used: one arm with Zentangle workshops, the other arm with Pastel Nagomi Art, while the control group received usual care (i.e., the university provides different mental health workshops and counselling to all students). Focus group interviews followed the interventions to assess the process evaluation and explore the intervention effects.

A sample size of 10–15 in a group is recommended for a feasibility study (Hertzog, 2008). Chinese-speaking undergraduate students from a university in Hong Kong were invited to participate in the study. Students who had attended or were attending Zentangle or Pastel Nagomi classes were excluded. Recruitment was conducted through emails and promotion posters. Both convenience sampling and snowball sampling methods were used. The application period lasted for 1 week, between 18 January 2021 and 26 January 2021.

Ethical approval was obtained from the Institutional Review Board before the commencement of the study. An online questionnaire was sent to the registered students (N = 41) with an information sheet and consent form. Details of the university counselling service were provided in the information sheet. After receiving the completed questionnaire, the students who agreed to participate in the study were randomly assigned to three groups: (1) the control group, (2) the Zentangle method, and (3) Pastel Nagomi Art, using computer-generated numbers [0 for the control group; 1 for Zentangle group; and 2 for Pastel Nagomi Art group]. The randomization was conducted by a colleague who was not involved in the study. The participants were then informed of their groupings and intervention schedule. Eight students withdrew at this stage due to timetable clashes, or through failing to respond. Finally, there were 33 participants: 12 in the Zentangle group, 11 in Pastel Nagomi Art, and 10 in the control group. The study took place between 3 February 2021, and 31 March 2021.

This mixed-mode intervention consisted of two face-to-face and six online workshops. A qualified instructor with certified qualifications for teaching both Zentangle and Pastel Nagomi Art was recruited for the workshops. Eight weekly 60-min workshops of Zentangle or Pastel Nagomi Art were held with groups of 3–8 participants. The participants were invited to have face-to-face workshops for the first and last sessions. Social distancing and infection control measures were maintained, and no food and drinks were allowed during the face-to-face workshops. The second to seventh sessions were held in online mode. Zoom, which is a free technology-mediated communication application, was used as the online platform for these meetings.

The participants in the control group received care as usual at the university. Two weeks after the completion of the study, they were invited to participate in one face-to-face Zentangle or Pastel Nagomi workshop of their own choice.

To determine the feasibility of the study, (a) consent rate; (b) attrition rate; (c) intervention compliance and acceptance were assessed. To determine the preliminary effect of the intervention, study questionnaire data were collected at the baseline, T0, (before the commencement of the session); T0.5, (immediately after the 4^th session); T0.75, (immediately after the 6^th session); immediately after the last session, T1; and 1 month after the last session, T2. The questionnaire consisted of eight parts:

Please refer to Table 1 for Cronbach’s alphas of the above-mentioned scales used in this study.

All participants from the two artwork groups were invited to participate in 45–60 min focus group interviews one month after the workshop sessions, to share their acceptance of and satisfaction with their experiences.

The Statistical Software Package for the Social Sciences (SPSS), version 26.0 was used for data analysis. Descriptive statistics, namely frequencies, means, and standard deviations, were used to analyze the study variables. The Kruskal-Wallis and Wilcoxon signed-ranks tests were used to analyze the time-point differences between groups and within groups, respectively. The significance level was set at p < 0.05. The statistical power to warrant hypothesis testing was not required due to the small sample sizes for the feasibility, pilot, and exploratory studies (Lee et al., 2014).

The focus group interviews were audio-recorded. The audio recordings were transcribed verbatim in Chinese and then imported into NVivo Pro 12 for data management and analyzed by two coders (EM, CH) using theoretical thematic analysis procedures (Braun and Clarke, 2006). The themes were identified by deriving from “the explicit meaning of the data … [with]the analyst not looking for anything beyond what a participant said” (Braun and Clarke, 2006, p. 84). The coding was guided by research questions related to intervention benefits, harm as well as explaining and supplementing quantitative results. The two coders coded the transcripts separately and discussed the coding to ensure agreement on a basic set of codes and categories.

Table 2 summarizes the participants’ demographic and characteristics at baseline. 90.9% (n = 30) were females and 84.8% (n = 28) were transfer students, who were community college graduates admitted to the undergraduate program (Cheung et al., 2020a,b). 51.5% (n = 17) were Year 1 students, 30% (n = 10) of them were studying in Year 2, 6.06% (n = 2) were Year 3 students and 8.25% (n = 4) were Year 4 students. Significant differences for BBF10-neuroticism (p < 0.05) and BBF10-openness (p < 0.05) were found between the participants in the Pastel Nagomi Art and control groups at baseline. Otherwise, there were no significant differences in demographic or other variables found among the group.

A total of 41 students registered for the study, 33 agreed to participate, and 31 completed the study. Thus, the consent rate was 80.5% while the attrition rate was 6.06%. Measurement completion rates at week 8 were 100% for the three groups. The attrition occurred only in the Zentangle group. Figure 3 shows the study consort diagram.

In the Pastel Nagomi Art group arm, 100% (n = 12) completed the 8-week program. Of the 12 participants in the Zentangle group arm, 83.3% (n = 10) attended the week-6 workshop, and 100% attended the other 7 workshops.

Twenty participants attended the focus groups, 11 from the Pastel Nagomi Art group and 9 from the Zentangle group. In the focus group, participants expressed great satisfaction with both art forms. All expressed having no strong preference for the modes of the workshops. 72.7% (n = 8) of the Pastel Nagomi Art group participants and 77.8% (n = 7) of the Zentangle group did say that the face-to-face mode provided better interaction with the instructor and classmates to build social networks and the feeling of being supported, and hence being able to receive clear instruction about the steps of the artwork. On the other hand, 27.3% (n = 3) of Pastel Nagomi Art group participants and 22.2% (n = 2) Zentangle group participants indicated that the online mode allowed them to use the art materials and tools more conveniently and in a more relaxed way at home. Most participants in the Pastel Nagomi Art group suggested that the workshop length could be extended to 75 min. The majority of participants (n = 16, 80%) agreed the workshop should start at the beginning of the semester but some of them (n = 4, 20%) suggested they could be held after the examination period.

To the best of our knowledge, this is the first study to investigate the feasibility and preliminary effects of a mixed-mode (face-to-face and online) intervention of Pastel Nagomi Art and Zentangle Art on improving the mental health of undergraduate students during the COVID-19 pandemic. The present study achieved an acceptable consent rate, demonstrating acceptable recruitment. In addition, high attendance rates in both intervention groups suggested excellent acceptability of the intervention. Timetable clashes were the main reasons for absence from workshops. Furthermore, the focus-group interview participants provided supporting evidence of their acceptance of both artwork interventions and the modes of the intervention. Compared with previous studies with the consent rate and attendance rate ranged from 56.3–70.2% (Coholic et al., 2018) and 50.3–95% (Lee, 2021) respectively, the present study has demonstrated its high feasibility to pursue a full-scale study.

In the current study, a mixed mode (face-to-face and online) was used which is probably the first of its kind. Most of the artwork interventions in previous studies were conducted merely face-to-face (Kim, 2013; Chen et al., 2016; Karpavičiūtė and Macijauskienė, 2016; Campenni and Hartman, 2020; Newland and Bettencourt, 2020; Hsu et al., 2021) or online modes (Sit et al., 2022). In terms of dose, frequency and duration, this present study is comparable with previous work. According to a recent systematic review of the effectiveness of mindfulness-based artwork therapy for managing symptoms of anxiety, depression, and fatigue, among 14 studies, the doses (i.e., lengths of sessions) ranged from 45 to 180 min, the frequencies (i.e., the time intervals between sessions) were weekly or bi-weekly, and the durations ranged from two to 13 weeks (Newland and Bettencourt, 2020). In this review, 11 studies (78.57%) conducted interventions weekly for eight weeks (Newland and Bettencourt, 2020), which the present study also adopted. There were some variations. Hsu et al. (2021) conducted a face-to-face Zentangle art study for healthcare workers with one four-hour session; while Sit et al. (2022) conducted Zentangle art in online mode with weekly 30-to 90-min sessions for 11 weeks. The doses of the interventions might have varied because of the types of artwork, for instance, 90 min for creative art-making (Burns and Waite, 2019) and 180 min for content-picture drawing (Monti et al., 2013). The doses of Zentangle art ranged from 60 (Chen et al., 2016) to 90 min (Chia et al., 2020). It is worth noting that in the interviews the participants suggested extending each session to 75 min. Future studies should explore the dose, frequency and duration of interventions further.

The results of this study support positive intervention effects, as both intervention groups improved in regard to depression, and positive affect, compared with the control group. Furthermore, as compared with the control group, the Pastel Nagomi art had a more sustainable effect on stress as noted in week 12. Karpavičiūtė and Macijauskienė, 2016 conducted an RCT and found that silk painting artwork could improve mental well-being, reducing stress and fatigue and increasing a sense of community at work among nurses and nursing assistants. Furthermore, previous studies also found art making with paint, markers, and collage materials on canvas material reduced stress in female counsellors (Ifrach and Miller, 2016). Thyme et al. (2007) indicated that painting and drawing helped to reduce depression symptoms in depressed women and relieved depression, perceived stress and parenting stress in Korean mothers of children with disabilities (Lee, 2021). Coloring mandalas can also help to reduce anxiety and negative affect and increase positive affect in undergraduates (Campenni and Hartman, 2020). Moreover, Zentangle art can improve self-esteem and reduce social interaction anxiety in patients with psychosis (Chen et al., 2016) and enhance psychological and family well-being (Sit et al., 2022). One previous study supported the present study’s finding, reporting that a four-week creative art program, including water-coloring, collage making, beading and knitting, was effective in enhancing the mental well-being among women (Fraser and Keating, 2014).

Although improvement trends were observed, there were no significant improvements in stress, anxiety, mindfulness and self-efficacy. One reason was the small sample size in this study. The improvement trends suggested that it might take time for the effects of both artworks on mental well-being through enjoyment of the artwork process, pride in the artwork and personal growth to develop. The qualitative exploration of this study supported this suggestion. The participants in both intervention groups experienced a relaxing and joyful atmosphere, being focused and establishing social networks. They pointed out that the eight artwork sessions provided them with regular weekly one-hour sessions away from stressors and allowed them to take a break and refresh their minds during the intervention. Regular practice of relaxation techniques can help individuals to manage their stress at moderate emotional levels (Pfeiffer, 2001). Moore (2013) reported that a repetitive act of drawing patterns or lines, similar to the Zentangle art, helped individuals get into a mindfulness state and generate self-healing processes to reduce stress. Karpavičiūtė and Macijauskienė, (2016) found that finger painting which was similar to the concept of Pastel Nagomi art could improve the state of mindfulness. Tactile sensation, mindfulness state and the experience of being present are related to cognitive processes (Grossman, 2011; Hinz, 2020). Developing a state of mindfulness could be effective in decreasing mood disturbance and stress (Brown and Ryan, 2003). In addition, Salzano et al. (2013) reported that group painting activities helped hospice caregivers to increase social support and hence reduce their stress, which is consistent with our findings.

Moreover, it was important that the participants were keen to share their artwork products with others or on social media, which serves as evidence they felt satisfied with their products. The participants expressed their enjoyment of receiving compliments from others on their artwork. Lin Brostrom et al. (2020) found that sharing the artwork could provide a positive new topic and an opportunity for the participants to interact with others via technologies such as text messages and social media. Choi and Choung (2021) also found that women who used social media for social connections had increased satisfaction with life during the COVID-19 pandemic. The present study supported previous findings that artworks can improve subjective well-being by providing positive self-images and hope, enhancing satisfaction with daily life, and expanding contact with the outside world (Reynolds et al., 2008). The satisfactory artwork products and sharing behaviours could further enhance self-esteem and self-efficacy.

Furthermore, the participants in both intervention groups encountered personal growth with a renewed sense of self and learned a new relaxation skill. They reflected that they had increased their awareness of their emotions, especially the changes that occurred to their negative affect, which added support to the finding that the Zentangle art group participants’ negative affect was higher than that of the control group participants at week 4. The Zentangle art group participants had increased their awareness of their emotions after the intervention, but they may still have tried to deal with the negative affect. Emotional awareness was found as an essential component of effective emotional regulation (Gratz and Roemer, 2004). The repetitive movement in both artworks could help the participants get into a state of mindfulness to increase their awareness of the present (Moore, 2013; Karpavičiūtė and Macijauskienė, 2016). This increased awareness helps individuals reconsider threatening situations and regulate emotional responses (Modinos et al., 2010; Prazak et al., 2012). Learning a new relaxation skill allowed the participants to acquire a new method to release stress for long life, not only for the short period of time of the study.

This study had some limitations. An imbalance number of online vs. face-to-face sessions might affect the study outcomes. This study was conducted during the COVID-19 pandemic, social distancing was reinforced. Thus, to minimize face-to-face contact, only two face-to-face workshops were conducted. Although the participants did not have a strong preference for the modes of the workshops, about three-fourths of them preferred face-to-face. Further studies can consider four online and four face-to-face workshops in the post-pandemic situation. The two-time points of data collection at weeks 4 and 6 during the 8-week intervention could be another limitation. As mentioned, a recent systematic review found that the duration of interventions ranged from two to 13 weeks (Newland and Bettencourt, 2020). With this reference, the intention of collecting data at these two-time points was to test the effect of the interventions in a short duration. However, repeated measures might induce survey fatigue. In addition, according to the qualitative results of this study, participants’ confidence has been improved. In this study, the Rosenberg (1965) 10-item self-esteem scale was used but low Cronbach’s alpha values at different time points were registered (0.01 < α < 0.66). These low values could be due to the small sample size. Due to the low values, the self-esteem scale was not included in the analysis. Future studies should re-assess the psychometric properties of this scale for this university student population, and other scales can also be used to verify the impact of these two studied artworks on self-esteem. This study involved students from one university which affects the generalizability of the study results.