'This bloody rona!': using the digital story completion method and thematic analysis to explore the mental health impacts of COVID-19 in Australia.

OBJECTIVE: To use the digital story completion method to prompt participants to describe thoughts, fears and mental health experiences in response to a story stem about COVID-19, to capture a specific sociohistoric moment.
DESIGN: We used digital story completion, a qualitative research method, to gather narratives from Australians coping with physical distancing and social restriction measures. Our reflexive thematic analysis of the data was underpinned by a constructionist approach to reflect the importance of social context in understanding health experiences.
SETTING: Australia. PARTICIPANTS: 52 people living in Australia (aged 18 years and over).
RESULTS: Four meta-themes were prevalent across 52 stories submitted: (1) expressions of mental distress linked to COVID-19; (2) various coping strategies offered by characters in stories; (3) narratives outlining social support offered to alleviate distress; and (4) specialised COVID-19 vocabulary.
CONCLUSION: We cautiously propose that points of convergence across stories indicate a level of shared experience among participants relating to COVID-19 in Australia. We suggest this is due to intensive media coverage of the pandemic, persistent public health messaging, engagement with social media and instant messaging technologies, and extended lockdowns that impacted the mental health of vast numbers of Australians. © Author(s) (or their employer(s)) 2022. Re-use permitted under CC BY-NC. No commercial re-use. See rights and permissions. Published by BMJ.

Story completion has the potential to elicit rich detail regarding individual and collective experiences of social phenomenon, such as the impact of a pandemic on health and well-being.

The story completion method invites people to think creatively and speculatively to reveal processes of sense-making and social experience.

This is the first study to report on data generated from story completion research relating to the COVID-19 pandemic.

There is a need for caution vis-à-vis assumptions about the relationship between the story content and participants’ realities.

Lack of significant ethnic and cultural diversity in our cohort signalled that we had not used story completion in appropriate ways for diverse cultural groups.

Background

Story completion is an innovative mode of qualitative enquiry that offers a unique means of creatively engaging research participants. The method encourages participants to respond creatively and speculatively to a scenario in a manner that does not require them to disclose personal details or their own experiences.1–3 It is adaptable to various epistemological frameworks but is most frequently used within a social constructionist framework.1 2 The story completion method has burgeoned in recent years.4 5 This relatively new research approach has been detailed in the introduction to a special journal issue on story completion.2 The method was also featured in a 2019 symposium with an introduction and three abstracts published in BMJ Open, 6–9 signalling a growing interest in its applications in health and social care.

Story completion originated in psychology, feminist theory and psychotherapy traditions.2 10 It was first adapted by Kitzinger and Powell11 for use in qualitative research using a social constructionist perspective and discourse analysis. Story completion is increasingly used to study patterns of social meaning and assumptions depicted in a set of stories responding to a ‘stem’ or cue.12 The method has recently garnered interest from qualitative researchers keen to pursue its possibilities as a narrative method,2 largely to stimulate social meanings and understand sense-making and lifeworlds.13 As such, story completion has untapped potential to elicit rich detail regarding individual and collective experiences of large-scale social phenomena such as the impact of a pandemic on health and well-being.

Research undertaken in various locations globally, including Australia, reveals that COVID-19 (and the unique socioeconomic impacts it has wrought) has adversely impacted the mental health and well-being of diverse cohorts of people.14–20 We used the story completion method with Australian participants coping with measures put in place to limit the spread of COVID-19 to gain insights regarding the impact of the pandemic on mental health and well-being.

Methods

Design

We used digital story completion to gather narratives from Australians coping with physical distancing and social restriction measures in place from March 2020. We were particularly interested in the pandemic’s impact on mental health and well-being as, at the time, government and media discourse often focused on mental health concerns resulting from the management of COVID-19.21–24 For example, in April 2020 the New South Wales Government announced a $73 million support package intended to support mental well-being during the pandemic.25

Story completion was appropriate to explore how people were responding to rapidly changing situations, given its use as a tool for ‘accessing social meanings and discourses, and dominant assumptions and norms’.26 The method was used to archive unusual global circumstances affecting Australia as a distinct context. Our interest in capturing how respondents coped with public health restrictions and their impact on health and well-being informed our choice of stem: an 11pm phone call is received from Ali’s elderly parent who is distressed due to COVID-19. What happens next?

We discussed the ‘hypothetical’ nature of the stem, in line with past applications of story completion. Given the uniqueness of the COVID-19 situation and attendant collective experiences of uncertainty, we acknowledged that the stem might also echo participants’ reality. Indeed, one participant wrote that the stem was an exact reflection of their life during the pandemic. We hypothesised that respondents might use the stem to creatively explore and represent their lived experiences of the pandemic.

Patient and public involvement

Members of the public were not involved in the development of this research. As discussed in the Strengths and Limitations section, we believe the lack of significant ethnic and cultural diversity in our participant cohort signalled that we had not used story completion in culturally appropriate ways nor paid enough attention to the recruitment strategy to engage more diverse respondents. This led us to develop recommendations to decolonise story completion. A key recommendation is that story completion research be codeveloped in collaboration with individuals with lived experience relevant to the research topic.27

Data collection

We set up a submission point using Qualtrics. We outlined consent and ethics information, followed by the stem, and demographic questions. We began participant recruitment in May 2020. At the time, various COVID-19-related restrictions in place across all states and territories were beginning to ease slightly. For example, in Australia’s most populous state, New South Wales, strict restrictions aimed at limiting the spread of COVID-19 began to relax in May 2020. For instance, a household was allowed up to 5 visitors at one time, up to 10 people could gather outdoors, and cafes and restaurants could seat a maximum of 10 people. However, various restrictions on physical distancing, travel across state and territory borders, and international travel remained in place.25

We shared the Qualtrics link via Twitter, seeking to recruit people living in Australia aged 18 years or older. Participation was not incentivised. We also posted a Facebook invitation on the Black Dog Institute web page (a specialist mental health research institute), after which we saw a net increase in submissions. We chose these recruitment strategies for a rapid response and to attract a diverse cross-section of respondents. When the survey was closed after 4 weeks, we had 52 stories. We did not set a word limit, hoping that it would encourage respondents to be creative with the stem.

To frame each participant’s storytelling choices in our analysis, we collected demographic data using a mix of predetermined and open responses. We wanted to understand who responded to the call to participate and what specific markers of identity might reveal about how people told stories about COVID-19. We provide demographic information in the Results section.

Sample size

Because of likely variation in quality and length of stories, recruiting at least 10 participants per story stem is recommended.28 However, story completion studies typically recruit sample sizes of 40–60 participants.29 30 We are confident that our 52 stories represent a robust corpus of data for analysis.

Analysis

We used reflexive thematic analysis,28 an approach widely used in story completion analysis and in qualitative research more generally.12 31 This analytical lens positions researchers as active agents in the process and requires reflection on the assumptions and experiences they bring to the analysis.32 CL and KB are senior researchers and have extensive experience using arts-based qualitative methods in the context of participatory sociological and psychological research on mental health and social inclusion. PV is an early-career, postdoctoral researcher who trained as a social anthropologist and has also used arts-based research methods to collect data on mental health. This was the first time members of the research team had used story completion as a research method. We have used thematic analysis extensively in previous work.33–35 We interpreted the data using a constructionist approach to story completion, as each story reflects the participant’s social context.11 36 37

The data analysis process involved the following phases. A detailed overview of this process is described elsewhere.38

Familiarisation: all researchers reviewed all stories to get a sense of the data set as a whole.

Initial coding: PV read all the stories and CL and KB read half each, so that each story was read at least twice. We recorded our initial thoughts using two headings: (1) ‘Are there particular storylines? What are the ‘turning points’ or what Fels calls ‘tug on the sleeve’ moments?39 Are there repeated patterns of meaning?’ and (2) ‘Key themes’.

Identification of themes: we identified the most prominent or important elements in the stories before engaging in collaborative analysis to produce an initial set of themes (or codes).

Refining themes: we collaboratively reviewed themes and iteratively refined our coding structure to identify the overarching themes which characterised the data set. This led to the identification of four meta-themes described in the Results section.

Results

Participants produced 52 stories. The length of each story varied from one sentence to a lengthy paragraph. Typically, short stories covered the next step or action the protagonist would take in immediate response to the situation outlined in the stem. Longer stories outlined a narrative arc with a beginning, middle and end (orientation, complication, resolution), or contained reflections on the situation the protagonist faced.

Demographic data

Demographic data collected indicated a relatively homogenous participant composition, particularly regarding ethnic or cultural background (see table 1 for details). Not all participants responded to all demographic questions. Given the relative homogeneity of the cohort and small sample size, demographic data did not significantly enhance our analysis but provided contextual information about who participated and who had not, which impacted our thinking regarding future story completion research. We discuss this in greater detail in the Strengths and Limitations section.

Table 1

Participant demographics

Demographics	Participants, n (%)
Age
18–24	2 (4)
25–34	20 (39)
35–44	14 (27)
45–54	7 (14)
55–64	6 (12)
65–74	2 (4)
Gender
Female	42 (84)
Male	5 (10)
Non-binary	1 (2)
Trans	2 (4)
Highest education level
Completed some postschool technical training or university study	4 (8)
Completed university degree	24 (47)
Postgraduate degree	23 (45)
Marital status
Married/de facto	30 (67)
Single	15 (33)
Race/ethnicity*
Aboriginal	1 (2)
Anglo	10 (20)
Asian	2 (4)
Australian (including Asian Australian, non-Indigenous Australian, Irish-Catholic Australian)	11 (22)
Caucasian	13 (26)
Chinese	1 (2)
European	3 (6)
Jewish Anglo	1 (2)
Mixed (including Anglo/Sri Lankan)	3 (6)
The human race	1 (2)
White	4 (8)
Country of origin
Australia	39 (78)
Other countries†	11 (22)
Language(s) spoken at home
English only	47 (94)
English and other language(s)‡	3 (6)
Employment status
Full-time employment	27 (53)
Part-time employment	16 (31)
Unemployed	7 (14)
Volunteer	1 (2)
Residency
Major city	42 (91)
Regional	4 (9)

*Categories based on language used by participants.

†Other countries reported: China, Germany, Malaysia, New Zealand, Pakistan, Sweden, UK and USA.

‡Other languages reported: Dutch, Polish, Punjabi, Spanish and Urdu.

Meta-themes

We identified four meta-themes across the stories:

Expressions of mental distress linked to COVID-19.

Various coping strategies offered by Ali and other characters in stories.

Narratives outlining social support offered to alleviate distress.

Specialised COVID-19 vocabulary.

In the following overview of these meta-themes, we explore their prevalence across stories. Moller et al5 note that reporting on ‘frequency counts or percentages’ when using story completion ‘may be seen as controversial (eg, antithetical to qualitative values)’. However, given our interest in archiving a specific moment in Australia relating to a public health crisis, we felt that a top-down focus, which captured commonalities and differences in experiences, was imperative.

Mental distress

We tracked expressions of mental distress that Ali, the parent or others conveyed in the stories. Almost all (96%) stories included implied or explicit discussions of experiences of mental distress. While the stem mentioned that Ali’s parent was distressed, participants could choose how they developed their narrative and how much emphasis they placed on this element. We identified the language used in stories rather than assigning diagnostic categories to the texts to understand how participants described a character’s state of mind. Some participants explicitly named the type of mental distress a character suffered, while in other instances we made inferences about mental distress experienced. For example, we inferred that the following was an expression of anger: “F*** off Mum, she thinks. F*** off Everybody. F*** off COVID”. We used an iterative process to classify these expressions by counting mentions associated with mental distress. Many stories featured more than one form of distress. We then assigned these to 19 top-level categories. We sought to retain participant language when creating meta-categories and combined only closely related terms under each overarching category.

The most expressed concern was worry (41% of concerns), mostly related to the possibility that Ali or their parent would contract COVID-19. Other expressions of distress related to government-imposed physical and social distancing provisions, leading characters to feel lonely, isolated and angry. Table 2 summarises expressions of mental distress that account for more than 5% of those counted. Expressions of distress appearing in fewer than 5% of stories were upset, frustration, panic, anger, anguish, trapped, helplessness, isolation, paranoia, stress, suicidal ideation, uncertainty and unease.

Table 2

Mental distress

Mental distress	Example	Percentage of count
Worry	‘Ali is tired, it’s been a long day trying to work from home with small kids, but they listen carefully to all the worries of their dad’.	41
General implication of psychological discomfort*	‘Ali remains calm and tries to get to the root of the problem and lends a listening ear. She redirects one of the concerns towards practical solutions to calm the parent and reminds the parent of some trusted neighbours and friends nearby and commits to checking back in one hour’.	10
Fear	‘She lives in an assisted care home and has been reading the news about the deaths in other nursing homes. She is understandably scared…’.	8
Loneliness	‘…She’s lonely, none of her children are there, this is not something she says often but Ali knows this to be true. The call is about COVID-19 on the face of it but she just wants to talk to someone’.	8
Anxiety	‘…his mother is anxious about running out of toilet rolls and not getting basic things from the grocery store’.	6
Distress	‘Ali gets distressed with his mother’s distress…’.	5

*This refers to instances when no specific details of mental distress were articulated in stories, but distress was implied.

Coping strategies

Seventy-nine per cent of stories featured coping strategies of some kind and 66% featured more than one strategy. Strategies were articulated by characters in the stories, typically to allay the distress of Ali’s parent. We used an iterative process and counted mentions of coping strategies, then assigned these to a set of top-level strategy types. The most common strategy was talking (29% of strategies). Talking as a coping strategy underpins the story stem, which begins with a phone call to Ali. Many stories began with Ali verbally reassuring their parent. Table 3 outlines the coping strategies accounting for more than 5% of strategies counted. Strategies accounting for less than 5% were cognitive (eg, changing and challenging patterns of thinking), food and drink, limiting news consumption, planning for the future, caring for others, exercise, professional support (eg, consulting a general practitioner or therapist), and sleep.

Table 3

Coping strategies

Coping strategy	Description of strategy	Percentage of count
Talking	Talking through worries; Ali providing verbal reassurance; talking about things beyond COVID-19.	29
Socialising	Planning to, or undertaking, social interaction.	12
Prevention strategies	Discussing strategies to limit risks of catching COVID-19 to limit parental worry.	11
Information	Getting information to combat worry and anxiety, including engaging with reputable sources to bust COVID-19 myths (eg, COVID-19 is caused by 5G network).	10
Instrumental	Practical support Ali proposed to alleviate stress (eg, offering to shop for parents).	10
Relaxation activities	Watching movies, reading books, gardening or getting fresh air.	8

Social support

Eighty-five per cent of stories included Ali providing or offering social support to their parent. We did not track support offered to Ali by their parents as these appeared only infrequently in stories. Arguably, Ali’s ability to provide support underpins the story stem, as Ali’s parent calls while distressed. We counted explicit mentions of support articulated in stories. In health research, various social support categories are used, including instrumental, emotional, tangible, affectionate, positive interactions or appraisal.40–43 We used three broad social support categories: (1) emotional: expression of empathy, love, trust, care and support; (2) instrumental: tangible service or help; and (3) informational: advice, suggestions and information. Table 4 records the presence (or copresence) of social support in stories. Emotional support was by far the most prevalent. We note that coping strategies and social support sometimes overlapped.

Table 4

Social support

Support type(s)	Example	Percentage of stories type appears
Emotional	‘…. Ali speaks to her mother for a few minutes, calming her down and reassuring her that she will be ok…’.	30
Emotional and informational	‘Ali assures her Mum that she is listening and encourages her Mum to share what is worrying her… Ali sits at her laptop while she talks to her Mum and looks up trusted sources of medical information (WHO, Aus Govt, mydr.com.au) to help her Mum, and herself, understand what we know’.	18
Emotional, instrumental and informational	‘Ali seeks to reassure her parent that if he/she follows current health advice that the risks to this parent can be minimised… Ali is mindful of his/her parents’ social isolation during COVID-19 and is conflicted by the need to visit to help and reassure his/her parent and to remain isolated to ensure risks are minimised… Ali contacts the Government COVID-19 assistance line for advice. The following day, Ali arranges a Telehealth consult with his/her elderly’s parent to ensure parent is informed on how best to look after themselves, and to arrange a flu shot’.	16
Emotional and instrumental	‘It will be okay, we can help. What do you need, we can get you groceries, there will be enough food, what’s in your pantry, we can send you toilet paper and hand sanitizer… We will take care of you. We love you’.	11
Informational	‘…Ali’s mum then goes on to explain she saw on Facebook how 5G is linked to increased cases… Ali rubs his eyes and explains that you should never believe what you see and read on social media…The next morning he finds some credible sources explaining there is absolutely no link between Corona Virus and 5G and send them to his mum’.	11
Instrumental	‘…Ali drove to his parent’s house, bringing with him some medicine. After looking at the nearest drive-through testing centre, he brought them to be tested the next day’.	9
Instrumental and informational	‘…Ali, now wide awake, opens her laptop, googles the symptoms of Corona and asks her mum if she is suffering from any of these symptoms…Ali will check on her in the morning’.	5

COVID-19 vocabulary

Stories contained specific language and terminology that became more prominent among the general public due to the COVID-19 pandemic. We compiled a COVID-19 vocabulary list drawing on news and web articles on COVID-19 terminology and vocabulary.44–50 We counted 24 COVID-19 vocabulary words or expressions (and variations) in 65% of stories. We counted the number of stories which featured specific words from the COVID-19 vocabulary list (eg, if a single story used the word Covid three times, this was counted as one story featuring that word). The authors took context into consideration when tracking words use. For example, isolated was taken to be an example of COVID-19 vocabulary if it related to public health orders about isolating to prevent the spread of COVID-19: ‘she’s in a place where she can self-isolate’. However, we did not count isolated COVID-19 vocabulary if it appeared in contexts like the following: ‘Ali’s mother is feeling isolated and lonely’. Table 5 lists the terms that appear in 5% or more of the stories. Other words counted (that appeared in less than 5% of stories) were face mask, restrictions, washing hands, 1.5 metres, hand sanitiser, telehealth, Zoom, asymptomatic, cluster, COVIDsafe app, death tolls, drive-through testing, essential activities, front-line healthcare worker, incubation period, intubate, panic buying and safety measures.

Table 5

COVID-19 vocabulary

COVID-19 term	Related variations	Percentage of stories term appears
COVID-19	COVID, corona virus, corona, rona	30
Isolation	Isolate, self-isolate, self-isolated, self-isolation	10
Social distancing	Socially distanced, socially distancing, distancing, distance from people, kept our distances	8
Lockdown		5
Pandemic		5
Vulnerable group	Vulnerable age group, vulnerable age bracket, high-risk category	5

Discussion

Story completion has been used in psychology, health and social research to learn about personal crisis and health concerns. It is emerging as an effective method to engage with COVID-19, ‘a crisis on a much larger scale’.26 To our knowledge, this is the first study to report on data generated from story completion with a COVID-19 scenario. We acknowledge Braun et al’s26 work describing their methodological approach using story completion during the pandemic. In our research, the story completion method garnered a rich array of engaging and moving stories, with insights into individual knowledge about, and experiences of, pandemic-related language, mental distress and strategies to alleviate discomfort and offer social support. The stories stand as a creative, textual archive of the pandemic and mental health and well-being concerns as experienced in Australia in early 2020.

Mental distress

Preliminary research in Australia indicates that the pandemic and associated health risks, social restrictions and economic impacts have adversely impacted the mental health of the general public and those with pre-existing mental health diagnosis.19 20 The full mental health impact of the pandemic is yet to become fully apparent. Preliminary research on the topic should be followed up with rigorous longitudinal studies.51 In our research, creative, narrative expressions seemed to imitate real life, with most stories centring around experiences of psychological anguish. Mental distress caused by fear of COVID-19 and the emotional and psychological impact of hearing from a loved one in distress formed the narrative core of most stories. This correlates with findings from narrative qualitative research undertaken in Australia which documented feelings of vulnerability and anxiety due to the threat of catching COVID-19 and hearing about the impact of COVID-19 on others.52 The stories clearly conveyed (sometimes viscerally) the psychological suffering wrought by COVID-19 due to its highly infectious nature and the fatalities that resulted from the virus.

Social support and coping strategies

While mental distress permeates most stories, respondents also included various social support and coping strategies. Ali and their parent offered, recommended and practised diverse coping strategies. Most were associated with positive self-care approaches such as recreation, socialising or exercise, rather than maladaptive coping strategies such as shouting, smoking or binge drinking.53 This suggests that participants had a relatively high level of mental health literacy and were aware of healthy approaches to managing psychological discomfort. This may be due to our recruitment approach (ie, distribution via the Black Dog Institute and our social media networks).

There is an emerging body of small-scale research assessing coping strategies in response to COVID-19. Socially oriented coping strategies, as well as mindfulness or positive thinking, emerge across this literature as productive and healthy approaches to cope with COVID-19-related distress.54–56 While we identified these coping strategies across stories, respondents also suggested additional practical approaches including strategies to limit the risk of contracting COVID-19, reading reputable material to learn facts or combat myths about the virus, or engaging in practical actions to alleviate anxiety (such as Ali shopping for their parent). The characters sought out or offered a combination of cognitive, emotional and practical support or coping strategies to address distress. Again, this suggests a good level of mental health literacy and shows awareness of stressors beyond the threat of catching COVID-19, such as increased news consumption and ‘doom scrolling’;57 the flourishing of COVID-19 myths, misinformation and conspiracy theories;52 58 59 and panic related to supermarket shortages.60

COVID-19 vocabulary

The language used in stories largely reflected public health messaging at the time of data collection. New words had entered the lexicon among the general public, likely due to increased exposure and attention to the 24-hour news cycle. Indeed, news media was mentioned in 13 stories (~8%) as a source of both information and anxiety. It is likely that constant engagement with media caused new terms to permeate everyday conversations. This uptake of a COVID-19 vocabulary assisted respondents to express their fears using commonly understood language that can unite people ‘around a set of collective cultural references points [like a]…lexical ‘social glue’’.48 The ubiquity of COVID-19 vocabulary in stories reflects community engagement with media and health messaging. As Rudd and Baur61 report, ‘communication messages and materials related to COVID-19 have introduced unusual vocabulary and phrases into common speech over the past several months…over a short time period these previously rare words have become commonly understood’.

Conclusion

We cautiously suggest that points of convergence across stories indicate a shared experience among participants resulting from intensive media coverage, persistent public health messaging, engagement with social media and instant messaging technologies, and extended lockdowns that impacted the mental health and well-being of vast numbers of Australians. However, we cannot do more than cautiously speculate. The reason for this caution is outlined in the following section.

Strengths, limitations and directions for future research

A strength of the story completion method and what makes it a potentially empowering research tool is that it engages participants in a hypothetical, creative and speculative manner to reveal patterns and processes of sense-making and social experience. It asks participants to imagine themselves in a situation but does not demand that stories only reflect their lived realities.5 In our research, this methodological strength has resulted in a limitation regarding how much we can infer about data collected. Researchers must be cautious about making sweeping assumptions regarding the relationship between story content and respondents’ realities. While we agree that story completion has the potential to reveal much about participants’ social, personal and psychological worlds, more information was needed to undertake this form of extended interpretative analysis within our data set. This does not diminish the method’s potential in illustrating how narrative enquiry can elicit and capture social knowledge, mores, vocabulary and experience. In future research, it may be effective to pair story completion with complementary methods such as one-on-one interviews10 to examine, for example, the gap between knowledge of positive mental health coping strategies and their uptake or practice. The combination of arts-based or narrative-based methods with oral qualitative data collection methods has worked successfully in our previous research.62 63

Another limitation of this study was the lack of significant ethnic and cultural diversity in our cohort. This perhaps signals that we did not use story completion in culturally appropriate ways nor paid enough attention to the recruitment strategy to engage more diverse respondents. For example, our stem (which we developed in line with the accepted story completion approach) privileged Western modes of story construction, with participants being asked to complete a story by providing a middle and end to the beginning offered by the stem.64 This approach may not have appealed to those familiar with other storytelling approaches grounded in distinct sociocultural contexts.65 A lack of diversity in our sample may also have been the result of our recruitment strategy via social media.66 In the future, these shortcomings could be addressed through the collaborative co-creation of a stem with a diverse participant advisory group and via deliberative recruitment to ensure cultural diversity. These considerations have led us to develop recommendations to decolonise story completion.