Unpacking qualitative data in organizational trust research: an application to community appraisal of COVID-19 management in Scandinavia Open Access

Extensive advances have been made in organizational trust research in recent decades. There have been developments in the understanding of intra-organizational trust involving management and employees (e.g. Gillespie and Dietz, 2009; Gustafsson et al., 2021; Gillespie et al., 2021; Grey and Garsten, 2001), inter-organizational trust involving collaborating organizations (Kostis et al., 2022; Owen and Currie, 2022; Zaheer and Harris, 2006), and extra-organizational trust, such as community trust in institutions and other stakeholder relationships (Eberl et al., 2015; Spicer and Okhmatovskiy, 2015; Liu et al., 2022). Across these areas of organizational trust, there is research asserting that trust is communication-based (e.g. Shockley-Zalabak et al., 2000; Candlin and Crichton, 2013; Kosonen and Ikonen, 2019; Vokić et al., 2020). Thus, an emerging approach to organizational trust emphasizes its dynamic, interactive nature, conceptualizing it as an ongoing process shaped and negotiated through communication, rather than a static trait or solely a cognitive judgement.

This communication- and interaction-cantered research on organizational trust employs diverse qualitative approaches – such as interpretive, thematic, narrative, and grounded theory analyses – offering detailed insights into trust formation, often through case studies (Kodish, 2014; Kosonen and Ikonen, 2019; Kähkönen, 2021; Hasche et al., 2021). However, qualitative studies that focus on case-specific thematization and thick description limit their engagement with shared methodological questions, complicating efforts to synthesize findings across studies and potentially reducing the broader impact of qualitative research on organizational trust (Guo et al., 2017). Common methods, such as qualitative coding, often prioritize clear examples over ambiguous statements, highlighting the challenge of analysing complex data, where identifiable patterns tend to take precedence over nuance (Williamson et al., 2020; Gjerde and Alvesson, 2019). Moreover, while the existing literature acknowledges the interplay between communication and trust, the systematic application of communication frameworks to analyse trust dynamics is limited. These challenges highlight the need for more active discussions on methodological issues and collaborative efforts to develop systematic approaches for examining communicative realizations of trust.

Theoretical developments likewise warrant methodological refinement, such as empirically understanding how we should “capture” individuals simultaneously affirming and critiquing an organization and constructing both trust and distrust (Lewicki et al., 1998; Kostis et al., 2022). Moreover, trustful relationships can facilitate control and power, resembling dynamics like loyalty and groupthink, which demonstrate that trust and power are not opposites, nor is trust always inherently positive (Grey and Garsten, 2001; Stevens et al., 2015). Additionally, the sheer volume and complexity of qualitative data often make it appear “messy” and difficult to manage (Parkhe, 1993). This is particularly evident in issues of agency and responsibility, which are frequently vague or obscured in discourse, complicating their analysis (van Leeuwen, 2008). These challenges also emphasize the need to articulate and refine methodological approaches to address such complexities effectively.

In making a methodological contribution in this direction, this paper draws on appraisal analysis, a strand of discourse analysis that frames stances on trust as expressed through evaluative language and communicative appraisal (Martin and White, 2005; Thompson and Hunston, 2000). It helps the paper respond to a call for deeper engagement with the discourse of trustors – those who evaluate trustworthiness and construct trust – addressing a gap that can sometimes lead to conflating trustworthiness with trust or assuming that trust naturally follows from trustworthiness (Möllering, 2019). By shifting the focus to trustors, the study also complements prior organizational research that has used discourse analysis to investigate the trust-building strategies of trustees (e.g. Pelsmaekers et al., 2014; Fuoli and Hart, 2018; Fuoli and Paradis, 2014; Fuoli, 2012; Jackson, 2014).

Building on these foundations and logics, the main research question driving this paper is: How can appraisal analysis contribute to a framework for investigating how trustors construct attitudinal positions on trust through evaluative language, with particular attention to trust-distrust dynamics and power? Consequently, this study aims to make a methodological contribution to qualitative research on organizational trust by focusing on the conceptualization and demonstration of analytical tools for extracting meaning from qualitative data – primarily text, spoken language, and transcripts of recorded speech (Phelps et al., 2007).

In the following, brief reviews of organizational trust research and appraisal analysis are provided. Next, the Scandinavian case of COVID-19 management is described, which the analysis section uses to outline key methodological points. The discussion and conclusion section summarizes the main methodological takeaways and suggests avenues for further qualitative research into organizational trust.

This section highlights four developments in organizational research on trust and distrust: (1) the increasing attention to multilevel and cross-level perspectives; (2) the understanding that trust and distrust can coexist in complex and interrelated ways; (3) the acknowledgement of how discourse and power relations influence these dynamics; and (4) the critique of the view that trust is inherently positive and distrust inherently negative.

To begin with, there have been extensive efforts in the past few decades to develop a more cohesive and multilevel understanding of organizational trust. Scholars have noted that trust research has often been isolated within specific domains – such as psychology, sociology, political science, organizational behaviour, and economics – and that contributions on trust within organizations, between organizations, and with external stakeholders have remained separate, without being effectively integrated to address organizational trust in a more unified and comprehensive manner (Mayer et al., 1995; Rousseau et al., 1998; Schoorman et al., 2007). This paper’s focus on contributing methodologically to the study of organizational trust dynamics aligns with the ambition of the aforementioned research to integrate multiple research streams, rather than focusing on just one. Hence, although this paper is based on data that pertains to extra-organizational trust – a subfield that has demonstrated how stakeholder trust in organizations underpins supportive behaviour and goodwill (Shahid et al., 2021) and enhances collaborative partnerships, thereby mitigating weaknesses and maximizing opportunities for growth (Heugens et al., 2002) – the aim is to identify patterns in the discourse of external stakeholders that could also emerge among employees within an organization or in the collaboration between two organizations. Indeed, the goal of methodology is to create frameworks or approaches that can be applied across a wide range of organizational trust contexts, providing insights that extend beyond specific cases or narrow domains.

Second, it is also notable that conceptualizations of trust and distrust are becoming more complex, with an increasing emphasis on understanding how both trust and distrust can co-occur. This marks a departure from the common binary view of trust and distrust as opposite ends of a spectrum, where the presence of one implies the absence of the other, and where trust increases as distrust decreases and vice versa (Walle and Six, 2013). While useful for studies using survey scales, it has been argued that this binary view oversimplifies the complexities of human sensemaking and social interactions (Lewicki et al., 1998). In response, some scholars focus on the dynamic interplay and potential coexistence of both trust and distrust (Walle and Six, 2013; Woolthuis et al., 2005).

This nuanced understanding aligns with the observation by Guo and colleagues (2017) that trust and distrust encompass distinct beliefs and practices. According to their analysis, trust is characterized by the belief that leaders, colleagues, or organizations will act in beneficial and reliable ways, reflected in practices such as open communication, collaboration, and behaviours fostering support, safety, and success. Conversely, distrust involves an active belief that others may act against one’s welfare and progress, evident in practices that imply negligence, incompetence, recklessness, or dishonesty. Thus, trust and distrust are multidimensional constructs that may overlap or remain distinct, depending on the context.

In line with this, Kostis and colleagues (2022) have shown how stakeholders can express trust in an organization, team, or individual in certain respects but mistrust them in others or readily modify a standpoint once they have reason to do so. A trustee’s abuse of trust results dynamically in a more critical stance on the part of the trustor, but distrust proved wrong also calls for a reorientation. This interplay of trust and distrust shows that professional relationships are conditional and imbued with demands on the participants.

Third, it is also notable that more research assumes that organizational trust and distrust take shape and are articulated through discourse and in power relations. According to Fuoli and Paradis (2014), discourse includes the structured ways language is used to construct, communicate, and negotiate trust dynamics, encompassing explicit expressions, narratives, conversations, and underlying assumptions that shape beliefs and behaviours. This discourse-centred approach aligns with Grey and Garsten's (2001, p. 229) definition of trust as “a precarious social accomplishment enacted through the interplay of social and discursive structures.” Building on this foundation, qualitative studies can explore how trust-distrust dynamics are continually constructed and reshaped through organizational discourses, information flows, and discourses available in the broader society. Research in this vein has explored, for instance, how organizational members are “doing” trust through the discursive practices of believing, relying, committing, reframing, recognising, and disclosing (Kostis et al., 2022) and by attesting that there is alignment between words and actions in the areas of competence, benevolence, and integrity, with actions validating the truthfulness of the words (Kodish, 2014).

Closely linked to a discourse-oriented understanding of trust and distrust is the recognition that these dynamics are shaped by power relations, and that trust may not stem from conscious, rational choices (Grey and Garsten, 2001; Hardy et al., 1998). This departs from research that ties trust to rational motives, such as perceptions of ability, integrity, and benevolence (Mayer et al., 1995). Ability reflects the trustor’s view of another’s knowledge, skill, and power to complete a task, while integrity involves perceptions of work ethic and honesty, and benevolence requires the belief that the trustee acts with care, not selfishly or recklessly. While these assumptions are valid – acknowledging that trust and distrust are grounded in cognitive patterns – there are also reasons to take a more critical stance. As Grey and Garsten (2001, p. 233) state, “trust is that form of control in which people behave predictably not despite, but by virtue of, the choices they make.” They thereby align with the Foucauldian stance that power is not only an exercise of strong-arm tactics, but also the ability to make things happen by sharing moral codes, knowledge, and aspirations with others so that they take on the responsibility of striving toward those same objectives and worldviews (Foucault, 1991). Trust is thus seen as a form of control in that it represents and assists predictable action and conformity.

Finally, this research moves away from the notion that trust is inherently good and distrust inherently bad. It aligns with earlier insights, highlighting that trust and distrust (1) can and must coexist, and (2) emerge within power dynamics. For example, Kostis and colleagues (2022) warn against uncritically accepting high organizational trust, noting that excessive trust may lead to inertia, path dependency, and unwarranted optimism. However, both excessive and insufficient trust carry risks. Stevens et al. (2015, p. 1237) assert that “reorientation is necessary when an imbalance occurs between the powerful and opposing forces associated with excessive trust (such as faith, favouritism, contentment, and loyalty) and insufficient trust (including scepticism, impartiality, exigency, and opportunism).” While institutional research often highlights the risks of low trust, such as economic and societal losses (Habibov et al., 2018), these do not justify a naïve belief in high trust. As Stevens et al. (2015) argue, flexibility in problem-solving, allowing for diverse orientations toward trust and distrust, is more likely to engage broader networks and competencies, increasing opportunities for success.

The study uses appraisal analysis to examine how trust and distrust are expressed in qualitative conversational data, focusing on evaluative language. Appraisal analysis is a linguistic framework that explores “how writers/speakers approve and disapprove, enthuse and abhor, applaud and criticize, and how they position their readers/listeners to do likewise” (Martin and White, 2005, p. 1). Key concepts in appraisal analysis are outlined below and can be applied in a stepwise approach (presented in the next section) to uncover layers of discourse and meaning worthy of investigation.

A key aspect of appraisal analysis is the distinction between three types of evaluation: attitude, graduation, and engagement (Martin and White, 2005). The analysis of attitude examines how speakers articulate their feelings, assessments, and emotions about people, actions, and situations, making it a valuable methodological asset given the emphasis organizational research places on the emotional dimensions of trust and distrust (Lee et al., 2023). It is divided into three subcategories: appreciation, judgement, and affect. Appreciation does not directly evaluate people but instead focuses on assessing actions, objects, or events based on their functional, aesthetic, or ethical qualities. For example, describing a leader’s speech as “crystal clear” or “inspiring” implies trust without explicitly evaluating the individual, with the person’s trustworthiness inferred from the context. Even though appreciation may seem to focus on actions rather than individuals, it still serves as a resource for conveying trust or distrust.

In contrast, judgement is the explicit evaluation of individuals, groups or organizations, based on personal or cultural standards. Judgement features assessments of people on dimensions of social esteem (normality, capacity, tenacity) or sanction (veracity, propriety). For instance, a statement that praises a person’s competence or criticizes their truthfulness directly signals trust or distrust. Judgement, therefore, plays a central role in determining how social actors are evaluated in terms of their integrity, capability, or ethical standards.

The third subcategory, affect, refers to the emotional responses elicited by social actors, actions, or events. Affect is expressed through language that conveys emotions like happiness, sadness, fear, or hope. For example, describing someone else’s statement as “heart-breaking” or “uplifting” reflects the speaker’s emotional engagement and can influence how trust or distrust is constructed. Affect enhances both judgement of people and appreciation of actions or events, further shaping the communicative dynamics of trust.

Beyond attitude, appraisal analysis also examines graduation and engagement, two critical resources for understanding that discursive invocations of trust and distrust can themselves be complex and ambiguous acts. This area has received little attention in management and organization studies, where research typically asserts that trust and distrust reduce complexity and ambiguity by enabling social actors to establish mutual expectations rather than remain in a state of uncertainty (Guo et al., 2017). Attending to the discursive resources of engagement and graduation can therefore provide valuable insights. Graduation refers to the way speakers adjust the intensity or force of their evaluations to communicate varying degrees of certainty, positivity, or strength. Graduation includes gradable evaluations (e.g. somewhat, quite, very, extremely) quantifiable evaluations (e.g. few, some, many, all) and non-gradable evaluations (“she’s sort of a friend” or “she’s a real friend”). By modulating the intensity of their language, speakers can signal the strength of their trust or distrust, indicating how strongly they feel about a particular actor or action.

Engagement, on the other hand, relates to how interlocutors incorporate and engage with voices, viewpoints, and positions within their discourse. Communication is inherently dialogic, as speakers use engagement resources to construct their positions while accounting for others’ opinions or previous discussions on the subject. Engagement allows speakers to either contract or expand the range of perspectives they acknowledge. Monoglossic statements present a single, definitive viewpoint, leaving little room for alternative interpretations, while heteroglossic statements introduce multiple possible positions, allowing for a broader discourse. In the context of trust, monoglossic utterances may assert one clear stance on a person’s trustworthiness, while heteroglossic statements might recognize the complexity of trust dynamics and open the conversation to multiple interpretations.

Appraisal analysis thus provides a valuable framework for understanding how trust is articulated and negotiated in discourse. By utilizing attitude, graduation, and engagement, speakers convey evaluations of ability, benevolence, and integrity, calling on others to agree, and navigating the social context. Although underused in organizational trust research, some studies have applied it to explore how corporate discourse shapes perceptions of credibility, responsibility, and identity (Fuoli and Hart, 2018; Fuoli and Paradis, 2014; Fuoli, 2012).

However, while existing research focuses on trustees, more attention is needed on trustors – those making judgements about trust (Möllering, 2019). Additionally, organizations are influenced by both internal processes and extra-organizational trust, where stakeholders assess the trustworthiness of organizations and leaders (Kuhn, 2008). This paper responds to these calls, illustrating how qualitative analysis can integrate recent theoretical insights into empirical observations.

The COVID-19 crisis prompted this study’s focus on responsible public institutions, where extra-organizational trust is essential for maintaining legitimacy and implementing safety measures requiring cooperation. As Schoorman and colleagues (2007) suggest, such cases offer lessons across disciplines and sectors. Public health agencies and government actors, while distinguished by their socio-economic missions and historically persistent structures, share similarities with other organizations in their reliance on hierarchical personnel, strategic communication, and trust from internal and external audiences (Scott, 2014). A limitation of this case, however, is the inherent distance between communities (trustors) and institutions (trustees), with community appraisals often reliant on media and news. Although this contrasts with studies on trust within organizations, large multinational corporations face comparable challenges of distance and indirect access, as their top management also primarily interacts with employees and stakeholders through media in a globalized and digitalized world (Cortellazzo et al., 2019). So, while this is a limitation, it is not entirely unique in organizational trust research.

The COVID-19 crisis provided a compelling case for studying extra-organizational trust, as it highlighted how public institutions navigated trust challenges during a global emergency (Liu et al., 2022). To explore these dynamics, focus group interviews were conducted across Scandinavia (Norway, Sweden, and Denmark), capturing a range of community perspectives on institutional trust. These countries share similar political systems, welfare services, and cultural norms, along with high levels of institutional trust, but differed in their pandemic strategies: Sweden implemented fewer collective prevention measures, relying on behavioural recommendations, while its neighbours adopted suppression strategies like much of Europe (Rasmussen et al., 2023). These differing approaches offered a unique opportunity to examine varied public reactions.

The focus group method was chosen for its ability to foster interaction and encourage the exploration of diverse value positions (Bloor, 2001), making it particularly suitable for combining with appraisal analysis. It has also been effectively used to examine expressions and realizations of trust. For example, White-Cooper et al. (2007) highlight that focus groups can gather comprehensive data on beliefs and values, enabling the observation of collective opinion formation. Empirical evidence from focus groups shows their capacity to inform our understanding of trust, as they allow for the exploration of participant attitudes toward trust and its determinants in collaborative settings (White-Cooper et al., 2007). However, as George (2012) points out, concerns about the reliability and validity of data obtained through focus groups remain, with moderators’ influences potentially skewing interactions, as well as challenges such as dominant viewpoints and participant hesitancy (Bloor, 2001). To mitigate these risks, we took care to structure the focus group discussions (see  Appendix, for the interview guide) in a way that encouraged equal participation and minimized the potential for bias, utilizing open-ended questions to further promote a balanced and inclusive dialogue.

The interview guide was collaboratively developed by a team of seven project researchers, with questions focused on pandemic risk and crisis management and not on sensitive issues like personal health. Due to social distancing restrictions, synchronous online focus groups were conducted (Bouvier and Rasmussen, 2022). To capture context-specific meanings, professional research assistants fluent in each country’s language conducted 21 focus groups with 168 participants across three Scandinavian countries. As seen in Table 1, the focus groups were organized with participants grouped closely by age (roughly young adults, adults, middle-aged, and seniors) while maintaining a mix of genders.

Participants also varied in employment, education, and social class. Group size ensured diverse opinions while allowing enough time for active engagement. Each interview lasted about two hours. The analytical examples were translated into English by the author. Participants provided informed consent and were anonymized to ensure their views could not be directly or indirectly linked to any individuals, in accordance with research ethics protocols. The names used in the findings section are fictional.

With the focus group interviews completed, the analysis cantered on how citizens appraise risk-governing bodies’ attempts at managing risk and how they signal trust, distrust, or combinations thereof. The analysis involved the author of this paper repeatedly and closely reading the interview transcripts while applying all main concepts from appraisal analysis, which are described in more detail in the previous section. Additionally, other insights from social semiotics on the analysis of social actors were considered, such as the use of categories and names as resources for constructing agency and responsibility, as well as agentless discourse (nominalisations) which may obfuscate those same issues (van Leeuwen, 2008). When sections relevant to an analysis of trust were identified, notes were made line by line, completing each step listed below:

1. Identify attitude: Determine the expressions of attitude in the text, categorizing them into affect (tokens of emotions), judgement (evaluation of people), or appreciation (evaluation of objects, actions, or events).

2. Examine graduation: Analyse how the intensity or force of the evaluations is modulated. Look for discursive intensifiers or downtoners that amplify, reduce, or scale the strength of the expressed attitudes.

3. Analyse engagement: Investigate how the speaker positions themselves in relation to other voices or viewpoints using monoglossic (single stance) or heteroglossic (multiple stances) statements.

4. Analyse agency: Examine how the use of categories, names, and agentless discourse constructs or obscures agency and responsibility.

5. Contextual Interpretation: Interpret the interaction of these elements within the broader social and communicative context, exploring how they construct trust or lack thereof and embedded issues of power dynamics and responsibility.

Drawing on the above-mentioned concepts, the analysis procedure addressed theoretical insights in organizational trust research empirically, introducing methodological tools to address power dynamics, avoid rationalist leanings, and help disentangle the ambiguity and “messiness” related to agency and responsibility in qualitative transcripts. This procedure led to the conceptualization of six types of evaluation, which were shown to underpin stances on trust and distrust.

This analysis section addresses six types of evaluations used by social actors to form stances on organizational trust and distrust: simple-assertive, prescriptive, and mixed evaluation and appraisal centring on the qualities of actions, social actors, and outcomes.

The first point to note is that not all instances of trust or distrust involve justifications for the position taken. Therefore, when segments of communication are coded, as is common in qualitative analysis (Lee, 1998; Williams and Moser, 2019), not all relevant material fits neatly into categories based on rational justifications such as capability, integrity, or benevolence.

This observation is supported by the focus group conversations, where trust and distrust were occasionally expressed without explanations or additional information about actors’ characteristics, actions, or achievements. While such simple assertions may seem devoid of intersubjective meaning due to the lack of evaluative language, they frame trust or distrust as unquestioned, taken-for-granted positions. For example, when asked which risk-governing actors they trusted and why, one group of respondents in Sweden replied as follows:

As seen in Excerpt 1, these respondents align attitudinally with trusting the PHAS, and in one case, the government, using simple assertions to form their stance. It is also noteworthy that most invest quite strongly in the value positions they profess via resources of engagement and graduation. None signify competing positions, most add force and upscale their positive evaluation of the responsible authorities in Sweden (i.e. “great trust” and “full trust”), forming markedly monoglossic statements, and only Elias offers some kind of attitudinal justification (“as they are experts.”) Through these declarative statements, participants straightforwardly profess and reproduce trust in government authorities and invite others to share this stance.

Other respondents, however, align attitudinally with distrust in health authorities and treat such positions as self-evident instead. For example, one Danish respondent candidly stated, “I trust none of them.” A more extended discussion articulating variations of distrust took place during a Swedish focus group:

As highlighted previously, the respondents articulate their attitudes straightforwardly and without offering justifications. Adam begins with an affective statement suggesting trustworthiness is a fundamental expectation of the health authorities that many could share. However, he contracts this openness by countering with his own position of lacking trust (“but I can’t fully trust them either”). This juxtaposition opens space for agreement initially but ultimately reinforces his critique of the PHAS. Then, he engages dialogically with further voices, namely researchers, bringing them into the discourse, which bolsters his stance (“researchers who have criticized Tegnell and colleagues”). Drawing on resources of graduation, he scales his judgement, contrasting researchers with Tegnell and the PHAS, creating a hierarchy of trustworthiness (“The ones I trust the most” etc.)

What then follows is Jane boosting the perceived sincerity of her statement and rejecting trust outright, broadening the scope to encompass all potential sources of authority (“Honestly, I don’t know if I trust anyone”) leaving little room for nuance. Alice then amplifies the issue as pervasive rather than limited to specific contexts or relationships, attributing distrust broadly across institutions and groups, using metaphorical graduation which evokes a sense of dramatic decline (“Trust … has hit a low point”).

It needs to be clarified, however, that the use of simple assertions by respondents – or organizational members in another case study – does not suggest that these individuals are incapable of explaining or justifying their positions or that they refrain from doing so in other parts of an interview or conversation. The aim here is to emphasize that simple assertive evaluations represent a distinct category of trust/distrust discourse, comparable to justifying discourse, with unique characteristics, affordances, and meaning potential that warrant attention and analysis.

Another way respondents form positions on trust or distrust is by asserting a general obligation to adopt certain stances and actions. As Martin and White (2005) explain, these linguistic acts are prescriptive social acts in that they articulate that human behaviour ought to be directed and controlled. They endorse certain actions beyond the verbal exchange, constituting “actional” rather than “informational” discourse. This prescriptive evaluation can influence trust or distrust depending on the context, though in this study, distrust was more commonly observed. In this case, many citizens voluntarily participated in the government’s COVID-19 management efforts, reflecting the citizen–state relationship that Foucault (1991) describes as governmentality, where the governed partner with governing bodies. An example of this can be seen in an interview with Norwegian participants:

Similar assertions to trust and abide by authorities are directed at the political power in the following excerpt from a focus group conducted in Sweden:

These excerpts illustrate assertions of the obligation to trust and follow public authorities. In Excerpt 3, trust in authority is framed as a universal expectation for morally responsible citizens, using generic pronouns (“you,” “we all”) and assessments of obligation (“have to trust,” “must do what you’re told”). Affect is invoked as counterarguments are anticipated and refuted (e.g. “I hope people don’t get lazy”). In Excerpt 4, Dan similarly asserts that political leaders must trust health authorities, contrasting with Tim’s critique of perceived inaction by political leaders, which shifts responsibility to public health authorities in Sweden.

In Excerpt 3, participants form a trusting attitude by engaging dialogically with a presumed audience they position as susceptible to “false” perspectives and in need of convincing. This convincing involves articulating trust through prescribed attitudes, a form of discourse that can occur in various organizational or issue contexts. Actors position themselves on trust by urging others to trust leadership, follow prescribed behaviours, and expand the trustee to include the institution of democracy, where differences of opinion are less important if faith in common norms is upheld. In other sectors, this monolithic system may be the culture, brand, or esprit de corps.

Indeed, these prescriptive evaluations represent a distinct class of trust discourse, differing from the commonly studied justifying discourse but still contributing to the negotiation of trust, distrust, and authority.

Other manifestations of trust and distrust that risk being under-analysed in organizational research include statements with mixed evaluations. In qualitative analysis, where discursive segments are often coded into themes or topics (Williams and Moser, 2019; Lee, 1998), clear examples of meaning-making may take precedence over ambiguous ones, risking oversimplification and overlooking social dilemmas (Billig, 1996/1987). To address this, “codes” or “sub-themes” could be created to highlight these ambiguous segments. In the current study, some respondents, like Anja in Norway, expressed mixed attitudes toward risk management, suggesting that two contrasting COVID-19 strategies could both be valid: “I agreed with the measures, but at the same time, I also wondered if Sweden could be right in its assessments.” Semantic choices such as “but,” “at the same time,” and “on the other hand” signal these mixed evaluations.

Moreover, a recurring topic in this context is whether and how organizations can provide clarity and effectiveness in communication, both of which are closely linked to organizational trust (Zeffane et al., 2011). For example, one focus group participant said that the official information had been “both good and bad, inconsistent but sometimes informative” (Hilda, Sweden), and another participant claimed that official risk messages had been “… sometimes clear, sometimes very vague” (Lisa, Sweden). The statements point neither to complete trust nor distrust. Danish and Norwegian participants also substantiated complex positions on trust via positive and negative tokens of attitudes towards the actions of authorities:

Both respondents exhibit mixed attitudinal positions toward public organizations, and in terms of engagement they convey their own point of view with little dialogical expansion. First, Trine employs evaluative judgement, signalling moderate dissatisfaction with the diversity and potentially overwhelming nature of communication sources (“I think it is unfortunate …”), using graduation to scale the issue in terms of scope (“so many different messages …”). At the same time, she acknowledges the actors’ intentions, which mitigates potential distrust through an affirmative, graded evaluation (“everyone is acting to the best of their knowledge and understanding”). Similarly, Kaya also articulates both satisfaction and dissatisfaction, employing intensification to emphasize clarity in one aspect of her experience (“it was very simple […] Mette gave a clear announcement”), while contrasting this with the articulation of difficulty and widespread ambiguity (“so it’s up to each person to interpret it”), highlighting the potential for varied lay interpretations, which may undermine the effectiveness of a coordinated crisis response.

These tokens of mixed attitudes and contrastive evaluation represent important qualitative data that should not be overlooked. Beyond ensuring the validity of a thorough analysis, they support discussions of strategic choices and measures by discouraging purely black-and-white thinking.

Another concern for qualitative research appears to be that trust- and distrust-related interactions can be vague as regards the responsible (dis)trustees. In many cases, people only indirectly evaluate social actors because they talk instead about the quality of actions or processes, what Martin and White (2005) classify as appreciation. Statements from a focus group in Norway may exemplify:

The participants address actions – measures – and fortify a trusting attitude via many affirming adjectives (good, reassured, less worried) and verb clauses (the measures work and have worked). Only at the end is some hesitation introduced, as one participant entertains a dialogical alternative path to health through herd immunity.

Interestingly, an analysis of the focus groups reveals that respondents in Denmark and Norway rarely assessed the safety measures as delayed or insufficient, whereas respondents in Sweden did, aligning attitudinally in the direction of distrust. The excerpt below shows how such a negative assessment of ability could take shape via impersonal discourse; only at the end is any mention made of responsible health authorities via the somewhat imprecise “they”.

The opening impersonal evaluation is realized grammatically through nominalisation, where nouns – such as recommendations and restrictions – are made central, and responsible actors and verbs are omitted. There are noticeably also efforts to discursively articulate that inadequate action involves a great deal of risk and serious consequences. The negative evaluation is boosted through graduation, and more specifically quantification, with so many (“a hundred customers”) simultaneously on the same premises, so few wearing protective equipment (“no one was wearing a mask”), and so many showing symptoms (“everyone was coughing”). In terms of engagement, the respondent claims to speak on behalf of his colleagues, adding a dialogical element that strengthens the critique (“we who worked in the shop were like ‘well … ’”). Linus furthermore conveys that the Swedish official goal of “the right action at the right time” (Irwin, 2020, p. 8) was sorely missed via assertions that careful restrictions were introduced only when the greatest need for them had passed. Finally, there is an instance of direct judgement of social actors (“they start thinking …”), but there is more on this point in the next section.

Omitting the type of “impersonal” discourse addressed here from organizational trust analysis would imply that trust- and distrust-relevant information is overlooked. Appraisal analysis may assist by distinguishing between action appreciation and actor judgement to assess overall trust.

Trust and distrust are most easily identified in judgements of individuals or groups. However, analysing these judgements is complicated by the diverse ways social actors are represented, sometimes obfuscating their responsibility. To clarify, I will address both collectivizing and individualizing discursive moves and the use of categorization and nomination to represent social actors (van Leeuwen, 2008).

One common form is judgement via categorization, where participants often use us/them pronouns. In Excerpt 8, for example, judgement is directed at social actors via the attribution of delayed action (e.g. “well, now … they start thinking about us”) and affective inscriptions of the consequences for shop assistants (e.g. “I felt like those of us who work in stores were not worth shit”), which casts those responsible for COVID-19 restrictions in Sweden as deficient with respect to at least two domains of trust – ability and benevolence. Categorization, especially with the vague “we,” also serves to collectivize, as seen when Anna in Sweden says, “we chose not to shut down completely,” or John states, “the balance in the country has been good. I choose whom I listen to, experts!” Additionally, judgements often involve nominations, such as referencing an organization or country, further realizing collectivization (e.g. Sweden didn’t close down). Both categorization and nomination can suppress variation in views and responsibilities – whether within a “we” community, among experts, or in a country – presenting trust or distrust as more widespread than it may be, thereby persuading others to adopt certain positions. As van Leeuwen (2008) notes, this assimilation makes heterogeneous individuals appear similar. In the following example, transparency, which is a foundation for organizational trust (Grimmelikhuijsen and Meijer, 2012), is discussed in a Norwegian focus group. First, they assess their health authorities positively and then make comparisons using collectivizing nomination:

Prompted by the moderator to evaluate the authorities’ transparency, the group affirmed that Norway’s leaders were open about uncertainty, mitigating their statements via resources of engagement and graduation (“I think so” and “they were at the start”) while also validating them by dialogically revoicing the health authorities (“They kept saying ‘this is new … ’”). Lukas then draws on quantitative graduation (“less open” and “more assured”), constructing a comparative hierarchy of openness and certainty, positioning Norway’s health authorities as more transparent than those in the “USA”, and “Sweden” where the COVID-19 strategy was considered evidence-based (Rasmussen et al., 2023). Through nomination and collectivisation, countries, rather than individual decision-makers, are attributed actions. Similar collectivizing discourse can also frame positions on trust or distrust toward other organizations, where the name of a country or city may represent specific business divisions or entities.

However, given that so much value is placed on leadership and the qualities of individual leaders in today’s society (Alvesson and Spicer, 2012), it is not surprising that organizational trust/distrust-discourse will not just feature collectivising discursive moves, but also individualising ones. Besides the collectivising variations, nomination features the use of personal names (first name, last name, or both). Examples of negative evaluation may then involve scapegoating, where an individual is made the owner of decisions that have actually involved institutional procedures and more than one decision-maker. If the evaluation is positive, an individual may be credited in such a way that s/he is cast as a “professional hero” (van Leeuwen, 2008). In the following excerpts, respondents construct markedly different attitudinal stances toward a leader and spokesperson for the Public Health Agency of Sweden:

In the first extract, a Swedish interviewee positively evaluates Anders Tegnell’s leadership, highlighting his ability and tenacity over time while suggesting that “others” (e.g. the government) avoided responsibility and may have exploited his willingness to lead. The participant uses graduation to strongly emphasize both positive judgements (e.g. “the whole time, whatever happened” and “he was always there”) and negative ones (e.g. “others didn’t take any responsibility”). Trust in Tegnell is further built through the token of intent to vote for him, had he been in politics, with high intensity added through an assessment of high probability. In contrast, Excerpt 11 features discursive work that expresses distrust in Tegnell and his organization. The respondent dialogically engages with Tegnell’s reasoning – that COVID-19 management is going well while many are dying – to highlight a contradiction if saving lives is assumed to be the top priority, raising doubts about Tegnell’s moral compass and benevolence – key aspects of trustworthiness.

Since performance is a precursor to trust in several organizational contexts (Xu et al., 2019), results speak volumes. In the pandemic context, focus group participants often cited infection control outcomes, comparing COVID-19 management strategies. Participants in Norway, in particular, drew on results as a resource for evaluating the effectiveness of safety measures in their own country and others, thereby reinforcing perceptions of ability and trust in Norwegian government organizations:

By contrast, respondents in Sweden hardly relied on results as a resource when forming trusting positions. In fact, only those critical of the Swedish leadership and strategy did so:

Notably, excerpts 12 and 13 feature no explicit evaluative language, such as adjectives or inscriptions of affect. Both Hanna in Norway and Josef in Sweden construct an attitudinal stance by engaging dialogically with external realities with strong positive/negative connotations (statistical patterns and public discourse) which lend their statements a sense of credibility (e.g. “you saw that the numbers started to go down”). They both introduce juxtapositions (e.g. “At the same time …”, “six times more …”) between the observed fatality numbers in Sweden and post-shutdown in Norway and other Nordics. Graduation is an apparent discursive resource here, as the quantifications that the respondents draw upon scale their evaluations of successful COVID-19 management in several Nordic countries, while indirectly critiquing Sweden’s approach.

Respondents in Sweden expressing high trust navigated the high fatality rate by offering rationalizations similar to those from authorities – namely, that losses were difficult to evaluate, any assessment premature, and that losses would likely balance out over time (Björk, 2021). For example, Agneta asserted that “all statistics from different countries must be taken with a grain of salt,” while Anette, giving the authorities the benefit of the doubt, stated: “Of course, unfortunately, many have died, but the question is whether the numbers will even out, and we won’t have so many more deaths here in the end.” Those defending high trust employed affective language, hoping the situation would improve. In contrast, those expressing distrust focused on the official death toll, assessing results along the way.

As organizational research increasingly identifies human communication as both a cause and a means of expressing trust and distrust, it becomes essential to explore effective ways of analysing how these stances are realized in interaction and discourse. To undertake this task, one must draw not only on theories of trust and distrust, but also on well-established theories of communication – something that, unfortunately, is rare. In this regard, the present paper contributes to qualitative research on organizational trust and distrust, including public institutions, by demonstrating how the widely recognized framework of appraisal analysis (Martin and White, 2005) can provide new insights into qualitative research on these topics.

Drawing on appraisal analysis, the paper took as a point of departure that stances on trust feature discursive appraisal or evaluation, and the findings suggested six ways that social actors use discursive resources to construct stances on organizational trust: simple-assertive evaluation, prescriptive evaluation, activity-, agent-, and results-centred evaluation, and mixed evaluation. This conceptualization – soon to be elaborated further – may be applied during the stage where empirical data are analysed and themes or categories are formed, a procedure common to most strands of qualitative research (Lee, 1998). It may also be valuable when subject-specific themes or categories are gradually analysed in more depth (Williams and Moser, 2019). Indeed, an advantage of drawing on a general communication theory such as appraisal analysis is that it can be applied in all variations of qualitative organizational trust research to reach a more nuanced understanding of the meaning of participants’ discourse, and as an “add-on” to the coding procedures of thematic or qualitative content analysis.

More specifically, there may be an advantage in drawing on the suggested concepts in order to (1) more exhaustively make sense of qualitative data and (2) avoid rationalist tendencies that may otherwise be a possibility. While offering reassurance about the importance of fundamentals such as assessments of ability, integrity, and benevolence for trusting relations (Schoorman et al., 2007; Mayer et al., 1995), the study points to additional ways in which the trust–distrust nexus is realized. Perhaps most interestingly, the study identified several instances of trust positioning that did not contain justifying or qualifying statements, making them difficult to categorize based on research that presumes that trustors act as rationally arguing decision-makers. In some of these instances, participants took a stance on trust straightforwardly via simple assertions of trust and distrust, or by prescribing which attitudes and behaviours concerning trust that others should adopt, while simultaneously boosting an organization’s trustworthiness and authority.

The seemingly simple, bare assertions also add support for the propositions made in previous research on how trust is intertwined with the exercise of power and control (Hardy et al., 1998; Grey and Garsten, 2001). By not adding arguments and not taking into account alternative positions, the use of simple assertions contributes to intersubjectively framing trust as a question that is not at issue. In addition, because the positioning lacks explicit whys and wherefores, the authority of an organization or leader becomes imbued with taken-for-grantedness, which could indicate the hegemonic status of certain attitudes and positions of trustworthiness (see also, Hardy et al., 1998).

The analysis also found that participants endorsed trust through prescriptive discourse, asserting obligations to trust authorities and follow their guidance. This allowed them to not only assert their own stance but also endorse that others should think and act similarly. Such sensemaking extends beyond verbal exchange, constituting “actional” discourse (Martin and White, 2005) and help realise power relations (Grey and Garsten, 2001). Specifically, citizens assist public institutions in articulating conduct of conduct (Foucault, 1991), a form of governing through which social actors align with state goals like health, safety, and social order. These uses of discursive resources illustrate sensemaking in countries with strong institutional trust, but similar evaluations could also align with distrust, encouraging independent and critical thinking.

Thus, so far two discursive moves have been demonstrated with which actors substantiate both positions on trust and distrust and positions in power relations. The first concerns what was called a simple-assertive evaluation, a kind of confession without qualifications to a stance on trust or distrust, signalling that arguments are redundant and that the position is not expected to be challenged. The second featured prescriptive evaluation, whereby social actors substantiate norms and articulate the kinds of attitudes and behaviours that people ought to adopt. Regardless of the organization or issue being studied, these modalities may be added to a qualitative approach, while also considering, for instance, trustors’ stances toward a trustee’s ability, integrity, and benevolence.

Moreover, some of the “messiness” of qualitative data, noted in prior organizational research (Parkhe, 1993), was addressed here. Trust and distrust are expressed through complex, entangled discursive moves, and positions on trust are not always clearly defined. A key challenge arises from agentless language, which creates vagueness around responsibility.

Such instances of vague language can be anticipated analytically. Drawing on variations of appreciation from appraisal theory to denote agentless evaluation (Martin and White, 2005), the analysis identified both agentless activity-centred and results-centred evaluations that contribute to stances on trust and distrust. For example, statistics on deaths during the pandemic serve as evaluative tools, while in other contexts, sales or revenue figures may serve the same purpose. When results carry strong connotations, they are crucial for trust-related sensemaking. Attention to agentless discourse is also especially important in contexts of power asymmetry, where clarity on responsibility and accountability matters (Hardy et al., 1998).

Furthermore, some qualitative data can be difficult to code or categorize, particularly when the same person expresses multiple attitudes about the same phenomenon, a common occurrence in professional settings where trust is conditional and multifaceted (Lewicki et al., 1998; Kostis et al., 2022). This type of multifaceted sensemaking was evident in the present analysis, for example, when participants evaluated the effectiveness and clarity of pandemic communication management, an area known to affect trust (Zeffane et al., 2011). The study suggests that managing such mixed evaluations involves anticipation, naming, and inclusion in coding. The concept of mixed evaluation captured instances where participants expressed multiple value judgements in one statement. Tokens of contrastive evaluation typically stand out through certain grammatical structures – such as “on the one hand X applies, and on the other hand Y” – and in concede-and-counter pairs – such as “I agree with proposition A (concede), but proposition B is perhaps even better (counter).” Analytically, attention to conjunctions like “though,” “but,” “nevertheless,” “whereas,” and “even if” can also help identify mixed evaluations.

Finally, the analysis suggests that more nuanced understandings of organizational trust discourse may be reached through careful observation of the types of categorizations and nominations that are used when the qualities of organizations and social actors are assessed. For instance, pronominal categorization (such as us or them) helps articulate responsibility quite broadly and vaguely, while some categorizations signal a more elevated, or lower, ethos for certain actors (e.g. experts vs bureaucrats). Likewise, people can, via nomination, have an entire country represent the actions of an institution (e.g. Sweden failed/succeeded) or, alternatively, highlight a single individual (e.g. Tegnell decided). People thus realize trust–distrust relationships via linguistic moves that bring about both collectivisation and individualisation, resulting in broad, perhaps even communal or national responsibility and identity constructions, but sometimes also, as van Leeuwen (2008) suggests, professional scapegoats or professional heroes.

With a broad range of studies on organizational trust emphasizing that trust is expressed through communication (Shockley-Zalabak et al., 2000), it seemed valuable to apply the specialized insights of appraisal analysis in this study, offering methodological fine-tuning. Different strands of organizational trust research may benefit from the observations analysed and summarized here. Yet, a limitation of this paper is that it uses appraisal analysis to study focus group data and the positioning of so-called trustors in the context of a single crisis-related issue. Naturally occurring data – collected in real-world settings without researcher intervention – would perhaps provide an even more appropriate context for applying appraisal analysis. Including the discourse of those in leading roles – the trustees – would also be ideal to analyse the interplay between trustee behaviour and trustor evaluation. However, such data, particularly in conversational form, would be difficult to obtain. Additionally, in major crises like the COVID-19 pandemic, numerous organizations and leaders are involved, causing stakeholders to evaluate a consortium of organizations and politicians during qualitative interviews. This complicates research aimed at discerning trust or distrust in a specific organization. However, appraisal analysis is often applied to less dramatic situations than COVID-19 management. It is a framework designed to identify and assess emotional responses and evaluations individuals make in response to various social, organizational, or interpersonal dynamics. By focusing on how individuals perceive and evaluate trustworthiness or distrust, appraisal analysis enables researchers to capture subtle, nuanced expressions of trust or distrust that may not be evident through more traditional, quantitative methods. This versatility makes appraisal analysis adaptable to a wide range of contexts beyond crisis situations. In the future, this approach could be used to analyse more naturally occurring communication, involving both trustors and trustees, across varying issue contexts and organizations.

1. Introduction: The moderator introduces themselves and outlines the focus on risk/crisis management and issues of trust in responsible institutions.

2. Anonymity and respect: Participants are assured of anonymity and encouraged to share openly, understanding that sensitive topics are not required.

3. Discussion guidelines: Participants can engage with each other, respond in any order, and are encouraged to stay focused.

4. Duration: The conversation is scheduled for two hours but may not take the full time.

5. When did you first hear about COVID-19? What did you hear, from whom/where, and what did you think about it?

6. When did you realize the situation was serious, and what made you see it as a crisis?

7. How easy or difficult was it to stay informed about the crisis? Was the information consistent or were there contradictions?

8. Are there any actors you feel confident relying on in this situation, and why? Is there anyone you would hesitate to rely on, and why?

9. If not already mentioned, the moderator names key organizations and leaders (Norway, Denmark, or Sweden) and repeats question 8.

10. Can you recall situations or actions that caused you to gain or lose trust?

11. How would you describe the information from the authorities? What adjectives come to mind?

12. How would you describe your experience with the institutions involved in COVID-19 management? Do you think they have different roles in this context? Do you see any differences in how they handled the situation? Did they speak with one voice, or did you perceive different messages from them?

13. How did these actors deal with the uncertainty of the pandemic? Were they open enough about it?

14. How do you view the measures that were implemented over time?

15. What infection control measures have you personally taken?

16. What influenced your decision to take these measures? Who influenced you, and why?

17. Have your measures changed over time, and if so, why?

18. Which issues in the crisis have concerned you the most?

19. How do you think institutions have handled and communicated these concerns?

20. Is there anything important we haven’t covered yet that you think should be discussed?