A preliminary study examining terminology used by individuals with lived experience describing beliefs about being targeted or harmed by others

Abstract

While recent studies have prompted re-evaluation of the term “schizophrenia,” few have examined the use of terms to describe persecutory ideation (PI) or paranoia. This study examines the preferences and terms used by a cross-diagnostic population of individuals (N = 184) with lived experience using an online survey. Participants most commonly described their PI in terms of the perceived source of threat, followed by clinical language, most commonly variants of “paranoia” and “anxiety.” Of five selected terms assessed quantitatively – “anxiety,” “paranoia,” “persecutory thoughts,” “suspiciousness,” and “threat thoughts” – participants were more likely to report that “anxiety” aligned with their experience of PI, followed by “suspiciousness.” Endorsement of terms more specific to PI was associated with self-report PI severity, while a preference for “anxiety” over other terms was both associated with less severe PI and lower scores on a measure of stigma. These results suggest that the heterogeneity of terms used by individuals with lived experience support a person-centered approach to language describing such experiences.

Introduction

Language impacts public perceptions of mental illness. In its 2021 Stigma and Discrimination Research Toolkit, NIMH emphasized that “the language we use to talk about health plays a crucial role in shaping opinions and beliefs about a disease or condition and the people affected.” (National Institute of Mental Health Information Resource Center, 2021). Recent studies have provided several examples of this in the mental health and substance use literature. For example, of the term “substance abuse” is linked with increased perceptions of personal culpability for substance use relative to the term “substance use disorder” (Kelly & Westerhoff, 2010). Person-centered language (e.g. “person with mental illness”) is linked with improved tolerance relative to language that emphasizes the condition (e.g. “mentally ill person” Granello & Gibbs, 2016).

Recently, some have argued that the term “schizophrenia” is not descriptive of etiology or most common symptoms, and that it falsely indicates that psychosis derives from a unitary disorder (Keshavan et al., 2011, 2013). A recent survey of the nomenclature section of the World Psychiatric Association found that a majority of members believe both that the term “schizophrenia” is stigmatizing, and that the disorder ought to be renamed (Maruta & Iimori, 2008). Psychiatric associations in some countries have officially renamed schizophrenia, for example, in Japan (togoshitcho-sho, or “integration disorder”), in South Korea (johyun-byung or “attunement disorder”), and in Taiwan (Si-jué-shi-tiáo-zhèng or “dysfunction of thought and perception”; Yamaguchi et al., 2017). There is a small but growing body of evidence linking these name changes to reduced public stigma toward individuals experiencing psychosis (Yamaguchi et al., 2017). One recent large study of individuals with lived experience, their family members, and mental health professionals demonstrated broad and strong support for changing the name of schizophrenia, with preferences highest for terms like “altered perception syndrome,” or “psychosis-spectrum syndrome” (Mesholam-Gately et al., 2021).

Similar questions around terminology may also apply to the symptoms of psychosis, and in particular, beliefs about being negatively targeted or harmed by others. This experience – most commonly referred to as “paranoia” or “persecutory ideation” in the research literature (for simplicity, we will use “PI”) – are common in psychosis (Paolini et al., 2016) and associated with poor physical health, low social support, and multiple psychiatric conditions (Freeman et al., 2011). Historical conceptions of persecutory delusions have focused on defining delusion broadly, rather than persecutory experiences specifically. This approach focused largely on identifying delusions according to the extent to which such beliefs are embody several characteristics, for example, whether they are unfounded, firmly held, and resistant to change (Freeman, 2007). However, more recent models of delusional beliefs – including persecutory ideation – have focused on categorizing such beliefs phenomenologically (e.g. persecutory beliefs, grandiose beliefs, etc.) and measuring each of the defining traits of delusional beliefs dimensionally. And indeed, persecutory ideation is a characteristic not only of individuals with schizophrenia-spectrum and other psychotic disorders, but also present in individuals with other or no clinical diagnoses at all (Freeman, 2006). And indeed, more research supports the existence of a continuum of persecutory ideation varying across individuals and contexts, ranging from mild worries about threats to severe, persistent, and fixed beliefs about dangers that are not shared by others in the environment (Bebbington et al., 2013; Elahi et al., 2017; Freeman et al., 2010).

One of the most common terms for these experiences, “paranoia,” derives from the Greek roots “para” (meaning “irregular”) and “noos” (meaning “mind”). The term “paranoia” connotes that beliefs are inconsistent with reality (Merriam-Webster, 2022). Worries about harm may be maladaptive in other ways than simply being false; for example, they may be true but disproportionately affect behavior, partially true but exaggerated (Bebbington et al., 2013; Kaymaz & van Os, 2010), or may change in regard to level of conviction over time (Ben-Zeev et al., 2012; Buck et al., 2019). Some persecutory beliefs could even be adaptive under certain circumstances but cause dysfunction in others (Raihani & Bell, 2019). The term “paranoia” could lead to discord with providers and potentially discourage help-seeking (Dell et al., 2021; Wang et al., 2018). Terms about other psychological phenomena – e.g. anxiety or depression – describe phenomenology and are ostensibly agnostic to whether the beliefs underlying the experience are true or false. Concerns about negative interactions with providers are one of many barriers that prevent help-seeking and prolong the duration of untreated illness (Gronholm et al., 2017).

Other commonly used terms raise their own questions. “Persecutory ideation” captures phenomenology and continuum thinking while remaining less agnostic about truth or falsehood. However, this term is not commonly used; Google Ngram viewer in 2019 showed “paranoia” to be used over eight times more frequently than “persecutory ideation” in published materials (Google, 2022). Other terms – e.g. “suspiciousness” or “anxiety” – similarly capture phenomenology and are agnostic to validity of beliefs, but may be inappropriately general; individuals may be anxious or suspicious in a range of contexts beyond feeling targeted for harm by others.

Little is known about the attitudes of individuals with lived experience toward varying terms for PI, and also the factors associated with preferences for each term. Emerging technologies allow for large scale data collection from individuals with lived experience, including both those who have engaged with traditional mental health services as well as those who have not, either because they have chosen not to or if they don’t have access (Buck et al., 2021). The present study uses remote research methodology to conduce a preliminary examination of the attitudes of individuals with cross-diagnostic PI toward terminology to describe their experiences. Specifically, we explore (1) what terms those with PI use to describe their experiences, (2) their reactions to a range of common terms, and (3) the extent to which they view each as stigmatizing.

Methods

Participants.

Data were drawn from a larger study examining PI using a multimodal smartphone system. Ethical approval was obtained from the institutional review board of [DEIDENTIFIED] (IRB#STUDY00001321); all participants provided consent for their data to be used in this research. Inclusion criteria included (1) being 18 years or older; (2) speaking English; (3) owning and able to use an Android smartphone, and (4) providing a score ≥11 on the 10-item ideas of persecution subscale within the Revised Green Paranoid Thoughts Scale (R-GPTS), an abbreviated version of the original GPTS (Green et al., 2008). We followed cut-off recommendations described in the psychometric work describing the development of the R-GPTS, according to which participants were included only if they met cut-off criteria for moderate-severe PI (Freeman et al., 2021). Participants were excluded if they (1) did not live in the US, (2) had already participated in the study, or (3) were unavailable for the 30 days of data collection. Participants were recruited through targeted Google Ads delivered in participants’ search results when they used pre-determined search terms. These search terms were generated both by the research team and the Google “broad match” algorithm based on initial ad performance results. This resulted in terms related both to clinical (e.g. delusion, schizophrenia symptoms) and non-clinical or related terms (e.g. seeing ghosts, conspiracy). Prospective participants who clicked on these ads were linked to an online landing page describing a study focused on understanding more about “experiences of worries about being harmed, followed, or spied on by others.” Participants who completed study questionnaires agreed to complete multiple daily assessments of their affect, cognition, and behavior, share data about their physical and social activity collected through passive sensors on their smartphones, and complete a month of active data collection.

Procedure.

On the final day of their 30-day data collection period, participants were offered the opportunity to provide additional insights about their opinions on PI nomenclature. This questionnaire was delivered through the study smartphone application that users had already used for the previous 30-day data collection period. Out of the full sample of study completers (N = 231), 43 (18.6%) did not complete the additional questionnaire, and four additional participants opened the questionnaire but skipped all items (1.7%). All data provided by the remaining 184 (79.7%) are reported here.

Preferred PI terminology and associated impressions were assessed with three sets of items. First, participants were asked to attend to their experiences with “thoughts about someone or something wanting to harm you” (i.e. experiences about which they had been responding for 30 days) and to explain how they would describe these to a friend through an open free-response field. Second, following this free-response item, participants were asked to respond, on a five-point Likert scale (“not at all” = 1 to “definitely = 5), how much each of the five terms – anxiety, paranoia, persecutory thoughts, suspiciousness, and threat thoughts – described what they experienced. Third, participants were asked to provide a response related to their attitudes toward an individual who experienced each of the same terms through a hypothetical scenario. Specifically, they were told that a person (called “Larry”) experienced each of the five presented terms, and then were asked to report whether knowing this about Larry would change how they felt about Larry (“much more negative” = 1 to “much more positive” = 5).

Measures.

As mentioned, persecutory ideation was assessed with the persecution scale of the Revised Green Paranoid Thoughts Scale (R-GPTS; Freeman et al., 2021). Beliefs about mental health were assessed with Endorsed and Anticipated Stigma Inventory, Mental Illness subscale (EASI-MI; Vogt et al., 2014). This eight-item subscale of the full Endorsed Anticipated Stigma Inventory is focused on beliefs related to mental health conditions and what they signify about individuals experiencing them (Vogt, Fox, et al., 2014). Mean imputation was used to replace missing values.

Data analysis.

Qualitative responses were coded by a team member using conventional content analysis. First, all responses were read and initial codes categorizing all responses were applied. First level codes were then reviewed and aggregated to generate second-level codes. Categorization of responses was non-exclusive such that when participants provided multiple terms for their PI, these could be coded into multiple categories. The primary qualitative coder then repeated the process of applying all codes; however, this time with a second coder. Coders first evaluated agreement, finding Kappa statistics ranging from 0.43 to 1, with an average agreement across categories of 0.78, indicating moderate to high agreement, then resolved discrepancy through consensus.

Responses to quantitative items were assessed to address three research questions, namely, which terms are (1) consistent with self-experience, (2) viewed as most stigmatizing, and (3) associated with measures of stigma and PI severity. For the first two research questions, we reviewed item response frequencies and mean scores. To test specifically whether a term differed from average at a statistically significant level, we conducted analyses in two steps, (1) generating five variables representing the average score provided to four of the five terms, excluding one of the five terms each time, and then (2) conducted a paired samples t-test comparing the score provided to one item (e.g. “paranoia”) against the average of the other four items. This allowed for a comparison for whether participants tended to see a particular term as more fitting or more stigmatizing than the average for all other terms. Given the difference in specificity regarding “anxiety” compared to the other four terms, which more specifically describe PI, we conducted this two-step analysis twice; once for all five terms, and another time for the four terms other than anxiety. To assess the relationship of terminology to PI severity and stigmatizing attitudes we examined Pearson correlations of each outcome with (1) the item score, and (2) a preference score representing the difference of one particular item from the participants’ average score of the other four items. In other words, preference scores represented whether the participant’s pattern of responding demonstrated a preference for that particular term (i.e. rated as more consistent with self-experience than the average of the ratings they provided to other items).

Results

One hundred and eighty-four (N = 184) individuals with persecutory ideation responded to the terminology questionnaire. Demographics of the sample are reported in Table 1. The sample was predominantly white (n = 129, 70.1%), identified as cisgender women (n = 135, 73.4%), and were non-Hispanic/Latino (n = 171, 92.9%). The average age was 38.64 (SD = 11.97). The most commonly self-reported diagnosis was a depressive disorder (n = 121, 66.5%), followed by PTSD (n = 92, 50.5%), and bipolar disorder (n = 73, 40.1%). A smaller proportion of participants reported a diagnosis of schizophrenia (n = 24, 13.2%) or schizoaffective disorder (n = 25, 13.7%), though four (2.2%) additional participants wrote in variants of “psychosis” in the other specified diagnosis selection. 23 participants (12.6%) reported no previous diagnosis.¹ Participants could endorse as many historical diagnoses as applied to them. With regard to persecutory ideation, 175 (95.1%) of participants reported persecutory ideation consistent with “severe” PI (i.e. indicative of persecutory delusions and reported in 7% of the non-clinical population) and 139 (75.5%) consistent with “very severe” PI (i.e. reported in 1% of the non-clinical population) as determined by R-GPTS cutoffs determined in psychometric work conducted by Freeman and colleagues (Freeman et al., 2021). About half reported that they were currently engaged in treatment (n = 94, 51.1%).

Table 1.

Sample descriptive statistics.

	M or N	(SD or %)	Minimum	Maximum
Age	38.64	(11.97)	18	75
Race
White / Caucasian	129	(70.1%)
Black / African American	31	(16.8%)
Pacific Islander	1	(0.5%)
American Indian / Alaskan Native	2	(1.1%)
Asian	4	(2.2%)
More than one race	17	(9.2%)
Gender
Female	135	(73.4%)
Male	43	(23.4%)
Transgender man	0	(0.0%)
Transgender woman	2	(1.1%)
Other / Non-binary	4	(2.2%)
Ethnicity
Hispanic / Latino	171	(92.9%)
Non-Hispanic / Non-Latino	13	(7.1%)
R-GPTS persecutory (total; possible range = 0-40)	31.39	6.89	13	40
EASI-MI (total; possible range = 8-40)	17.26	6.49	8	40

Open responses.

The most common framing for responses related to PI (see Table 2) included those involving the actions of other people, entities, or institutions (n = 42, 27.4% of responses). Of these responses, 23 (15.0%) simply referred to general terms for people (e.g. “being around people,” “people in general”), while 22 (14.4%) involved specific targets (e.g. “doctors,” “police,” “evil spirits”). Another 23 (15.0%) responses were categorized as referencing general difficulty, whether those difficulties be attributed primarily internally (e.g. “myself and not being able to shut my mind down,” “my issues”) or externally (e.g. “dealing with everyday things,” “dealing with everything”). Seventeen participants (11.1%) provided responses involving more general terms for thoughts and feelings (e.g. “big emotions,” “thoughts of people being jealous of me”).

Table 2.

Response frequencies to label and stigma items.

Term	N (%)^	Examples
Other people or institutions	42 (27.4%)	---
General targets	23 (15.0%)	“I have difficulty with people in general” / “Being around people” / “Dealing with people” / “Harmful people and thoughts of them harming me”
Specific targets	22 (14.4%)	“Doctors” / “Police” / “He’s back at it again” / “Thinking that one of my old boyfriends or a stranger may get me” / “Evil spirits”
Terms denoting general difficulty	23 (15.0%)	---
External difficulty	18 (11.8%)	“Dealing with everyday things” / “Dealing with everything” / “Having a bad day” / “Life” / “I’m having a difficulty day with everyone and everything”
Internal difficulty	5 (3.3%)	“Myself and not being able to shut my mind down” / “Me” / “My head” / “My brain” / “My issues”
Variant of “Anxiety”	22 (14.4%)	“Anxiety” / “Worrying and overthinking” / “Anxious thoughts” / “Anxiety – I feel like the whole world is bearing down on me” / “Stress and anxiety”
Description of (rather than a term for) persecutory ideation	21 (13.7%)	“Feeling like people are against me” / “Feeling like people are out to get me” / “Feelings or someone or someone trying to harm me” / “Harmful people and thoughts of them harming me”
Variant of “Paranoia”	19 (12.4%)	“Paranoia” / “Being paranoid” / “My paranoia” / “I’m sorry, I think my paranoia is getting to me today” / “Paranoid thoughts”
Generic term for thoughts or feelings	17 (11.1%)	“Big emotions” / “Difficult day with my thoughts” / “My mood and feelings” / “Thoughts in my head” / “Thoughts of people being jealous of me”
Other psychological phenomena	12 (7.8%)	“Voices in my head” / “Hallucinating” / “Keeping my attitude and temper in check” / “My feelings of insecurity and lack of motivation”
Novel term	12 (7.8%)	---
Idiosyncratic	8 (5.2%)	“A battle in my mind that every day I try to overcome” / “I’m a Christian being challenged” / “The bugs are really on me today” / “Difficult day with my check points”
Generalizable	6 (3.9%)	“Hyperarousal/Hyperawareness” / “Overthinking” / “Feeling targeted” / “Misleading thoughts” / “Irrational thoughts”
Other diagnostic term	8 (5.2%)	“My illness” / “PTSD” / “Bipolar thoughts” / “Schizophrenia”
Variant of “Depression” / “Sad”	8 (5.2%)	“Depression” / “Being sad” / “Feeling inadequate about the big things and even small things in life” / “Loneliness”
Variant of “Fear”	6 (3.9%)	“Handling my fears and discerning real or imagined” / Feeling scared” / “Feeling like I’m scared of my own thoughts” / “Fear”
Discerning reality	6 (3.9%)	“Sometimes I’m not sure what’s real and what’s my mind playing tricks” / “Deciphering between real or imagined fears” / “Determining what’s real and what’s a delusion”

^{^}Overall N = 153 participants who provided some kind of open-ended response to the question, also excluding participants whose response denoted that the question was unclear (e.g. “I don’t understand the question”) or not relevant to them (e.g. “I don’t talk to anyone about this”).

A number of participants provided responses that either referenced specific terms for PI-related experiences (e.g. anxiety, paranoia), described PI, used specific terms not typically used in the PI literature (i.e. generalizable novel terms), or descriptions that appeared specific to the individual’s framing of their experience (i.e. idiosyncratic novel terms). A similar number of participants provided responses that were a variant of or synonymous with “anxiety” (n = 22, 14.4%) or “paranoia” (n = 19, 12.4%). Twenty-one participants (13.7%) provided a description of persecutory ideation that could not be contained to a brief (i.e. 1-2 word) term (e.g. “feeling like people are against me,” “harmful people and thoughts of them harming me”). Of the novel terms provided (n = 12, 7.8%), generalizable terms (n = 6, 3.9%) included descriptions of the specific thoughts (e.g. “misleading thoughts,” “irrational thoughts”), other colloquial terms for anxiety or related symptoms (e.g. “hyperarousal/hyperawareness”). Idiosyncratic terms (n = 8, 5.2%) included ostensibly personal terms. Notably, zero participants referred to their experience as “persecutory ideation,” the term most commonly used by our research team.

Remaining participants used other emotions (e.g. sadness or fear), psychological phenomena or diagnostic terms to describe their experiences. Eight (5.2%) described a variant of “sad” or “depressed,” and six (3.9%) described a variant of fear. Twelve participants (7.8%) described other psychological phenomena (e.g. lack of motivation or hallucinations). Another eight (5.2%) used a diagnostic term not otherwise coded separately (e.g. “bipolar thoughts,” “schizophrenia”). Last, six participants (3.9%) described the process of struggling to discern reality (e.g. “sometimes I’m not sure what’s real and what’s my mind playing tricks”).

Quantitative evaluation of pre-selected terms.

Consistent with self-experience.

Response frequencies can be found in Table 2. “Anxiety” (M = 4.34 (SD = 0.94); 84.5% “very much” or “definitely”) was the term participants regarded as most consistent with one’s experience, followed by “suspiciousness” (M = 3.54 (1.18); 56.4%), “paranoia” (M = 3.36 (1.20); 48.1%), “threat thoughts” (M = 3.35 (1.20); 50.8%), and “persecutory thoughts” (M = 3.24 (1.20); 44.9%). Comparing responses within individuals, “anxiety” (M_difference = 0.97 (0.99); t = 13.16, p < .001) was described as more accurately capturing respondents’ experience, while “paranoia” (M_difference = −0.25 (1.06); t = −3.21, p = .002), “threat thoughts” (M_difference = −0.28 (0.97); t = 3.91, p < .001), and “persecutory thoughts” (M_difference = −0.42 (0.97); t = −5.72, p < .001) significantly differed in the opposite direction, signifying that they were regarded as less fitting to participants’ experiences.

We repeated these analyses focusing only on the four terms most specific to PI – suspiciousness, paranoia, persecutory thoughts, and threat thoughts – and found that relative to the within-individual mean of the other three PI-related terms, participants reported that suspiciousness more appropriately described their experience (M_difference = 0.24 (0.96); t = 3.32, p = .001), whereas “persecutory thoughts” was regarded as less appropriate (M_difference = −0.19 (0.99); t = −2.53, p = .01). The other comparisons were non-significant.

Stigma.

The second item asked participants to respond with how their feelings about a person would change if they were to learn that he experienced each of the pre-selected terms. Response frequencies for this question can also be found in Table 2. Importantly, a relatively high proportion of participants reported that their impression of the individual would change (either positively or negatively) if any of the related terms were associated with them. “Persecutory thoughts” was regarded as having the most negative impression on average (M = 2.84 (0.85); 27.8% “a little more negative” or “a lot more negative”), followed by “threat thoughts” (M = 2.92 (0.96); 32.0%), “paranoia” (M = 2.97 (0.90); 25.6%), “suspiciousness” (M = 3.01 (0.86); 22.8%) and “anxiety” (M = 3.37 (0.90); 36.5%). Using the same within-subjects comparison, “persecutory thoughts” (M_difference = −0.24 (0.67); t = −4.71, p = <.001) and “threat thoughts” (M_difference = −0.12 (0.84); t = 2.04, p = .04) were regarded as significantly making one’s impression more negative, while “anxiety” (M_difference = 0.43 (0.74); t = 7.85, p < .001) was scored as making one’s impression more positive.

We again repeated these analyses focusing only on the four terms most specific to PI (suspiciousness, paranoia, persecutory thoughts, and threat thoughts) and found that relative to the within-individual mean of the other three PI-related terms, participants reported a more negative impression of the term “persecutory thoughts” (M_difference = −0.14 (0.67); t = −2.75, p = .01) relative to the other options. The other comparisons were non-significant.

Relationship to persecutory ideation severity.

We examined relationships of responses indicating whether a term was consistent with self-experience to reported scores on measures of PI and anxiety. Identification scores of the five terms – suspiciousness (r = .25, p = .001), paranoia (r = .25, p = .001), persecutory thoughts (r = .19, p = .011), and threat thoughts (r = .20, p = .006) – were significantly related to R-GPTS total scores. We repeated these analyses using participant preference scores (i.e. centered on mean response across all terms) to examine whether preference for particular terms was associated with symptom severity, and found that a preference for anxiety was negatively associated with R-GPTS scores (r = −.16, p = .03); in other words, participants with a higher preference for anxiety relative to the other terms had lower scores on the R-GPTS. No other preference scores were associated with R-GPTS.

Relationship to stigma.

We examined relationships of label identification to stigmatizing attitudes toward mental health conditions through correlations with the EASI-MH. None of these correlations were statistically significant, though the relationship of the EASI-MH to identifying with the term anxiety approached significance (r = −.14, p = .057). We also repeated these analyses using participant preference scores to examine whether preference for particular terms was associated with stigmatizing attitudes, and found that a preference for the term “anxiety” over the four other terms was associated with EASI-MH scores (r = −.18, p = .02); in other words, participants with a higher preference for “anxiety” relative to the other terms had lower negative attitudes toward mental illness.

Discussion

This article reports a preliminary survey examining terminology used by individuals with cross-diagnostic PI to describe their own experiences. First, overall results suggest a disconnect between terminology used by individuals with lived experience and in research. This was most evident in the open-response item, where, for example, zero participants used the term “persecutory” – a term commonly used in the academic literature, including by the authorship team – to describe such experiences. Participants most commonly described their experience by naming other people or institutions that were the source of threats to them, rather than terms to describe psychological phenomena. Next most commonly, participants described PI in their own words, used generic terms about difficulty or stress, or used variants of the terms “paranoia” or “anxiety.” This pattern of results suggests that many individuals with PI use individualized terms to describe their experiences, rather than terms derived from psychiatric labels. Of those who used symptom-related labels, those that are more commonly used in clinical settings (e.g. anxiety and paranoia) were more preferred than those more common in academic studies (e.g. persecutory thoughts).

Second, of the five commonly-used terms our team examined – anxiety, paranoia, persecutory thoughts, suspiciousness and threat thoughts – the sample varied both in (1) identification with each term, and (2) strength of positive or negative connotations. Participants reported a marked preference for the term “anxiety.” At the same time, scores denoting identification with all terms except “anxiety” were significantly correlated with participants’ self-reported levels of PI, and a relative preference for use of anxiety over other terms was associated with reduced PI. It’s possible that this simply reflects a preference for this term over the others. It also may suggest that individuals with PI more frequently experience anxiety than worries related to threats or that preference for the term “anxiety” over others may indicate a greater frequency of general worries rather than those about intentional harm from others. A drawback of the use of “anxiety” to describe PI is its lack of specificity. Of the terms more specific to worries about threats from others acting intentionally, suspiciousness appeared both to be viewed as more consistent with participants’ experiences and was associated with a validated measure of PI. “Suspiciousness” may be useful because it provides a phenomenological description (similar to “depression” or “anxiety”) while at the same time not emphasizing maladaptive qualities (as in the case of “paranoia”).

These considerations have potential implications in clinical practice and health messaging. First, results suggest clinical interactions would benefit from providers’ use of a person-centered approach (Stanghellini et al., 2013). Descriptions of PI reported by participants were heterogeneous and often idiosyncratic. Taking time to understand service users’ selection of terminology could help providers build rapport and gather additional relevant information about patients’ clinical presentation or background. This personalized approach might also prove useful in the development of client-facing intervention content, for example, descriptions of clinical phenomena in digital interventions (Pathak et al., 2021). Second, our study adds to research and clinical developments that support person-first language (Dunn & Andrews, 2015). Given the fact that some terms were associated with negative impressions, the use of person-first language might counteract notions that link psychiatric experiences with essential personality traits or judgments of character. Third, attempts at outreach and engagement in help-seeking might be more successful if they more frequently align clinical and treatment descriptions with language used by those with lived experience. About a quarter of participants thought that the terms “paranoia” or “persecutory thoughts” either didn’t capture their experience or did so only “a little bit.” These terms are commonly used in the academic literature and in assessment instruments. Mental health service users commonly perceive providers to hold stigmatizing or other disempowering attitudes (Wang et al., 2018). Bifurcated terminology between providers and patients could impede effective communication, demoralize individuals who seek care, or discourage them from seeking care altogether (Clement et al., 2015; Corrigan, 2004).

This study has several limitations, and provides one early examination of preferences of individuals with PI in describing their experiences. The first set of limitations pertain to representativeness of the study sample. Data were collected entirely online and remotely, and thus may underrepresent individuals without access to the Internet and smartphones. Further, the study sample was predominantly white and female. Terminology for a number of physical and mental conditions vary across cultural groups (Kreuter & McClure, 2004), and PI may be no exception. It is important also to note that the study recruited participants cross-diagnostically, and the most commonly reported diagnoses were major depressive disorder and PTSD. While PI is elevated in each of these presentations, it is regarded as a hallmark symptom of schizophrenia-spectrum disorders. Given that diagnoses were self-reported, it is possible – particularly given the high severity of PI reported in our sample per the R-GPTS – that schizophrenia-spectrum disorders were under-reported. Still, a population of participants that represented a greater number of individuals with SSDs may have expressed differing preferences. Another set of limitations pertains to the items assessed. Additional terms describing the experience of PI and more methods of evaluating these terms may have yielded further and more detailed insights. This study could serve as a first step in a program of research continuing to examine possible terms, with our qualitative results as one basis for selection of future terms. Finally, PI and paranoia-related terminology might also be shaped by implicit or explicit assumptions about whether such phenomena are beliefs (i.e. persecutory ideation) or emotions (i.e. anxiety) or some combination of both. Some clinical approaches (e.g. Social Cognition Interaction Training or SCIT; Roberts et al., 2015) encourage individuals to conceptualize paranoia as an emotion to aid in normalization and practices increasing cognitive flexibility. Participants’ perception of such experiences as cognitive or emotional may have affected their preference for specific terms.

Overall, this study provides an important first step to developing an understanding terminology used by individuals experiencing PI that lays the groundwork for continued quantitative and qualitative work. Results suggest that individuals with lived experience are heterogeneous with respect to how they describe these PI, and further, that terms vary with regard to how fitting and how stigmatizing they are perceived to be.

Table 3.

Response frequencies to label and stigma items.

	Appropriateness “How much does each of these describe your experience?”					Associated stigma “Larry experiences ______. Do you feel more positive about Larry, more negative, or exactly the same?”
	1	2	3	4	5	1	2	3	4	5
Anxiety	4 (2.2%)	6 (3.3%)	18 (9.9%)	50 (27.6%)	103 (56.9%)	3 (1.7%)	18 (9.9%)	94 (51.9%)	41 (22.7%)	25 (13.8%)
Paranoia	11 (6.1%)	37 (20.4%)	46 (25.4%)	49 (27.1%)	38 (21.0%)	10 (5.7%)	35 (19.9%)	91 (51.7%)	31 (17.6%)	9 (5.1%)
Persecutory thoughts	19 (10.7%)	26 (14.6%)	53 (29.8%)	53 (29.8%)	27 (15.2%)	12 (6.7%)	38 (21.1%)	104 (57.8%)	19 (10.6%)	7 (3.9%)
Suspiciousness	10 (5.6%)	28 (15.6%)	40 (22.3%)	57 (31.8%)	44 (24.6%)	7 (3.9%)	34 (18.9%)	99 (55.0%)	30 (16.7%)	10 (5.6%)
Threat thoughts	15 (6.5%)	32 (17.7%)	42 (23.2%)	59 (32.6%)	33 (18.2%)	10 (5.5%)	48 (26.5%)	82 (45.3%)	28 (15.5%)	13 (7.2%)
	1 = Not at all, 2 = A little bit, 3 = Somewhat, 4 = Very much, 5 = Definitely					1 = Much more negative, 2 = A little bit more negative, 3 = Exactly the same, 4 = A little bit more positive, 5 = Much more positive

Highlights.

• Individuals with lived experience of persecutory ideation use a variety of terms to describe their experience.

• Most commonly used terms focused on the perceived source of threat, general terms for difficulty, or clinical language, typically variants of “paranoia or anxiety.”

• Of five selected terms, participants were more likely to report that “anxiety” aligned with their experience, followed by “suspiciousness.”