Abstract
Language plays a central role in how neurodivergent individuals express identity and belonging. Although terminology such as neurodivergent and neurodiverse has become increasingly common, little empirical research has examined how these and related terms are perceived within and across neurodivergent communities. This exploratory study surveyed 740 adults with a diagnosis of autism (n = 189), attention-deficit/hyperactivity disorder (ADHD) (n = 190), dyslexia (n = 179), or co-occurring autism and ADHD (n = 182) from the United Kingdom and United States to assess preferences, offensiveness, acceptability, and self-use of six neurodiversity-related terms. Participants rated neurodivergent and neurodiverse as the most preferred, least offensive, and most acceptable, whereas neurominority and neurospicy were least endorsed. The overall order of preferences was largely consistent across diagnostic groups, gender identities, and countries, though group-level differences emerged for specific terms. Dyslexic participants rated neurodivergent and neurodiverse less positively than other groups, while the Autistic + ADHD group expressed the strongest endorsement of neurodivergent. Participants reported infrequent use of any of these labels to describe themselves, suggesting that while certain terms are broadly acceptable, they may not yet serve as primary identity markers. Overall, findings indicate strong agreement, with neurodivergent emerging as the most accepted term.
Lay Abstract
Language is important to how neurodivergent people express who they are and how they connect with others. Although words like neurodivergent and neurodiverse are now common in education, workplaces, and policy, we still know little about how people in these communities feel about them. This study explored the views of 740 adults with lived experience of autism, attention-deficit/hyperactivity disorder (ADHD), dyslexia, or both autism and ADHD. Participants were asked which terms they preferred, which they found offensive, and how acceptable they found six commonly used terms describing neurodiversity: neurodiverse, neurodivergent, neurodistinct, neuroatypical, neurospicy, and neurominority. Overall, neurodivergent was rated as the most positive, least offensive, and most acceptable term, followed by neurodiverse. In contrast, neurominority and neurospicy were viewed less favourably. Participants were also asked how often they use neurodiversity terms to describe themselves. These terms were generally used infrequently, with people identifying as both Autistic and ADHD more likely to use them, and people with dyslexia less likely to do so. Most participants agreed that neurodivergent was the most acceptable term overall. However, more work is needed to understand why people prefer certain words and what influences whether they use them in daily life. This study provides useful guidance for choosing respectful and inclusive language when describing neurodivergent people in policy, education, and practice.
Language can be an important means by which people construct and communicate their identity, particularly for members of groups that experience marginalisation or stigma. The words individuals choose to describe themselves are not simply descriptive; they can signal values, community membership, and sometimes resistance to externally imposed labels (Kapp et al., 2013). For neurodivergent communities, language can be closely tied to the rise of the neurodiversity paradigm, which reframes conditions such as autism, attention-deficit/hyperactivity disorder (ADHD), and dyslexia as expected and valued forms of human variation rather than deficits requiring correction (den Houting, 2019). Emerging from Autistic self-advocacy movements, the neurodiversity paradigm has become increasingly linked with social justice agendas that emphasise rights, inclusion, and autonomy (Botha et al., 2024; Dwyer et al., 2024). Yet, despite the centrality of language to both identity and advocacy within this movement, little research has examined which terms people who identify under the neurodiversity banner prefer, a significant gap given the implications for inclusive policy and respectful public discourse.
Understanding the use and preferences for neurodiversity terms extends prior work on studies seeking to understand identity displays within neurodivergent subidentities. Autism research has led much of the discussion on language preference (see Archibald et al., 2024, for a recent review), often framed around the distinction between identity-first language (e.g. ‘Autistic person’) and person-first language (e.g. ‘person with autism’; Buijsman et al., 2022; Bury, Jellett, Spoor, & Hedley, 2023b; Kenny et al., 2016). While no single global consensus has emerged, research shows increasing support for identity-first language, particularly among Autistic individuals and within English-speaking contexts (Archibald et al., 2024; Schuck et al., 2025). Research in terminology preference for other groups (e.g. dyslexia, ADHD) is not as well established, focusing mostly on opinion pieces or qualitative discussion (Evans, 2014; Mackenzie, 2017; Wissell et al., 2025); but the effort to seek meaningful and respectful language in identity formation is also important in these groups.
Social and political dimensions of identity appear to influence language preferences (Dwyer et al., 2024). Autistic individuals with stronger identification with the Autistic community, greater disability pride, or greater alignment with the values of the neurodiversity movement are more likely to prefer identity-first terminology (Bury, Jellett, Haschek et al., 2023a; Kapp et al., 2013). Later age of diagnosis is also associated with a greater preference for identity-first language (Buijsman et al., 2022; Bury, Jellett, Haschek et al., 2023a), which could be related to how learning about one's autism at older ages is associated with experiencing more positive emotions, such as relief (Oredipe et al., 2023). It may also indicate that individuals diagnosed more recently are doing so within a cultural context increasingly shaped by the neurodiversity movement, which emphasises strengths-based and less medicalised understandings of autism (Bury, Jellett, Haschek et al., 2023a), with similar patterns seen in other disability groups (Dunn & Burcaw, 2013). These findings underscore that language preference is not merely a matter of semantics, but can be closely tied to individuals’ experiences of empowerment, belonging, and resistance to medicalised or deficit-based framings.
The rise of the neurodiversity paradigm and movement has created a need for collective or superordinate terms to refer to everyone the movement advocates for. As with autism-related language, one key challenge is understanding and respecting individuals’ preferences for such terms, rather than imposing externally defined or standardised terms. One potential term is neurodivergent, an identity-first term coined by Asasumasu (2015) to refer to neurological divergence from typicality, or more precisely, neurocognitive divergence from societal norms (Walker, 2014). Although the term is sometimes misunderstood as referring specifically to diagnosable neurodevelopmental differences or conditions (e.g. Barkley, 2024; Royal College of General Practitioners, 2025), it aims to be broadly inclusive of marginalised neurocognitive minorities (Asasumasu, 2015). On the other hand, qualitative accounts suggest some students registered for neurodevelopmental disability supports reject the validity of the concepts of ‘normal’ and ‘norms’, and thus may reject the term neurodivergent (Quigley et al., 2024).
Another key term is neurodiverse, which is often used in a universal sense to describe the full range of neurological variation, including those considered neurotypical (Chapman & Carel, 2022). However, others use the terms neurodiverse as an alternative to neurodivergent: as a term for referring to marginalised neurocognitive minorities (e.g. Jenson et al., 2023; Quigley et al., 2024). Those preferring to use the term neurodiverse in this manner may perceive it as more fluid than alternatives and value that it does not create a clear distinction between neurodiverse and neurotypical people (Quigley et al., 2024). Still, while some say it reflects ‘spiky’ cognitive profiles common in neurodivergent people (Doyle, 2020), others argue it is impossible for an individual to be ‘diverse’, thus this usage is often considered fallacious and incorrect (Conner & Brown, 2025), or indicating lack of familiarity with the neurodiversity movement (Grant et al., 2025).
The alternative term neurodistinct, which circumvents the problem of individuals not being ‘diverse’, was coined to provide another alternative to neurodivergent, as the coiner perceived neurodivergent as ‘negative, separating, [and] divisive’ (Goldstein, 2020). However, this term has received little empirical attention, and existing discussion is largely conceptual or anecdotal. Neurodistinct has been criticised precisely because it fails to offer a clear distinction between neurotypical and neurodivergent people (Wise, 2023), limiting its applicability in identity or advocacy. Nor are these the only options. The term neurospicy is often used online to refer to neurodivergent people, especially those with multiple forms of neurodivergence. However, anecdotal evidence suggests it is controversial; for example, some might perceive neurospicy as informal or as dismissive of the disability and challenges faced by neurodivergent people (Annear, 2024). While others have expressed discomfort with the term neurospicy, noting its racialised and gendered connotations for women of colour and how its reclamation by others can overlook these histories (Boren, 2024; Menon, 2024).
The term neurominority has been proposed as a neutral alternative (Walker, 2021). Doyle (2020) argues that it provides a statistically and socially accurate descriptor for neurodivergent groups. Framing these groups as neurominorities also aligns with social identity theory, allowing for the conceptualisation of individuals without neurodivergent identities as the neuromajority. However, some scholars note that those who prefer not to delineate between neurominority and neuromajority groups may find this terminology divisive (Quigley et al., 2024). However, there is little research into this term.
Recent work by Dwyer et al. (2025) found that most Autistic adults supported terms describing neurodiversity, such as neurodivergent and neurominority, although preferences varied depending on participants’ alignment with the neurodiversity paradigm. Other recent research suggests that while neurodiversity terminology is widely recognised and used amongst neurodivergent groups, its uptake is shaped by tensions between stigma reduction, identity clarity, and perceived over-generality (Grant et al., 2025). However, there is limited research that articulates preferences for terms that seek to capture the identity of individuals within the neurodiversity paradigm. Moreover, we are unaware of any quantitative research examining views on neurodiversity-related terms across multiple neurodivergent populations, such as autism, ADHD, and dyslexia.
Current Study
The current study examined how individuals with different neurodevelopmental identities – autism, ADHD, dyslexia, and co-occurring autism and ADHD – evaluate a range of terms used to describe neurodivergence. While the broader research programme was informed by Social Identity approaches (see Bury et al., 2025), this study was intentionally exploratory, reflecting the limited prior empirical work on neurodiversity-related terminology and the absence of clear theoretical predictions. These groups were selected because they are among the most commonly recognised within the neurodiversity paradigm and frequently included in both research and community discussions of neurodivergence. Including these groups allows exploration of potentially diverse perspectives and experiences in how neurodivergent language is understood and valued. Specifically, we investigated which terms participants preferred, and which they found offensive or inappropriate. By comparing preferences across diagnostic groups, this exploratory study aimed to identify patterns and differences not only between neurodivergent groups but also across country (United Kingdom (UK) vs United States (US)), gender identity, and race. In doing so, it provides novel insight into how language shapes identity and belonging within and across neurodivergent communities.
Method
Participants
A total of 740 neurodivergent adults were recruited through the online research platform Prolific. Eligibility required participants to self-report a formal diagnosis of autism (n = 189), ADHD (n = 190), dyslexia (n = 179), or a dual diagnosis of autism and ADHD (Autistic + ADHD; n = 182). Table 1 provides demographic details for each diagnostic group. In terms of gender, slightly more participants identified as women (47.9%) than men (46.4%), with the remainder identifying as non-binary (3.5%), gender fluid (1.2%), or another non-binary identity (0.5%). 1 Gender distribution varied significantly across diagnostic groups, χ2(15, 733) = 35.29, p = .002, Cramer's V = .127, with higher representation of non-binary individuals observed in the Autistic and Autistic + ADHD groups. The participant mean age was 34.58 (SD = 10.51), which differed by group, as indicated by a Brown-Forsythe F(3, 712.89) = 8.20, p < .001, η2 = .032. Specifically, individuals in the dyslexia group were significantly older than those in the Autistic and Autistic + ADHD groups. The sample was predominantly White (78.1%), with smaller proportions identifying as having multiple racial identities (8.2%) or as Black (6.6%). No significant differences in race were observed across diagnostic groups. Participants were recruited from both the UK (52.6%) and the US (47.4%), ensuring a sufficient pool of potential participants while minimising cultural variation. Group distributions were balanced, except for dyslexia, where fewer US participants were available, leading to proportionally more from the UK and a significant country difference, χ2(3, 740) = 11.89, p = .008, Cramer's V = .127.
Demographic Characteristics and Group Comparisons Across Neurodivergent Identity Groups.
Notes: ADHD: attention-deficit/hyperactivity disorder.
*Brown-Forsythe statistic and Games-Howell post hoc analyses reported due to violation of homogeneity.
Values in the same row sharing superscripts do not differ at p < .05.
Procedure
This study formed part of a larger project on neurodivergent identity, stigma, community, and well-being, approved by the La Trobe University Human Research Ethics Committee and conducted between August and October 2024. Prolific pre-screening questions were used to target advertisements to adults who self-reported a formal diagnosis of autism, ADHD, dyslexia, or a combined diagnosis of autism and ADHD, without any additional neurodevelopmental conditions that they were aware of. We aimed to recruit roughly equal numbers across diagnostic groups. After responding to the advertisement on Prolific, participants were directed to a survey hosted on REDCap (Harris et al., 2019). Informed consent was obtained electronically via a checkbox after presentation of the participant information statement. Each participant received £4.50 as reimbursement for their time.
Materials
Variables for this study are taken from a broader study on neurodivergent identity. Participants first completed demographic questions (e.g. age, gender, race 2 ), followed by the primary measures of interest, with the following relevant to this project:
Autism traits
The Comprehensive Autistic Trait Inventory (English et al., 2021) is a 42-item self-report measure of Autistic characteristics. Responses are made on a five-point scale ranging from definitely disagree to definitely agree. Higher scores indicate greater endorsement of Autistic traits (McDonald's ω = .949). A total score of 134 or above suggests a higher likelihood of autism (English et al., 2021, 2024).
ADHD traits
The Adult ADHD Self-Report Scale (ASRS; Kessler et al., 2005) is a widely used six-item screener of ADHD traits in adults. Items are rated on a five-point scale from never to often, with higher scores reflecting greater ADHD symptomatology (McDonald's ω = .949).
Dyslexia traits
The Adult Reading Questionnaire (Snowling et al., 2012) is a 15-item screening tool for dyslexia in adults. In the current study, only the six items assessing reading and spelling difficulties were used, as the remaining items on attention and hyperactivity overlapped conceptually with the ASRS. Based on participant feedback and the low rates of employment among neurodivergent adults (e.g. Bury et al., 2024) one employment-related item was omitted after collecting data. Items are rated on Likert scales (e.g. 0 = never; 4 = always), with higher scores indicating greater dyslexia-related difficulties (McDonald's ω = .78).
Preference and offensiveness
Participants were asked, ‘Please indicate your preference for the following terms used to describe neurodiversity?’ and responded using a 7-point Likert-scale (1 = strongly dislike, 7 = strongly like) to each item. They were also asked ‘How offensive do you find the following terms used to describe neurodiversity?’ (1 = not at all offensive, 7 = strongly offensive), to each item. Scores were averaged across participants for each term, with higher mean scores indicating greater preference for, or perceived offensiveness of, each individual term, respectively.
Utilising the approach from Keates et al. (2025) participants were also asked ‘Are these ways of talking about neurodiversity okay?’ and were given three response options, ‘Yes, this term is ok’, ‘I do not have an opinion on this term or phrase’, or ‘No, this term is not ok’.
Terms included were those most commonly used in research and practice settings, as well as those identified within a short scoping study, 3 namely: neurodiverse, neurodivergent, neuroatypical, neurodistinct, neurospicy, neurominority.
Term usage
Finally, with a self-made measure, participants were asked to identify how frequently they use any of the terms to describe themselves on a five-point scale (1 = never, 2 = sometimes, 3 = often, 4 = frequently, 5 = always).
Analyses plan
Quantitative analyses
A repeated-measures analysis of variance (ANOVA) assessed differences in participants’ overall rankings of neurodivergent-related terms. ANOVAs examined group differences on continuous variables, using a Bonferroni-corrected significance threshold of p < .0029. Significant main effects were followed by Bonferroni post hoc tests, or Games–Howell tests where homogeneity of variance was violated. Chi-square tests examined group differences on categorical variables. For analytic power, we collapsed participants endorsing non-binary, gender fluid, or not listed gender identities into a single gender-diverse group (i.e. identities outside the man/woman binary).
Results
A one-way ANOVA showed that groups differed on individual diagnostic traits as would be expected (Table 1).
Language preference
Pearson correlations were calculated separately for preference ratings and offensiveness ratings. Specifically, correlations between preference ratings for each term and age are shown in the upper right triangle of Table 2, while correlations between offensiveness ratings and age are shown in the lower left triangle. Within each rating type, all neurodiversity terms were positively intercorrelated for both preference and offensiveness, indicating general agreement in how terms were evaluated. The biggest association in magnitude was between neurodivergent and neurodiverse, suggesting these terms are perceived similarly. A small negative correlation emerged between age and preference for neurodivergent, with no other significant relationships with age observed.
Intercorrelations Between Neurodiversity Terms and Age, With Label Preference Shown in the Upper Right Triangle and Offensiveness (in Italics) Shown in the Lower Left Triangle.
Note: *p < .05, **p < .01, ***p < .001.
A repeated-measures ANOVA was conducted to compare total mean preference scores across the six different neurodivergent terms (Table 3). Mauchly's test indicated that the assumption of sphericity was violated, χ2(14) = 548.23, p < .001, so Greenhouse-Geisser corrections were applied. Results showed a significant effect of terminology type, F(3.89) = 519.48, p < .001, η2p = .419. Post hoc Bonferroni comparisons revealed that neurodivergent was the most preferred term, which was rated significantly higher than all other terms. This was followed closely by neurodiverse, which also differed significantly from all other terms. Neurominority was rated least preferred and significantly lower than all other terms.
Mean Ratings of Label Preferences and Offensiveness and Between-Group Differences Among Neurodivergent Groups.
Notes: *Brown-Forsythe statistic and Games-Howell post hoc analyses reported due to violation of homogeneity.
Means in each row sharing superscripts do not differ at p < .05.
Means in each column sharing the same superscripts do not differ significantly at Bonferroni correct p < .05.
Patterns of preference between groups were relatively stable across the ratings of the six terms, with neurodivergent rated highest and neurominority lowest. However, there was some variability in preference for the neurodivergent term, with a significant difference between all groups, F(3, 739) = 10.90, p < .001, η2 = .043, with the dyslexia group rating neurodivergent significantly lower than all other groups.
The results of the total perceived offensiveness of terms were the inverse of the preference scores, with a repeated-measures ANOVA conducted to compare mean scores of offensiveness ratings (Table 3). Mauchly's test indicated that the assumption of sphericity was violated, χ2(14) = 1141.00, p < .001, so Greenhouse-Geisser corrections were applied. Results showed a significant effect of terminology, F(3, 43) = 299.68, p < .001, η2p = .293. Post hoc Bonferroni comparisons revealed that neurominority was considered the most offensive label, but was not significantly different from neurospicy, which was also considered relatively offensive. Neurodivergent was considered the least offensive by participants, but not significantly different from neurodiverse. All scores were below the midpoint of the scale, suggesting that overall, participants did not find any of the terms overly offensive. The general pattern of results was stable across groups. However, there was a significant difference between how groups rated the term neurodivergent, Brown-Forsythe F(3, 701.20) = 7.19, p < .001, η2 = .029, and neurodiverse, Brown-Forsythe F(3, 699.27) = 8.20, p < .001, η2 = .032, with the dyslexia group finding these terms significantly more offensive than the other groups.
Overall results for the acceptability (i.e. ‘OK to use’) of terms were similar to the other findings (Table 4). Participants largely endorsed the term neurodivergent as the most acceptable to use, with 87.4% of the sample finding that this term is okay to describe neurodiversity. There was a significant difference between groups, χ2(6) = 25.77, p < .001, Cramer's V = .132, but this reflected a difference in proportion of acceptability between groups, with neurodivergent the highest rated across groups. There were similar findings for neurodiverse, with an overall endorsement by 83% of the sample, with significant differences between groups, χ2(6) = 26.55, p < .001, Cramer's V = .134. This reflected different proportions of support for this term between groups, however neurodiverse remained second second-highest overall across participants. There were no other significant differences between groups, with neurominority and neurospicy being the least endorsed, at 44.9% and 46.1%, respectively, largely suggesting these terms are not as acceptable. More participants chose the ‘I do not have an opinion’ option than endorsed the neurominority term.
Ratings of Acceptability (“Ok to Use”) of Labels and Group Differences Among Neurodivergent Participants.
Overall, participants reported low usage of neurodiversity terms to describe themselves, with 41.7% of participants indicating they ‘never’ use these terms. There was also a significant difference between groups in how frequently they use any neurodiversity term to describe themselves, Brown-Forsythe F(3, 635.70) = 38.97, p < .001, η2 = .143. Post hoc analyses suggested that the Autistic + ADHD group used these terms significantly more frequently than the other groups, with the dyslexia group using these terms the least. On average, the dyslexia and ADHD group scores were between the ‘never’ and ‘sometimes’ scores on the scale. This suggests that participants who were ADHD or Dyslexic largely do not use any of the terms to describe themselves.
Differences by country, gender, and race
Analyses between countries (Table 5) indicated some variation in term preferences; however, the overall order of preference based on mean ratings remained consistent. Participants in the US expressed a stronger preference for the term neuroatypical compared to U.K. participants, whereas U.K. participants rated neurodistinct as more offensive than their counterparts in the US. Despite these differences, the broader pattern of preference and offence across terms was similar across countries. Although U.S. participants reported using neurodiversity-related language more frequently, this difference was not statistically significant once corrections for multiple comparisons were applied. Likewise, while country differences emerged regarding whether certain terms were judged as acceptable (Table 6), the overall pattern of results was consistent. For example, while neuroatypical, χ2(2) = 18.51, p < .001, Cramer's V = .159, and neurodistinct, χ2(2) = 21.70, p < .001, Cramer's V = .172, were more often considered acceptable in the US than in the UK, their order in overall acceptability of terms did not differ.
Mean Ratings of Label Preferences and Offensiveness and Between-Group Differences Among Country of Residence.
Notes: *Brown-Forsythe statistic and Games-Howell post hoc analyses reported due to violation of homogeneity.
Means in each column sharing the same superscripts do not differ significantly at Bonferroni correct p < .05.
Ratings of Acceptability (“Ok to Use”) of Labels and Group Differences Across Counties.
Analyses between gender groups (Table 7) revealed variability in responses to individual terms, though the overall pattern of preferences and offensiveness was consistent. Across all genders, neurodivergent was rated most preferred, followed by neurodiverse, with a substantial gap to other labels. Despite being the most preferred overall, ratings of neurodivergent differed significantly between gender groups, F(2, 294.14) = 17.83, p < .001, η2 = .038. Gender-diverse participants rated it most preferred, followed by women, with men rating it lowest. A similar effect was observed for neurospicy, F(2, 126.66) = 8.10, p < .001, η2 = .002, where women and gender-diverse participants rated it more positively than men. Differences also emerged in offensiveness ratings. For neurodivergent, F(2, 577.65) = 7.05, p < .001, η2 = .013, gender-diverse individuals rated the term as less offensive than both men and women. However, after correcting for multiple comparisons, no significant gender differences remained in the overall acceptability of terms (Table 8).
Mean Ratings of Label Preferences and Offensiveness and Between-Group Differences Among Gender Groups.
Notes: *Brown-Forsythe statistic and Games-Howell post hoc analyses reported due to violation of homogeneity.
Means in each row sharing superscripts do not differ at p < .05.
Means in each column sharing the same superscripts do not differ significantly at Bonferroni correct p < .05.
Ratings of Acceptability (“Ok to Use”) of Labels and Group Differences Among Genders.
There was, however, a significant gender difference in self-use of neurodiversity language (Table 7), F(2, 691) = 25.44, p < .001, η2 = .069. Gender-diverse participants reported using neurodiversity terms more frequently than both women and men, who did not differ significantly from each other.
Analyses across racial groups revealed a broadly similar pattern of results. Although initial analyses (see Supplemental Table S1) indicated significant differences between racial groups in preferences for the terms neurospicy (p = .040) and neurominority (p = .011), these effects were no longer significant after adjusting for multiple comparisons. No significant race-based differences were observed in perceived offensiveness (see Supplemental Table S2) or acceptability of any of the six terms, nor in the frequency with which groups used any of the terms.
Qualitative Considerations
Although no free-text question directly asked participants to explain their label preferences, 35 individuals (Mage = 33.46, SD = 10.81; women = 61.8%, men = 32.4%, gender fluid = 2.9%, prefer not to say = 2.9%) provided unsolicited comments in free-text survey items immediately following. These respondents included 10 Autistic (28.6%), 10 Dyslexic (28.6%), 10 ADHD (28.6%), and five Autistic + ADHD (14.3%) participants. Given responses were short free-text written responses, we did not consider them as sufficient for a qualitative analysis. However, we have summarised these comments here, as we felt that they offer some useful context.
Several participants highlighted that context matters – for example, stating that they preferred terms like neurodivergent or neurodiverse in formal settings, but diagnostic labels informally (e.g. ‘if I'm at work, I prefer the more formal neurodivergent. casual settings [I’ll] just say I have ADHD’ (ADHD, Male, 40)). Another example was noting that neurodiversity felt too broad (e.g. ‘if I'm honest I would rather describe myself as having ADHD as neurodiverse is very broad’ (ADHD, Female, 34)), or not representative (‘I don't like a blanket expression “neuro” it's part of my brain yes but a very small part and find it slightly offensive’ (Dyslexia, Male, 49); ‘I will be pissed off though if you lump me in with every other “neuro” crap’. (Autism, Male, 23)). Others reported being unfamiliar with some of the presented labels (e.g. ‘Never heard of these term[s] before. Indifferent to them’ (Autism, Male, 34)).
Most free text responses, however, clustered around the contested status of neurospicy. Some participants described the term as acceptable in peer contexts or as an in-group joke (e.g. ‘“Neurospicy” isn’t a generally bad term if it's coming from a neurodiverse individual; however, it can be offensive if used by neurotypical individuals’ (Autism, Woman, 23)) Others viewed it as inappropriate outside informal settings (‘I think this term is not offensive and ok to use but wouldn’t be ok … in a professional setting’ (Dyslexia, Female, 25) or ‘I would be absolutely gobsmacked and die of laughter if there was an Elsevier published research paper on autism that used the word “neurospicy”. It would be inappropriate because it's considered a post-ironic in-joke…’ (Autism, Female, 22)).
Conversely, several rejected the label outright, calling it ‘cringeworthy’ (ADHD, Female, 35) or trivialising (‘I would like it if people took neurological/psychological differences seriously and approached these concepts respectfully instead of treating it like a silly cool thing’. (Autism, Female, 19)).
Discussion
This study provides the first systematic examination of preferences for terminology used to describe a superordinate neurodivergent identity across multiple neurodivergent groups. Neurodivergent emerged as the most widely supported label by participants, being both the most preferred and least offensive. Importantly, only 4% of participants reported that it was not appropriate, highlighting its broad acceptance as a descriptor of collective identity. The term neurodiverse was also highly endorsed yet was overall less popular than neurodivergent. The relatively higher preference for neurodivergent might suggest that participants valued terminology that both acknowledges difference and signals membership in a shared community, reflecting the role of language as a tool for belonging, resistance, and empowerment (Kapp et al., 2013), and aligning with optimal distinctiveness theory (Leonardelli et al., 2010).
In contrast, neurominority was the least preferred and considered the most offensive of all presented terms. Although some theorists have suggested that neurominority offers advantages by framing neurodivergence in terms of social identity rather than pathology (Walker, 2021), or being more accurate (Doyle, 2020), its limited endorsement here suggests it may not yet resonate strongly within neurodivergent communities. One possible explanation could be neurominority invokes minority-group status and, by extension, power differentials and structural exclusion. For some participants, this framing may have felt less ‘neutral’ because it foregrounds outgroup relations, whereas neurodivergent can be interpreted as a softer descriptor that does not necessarily imply outgroup differentials.
Similar language, such as neurodistinct or neuroatypical, was also rated relatively lowly by participants. Interestingly, this pattern suggests that ongoing debates about whether terms should clearly distinguish neurodivergent from neurotypical people or reference to concepts like normality and social norms (e.g. Goldstein, 2020; Wise, 2023), may have been relatively non-salient to our participants. It is possible that less-favoured terms, and even the broader neurodiversity concept, remain unfamiliar, as reflected in some qualitative responses, or that participants perceive them as overly clinical or reductive compared to terms that emphasise diversity and identity. Moreover, concepts such as hermeneutic justice (Fricker, 2007), ensuring that marginalised groups have the interpretive resources to define their own identities, may influence how terminology evolves over time. As shared understanding grows, the familiarity and popularity of certain terms may shift. Thus, we should be cautious about interpreting low levels of support for some of the terms in this study as evidence that the terms should never be used; advocates may have compelling arguments for attempting to popularise alternative and novel terminology.
Importantly, the general pattern of findings was consistent across demographic variables. While some differences emerged, for example, U.S. participants endorsed neuroatypical more strongly than those in the UK, and gender-diverse participants rated neurodivergent more positively than women, who in turn rated it higher than men, the overall ranking of terms and their relative offensiveness remained stable. Similarly, while preferences fluctuated across terms for racial groups, the overall ranking of terms was largely consistent, with no differences between groups after adjusting for multiple comparisons. Together, this suggests that certain terms, particularly neurodivergent and neurodiverse, may be broadly safe to use across contexts, even while individual preferences should always be respected.
These results should not be interpreted as prescribing how individuals should identify. As with autism-related terminology (Buijsman et al., 2022; Sinclair, 1999), language preferences are deeply personal, and the most affirming approach remains to follow an individual's stated choice. However, at a population level, this study provides a snapshot of current preferences, which can inform more inclusive practice, policy, and public discourse by identifying terms that are less likely to cause offence.
Interestingly, although there was broad agreement on which terms were considered acceptable, participants reported that they did not regularly use these terms for self-description, with a significant proportion (41.7%) reporting never using such terms. This contrasts with findings from Grant et al. (2025), who reported that 74.8% of participants used neurodiversity-related terms. This discrepancy may reflect differences in sampling. Grant et al.'s study included a broader range of neurodivergent groups and recruited primarily through community and advocacy networks, where engagement with neurodiversity terminology may be higher. In contrast, our prolific-based recruitment may have yielded a less self-selected and less terminology-engaged sample. Within our sample, the Autistic + ADHD group reported higher usage of neurodiversity terminology, likely reflecting the presence of multiple diagnoses. However, this pattern may also point to broader dynamics of identity and identity endorsement. Consistent with other findings from this sample (Bury et al., 2025), mean endorsement or identification with a neurodivergent identity was low (at or below the midpoint of identity measure), suggesting that the umbrella category of ‘neurodivergent’ may not fully capture participants’ lived experiences or processes of identity formation. At the same time, several free-text responses indicated that context shaped terminology use, with participants sometimes adopting broader neurodivergent labels in formal or collective settings, while relying on diagnostic identities in more personal or informal contexts.
Limitations and Future Directions
While this study identifies which terms are more or less popular, it does not fully explain why. Prior research on autism terminology suggests that preferences are shaped by diverse factors such as stigma, empowerment, and the centrality of identity (Bury, Jellett, Haschek et al., 2023a; Kapp et al., 2013). Similarly, endorsement of the neurodiversity movement has been shown to predict greater acceptance of neurodiversity-related language among Autistic people (Dwyer et al., 2025). It is plausible that similar mechanisms underpin the preferences observed here, but this requires further investigation. Although we reported qualitative explanations to illustrate support for certain responses, these comments were unplanned and opportunistic; more targeted qualitative research would be needed to capture the full nuance of participants’ perspectives. Future research should therefore employ mixed-method approaches, combining quantitative surveys with qualitative inquiry, to explore the meanings and motivations underlying terminology choices. Such work could also examine how language preferences intersect with experiences of identity, community, and stigma across different neurodivergent groups. Furthermore, because the sample was predominantly White and the race data relatively coarse, findings related to race should be interpreted with caution. Future research would benefit from more targeted designs to examine how the intersectionality of different identities, such as race and gender, influence perceptions of neurodiversity terms (e.g. neurospicy; Boren, 2024; Menon, 2024), which could not be examined in detail in the present study due to limited diversity in race and gender in the sample.
Conclusion
In conclusion, this study highlights the central role of language in shaping identity and belonging for neurodivergent communities. Neurodivergent and neurodiverse emerged as the most accepted terms, whereas neurominority and neurospicy were met with scepticism or rejection. The contrasting ways participants engaged with terms like neurospicy illustrate the dynamic, contextual, and contested nature of language. These findings reinforce that not all terms – even those developed with inclusive intent – are created equal, and that centring the voices of neurodivergent people themselves is critical in shaping respectful, affirming, and socially just discourse.
Supplemental Material
sj-docx-1-ndy-10.1177_27546330261430009 - Supplemental material for Divergent, Minority or Spicy? Neurodiversity Language Preference for Autistic, ADHD, Dyslexic, and Autistic+ADHD People
Supplemental material, sj-docx-1-ndy-10.1177_27546330261430009 for Divergent, Minority or Spicy? Neurodiversity Language Preference for Autistic, ADHD, Dyslexic, and Autistic+ADHD People by Simon M. Bury, Patrick Dwyer, Rebecca L. Flower, Ellen K. Richardson and Jennifer R. Spoor in Neurodiversity
Footnotes
Ethical Approval and Informed Consent Statements
This study was approved by the La Trobe University Human Research Ethics Committee (Reference: HEC24242). All procedures complied with the ethical standards of the institutional and/or national research committee and with the 1964 Helsinki Declaration and its later amendments. Participants were recruited through the online research platform Prolific and provided informed consent electronically prior to participation. They were informed about the purpose of the study, the voluntary nature of participation, and their right to withdraw at any time without penalty. No identifying information was collected, and all data were stored securely in accordance with institutional data management and privacy policies.
Funding
The authors disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This project was supported by internal funding from the Olga Tennison Autism Research Centre at La Trobe University.
Declaration of Conflicting Interests
The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Supplemental Material
Supplemental material for this article is available online.
Notes
References
Supplementary Material
Please find the following supplemental material available below.
For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.
For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.
