Neurodiversity describes the range of differences in how people’s brains function, their behavioral traits and how they interact with the world around them. The term “neurodivergent” is commonly used to describe the experiences of people with autism, attention-deficit/hyperactivity disorder, dyslexia, dyspraxia and Tourette syndrome, among others.
According to some estimates, between 10% and 20% of the global population is neurodivergent. However, little research exists on the experiences of neurodivergent respondents when they take part in public opinion research, employee surveys and personality assessments. This void means researchers and organizations that may be concerned about inclusivity in surveys are largely unaware of how neurodivergent participants experience and respond to survey questions.
Because Gallup is committed to improving the workplace or educational experiences of all individuals, including those who are neurodivergent, we conducted a survey to explore how respondents with neurodivergent conditions experience common workplace or school situations such as working from home, applying for jobs and collaborating with others.
This article describes the methodological characteristics of those respondents, such as how they responded to the survey and when they responded to it. We compared individuals who self-reported neurodivergent conditions with those who did not.
Future Gallup publications will provide insight into the specific findings from the survey responses.
Research Method
In August 2023, Gallup fielded an online opt-in survey of 6,305 adults who had previously taken the CliftonStrengths® assessment and expressed an interest in participating in future research.
Respondents in this sample previously completed a talent assessment containing paired-comparison items, indicating that the neurodiversity of these respondents is likely not a total barrier to responding to surveys or assessments, and therefore introducing range restriction to our results. It’s important to note that our results are not representative, and Gallup did not weight the data.
In general, the sample skewed female (60.9%), older (M=33.9, SD=11.5) and more highly educated (82.1% had a bachelor’s degree or higher). Additionally, 30.4% of respondents who completed this survey resided in a country other than the United States.
Given the constraints of our sampling frame, this study should not be interpreted as conclusive research on neurodiverse respondent behavior. Instead, it contributes valuable insights for organizational leaders who want to understand survey experiences across different types of employees and aims to encourage ongoing research in this area.
How Respondents Identified Themselves in Terms of Neurodiversity
Respondents self-reported neurodiversity by answering this survey question:
Neurodiversity describes the range of differences that humans have with brain functioning, behavioral traits, and interacting with the world around them. This term is used commonly to describe the experiences of people with autism, Asperger’s syndrome, ADD/ADHD, dyslexia, dyspraxia, Tourette syndrome, and others.
Have you ever been identified or diagnosed as having any of the following conditions? Please select all that apply:
- Autism
- Asperger’s syndrome
- ADD/ADHD
- Dyslexia
- Dyspraxia
- Tourette syndrome
- Other type of neurodiversity, please specify: [open-ended, 100 characters]
- None of the above [exclusive]
- Prefer not to answer [exclusive]
Researchers opted to ask whether respondents have been identified or diagnosed with a neurodivergent condition to reach a broader group of people who may have been told they have one of the aforementioned conditions but have not been clinically diagnosed by a healthcare professional.
Gallup researchers reviewed the open-ended responses and recoded them according to the lists of neurodivergent conditions provided by Cleveland Clinic and Oxford University. Examples of write-in conditions that were recoded as neurodivergent include synesthesia, hypersensitivity and dysgraphia. Examples of write-in conditions that were not recoded as neurodivergent in our analysis include anxiety and depression.
Of the 6,305 individuals who completed this survey, 1,051 respondents (16.5%) were categorized as having a neurodivergent condition, 5,254 respondents (82.7%) were categorized as not having a neurodivergent condition (that is, neurotypical), and 48 respondents (0.8%) did not provide enough information in the survey to be classified as neurodivergent or neurotypical.
For this analysis, we compared respondents who disclosed any neurodivergent condition with respondents who disclosed no neurodivergent condition (i.e., neurotypical). We acknowledge there may be variability in how or when a respondent completes a survey based on neurodiversity, and we encourage others to consider this potential variation in future studies.
Respondents who were not classified as neurodivergent or neurotypical due to missing data were excluded.
Survey Structure and How People Responded
Straightlining occurs when a respondent gives the same answer to multiple questions in succession. The survey included four item sets that respondents could potentially straightline.
Respondents were presented with these sets of items if they were employed part time or full time, were enrolled in school full time, or were recently retired. Each group was presented with nearly identical items, with minor variations according to their workplace or school situation.
For example, employed respondents were presented with an item that read, “I am always treated fairly at work.” Students and retirees read a similar item such as, “I am always treated fairly at school” or “I was always treated fairly at work.” It should be noted that respondents were asked about neurodivergent conditions after they answered the items.
The first set consisted of five items (presented on a single screen) that measured respondents’ level of agreement with different statements, using a five-point scale with a “Don’t know/Not applicable” option. The statements covered topics such as respect, fairness and wellbeing.
The next set consisted of 21 items that asked about the ease or difficulty of completing various tasks in the workplace or at school, using a four-point scale ranging from “Extremely easy” to “Extremely difficult,” including a “Don’t know/Not applicable” option. The 21 questions were distributed across three screens. Each page contained the same subset of seven items for all respondents, with minor variations in wording due to respondents’ workplace or school situation.
For our straightlining analysis, the second section was further subdivided into three mini-batteries, each corresponding to one page of questions. In total, respondents had the opportunity to straightline on four distinct question batteries.
For respondents to be counted as having straightlined, they had to provide the same answer for every item on a single screen of the survey. Respondents may genuinely want to provide the same response for every question in a battery, making this an imperfect measure of inattention on surveys but a valuable metric for understanding how respondents engage with surveys nonetheless.
Results suggest that neurodivergent respondents straightline less frequently than neurotypical respondents do; however, rates of straightlining among both groups are relatively low among this sample. On three of the four item batteries, neurodivergent respondents were significantly less likely than neurotypical respondents to straightline their responses. Specifically, neurodivergent respondents were significantly less likely to straightline on all three of the seven-item ease/difficulty batteries (see table below).
However, neurodivergent and neurotypical respondents did not differ significantly in the rate of straightlining on the first (five-item) battery. Overall, these patterns suggest that neurodivergent respondents have approximately similar or lower rates of straightlining compared with neurotypical respondents.
The same batteries from the straightlining analysis were used to total the number of “Don’t know/Not applicable” (DK/NA) responses across the 26 items.
There is no difference between neurodivergent and neurotypical respondents in their propensity to give DK/NA responses. This suggests that on this particular survey, neurodivergent and neurotypical respondents were similarly willing to provide valid responses to the survey items.
Response Timing
The survey contained more items for neurodivergent respondents than for neurotypical respondents. The survey also did not use time stamps to measure the exact time spent on each item, so we were unable to compare mean or median response times across the two groups. Depending on work experience and follow-up items, neurodivergent respondents answered up to 12 additional items (seven open-ended and five closed-ended) compared with neurotypical respondents.
To gauge whether survey duration differed across groups, we calculated duration as the amount of time that passed between when respondents were recorded as having opened the survey and when they submitted the completed survey. We identified some respondents as having extreme survey response duration, which we hypothesize occurs when respondents start the survey at one point in time, pause their participation and then resume later.
Based on the median (542.0 seconds), 75th (919.0 seconds) and 90th (3,500.6 seconds) percentiles for survey duration, we determined that respondents with a survey duration greater than 3,500.6 seconds (or 58 minutes), the 90th percentile, were outliers.
Results show that there is no significant difference between the rate at which neurodivergent and neurotypical respondents had an extreme response duration (10.8% and 9.9%, respectively) This suggests that neurodiversity did not affect a respondent’s tendency to pause the survey and resume it later.
Reasonable survey duration may vary by respondent characteristics such as education level and survey-taking experience, and decisions about what is “too fast” involve some level of subjective input. To operationalize “speeding” or responding to the survey at an unusually fast pace, we used a cutoff of four seconds per item, assuming that to adequately read, process and respond to each item should take four or more seconds per item. This allowed us to proceed with the analysis even though neurotypical and neurodivergent respondents answered a different number of survey items.
The maximum number of items that a neurotypical respondent answered was 42 (41 closed-ended and one open-ended), while the maximum of items a neurodivergent respondent answered was 54 (48 closed-ended and six open-ended). Therefore, the cutoff to be counted as speeding for neurotypical respondents is 168 seconds (2.8 minutes), and the cutoff for neurodivergent respondents is 216 seconds (3.6 minutes).
Just 10 respondents were classified under our threshold, thus indicating potential inattentiveness. So few respondents speed that the between-groups comparison is exceedingly limited and we cannot conclude based on this approach that one group is more inattentive than the other when completing surveys.
Implications
Based on this sample of respondents who had previously completed the CliftonStrengths assessment, preliminary evidence suggests that neurodivergent respondents engaged with the survey in a similar manner as neurotypical respondents. There was no evidence that the respondents in each group experienced the survey instrument itself distinctly.
Therefore, our evidence suggests that organizations may use employee survey instruments without concerns that neurodivergent respondents are adversely affected by the medium -- if employees have already demonstrated successful completion of another survey or assessment.
To stay up to date with the latest Gallup News insights and updates, follow us on X.
