The present study was funded by a Faculty Development Grant (FDG) intended to pilot improving curricular development for teaching students with disabilities by determining which accommodative technologies would help a student to learn best given their unique disability. The pilot study was conducted on a subset of N = 155 4-year undergraduate college students (n = 111 students without disabilities and n = 44 students with disabilities) that were mostly majoring in Psychology. The study was approved by The State University of New York (SUNY) Old Westbury Institutional Review Board (IRB) and the SUNY Old Westbury Office of Services for Students with Disabilities (OSSD). The n = 44 students with disabilities consisted of learning disabilities, emotional or psychiatric conditions, orthopedic or mobility impairments, attention-deficit disorder/attention-deficit hyperactivity disorder (ADD/AD/HD), health impairments (HI), and multiple disabilities which included students with more than one psychiatric condition and requiring both different and high support needs. The Director of the OSSD was the only individual privy to the student’s disability status as they coordinate their services on campus. As students volunteered for the study, they were given coded numbers to conceal their disability status and were triple-blinded to the researchers running the actual Flanker Task. When students signed up to take the Flanker Task, they presented their code to the researchers, and this was used as the only identifier in the data. At the end of the study, the OSSD Director grouped the data identifiers and recoded them to then be sent to the researchers for further analyses. These unique codes could not be traced back to the students and were only known by the OSSD Director to ensure the confidentiality of the disability status of all participants. Both students with and without disability status were coded in this same manner. The students were then assessed using a Flanker Task, specifically to determine how sensitive it was in detecting differences in their visual attention performance. This information could be used to predict whether the student would benefit from specific accommodative technologies.

The participants consented to voluntarily take part in the pilot study to assess whether their visual attention performance could be both sensitive and predictive of the kinds of accommodative technologies that would best match their learning needs for their specific disability. The participants volunteered and were randomly assigned to participate in the pilot study across a set of four experiments. All participants were triple-blinded from the main investigators to ensure anonymity and unbiased analyses from the study’s findings.

Participants were subjected to a classic Eriksen Flanker Task (Eriksen and Eriksen, 1974) that was adapted for selectivity and for investigating executive control (Lehle and Hübner, 2008; Stins et al., 2007; Kopp et al., 1996), and it was time locked with an eye gaze visual tracking system Gazepoint GP3 (Vancouver, BC) with the flanker stimuli (i.e., horizontal black arrows) presented visually on screen through Neurobehavioral Systems Inc. (Albany, CA). In brief, the flanker task was designed with a set of three arrows in which the target arrow remained in the middle of the screen and two flanker arrows were displayed to the left and right of the target arrow. The two flanker arrows were displayed in either congruent (i.e., facing the same direction as the target arrow) or incongruent (i.e., facing opposite directions when compared to the target arrow; see Figure 1). Participants were tasked with behaviorally selecting key responses to indicate either a congruent or incongruent flanker for each trial. Participants were given 12 practice trials before starting the experiment to familiarize themselves with the task. After the practice trails, participants entered the experiment which contained 50 trials that were randomized across 5 test conditions (i.e., 10 trials of Control, Left Congruent, Left Incongruent, Right Congruent, and Right Congruent arrows). The experiment lasted approximately 10–15 min. Participants percent (%) accuracy of correctly chosen trials across all experimental conditions, their reaction time (RT) measured in microseconds (μsec), and their cumulative records for each experimental condition for both % accuracy and RT were used to infer their cognitive attentional processes. In addition, their eye tracking data were used to determine how their visual gaze as a function of external environmental distraction (i.e., noise), disability (i.e., when aggregated together), and specific type of disability (i.e., when disaggregated) would suggest which type of visual accommodative technology might best match a student with a disability to use as educational learning resources while in college.

Experiment 1: A subset of (N = 61) participants were randomly assigned to test the sensitivity of the Flanker Task in distracting conditions between a Control Environment and a Noisy Environment. A total of N = 11 males (i.e., n = 4 Control Environment and n = 7 Noisy Environment) and N = 50 females (i.e., n = 26 Control Environment and n = 24 Noisy Environment) were examined on their % accuracy and RT. This experiment was used as an internal negative control to ensure that the eye gaze technology was sensitive to external environmental factors, before assuming that it could be sensitive to internal human factors (i.e., disability).

Experiment 2: A second subset of (N = 40) participants were randomly assigned to test the sensitivity of the Flanker Task to detect any differences between students with and without disabilities. A total of N = 6 males (i.e., n = 4 students without disabilities and n = 2 students with disabilities) and N = 34 females (i.e., n = 26 students without disabilities and n = 8 students with disabilities) were examined on their % accuracy and RT.

Experiment 3: A third subset of (N = 40) participants were randomly assigned to test the sensitivity of the Flanker Task to detect any differences between students with and without disabilities. A total of N = 20 males (i.e., n = 10 students without disabilities and n = 10 students with disabilities) and N = 20 females (i.e., n = 10 students without disabilities and n = 10 students with disabilities) were examined on their % accuracy, RT, and cumulative records.

Experiment 4: A third subset of (N = 14) participants were randomly assigned to test the sensitivity of the Flanker Task to detect any differences between students with different types of disabilities. A total of N = 14 participants with disabilities (i.e., n = 3 identified as having ADD/AD/HD, n = 3 identified as having a learning/processing disorder, n = 1 identified as having an emotional/psychiatric disorder, n = 2 identified as having an occupation/motor impairment, n = 3 identified as having a neurological/health impairment, and n = 2 identified as having multiple disabilities) were examined on their % accuracy, RT, and cumulative records.

The participants signed a consent to volunteer to participate in the study. Participants were given a unique letter and number (e.g., A1 and B1) to be used as their data coding mechanism for their study and were triple-blinded throughout the study and even in their results. The participants were shown the Eriksen Flanker Task on a Dell laptop computer that was either in an uncontrolled noisy environment (i.e., a table in the main hallway of the university campus center that has the most foot traffic) or in the Psychology Department’s Neuropsychology Laboratory in a controlled sound-attenuated environment. The flanker stimuli (see Figure 1) were programmed using MATLAB (The MathWorks Inc., Natick, MA) and presented visually on screen through Neurobehavioral Systems Inc. (Albany, CA). The stimuli were programmed to display 12 practice trials before starting the experiment to familiarize participants with the task. After the practice trails, participants entered the experiment which contained 50 trials that were randomized across 5 test conditions (i.e., 10 trials of Control [target arrow presented alone either facing left or right], Left Congruent, Left Incongruent, Right Congruent, and Right Congruent arrows). The experiment lasted approximately 10–15 min. The Dell laptop computer was time locked with an eye gaze visual tracking system Gazepoint GP3 (Vancouver, BC) to track and map where the participant eyes were focusing during their response time for each of the trials prior to making a selection (Figure 2). Once the experiment ended, the participants were thanked for their participation and left the controlled or uncontrolled environment. Data were then downloaded from the computer into Microsoft Excel, reorganized, data visualizations graphed, and subsequently the data were transferred into SPSS version 24 (IBM^®: Armonk, New York, United States) for later statistical analyses.

As this was a pilot study to evaluate the potential to use and incorporate more visual attention screening tools to determine which students would benefit most or have the most difficulty with visual accommodative technologies offered to them to support their undergraduate learning in the field of psychology, the statistical analyses solely focused on the % accuracy and RT as the dependent variables of the participants visual attention on the Eriksen Flanker Task (Eriksen and Eriksen, 1974). Since the disability group was aggregated and their combined visual eye tracking heat map did not differ from the no disability group (data not shown), it did not add to the study findings and was therefore excluded. Furthermore, for the disaggregated disability sub-group datasets, since the samples sizes were small and not enough to provide a normalized/representative group average visual eye tracking heat map, these data were also excluded. Instead, the disability sub-group data were evaluated using a single-subjects design for cumulative records to illustrate their % accuracy and RT over the 10 trials for each Flanker Task test condition after pseudo-randomization. The statistics were conducted using a general linear model univariate analysis of variance (ANOVA) between-subjects design followed by a Tukey’s B honestly significant difference (HSD) test for pair-wise comparison along with a partial eta-square (η_p²) for determining effect sizes where applicable. The multi-factorial ANOVA consisted of Environmental Test Condition (Control vs. Noisy), Disability (no disability vs. disability), and Flanker Test Condition (Left Congruent, Left Incongruent, Right Congruent, and Right Incongruent) resulting in a 2X2X3 factorial design to assess the main effects and interactions between groups. This was followed by a partial eta-square for determining the Effect Size when appropriate to evaluate differences in Environmental Test Condition, Gender, and Disability as independent/quasi-independent variables between groups. All statistical analyses were conducted using SPSS version 24 (IBM^®: Armonk, New York, United States). The criteria for statistical significance were set at α ≤ 0.05 with a confidence interval of 95% (CI = 95%). Data are presented visually as bar graphs ± standard error of the means (SEM) for the % accuracy and RT, whereas the cumulative records for the % accuracy and RT are presented visually as line graphs ± SEM.

The results for males showed that the sensitivity of the eye gaze technology in distracting conditions between a Control Environment and a Noisy Environment for the % accuracy measure revealed a significant effect of Environment [F₍₁₎ = 8.784, p = 0.005**, η_p² = 0.196]. However, there were no significant effects of Flanker Test Condition [F₍₃₎ = 0.089, p = 0.966 n/s], nor was there any significant Environment X Flanker Test Condition interactions [F_(1,3) = 0.089, p = 0.966 n/s] for the % accuracy measure (Figure 3A). The results for females showed that the sensitivity of the eye gaze technology in distracting conditions between a Control Environment and a Noisy Environment for the % accuracy measure revealed a significant effect of Environment [F₍₁₎ = 20.660, p = 0.001***, η_p² = 0.097] with a significant pair-wise comparison between the groups on the Left Congruent Flanker Test Condition (p = 0.031^#), the Left Incongruent Flanker Test Condition (p = 0.014^##), and the Right Congruent Flanker Test Condition (p = 0.005^##). However, there were no significant effect of Flanker Test Condition [F₍₃₎ = 0.750, p = 0.524 n/s], nor was there any significant Environment X Flanker Test Condition interaction [F_{(1, 3)} = 0.268, p = 0.849 n/s] for the % accuracy measure (Figure 3B). The results for males showed that the sensitivity of the eye gaze technology in distracting conditions between a Control Environment and a Noisy Environment for the RT measure revealed no significant effect of Environment [F₍₁₎ = 0.018, p = 0.895 n/s], no significant effect of Flanker Test Condition [F₍₃₎ = 0.012, p = 0.998 n/s], nor was there any significant Environment X Flanker Test Condition interaction [F_{(1, 3)} = 0.187, p = 0.904 n/s] (Figure 3C). The results for females showed that the sensitivity of the eye gaze technology in distracting conditions between a Control Environment and a Noisy Environment for the RT measure revealed no significant effect of Environment [F₍₁₎ = 2.140, p = 0.145 n/s], no significant effect of Flanker Test Condition [F₍₃₎ = 0.290, p = 0.832 n/s], nor was there any significant Environment X Flanker Test Condition interaction [F_{(1, 3)} = 0.192 p = 0.902 n/s] (Figure 3D). Overall, when examining the Main Effect of Gender, it was observed to be statistically significant for % accuracy [F₍₁₎ = 3.626, p < 0.05*, η_p² = 0.097], and a Main Effect of Environment was also observed to be statistically significant for RT [F₍₁₎ = 13.162, p < 0.001***, η_p² = 0.055].

The results for males showed that the sensitivity of the eye gaze technology to detect any differences between students with and without disabilities for the % accuracy measure revealed a significant effect of Disability [F₍₁₎ = 122.909, p = 0.001***, η_p² = 0.885] with a significant pair-wise comparison between the groups on all the Flanker Test Conditions (p = 0.001^###). However, there was no significant effect of Condition [F₍₃₎ = 1.805, p = 0.187 n/s], nor was there any significant Disability X Condition interaction [F_(1,3) = 1.805, p = 0.187 n/s] for the % accuracy measure (Figure 4A). The results for females showed that the sensitivity of the eye gaze technology to detect any differences between students with and without disabilities for the % accuracy measure revealed a significant effect of Disability [F₍₁₎ = 84.145, p = 0.001***, η_p² = 0.397] and a significant effect of Flanker Test Condition [F₍₃₎ = 0.2.647, p = 0.05*, η_p² = 0.058]. However, there were no significant Disability X Flanker Test Condition interactions [F_(1,3) = 0.774, p = 0.511 n/s] for the % accuracy measure (Figure 4B). The results for males showed that the sensitivity of the Flanker Task to detect any differences between students with and without disabilities for the RT measure revealed a trend approaching significance for an effect of Disability [F₍₁₎ = 4.142, p = 0.059, η_p² = 0.206]. However, there was no significant effect of Flanker Test Condition [F₍₃₎ = 0.092, p = 0.964 n/s], nor was there any significant Disability X Flanker Test Condition interaction [F_(1,3) = 0.074, p = 0.973 n/s] (Figure 4C). The results for females showed that the sensitivity of the Flanker Task to detect any differences between students with and without disabilities for the RT measure revealed a significant effect of Disability [F₍₁₎ = 6.765, p = 0.01**, η_p² = 0.500]. However, there was no significant effect of Flanker Test Condition [F₍₃₎ = 0.077, p = 0.972 n/s], nor was there any significant Disability X Flanker Test Condition interactions [F_{(1, 3)} = 0.219, p = 0.883 n/s] (Figure 4D). Overall, when examining the Main Effect of Disability, it was observed to be statistically significant for % accuracy [F₍₁₎ = 73.441, p < 0.001***, η_p² = 0.338], and a Gender X Disability Interaction was also observed to be statistically significant for RT [F_{(1, 1)} = 11.158, p < 0.001***, η_p² = 0.072].

To assess and test the sensitivity of the Flanker Task to detect any differences between participants with and without disabilities in their ability to complete the Flanker task as before, the data were subjected to a pseudo-randomization of each test trial for each test condition. Thus, the change in test performance for each of the participants % accuracy is illustrated as cumulative records for males and females with and without a disability (Figure 5). The % accuracy cumulative record data show that the Flanker Task is sensitive in showing data separation as a function of trial as the participants advance through the test. Males without a disability typical showed a pattern of data separation at trials 6–7 (Figure 5; upper left panel), whereas males with a disability have a pattern of data separation that is shifted leftward at trials 3–4 for their % accuracy cumulative records (Figure 5; lower left panel). Interestingly, females without a disability (Figure 5; upper right panel) and with a disability (Figure 5; lower right panel) show similar data separation occurring at trials 4–5. However, the data separate more and become more variable for females without disabilities at trials 5–10 for their % accuracy cumulative records. With respect to the RT cumulative records, there were no differences in separation noted for males or females with and without disabilities and they began to show separation at trial 10 (Figure 6). Taken together, the Flanker Task was sensitive to detect shifts in males with disabilities when compared to males without disabilities and the lack of data separation in later trials with females with disabilities when compared to females without disabilities for % accuracy but not RT cumulative responses.

To further assess how a specific type of disability may affect a person’s visual attention when conducting the Flanker Task, the triple-blinded data were disaggregated by type of disability while maintaining anonymity to assess how sensitive the technology would be. The following groups were broken down based upon disability category (i.e., ADD/AD/HD, learning/processing disabilities, emotional/psychiatric disorder, occupation/motor impairment, neurological/health impairment, and multiple disabilities). First, the visual attention patterns for the neurological/health impairment sub-group were examined. Figure 7 illustrates the visual attention patterns for males (upper panels) and females (lower panels) with neurological/health impairments. In this disability sub-group, the males and females performed nearly identical on their cumulative records for % accuracy. However, their RT cumulative records were different whereby males had more variability at trials 1–5 and females had more variability at trials 6–10. Notably, in females with neurological/health impairments, the Left Congruent condition had longer and sustained RTs throughout the experiment over other test conditions. This suggests that the visual attention performance for RTs could be used as a screening tool for students with neurological/health impairments.

Second, the visual attentional patterns for the learning/processing disabilities sub-group were examined. Figure 8 illustrates the visual attentional patterns for males (upper panels) and females (lower panels) with learning/processing disabilities. In this learning/processing disabilities sub-group, the males and females performed nearly identical on their cumulative records for % accuracy. However, their RT cumulative records were different whereby only males showed more variability at trials 1–5 and females had stable RTs throughout the experiment over other test conditions. Notably, in females with learning/processing disabilities, the visual eye tracking data were insensitive to pick up any deviation in pattern. This suggests that the visual eye tracking system RTs cumulative records could be used as a screening tool for male students with learning/processing disabilities, but it might be difficult to determine whether that particular disability were the issue or if it were due to a neurological/health impairments as they shared the same pattern of RT cumulative records.

Third, the visual attentional patterns for the emotional/psychiatric disorder sub-group were examined. Figure 9 illustrates the visual attentional patterns for females with emotional/psychiatric disorders. In this emotional/psychiatric disorder sub-group, only females were found to participate in the experiment. The females’ % accuracy cumulative records were highly variable at trials 6–10 with a steep decline in accuracy (left panel). In contrast, the females performed with a more flattened cumulative record for their RT (right panel) with very fast response times for the last trials of each test condition. This suggests that the visual eye tracking systems % accuracy and RTs cumulative records could be used as a screening tool for female students with emotional/psychiatric disorders. Since no males with emotional/psychiatric disorder participated in the study, it is unclear whether their visual eye tracking data would be similar to or different from the females with emotional/psychiatric disorders datasets.

Fourth, the visual attentional patterns for the ADD/AD/HD sub-group were examined. Figure 10 illustrates the visual eye tracking patterns for females with ADD/AD/HD. In the ADD/AD/HD sub-group, only females were found to participate in the experiment. The females’ % accuracy cumulative records were highly variable at trials 4–10 with a steep decline in accuracy and clear separation of Left Congruent and Left Incongruent trials with worse accuracy than the Right Congruent and Right Incongruent trials (left panel). In contrast, the females performed with a more variable and flattened cumulative record for their RT (right panel). This suggests that the visual eye tracking systems % accuracy and RTs cumulative records could be used as a screening tool for female students with ADD/AD/HD. Since no males with ADD/AD/HD participated in the study, it is unclear whether their visual eye tracking data would be similar to or different from the females with ADD/AD/HD datasets.

Fifth, the visual attentional patterns for the physical occupation/motor impairment disorder sub-group were examined. Figure 11 illustrates the visual attentional patterns for females with physical occupation/motor impairment disorders. In this disability sub-group, only females were found to participate in the experiment. The females’ cumulative records % accuracy was absent of any variability across all trials (left panel). In contrast, the females had minor variability with a more flattened cumulative record for their RT (right panel). The visual eye tracking system was insensitive for both female’s % accuracy and RT cumulative records throughout the experiment.

Sixth, the visual attention patterns for the multiple disability sub-group were examined. Figure 12 illustrates the visual attention tracking patterns for females with multiple disabilities. In the multiple disability sub-group, only females were found to participate in the experiment. The females’ cumulative records % accuracy showed a parallel pattern that separated the Congruent from the Incongruent test conditions at trials 2–10 that remained across all trials (left panel). In contrast, the females had minor variability with a more flattened cumulative record for their RT (right panel). The visual eye tracking system was sensitive for only female’s % accuracy and not their RT cumulative records throughout the experiment.

In reviewing the literature, there are scant articles utilizing the Flanker Task in college-aged students with varying disabilities to determine whether it could be used as a prescriptive tool to assess visual accommodative needs. First, it is important to note that when implementing a Flanker Task, that researchers testing participants with and without disabilities are to ensure the congruent and incongruent trials are presented with equal frequency, to reduce the likelihood of an increased Flanker effect (Lehle and Hübner, 2008), and to be mindful of location stimulus selectivity as another by-product threatening internal validity (Lehle and Hübner, 2008). As failure to do so would limit interpretations of any Flanker Task outcomes on either population of college students. The present pilot study was able to parse populations of college students with varying disabilities and determined that it could be an effective tool in selecting certain populations of students with disabilities that might benefit from visual accommodative technologies. Prior work on large sample of children (N = 272) with and without AD/HD has reported that when presented with the Flanker Task and the Simon Task, they exhibited performance deficits in RT, percentage of errors, and efficiency in detecting incongruent trials when compared to control trials, thereby suggesting increased problems with working through interference test conditions (Mullane et al., 2009). However, studies on college-aged adults with AD/HD replicating this study have yet to be done to see whether this reduced interference skillset is recovered or exacerbated with age, and whether early interventions can serve as a mitigating intervening variable. In adults without disabilities, fronto-executive control can be correlated in the Flanker Task to an N200 event-related potential (ERP) (Kopp et al., 1996), but whether this same ERP would occur at the same time in individuals with disabilities and how it might shift either leftward or rightward relative to individuals without disabilities remains to be elucidated. A study in college-aged students in Cyprus University of Technology attempted to pilot correlates of EEGs and eye tracking and found that the eye tracking data correlated with different cognitive abilities in individuals without disabilities (Nisiforou and Laghos, 2016). Thus, if differing cognitive states could be differentiated in relation to eye tracking in individuals without disabilities, then logically it would be within reason to assume that further differentiation in these correlates may uniquely present with individuals of varying disabilities. Despite the gaps in the literature, the present pilot study proved promising as a next step to systematically begin to address this gap in the literature. What is also promising is that prior reports have shown that altered self-regulatory control of an individual’s own eye movements can be a key factor related to neurodegenerative movement disorders (Gorges et al., 2014) and could also be perhaps prescriptive in this same context. It is important to note that combining EEG or ERP with eye tracking may be the best approach to determine the cognitive strategy the person is using when they are attending to specific stimuli. Cooke (2006) aptly noted that relying on eye tracking alone that researchers can understand where the participants are looking but not “why”? It is through this framework that the present pilot study is suggesting using another correlate to overcome the “why” problem. If students can be grouped by a disability category as done herein and under these assumptions, which have their own limitations, that the individuals in a specific group would have its own distribution of their cognitive states, then we could perhaps infer from their eye tracking pattern and behavioral responses in the Flanker Task would correlate with their disability type. If future studies can repeat this work, build additional datasets that match using the same disability categories, then more aggregated data across studies, meta-analyses, and the like could be done to validate both small population samples to screen and larger population samples to run more complete studies. As most colleges are seeing an increase in enrollment of students with a range of disabilities having the ability to screen them with an eye-tracker and a simple Flanker Task in less than 20 min to determine whether they would benefit from a visual accommodative technology may have a critical positive impact on their education. If they were to benefit greatly from the visual accommodative technologies available at their college, they may find a new set of supports for them facilitating their persistence and timeliness to degree completion. However, this would require students to self-disclose to receive such support services in college, and there are many students that may not self-disclose making such prescriptive screening efforts self-limiting. Notably, if a greater number of students with disabilities were to benefit from such a prescriptive screening process, data collected, publicly shared on its usefulness, and student impacts, then perhaps more students with disabilities may come forward, self-disclose, and seek such supports and services in college. Thus, more work is warranted in exploring these factors. This pilot dataset will hopefully influence institutional policy changes to improve visual accommodative technologies that will match the needs of undergraduate students with disabilities to provide them with the resources they need to complete their psychology degrees in an optimal manner. Finally, this pilot study may become the beginning of a richer conversation on screening and assessment methods for undergraduate college students with disabilities to help them in selecting the best matched accommodative technologies to support them as early college interventions for adult learners.

The present pilot study is not without limitations. First, the sample size for the students with disabilities was quite small and the overall pool lacked equal representation for males and females. There can be several reasons for this and as such make it difficult for any comprehensive research study; hence, why the present pilot study is being used to start an important societal conversation regarding the matter. Students with disabilities may feel less capable and confident to enroll into college and may become less certain of the kinds of supports, assistance, and services that can be provided to them at the college level, which may make them less likely to enroll in college. In contrast, even if they enroll in college, as an adult is it their choice to self-identify as having a disability which is very different from their earlier ages in which their parents will lead such discussions. It is very possible that as students with disabilities that have successfully achieved the educational ability to enroll in college, may seek to reduce their stigma and labels to integrate more into the college experience without being detected as having a disability. They can cause two problems: (1) less participants with disabilities for studies such as the present one conducted and (2) students with disabilities that are unwilling to self-identify may then alter the students without disability data when aggregated together thereby counteracting such studies to be as clean as possible in their data findings. To this point, it may not be entirely possible to conduct full statistically sound sample sizes for such study purposes and may make it prudent for a series of pilot studies such as the present one to be published first. Then, perhaps a meta-analysis of these varied pilot studies could be used to attempt to resolve this issue, or retrospective analysis may also prove useful in overcoming the problem. It may take many years for the required comfortability and trust to take place in college students with disabilities to self-identify to be studied in an effort to help other students with similar and different disabilities at the college level. The unwillingness to self-identify as well as a fear of stigmatization is a reasonable attribution for the relatively small size of the pool of students with disabilities that were examined in the present study that corroborate with prior reports (Grimes et al., 2017). Furthermore, by National Science Foundation, National Center for Science and Engineering Statistics (2018) estimates, 78% of all psychology undergraduates are women, and this is consistent with the 9:1 ratio of females to males in this present study, since the participant pool was primarily comprised of psychology majors. It should be noted here that while there is a small gender difference between students with identified disabilities, approximately 64% of students with disabilities are female, as compared with 57% of students without disabilities who are female (U.S. Government Accountability Office, 2024), and this difference was washed out given the enormous gender differences in the population of students majoring in psychology. These two very real challenges have limited these findings to a pilot study. Future research, with a larger participant pool of undergraduate college participants with disabilities, should explore the data trends as indicated by this pilot study. This would permit future researchers to determine whether the six disability sub-groups exhibit different patterns of responses with respect to the visual attentional data and the visual eye tracking heat maps, as well the degree to which to those patterns can be used to characterize the response patterns across groups.

This pilot study was the first of its kind to evaluate college-age students at the undergraduate level and to see whether using cognitive psychology tests such as the Flanker Task could be used to detect differences in visual attention and eye-gaze patterns. Unfortunately, much of the literature on individuals with disabilities are limited to children and adolescents and there are less reports on adults with disabilities in college. Moreover, the literature is scant on examining differences in visual processes of students with disabilities, and perhaps it is due to the lack of clarity in grouping students in ways that may make it easier to identify similarities and differences in their abilities despite their disabilities. Much of these codification systems have evolved substantially over the years, and this has offered a more wide-spread understanding of the range of disability for a given condition as well as the increased inclusivity of people along a neurodiversity spectrum. The students with disabilities in the present study were codified into six main categories: learning disabilities (LD), emotional or psychiatric conditions (EPC), orthopedic or mobility impairments (OMI), attention-deficit/hyperactivity disorder (AD/HD), health impairments (HIs), and multiple disabilities (MDs). These groups proved to be useful when examining them for eye tracking and visual attention tests using the Flanker Task. What can be learned from this pilot study is that future studies that want to build upon the present work could attempt to replicate what was done herein to see whether the same findings arise or with larger sample sizes could build upon our population datasets (i.e., using statistical Jack Knifing, Boot Strapping, or Monte Carlo methods) to simulate and build upon the sample size to see how the data might look like if it approached a large sample simulating a normal distribution. However, cation must be exercised as mathematical modeling may serve to provide insight and direction but may not always accurately model how humans might respond to certain stimuli or conditions. Thus, the best approach would be to recruit more students with disabilities to participate in similar kinds of study in the future. This recruitment effort depends on a balance of very sensitive issues regarding student confidentiality, triple-blinded processes, collaboration with the college’s students with disability services department, and the willingness for students with disabilities to participate in the study. These factors are all predicated on whether students with disabilities enroll into college and self-identify, which can be a difficult challenge for many of the reasons discussed earlier. Thus, if the field wants to develop more supports for students with disabilities in colleges at the undergraduate level, perhaps more collaborative opportunities between psychological researchers, psychological teachers/educators, supports for students with disabilities, and the administration should develop strategic plans to afford a more neurodiverse population of students with new assessment tools to help them along their most formative educational years. The findings from the present study collectively show that the Flanker Task was sensitive to parse students with differing disabilities to be screened for visual or other disability accommodative software that can help them with their college studies (i.e., Kurzweil 3,000, ZoomText, Fusion, and/or Onyx). However, since this is a pilot study, more work will need to be done to replicate, approach more of a meaningful sample size, and help follow up with pre- and post-test analyses of using such accommodative software from these screening processes. This study also suggests that perhaps other classical cognitive tests, such as the Flanker Task, could also be used in a similar manner, but that remains to be elucidated. Notably, such efforts will take some time before we can truly screen and facilitate students with disabilities at the college level in the ways the present study began to explore.