Neurological disorders such as Attention-Deficit/Hyperactivity Disorder (ADHD), dyslexia, and Alzheimer’s disease often remain undiagnosed which limits opportunities for early intervention. This study investigates relationships between personal and familial histories of these disorders. Connections were also drawn between disorder history and the participant’s performance on a game-like cognitive/memorization assessment. Participants completed a task in which they had a small amount of time to memorize two randomized letters and determined whether they appeared in alphabetical order, with memorization time decreasing as they progress through the rounds to increase difficulty. Statistical analyses were performed using the Mann-Whitney test and the Chi Squared Test for Independence. The main findings emphasize known hereditary patterns of ADHD and dyslexia, and suggest potential connections between dyslexia and Alzheimer’s disease that may pave the way for as a tool in earlier intervention efforts.
Introduction
Introduction
Although detrimental, most neurological disorders seem to go undiagnosed for prolonged periods of time before becoming increasingly evident in daily life. While many modern drug-based solutions have been developed concerning these disorders, they are often designed to act as a coping/assisting mechanism as the disorder becomes progressively worse, meaning that the assistance does not lessen the issue but instead makes it tolerable. Existing drugs such as Adderall for ADHD or Donepezil for Alzheimer’s, while somewhat beneficial, often seek to treat already occurring symptoms, making them incredibly ineffective when it comes to early action. Fortunately, research reveals that levetiracetam, a decades-old, FDA-approved anti-seizure medication, could potentially stop Alzheimer’s before it starts. However, the key issue here is that the drug would need to be taken decades before memory loss begins, a detection issue that the drug alone cannot solve. The main disorders that will be looked at here is ADHD, dyslexia, and Alzheimer’s in the extended family. Such disorders affect both the motor skills and memorization capacity of the patient.
This essay analyzes relations between neurological disorders that the user or their family members (both immediate and extended) and to construct a game based tool that exhibits associations with the incidence of these neurological disorders. This game chose two randomized letters of the alphabet, and the user began with three seconds to memorize the letters before they disappeared and two options appeared. The options were “in order” (were the two letters given in alphabetic; order) or “out of order” (were the two letters given not in alphabetical order). The time that the user has to memorize the letters decreased each of the eight times that the user has to memorize the given letters. After collecting all the data, I compared demographics (age, ethnicity, gender, generational ADHD and Dyslexia risk, user dyslexia and ADHD risk) to the user’s respective game scores. I sought to identify correlations with and within disorders and hopefully aid their early detection. This game was specifically chosen because it simultaneously engages short-term memory, attention, information processing, and decision-making, cognitive functions that are often affected by ADHD, dyslexia, and Alzheimer’s disease.
Two main statistical analyses were used in order to quantify the extent to which these relations existed: Mann-Whitney and Chi Squared Independence Test. Firstly, the Mann-Whitney test was used to example relationships between categorical and quantitative variables. Secondly, the Chi Squared Independence Test was used to do the same using categorical and categorical variables.
Introducing the Functions
Introducing the Functions
First, I built helper functions that I could use in my game. They allow me to organize my code and manipulate variables in a cleaner method that I can efficiently integrate into the final DynamicModule.
spacedLetters
spacedLetters=StringSplit["ABCDEFGHIJKLMNOPQRSTUVWXYZ",""];letters=RandomSample[spacedLetters,2];visibleLetters=letters;
In the first round, you have 3 seconds to remember the letters before they disappear and you have to choose if the letters you just saw were in order or not. For example, if “A,B” is displayed, then the letters are in order and if “B,A” are displayed then the letters are not in order. will be manipulated throughout the code and will consistently decrease by a factor of 0.8, ultimately shortening the time that the letters stay until a minima of 0.67 seconds.
pauseTime
pauseTime
pauseTime=3;
record=<||>;buttonEnabledVar=False;startButtonEnabled=True;
Two random letters from the variable are displayed.
spacedLetters
By shuffling between turns in , this will always be randomized from turn to turn:
shuffleLetters[]
shuffleLetters[]:=(letters=RandomSample[spacedLetters,2];visibleLetters=letters;);
After the allotted time for memorization is finished, the letters need to be removed before the options show up.
clearLetters
pauseTime
clearLetters[]:=(letterCheck=Append[letterCheck,visibleLetters];visibleLetters={"",""};pauseTime=pauseTime*0.8;);
On each turn, the player’s choice needs to be checked against a correct answer. The player is awarded a point if they answered correctly, and this value should be recorded.
check
AlphabeticOrder
record
check[choice_]:=If[(AlphabeticOrder[letters[[1]],letters[[2]]]>0&&choice=="in")||(AlphabeticOrder[letters[[1]],letters[[2]]]<0&&choice=="out"),record[N@ToString[pauseTime]<>"secs"]=1,record[ToString[N@pauseTime]<>"secs"]=0];
Below are all the questions that I asked on the form in order to conduct effective data analysis. By comparing personal diagnoses, family histories, and game performance, I was able to determine whether inherited neurological risk factors were associated with measurable differences in the task. This data later enabled the identification of significant relationships between neurological conditions, generational risk factors, and game outcomes, helping evaluate whether performance on the game could contribute to future methods for early screening and detection. This data also allowed me to later draw connections between generational disorders and the users disorders, creating patterns that allowed for some sort of fabrication of a prediction of who would be at risk for one of the given disorders based off the results supplied.
The ID number of each user was requested to prevent duplicates:
Text["What is the red ID number on your Bentley Summer Guest Access Card? For example, mine is 800002011 (this is to prevent duplicate entries)."],InputField[Dynamic[idNumber],Number]
Age, gender, and ethnicity was collected to identify relations between surface-level demographics and the user’s results from the game:
Text["What is your Gender"],Control[{genderChoice,{"female","male","other","n/a"}}]
Text["What is your age?"],InputField[Dynamic[ageChoice],Number]
Text["What is your Ethnicity?"],Control[{ethnicityChoice,{"Asian","White","Black","American Indian or Alaska Native","Native Hawaiian","other"}}],
Questions regarding ADHD, dyslexia, and Alzheimer’s disease were included because these neurological disorders are known to affect cognitive functions that are directly relevant to the Alphabet Organization game.
ADHD is associated with differences in attention, memory, and functioning:
Text["Do you have ADHD?"],Control[{adhdYN,{"yes","no"}}]
Dyslexia is important for when the user answers questions concerning the organization of the letters:
Text["Do you have Dyslexia?"],Control[{dyslexiaYN,{"yes","no"}}]
In addition to information about the user themselves, I also collected information about surface-level family history for all three disorders.
ADHD is known to have strong genetic components:
Text["Does your immediate family have ADHD?"],Control[{adhdGenerationalYN,{"yes","no"}}]
Asked about generational ties to dyslexia:
Text["Does your immediate family have Dyslexia?"],Control[{dyslexiaGenerationalYN,{"yes","no"}}]
Alzheimer’s is characterized by declines in memory and cognitive processing. Since the game requires participants to memorize letters, process their order, and then make quick decisions, these conditions were expected to potentially influence performance.
Asked about generational ties to Alzheimer’s instead of the user themselves since it would be difficult to know if the student has a disorder not typically discovered until later in their lifetimes:
Text["Does your extended family have Alzheimer's?"],Control[{generationalAlzheimers,{"yes","no"}}]
There are numerous different disorders that may possibly affect this study.
otherNeuro
Text["Do you have another neurological condition not addressed above?"],Control[{otherNeuro,{"yes","no"}}]
Text["Does your immediate family have another neurological condition not addressed above?"],Control[{otherGenerationalNeuro,{"yes","no"}}]
Next, I needed to organize the game to make it visually simple to navigate for the users. This interface is quite simplistic in design, question, text box, question, two buttons. The two options for the letters is that they can be alphabetically “In order” or “Out of order”. The “Start Game” and “Submit Answer” button will coexist. When the user clicks “Start Game”, then the first two randomized letters will appear, after the letters disappear and the user choses one of the two options, the Start button will turn into a “Submit Answer” button, and when the user clicks it, the game will again give two new randomized letters for the next round.
Start Game button will begin the game and output a randomized pair of letters for the first round. After you select your choice, the Submit button will pop-up:
If[btnText=="Start Game",btnText="Submit Answer";,check[If[inORout=="In Order","in","out"]];buttonEnabledVar=False;];
The game will automatically stop itself if the memorization time drops below 0.7 seconds:
If[pauseTime<0.7,btnText="Click big shiny submit button below!";submitEnabledVar=True,shuffleLetters[];Pause[pauseTime];clearLetters[];buttonEnabledVar=True;];
Beneath the “Submit Button” as part of the game, there is a far larger Submit button which sends the answers to the entire form as well as your score on the game into the databin. After you click the button, the game will tell you submitted to indicate that your answers were sent to the databin. Each answer variable will store all the answers given to that question. (Ex: idNumberQ would store a large list of all the id Numbers that users inputted in the form, adhdQ would store all the Yes’ and No’s received as an answer to the question “Do you have ADHD?”). All of the data was stored under a variable “finalData”.
On submit, data is sent to a Databin and the game ends:
Button["Submit",(Echo["Submitted!"];submitEnabledVar=False;DatabinAdd["1FtnH9nvu",<|"idNumberQ"->idNumber,"adhdQ"->adhdYN,"dyslexiaQ"->dyslexiaYN,"otherNeuroQ"->otherNeuro,"gameRecord"->record,"genderQ"->genderChoice,"ageQ"->ageChoice,"ethnicityQ"->ethnicityChoice,"adhdGenerationalQ"->adhdGenerationalYN,"dyslexiaGenerationalQ"->dyslexiaGenerationalYN,"otherGenerationalNeuroQ"->otherGenerationalNeuro,"generationalAlzheimersQ"->generationalAlzheimers,"incorrectLetterPairs"->letterCheck|>]),Enabled->Dynamic[submitEnabledVar]]
Below is the form and the game that all users were given. All Yes/No questions were defaulted to Yes, the gender question was defaulted to female, and the ethnicity question was defaulted to Asian.
In[]:=
dataCollectionForm=
Out[]=
What is the red ID number on your Bentley Summer Guest Access Card? For example, mine is 800002011 (this is to prevent duplicate entries). | ||||
| ||||
What is your Gender | ||||
What is your age? | ||||
| ||||
What is your Ethnicity? | ||||
Do you have ADHD? | ||||
Does your immediate family have ADHD? | ||||
Do you have Dyslexia? | ||||
Does your immediate family have Dyslexia? | ||||
Does your extended family have Alzheimer's? | ||||
Do you have another neurological condition not addressed above? | ||||
Does your immediate family have another neurological condition not addressed above? | ||||
Data Pre-Processing
Data Pre-Processing
While each variable of the final data may say 71 entries, it’s important to note that there were duplicates for the people who may have been confused the first time filling out the form. The true number of entries was 62, which means that I needed to eliminate 9 duplicates of data from every variable before beginning any statistical analyses.
Data is retrieved from the Databin that game data was pushed to:
In order to remove duplicates, I first found the index for each of the duplicate rows of data that I had to remove from the dataset:
Started storing letters pairs from submission #17, so backdated entries are inserted:
I decided to put the idNumberQ column first simply so that it would be easier to differentiate between rows where the identification is in the front :
Next I had to decide by what metric (using the game data) I wanted to draw conclusions from. Originally, I chose to generate a list of the smallest time that each user got the answer correctly, meaning the quickest that they were able to memorize the letters. Unfortunately, everyone was able to get the smallest amount of memorization time (at 0.67 seconds) correct, which would not have led to any valuable conclusion. Instead, I decided to use the earliest time that each user got a question incorrect.
Picking the values with game records of 0 (which means that they got the answer incorrect) would return either none, one, or multiple values for each sample:
Since I wanted to find the earliest time that the answer was incorrect, I made a list that picked the first instance.
numIncorrectList is a list of all the earliest times in the game that each of the users got a question incorrect:
Statistical Significance Analysis
Statistical Significance Analysis
After collecting data using my form, I wanted to determine relations between the disorders themselves as well as the disorders and the possible effect of said disorders on the users game performance. I accomplished this analysis using the Mann-Whitney Test to determine if there is a statistically significant difference between two independent groups, and was able to determine if there was a significant association between the incidence of dyslexia and scores on the alphabet game.
Associates the earliest time that each person got a question wrong in the game with the person’s respective id number:
Dyslexia
Dyslexia
Diagnosed with Dyslexia
Diagnosed with Dyslexia
Associates whether each person has dyslexia or not with the person’s respective id number:
The significance level that was chosen was 0.05, the value returned by our test was 0.451542, which is greater than the significant level . This means that dyslexia independently in the user does not necessarily have a significant impact on the first time in the game that they made a mistake .
The following code tests the significance of dyslexia on the users incorrect game record:
The smooth histogram shows that the dyslexia group is bimodal, with two peaks around 1.2 and 3.2 seconds, while the non-dyslexia group is unimodal and more concentrated between 1.5 and 2.0 seconds. However, the data overlap between the two distributions suggests that participants with and without dyslexia performed similarly, supporting the Mann-Whitney result that there is no statistically significant difference between the groups.
Since the p-value is above the significance level of 0.05, and the concentrated areas on the histogram are quite dispersed, I came to the conclusion that Alzheimer’s in the user’s extended family does not necessarily have a significant impact on the first time in the game that they made a mistake.
Family History of Dyslexia
Family History of Dyslexia
Associates whether each person has dyslexia in the family or not with the person’s respective id number:
The users generational Dyslexia in the immediate family does not necessarily have a significant impact on the first time in the game that they made a mistake.
Alzheimer’s
Alzheimer’s
Associates whether each person has Alzheimer’s in the family or not with the person’s respective id number:
ADHD
ADHD
Diagnosed with ADHD
Diagnosed with ADHD
Associates whether each person has ADHD or not with the person’s respective id number:
Family History of ADHD
Family History of ADHD
Associates whether each person has a family history of ADHD or not with the person’s respective id number:
Users with ADHD in immediate family perform significantly differently on this test to those without. A stark difference from the personal diagnosed ADHD which was discussed earlier. A possible explanation for such a phenomenon is that those who have generational ADHD could have undiagnosed ADHD, meaning that the provided Alphabet test could aid in early detection.
Associations with Game Record
Associations with Game Record
Analysis of relations between two categorical variables (that is, between the incidence of two disorders) had to be done using a Chi-Squared Test for Independence.
P-Values
P-Values
Using the Chi-Squared Test on every possible pair, similar general conclusions were able to be drawn. Many results were seen to be significant (p < 0.05), it was found that the participant’s family having Alzheimer’s and the participant having dyslexia were correlated, indicating that dyslexia could be a precursor to Alzheimer’s later in life. This is essential to early identification since I was able to make the analysis that if a child has dyslexia at a young age, they should be closely monitored for symptoms of Alzheimer’s as they get older. Similarly, it was seen that there was a significant relation between participant’s family having Alzheimer’s and the participant’s family having dyslexia.
Histograms
Histograms
Display a row of all the histograms made. Each one compared a different variable from the form to the times of each incorrect answer:
A more simplistic noticing was the participant’s family having dyslexia and the participant also having dyslexia, effectively validating most literature surrounding the fact that dyslexia is a strongly heritable disorder. Much like the reasoning for the first noticing, I saw that there was a relation between the participant’s family member having ADHD and the participant having dyslexia, this could also indicate that if a child has ADHD, it may be critical for them to also get tested for dyslexia. While this is not always the case, numerous studies have found the interconnectedness of both disorders with cognitive and some neural overlap. In tandem with this study, there was a correlation between the participant having dyslexia and the same participant having ADHD, furthering their intertwined nature and giving more of a reason to use the knowledge of having one of the disorders to test for the other early on. The final two observations that the participant’s family having dyslexia and the participant’s family having Alzheimer’s as well as the participant’s family having dyslexia and the participant having ADHD act as further evidence for the above conclusions.
On the other hand, the values of pairs between the participant’s family having Alzheimer’s and the participant having ADHD as well as the participant’s family having Alzheimer’s and the participant’s family having ADHD were not significant (p > 0.05). There is a logical rationale behind Alzheimer’s and ADHD being completely independent of one another. It is found that ADHD, which is a neurodevelopmental disorder, affects approximately 5% of children and 3% of adults globally, while Alzheimer's, which is a neurodegenerative disorder, is the leading cause of dementia in older adults, highlighting a key difference between the two.
Conclusion
Conclusion
In this study, I built an Alphabet Organization game in which you have to memorize a randomized pair of letters at increasingly faster speeds. I then explored how demographic factors and neurological disorders influence performance in an Alphabet Organization game, using numerous statistical analysis methods to uncover deeper patterns and connections what would otherwise be overlooked. For instance, while the user themself may have ADHD, whether it be professionally or self-diagnosed, it did not significantly impact their performance. On the contrary, having an immediate family member with ADHD was associated with significantly different performance. This finding is incredibly notable as ADHD is one of the most heritable neurodevelopmental disorders. One possible explanation is that some participants with a family history of ADHD may be undiagnosed themselves, suggesting that cognitive tasks such as the Alphabet Organization game presented to them could potentially act as an early detection tool for identifying those who may benefit from further evaluation.
The analysis also revealed significant relationships involving dyslexia, which lines up with extensive literature researching and validating the heredity of dyslexia. Furthermore, significant relationships between ADHD and dyslexia variables may be explained by the well-documented comorbidity between the two disorders. Research suggests that ADHD and dyslexia share both genetic risk factors and overlapping neural mechanisms. One particular replication study (Liloia et al., 2022) analyzed patterns of gray matter alteration associated with ADHD and dyslexia, ultimately finding that both disorders occur together more frequently than expected by chance.
Future Directions
Future Directions
All things considered, while this study identified several significant relationships between neurological conditions and performance in my Alphabet Organization game, there are several ways that the study could be improved and expanded in future research. One of the most important modifications I would make would be to reduce the starting pauseTime, which was the period of time that the participants had to remember the pair of letters they were given. The relatively long memorization times sued in this study likely made the game easier for many participants, and I noticed some rehearsing the alphabet sequence even before the options were given, rejecting the main notion of the game. As a result, even individuals with neurological conditions were often able to complete the game with few or no mistakes, mostly obscuring the differences in cognitive processing and reducing the sensitivity of the test. Future versions of the game would also likely include larger sets of characters, time pressure, or distracting elements that place a greater emphasis on working memory and attention, both integral symptoms of the conditions studied with this game. The main parts that must be addressed are differences in the participants processing speed and cognitive awareness. Additionally, a larger and more diverse participant pool would improve the reliability of the statistical analyses. Collecting information about more unique data points such as severity of symptoms, age of diagnosis, and treatment history could provide a greater insight into how neurological conditions influence game performance. When combined with more nuanced datasets, these results could be used to train machine learning models capable of identifying subtle indicators of neurodevelopmental disorders. Although not professional, and thus not as trustworthy, such a tool could serve as an efficient accessible, low-cost screening method in minorities and other community settings, especially in a rapidly advancing digital age, safely enabling earlier identification and thus intervention for individuals who may otherwise go undiagnosed.
References
References
A new drug could stop Alzheimer’s before memory loss begins. (2025). ScienceDaily. https://sciencedaily.com/releases/2025/12/251222080119.htm
McGrath, L. M., & Stoodley, C. J. (2019). Are there shared neural correlates between dyslexia and ADHD? A meta-analysis of voxel-based morphometry studies. Journal of Neurodevelopmental Disorders, 11(1). https://doi.org/10.1186/s11689-019-9287-8
Riccardo Borgonovo, Nespoli, L. M., Martino Ceroni, Arnaud, L. M., Morellini, L., Lissi, M., & Sacco, L. (2025). Potential Genetic Intersections Between ADHD and Alzheimer’s Disease: A Systematic Review. NeuroSci, 6(4), 97–97. https://doi.org/10.3390/neurosci6040097
Liloia, D., Crocetta, A., Cauda, F., Duca, S., Costa, T., & Manuello, J. (2022). Seeking Overlapping Neuroanatomical Alterations between Dyslexia and Attention-Deficit/Hyperactivity Disorder: A Meta-Analytic Replication Study. Brain Sciences, 12(10), 1367. https://doi.org/10.3390/brainsci12101367
Early detection and diagnosis: how machine learning can bring healthcare’s holy grail within reach // van der Schaar Lab. (n.d.). Van Der Schaar Lab. Retrieved July 8, 2026, from https://www.vanderschaar-lab.com/edd/
Artificial intelligence was used as a search tool in early research for this project.
Acknowledgements
Acknowledgements
Firstly I would like to thank my mentor, William Choi-Kim, for his incredible guidance throughout the entirety duration of this project. I would also like to thank Jason Sonnenberg for helping me clear my interests to get a lovely project. Additionally, I would like to thank Aanya Gupta, Kaavya Majumder, and Eileen Lee for all helping me gain a deeper understanding of everything from the simplest of ideas to the most complex of concepts. Thank you to all the students in my advocate and mentor groups, and, above all, the friends in my TA group. Thank you to Stephen Wolfram and the entire WSRPi staff for supporting my ideas and creating such an exciting program and unforgettable opportunity.
Additional thanks to Stephen Wolfram for suggesting this project.
CITE THIS NOTEBOOK
CITE THIS NOTEBOOK
Developing a lexical diagnostic tool and examining correlations with neurological disorders
by Adrika Mehta
Wolfram Community, STAFF PICKS, July 9, 2026
https://community.wolfram.com/groups/-/m/t/3749400
by Adrika Mehta
Wolfram Community, STAFF PICKS, July 9, 2026
https://community.wolfram.com/groups/-/m/t/3749400