A review of executive function deficits in autism

Post Reply New Topic

Here's the blurb from my ASD report, about EF. I'll look at my ADHD report to see if it says anything.

The ADHD report just says I have extremely low EF. It doesn't break it down into any component parts.

Thanks. There are some new (to me) types of EF in there, and some new descriptions of some of them:

Shifting - Does that mean switching tasks?

Emotional control - That's an executive function thing?

Self monitoring - Don't know what that means.

Initiating (begin a task, fluidly generate ideas) - first part clear, don't understand second.

Task monitoring (check work, assess performance during or after finishing task to ensure attainment of goal, keep track of effect of own behaviour on others) - mostly clear, but don't understand why they have that last item there - effect of behaviour on others. I guess they're referring to teamwork of some kind.

Organisation of materials (keep workspaces, living areas and materials ordered) - that's clear to me.

Keep track of effect of own behaviour on others, attend to own behaviour in the social context - they seem to have written this in a muddled way, as they've already referred to the first item under Task Monitoring. Don't see how second term differs from first.

Working memory: stay with or stick to an activity - I always thought working memory was simply the ability to retain information in the short-term memory in the face of having to focus on other things between getting the information and having to recall it, so "sticking to an activity" surprises me.

Planning / organisation (anticipate future events, set goals, develop appropriate steps carry out an associated action, carry out tasks systematically, understand main ideas) - I guess that mostly makes sense.

Is their report slightly muddled in places or is it me? To me it seems a little confusing, though that might be because some of the terms are unfamiliar and because I often get hung up on the tiniest deviations from perfectly methodically and systematically presented data. And other people's ways of categorising information often confuse me and make me want to improve them so that they make more immediate sense to me.

I see your mother rated you better than you rated yourself on most things, especially emotional control and organising materials. Could you have been a little hard on yourself? You seem to rate yourself as the worst possible in every category.

Anyway, it looks like useful stuff to deepen the understanding of EF. No doubt my confusion will die down as I get more used to the ideas and their method of categorising the subject. I suppose different experts would have different systems so some of it probably would be kind of arbitrary.

Persistent Preoccupation With Parts or objects. Basically that is my whole career.
Also I am one of those people who feels he knows nothing about a subject until he knows EVERYTHING about a subject.

I once asked my ADHD coach "what is the FUNCTION of the executive function?"
She thought about it and eventually pointed me to an article.
The article contained this diagram:


(click on it to download)

As a computer scientist the article makes sense to me and helps me to think about executive function.
The section about "error correction especially sticks with me".
All the other descriptions of EF come down to the error correction working or not working.

Here is the full text of the article - sorry if this is TMI - I cannot find the article on line anymore

The Annals of the New York Academy of Sciences: Structure and Function of the Human Prefrontal Cortex, Vol. 769, 1995
A Computer Model of Prefrontal Cortex Function
Richard Levinson
Recom Technologies Inc.
NASA Ames Research Center
Computational Sciences Division
Mail Stop: 269-2, Mo‚ÄÄett Field, CA 94035
E-mail: [email protected]
This manuscript describes a computer model of prefrontal cortex function that was de-
signed by integrating the perspectives of neuropsychology and arti‚ÄÄcial intelligence. The
model shows how several neuropsychological theories of frontal lobe function can be com-
bined into a single computer model. The model also extends those component theories by
focusing on information processing details that glue the pieces together.
This work is motivated by the following points: Neuropsychology and Arti‚ÄÄcial Intel-
ligence (AI) describe complementary parts of a model for autonomous action. Neuropsy-
chological models provide descriptions of how planning and reaction must be integrated for
human autonomy, but the planning component is poorly understood and many information
processing details have not been
ushed out. In contrast, AI provides implementations of
independent planning and reaction modules, but their integration is poorly understood. A
computer model forces information processing issues to be addressed in detail. A computer
model can also be tested more easily than models based on verbal descriptions because it pro-
duces behavior that can be compared directly with clinically observed behavior. In contrast,
verbal models must be interpreted subjectively to predict and test their behavior.
NASA is interested in applying this model towards the development of autonomous in-
struments and spacecraft. Today's autonomous control technology is limited by executive
function de‚ÄÄcits that are similar to those found in frontal lobe patients. The control technol-
ogy works well in pre-programmed situations, but it cannot reprogram itself to handle novel
events. Our goal is to extend AI planning methods to simulate human executive functions
in real-time closed-loop control applications. We also intend to use the model for a cognitive
rehabilitation application. Earlier descriptions of the model can be found in [?] and [?].
An Information Processing Model
This model is a synthesis of several neuropsychological theories which will be reviewed
in the next section. We designed our model by disassembling the functional components
of the neuropsychological theories and then recombining them so that the functions are
grouped together based on the type of knowledge representation and information processing
involved. Thus the model components are organized from an information processing per-
spective. This organizing principle shows how the information processing requirements for
computer implementation have guided our model de‚ÄÄnition.
An overview of our model is shown in Figure 1. The prefrontal cortex stores symbolic
(linguistic) memories of conditioned action sequences called programs. In humans, these
programs correspond to functional neural circuits (also called pathways). The programs en-
code memories of long duration action sequences that remain active to direct behavior in the
absence of sensory stimuli. Routine sensory conditions activate default program pathways,
and novel sensory conditions activate deliberate program variations.
1
Prefrontal Cortex
Sensors Muscles
Goal
Management
Simulation
Management
Error
Detection
Error
Correction
Executive Function Programs
Daily Activity Programs
Eat Lunch Go Shopping
Other Programs
Sensory (posterior) cortex,
Motor and Premotor cortex,
and subcortical regions
Default (automatic)
control pathways
Deliberate (planned)
control pathways
Figure 1: Programs stored in the prefrontal cortex
Some prefrontal programs encode behaviors for daily activities such as cooking, shop-
ping, and bathing. Other programs encode executive function behavior for self-monitoring
and self-programming. The Executive Functions determine when and how to replace the au-
tomatic, default programs with deliberately planned variations. These executive functions
are summarized in Table 1.
Function Name Function De‚ÄÄnition De‚ÄÄcit
Goal Management: Maintain goal structures that map conditions into reward values
Add Add a new goal Inactivity
Remove Remove an old goal Perseveration
Change Change reward value associated with goal Perseveration
Simulation Management: Simulate programs to predict their e‚ÄÄects
Goal Triggers (re)planning when goals are modi‚ÄÄed Unawareness
Condition Triggers replanning when conditions change Unawareness
Execution Triggers replanning when execution failures occur Unawareness
Deadline Triggers replanning as deadline approaches Unawareness
Error Detection: Analyze simulations to detect program errors
Ineffective Routine preconditions fail due to sensory conditions Unawareness
Interfering Routine succeeds, but causes another program to fail Distractable
Irrelevant Routine is triggered by sensory conditions and Distractable,
does not fail, but it is unrelated to active goals Stimulus-bound
Error Correction: Replace default programs with deliberate variations
Inhibit Override a default start-condition that is currently true Disinhibition
Start Override a default start-condition that is currently false Poor initiation
Continue Override a default stop-condition that is currently true Poor persistence
Stop Override a default stop-condition that is currently false Perseveration
Choose Override a default choice point selection in program body Perseveration
Sequence Override a default subroutine order in program body Poor sequencing
Table 1: The Executive Functions. De‚ÄÄcit indicates the e‚ÄÄect if function is impaired.
2
The main focus of our model is on the de‚ÄÄnitions of the Executive Function programs,
and the four knowledge representations they process. These four representations are:
1) Condition structures that represent symbolic descriptions of environmental conditions.
2) Goal structures that map sensory conditions into positive and negative reward values.
3) Program structures that link conditioned action sequences together.
4) Simulation structures that record the e‚ÄÄects of simulating di‚ÄÄerent program variations.
The information processing performed by the Executive Functions can be summarized
as follows: First, the Goal Management program generates goal structures that identify a
person's preferred conditions based on experience. Second, a set of programs are learned
from experience that represent automatic conditioned action sequences that achieve goals
in routine situations. These default programs map conditions automatically into e‚ÄÄec-
tor commands in real-time without any deliberation. The main function of the Executive
Functions is to monitor the automatic conditioned responses, and to anticipate, detect and
correct any errors that occur due to novel conditions. The Simulation Management pro-
gram produces simulation structures that describe the predicted e‚ÄÄects of executing any
given program in the context of current conditions. The Error Detection program then
compares the simulation predictions with the goals in order to detect potential errors. If
errors are detected, then the Error Correction program overrides the default program in
order to maximize goal achievement.
An important feature of our model is that it is being implemented on top of an AI
planning system, called propel[?], that provides implemented methods for representing,
executing, and simulating programs. We have previously used propel to demonstrate
how a single program representation can be used to produce both default reactions and
deliberate plans[?]. This system combined both default and deliberate modes of action, but
it was lacking a theory of how to coordinate the two modes. The executive function programs
were designed to provide that capability.
We will now further describe how the executive functions process the above representa-
tions. To illustrate the concepts, we will use the following simpli‚ÄÄed program that repre-
sents the routine for doing laundry every Thursday. The Do-Laundry program represents a
conditional action sequence that will repeatedly wash the clothes, wait 20 minutes, and then
dry the clothes, until the laundry basket is empty.
Program Name: Do-Laundry
Start Condition: Today is Thursday and Detergent is not empty
Stop Condition: Laundry basket is empty or Detergent is empty
Program Steps: Put clothes and detergent in washer
Start washer
Wait 20 minutes
Put clothes in dryer and start dryer
The Goal Management program generates and updates the goal representations that
map conditions into positive and negative reward values. Goals can be added, removed,
or changed. An initial set of top level goals for safety and health are predetermined. New
(sub)goals are then generated when the Simulation Manager generates simulations with
hypothetical conditions that are associated with high rewards. Without a well de‚ÄÄned and
maintained set of goals and values, goal-driven behavior is impossible.
3
The Simulation Management program generates and updates the simulation repre-
sentations. When this program is triggered by one of the four events listed in Table 1, it uses
propel's program simulator to predict the e‚ÄÄects of a given program in the context of cur-
rent conditions. The simulation process is coupled with an AI search method that allows
di‚ÄÄerent program variations to be simulated and tested against the goals in order to ‚ÄÄnd
the most e‚ÄÄective behavior. The resulting simulations are used for Error Detection and
to identify predicted conditions that serve as plan assumptions which need to be monitored.
Simulation is triggered by the following events: A goal event occurs when the Goal
Manager modi‚ÄÄes a goal structure. Adding new goals will trigger the initial simulation of
the default response to the goal. Removing goals or changing their reward value will trigger
replanning that involves backtracking through the space of program variations. The con-
dition event triggers replanning if an updated sensory condition con
icts with a predicted
condition that was generated by the simulator (a plan assumption). For instance, it may be
necessary to delay the start of Do-Laundry if the washing machine breaks on Wednesday, the
day before laundry day. The execution event triggers replanning when a default program
fails during execution, as when someone drops a co‚ÄÄee mug. Unlike with the condition
event, this is the case when program execution has already begun and the motor pathways
have already been activated. The deadline event occurs when a goal's deadline moves
within some threshold temporal distance. This event could trigger replanning using updated
sensory conditions ‚ÄÄve minutes before execution.
Simulation Management allows the brain to detach mental activity from real-time events
in order to reason about simulations of the past and the future. This enables the mental
independence from sensory conditions that is required for abstract, non-concrete thought.
Implementation of this component will require resolution of many open issues. One issue
involves choosing the best order in which to try di‚ÄÄerent program variations when replanning.
Other issues involve deciding how much time to spend planning, and understanding how to
simulate actions at an abstract level without simulating every little detail.
The Error Detection program analyzes simulations to detect the three error patterns
described in Table 1. Ineffective routines are detected if the simulation record indicates
that a program's preconditions could not be satis‚ÄÄed. For example, if there is no detergent
to wash the clothes. This is called a program failure. Interfering programs are detected
when the simulation indicates that multiple programs con
ict over a shared resource,
resulting in at least one program failure. For example, doing laundry may interfere with
taking a hot shower if the hot water is limited, or a phone call may distract someone from
moving the clothes from the washer to the dryer. Irrelevant routines are detected if the
simulation does not show any program failures, but it doesn't show any goal achievement
either. For instance, taking the garbage out on the regularly scheduled night is irrelevant if
it won't be collected due to a holiday. A frontal patient who begins to bake cookies whenever
she sees an oven also illustrates the type of de‚ÄÄcit that can occur if this function is impaired.
If one of these error conditions is detected, then the Error Correction program is activated.
The Error Correction program analyzes the simulation and program structures in
order to generate program variations that override inappropriate default programs. Rou-
tine programs contain default start and stop conditions that automatically trigger the
start and stop of program execution. The ‚ÄÄrst four Error Correction methods in Table 1
are inhibit, start, continue and stop. These methods correspond to the 2 by 2 matrix
4
produced by overriding the start and stop conditions in the true and false cases. For example,
the start of the above Do-Laundry program can be deliberately postponed from Thursday
until Saturday by using inhibit on Thursday and start on Saturday. Also, the default stop
conditions of Do-Laundry can be overridden by using continue when the detergent runs
out or by using stop when there is not enough time to wash all the clothes in the basket.
Underlying Neuropsychological Theories
Our model combines the following neuropsychological theories, and it extends them by
elaborating on the information processing details described above. Luria's early description
of the frontal lobes as system for the \programming, regulation, and veri‚ÄÄcation of activity"[?]
is perhaps the best one sentence description of prefrontal cortex functionality. However, it
is too general to use as a design for computer implementation. We therefore looked for the
most detailed descriptions of executive function dimensions we could ‚ÄÄnd, and we selected
the models proposed by Sohlberg and Mateer[?], and by Lezak[?].
Sohlberg and Mateer developed the Executive Function Behavioral Ratings Scale
(efbrs)[?], to assess behavioral dysfunction after head injury. The efbrs includes three
main executive function components which are further divided into subcomponents. The
‚ÄÄrst component, Selection and Execution of Cognitive Plans, involves the ability to describe
goals and procedures, to determine appropriate action sequences, to initiate activity, to re-
pair plans, and to maintain persistent e‚ÄÄort until a task is completed. The second component
of the efbrs, Time Management, involves the ability to generate realistic schedules, and to
perform the scheduled activities within given time constraints. The third component, Self-
regulation, involves using feedback to control behavior and to inhibit inappropriate reactions.
Lezak describes four essential components of executive function[?]. First, Goal Formu-
lation is the ability to generate and select descriptions of desirable future states. Second,
Planning involves the selection of steps, elements, and sequences needed to achieve a goal.
This requires the ability to recognize and evaluate choices. Third, Carrying Out Activities
involves the ability to start, stop, maintain, and switch between planned actions. These
subcomponents led directly to several of the Error Correction methods shown in Table 1.
Fourth, E‚ÄÄective Performance involves the ability to monitor and repair activities.
We began the design of the executive function programs by making a list of the executive
function components as described by the efbrs and by Lezak. In order to facilitate computer
implementation, we then reorganized those functions based on the type of information being
processed. We have tried to capture all of the functionality of these two models in our own.
However, due to our functional reorganization, there is not always a one-to-one mapping
between their components and our own. Instead, some of their components can be found
distributed across several of our model components.
The Norman and Shallice model of the frontal lobes as a Supervisory Attentional
System (SAS)[?] corresponds strongly to our model's Executive Functions, and is also based
(loosely) on an AI information processing model. This model, like our own, proposes two
modes of action: automatic default responses for routine situations, and deliberate planned
responses for novel situations. Shallice and Burgess propose that the frontal lobes serve
as a Supervisory Attentional System that is required to: (1) inhibit undesirable automatic
responses, and (2) generate and execute desirable new responses. The SAS model helped to
de‚ÄÄne the role and function of our executive functions. It also supported our independently
developed method for using both default and deliberate action modes. Our work extends
5
the SAS model by providing an implemented program representation and by elaborating
on exactly how the SAS may detect and correct program errors in novel situations.
Grafman proposes that the prefrontal cortex contains representations called Structured
Event Complexes (secs)[?]. Secs encode memories of long duration action sequences that
guide behavior through well-learned activities. Secs correspond strongly to programs in
our model. Grafman describes sec hierarchies, where the highest secs are called Managerial
Knowledge Units (mkus) because they manage other secs. These mkus correspond to our
model's Executive Function programs. This sec-mku model led us to adopt the idea
that the daily activities and the executive functions in our model are all represented as
programs, rather than requiring a di‚ÄÄerent representation for the executive functions.
Our work extends the sec model by describing an implemented program representation
and a planning system that can be used to anticipate, detect and correct program errors.
Also, our Executive Function programs de‚ÄÄne a speci‚ÄÄc set of mkus that have not previously
been discussed in detail. Our model predicts that two other representation types, goals
and simulations, are stored in the prefrontal cortex in addition to the sec-like programs.
Grafman also elaborates on issues we have not addressed such as the e‚ÄÄect of di‚ÄÄerent neural
activation patterns on sec development and learning.
Stuss' view of self-awareness [?] is also an important part of our model. Frontal lobe
patients often have intact awareness of their sensory environment despite impaired awareness
of the interactions between themselves and the environment. Stuss proposes that frontal
system damage can impair the awareness of one's Self as a continuity from the past into
the future. This concept of self is represented in our system by the goals, which identify
a value system that is learned from experience. In our model, self-awareness is the process
of relating one's goals to external conditions and simulations. This is achieved by the
combined functions of our Simulation Management and Error Detection programs.
Evaluation Methodology
The model's de‚ÄÄnitions of the executive functions and the representations they manipu-
late identify hypothetical neurological functions and connections and mental representations
that can be tested for validity in various ways. Clinical studies can be designed to test for
the presence of these model elements. Our plan is to have frontal lobe patients perform daily
living tasks in a computer simulated world. We will then connect our model to the same
simulated world and compare the performance of our model with that of the patient. The
model will be \lesioned" to produce behavior that is similar to that of the patient. This will
allow us to compare the behaviors of the model and patient directly rather than subjectively
interpreting the behavior predicted by a verbal model. The predicted de‚ÄÄcits produced by
these lesions are shown in Table 1.
Another bene‚ÄÄt of our model is that we can test it on large scale tasks that last minutes,
hours, or even days. Propel's program representation is expressive enough that we can
represent complex daily living activities such as making dinner. In contrast, many frontal
lobe models are based on analysis of simplistic tasks such as the Towers of London (or
Hanoi), Block Design, or the Delayed Response Task. We propose that directly measuring
the model's performance on large-scale daily activities such as shopping and cooking will be
a better measure of the ecological validity of our model.
6
Conclusion
The propel substrate for our system is currently implemented, providing our basic abil-
ity to represent, simulate and execute propel programs. The Executive Function programs,
however, are still under development. Many di‚ÄÄcult technical design and implementation
issues remain unresolved. Our current e‚ÄÄorts are directed towards extending the system to
support the model's full functionality.
We have presented a computer model of prefrontal cortex function that combines several
frontal lobe and executive function models based on information processing principles. Before
the model is complete, many open research questions will have to be answered from both
the neuropsychological and the information processing perspectives. We hope our model
de‚ÄÄnition provides motivation and a common language for future interdisciplinary e‚ÄÄorts to
improve upon this start.
Acknowledgements
Thanks to McKay Sohlberg, Tom Boyd, Jordan Grafman, Je‚ÄÄrey Englander, Richard Del-
monico, Peter Robinson, and Steve Farmer for providing many useful comments that have
in
uenced the shape of this work.
References
[1] Levinson, R. 1994. Human Frontal Lobes and AI Planning Systems. Proceedings of the Second
International Conference on AI Planning Systems. AAAI Press, Menlo Park, CA.
[2] Levinson, R. 1995. An interdisciplinary theory of autonomous action. 1995 AAAI Stanford
Spring Symposium on Extending Theories of Action, Stanford, CA.
[3] Levinson, R. 1995. A General Programming Language for Uni‚ÄÄed Planning and Control.
Arti‚ÄÄcial Intelligence journal special issue on Planning and Scheduling. To appear 1995.
[4] Luria, A. R. 1973. The Working Brain. Basic Books Inc. New York.
[5] Sohlberg, M. M. & C. Mateer. 1989. Introduction to Cognitive Rehabilitation. Guilford Press,
New York.
[6] Lezak, M. 1983. Neuropsychological Assessment. Oxford University Press. New York, NY.
[7] Shallice, T., Burgess, P. 1991. Higher-Order Cognitive Impairments and Frontal Lobe Lesions
in Man. In Frontal Lobe Function and Dysfunction, edited by H. Levin, H. Eisenberg, and A.
Benton. Oxford University Press, New York.
[8] Grafman, J. 1989. Plans, Actions and Mental Sets: Managerial Knowledge Units in the Frontal
Lobes. Integrating Theory and Practice in Clinical Neuropsychology. E.Perecmen, Eds. Erlbaum
Press. Hilldale, NJ
[9] Stuss, D. 1991. Disturbance of Self-Awareness after Frontal System Damage. Awareness of
De‚ÄÄcit After Brain Injury. Prigatano and Schacter, Eds. Oxford University Press. New York

I thought I was able to upload the pdf of the full article - it didn't really work - just converted the first page to a PNG:


(click to see bigger image of first page of the article)

PM me is you want a copy of the article (click on "Profile" below, then the "PM" button on my WP profile page - the "PM" button below doesn't really work).

Shifting: Yes it means switching tasks, or multitasking. I see that word in my ADHD report too.

Emotional Control: I think the idea is that poor emotional control will affect our EF, or might be a predictor of it. Perhaps for some people with other mental health diagnoses, it's the only reason for poor EF. In that regard maybe they measure it as a variable to compare against the ASD-related scores?

Self-Monitoring: Are you able to notice when your EF is poor in a situation? Are you even aware of it? Can you help yourself or snap out of it when you discover it's happening? Are you able to track your own progress or set goals?

Initiating: Fluidly generating ideas would mean, if a problem or situation arises, can I brainstorm effectively? Can I think the issue through cognitively to generate ideas / plan what to do? For example if a big problem happened could I break it down into component parts and prioritise what to do first, second, third, or what needs to be done at all (big picture?)

Task Monitoring: Keep track of behaviour on others - I assume that means if I'm doing a task (or not doing a task, as the case may be), do I notice if I'm bothering / affecting others? Because, in a socially responsible world, I should be able to adjust what I'm doing accordingly. If I'm making a mess in a place where others will be inconvenienced, or I'm making a big noise at midnight which might bother my neighbours, do I stop and revise the plan?

Keep Track .... - Yes you're right, that's seems a duplicate.

Working Memory - I can open a tab on my computer to check something, and totally forget that I opened it. I'll see the tab hours later and realise I didn't complete the task. In that regard I didn't "stick to" my plan. Sometimes I'll remember "Oh man! I started to research xyz about a week ago, but I got distracted and forgot all about it for the last ... six days!" Then I'll open the tab again, or start my research again, but forget what I'm doing a second time. A fourth time. A fifth time. My inability to multitask causes me to forget if I'm juggling more than one thought at a time. Even if I'm only doing one thing, I can forget what it is because of poor working memory. I'll walk into a room to do one thing and forget why I'm there. I might have it written on a To-Do list, but most of the time I forget where the To-Do list is written. I think that all relates to working memory. It's not just a matter of whether I can remember a phone number or string of numbers for a few seconds. There are practical situations where poor working memory means I won't "stick to" what I'm doing physically. I start something but literally forget I'm doing it. It's hard to have good EF if you can't remember what you're doing.

Muddled - That's not how my whole report is written. That's just reference information that they seemingly cut / paste or autofill for all reports, to explain the purpose of the specific test (called Adult BRIEF-A). It's likely written by whoever created Adult BRIEF-A, and not my doctor. The other test descriptions / data charts in my report are presented like that as well, but the balance of my report is written in more descriptive language specific to me, with pertinent examples of what I did or said. Those comments were a little to personal to share on WP for privacy reasons, but I'll look for some excerpts that might be suitable to reproduce. Overall my report is very clear and thorough, even though these specific screenshots seem to lack clarity. I just wanted to give you an idea how they defined each category clinically.

My Mother / Informant - The informant wasn't my mother. It was my daughter who is in her 20's. I've lived with her longer than I ever lived with my mother. My daughter had to answer lengthy questionnaires about me by ranking "Always, Sometimes, Never" etc. and sending them in online, without asking me or asking my input. It wasn't a verbal interview or question / answer interaction with the doctor.

I asked the doctor about the fact my daughter's scores are different than mine. She said that's completely normal. Another person won't know how our EF is, except for what they perceive in relation to themselves. For example, my daughter knows I keep the house relatively clean and I take care of her needs, but she doesn't know what's in my mind or how my thought process works along the way. She wouldn't know if I made a plan for something, and changed my mind or forgot what I was doing midway - especially if it had no impact on her. She wouldn't know if I'm monitoring my own tasks, or how often I initiate / don't initiate plans for myself that have no relevance to her. She wouldn't know if I follow through on all the things I think about, or if I start things that I don't ever finish. She might see me doing Ancestry research for myself but it's not like she asks "Hey, did you finish that? Did you get distracted? Did you organise the information properly on your computer or is it really disorganised? Did you research the people you intended to, or did you get sidetracked and change your focus fifteen times? Did you remember to text your friend in the middle of it? Did you remember to eat?"

If I decide I'm going to rearrange my desktop files in a certain way at a certain time, she would have no idea whether I followed through or not. Likewise with organisation of materials. She knows the house is (somewhat well) organised, but how would she know the degree of struggle I went through to keep it that way unless I complained out loud? Again her knowledge only extends to what she can see, what she notices, or what affects her directly. How would she know if my paperwork files are organised properly, if she doesn't go through my files? How would she know if I'm shifting appropriately between thoughts, when she doesn't know all my thoughts? How would she know my stress level, or how hard a task was for me?

If I put food on the table at a decent time, that's all she sees. She wouldn't know it took me six hours of start / stop / confusion / stress to plan the shopping list, to get the food delivered, the fact I was tired or disorganised making the meal, the fact I might have changed ingredients ten times in my mind, lost the recipe, or the fact we're having chicken because I forgot to order lamb. Maybe I was supposed to phone someone when I was cooking but I couldn't do both things at once. She will only know that yes, I fed her -- and the fridge looks full every couple of weeks. I can't go around complaining that it's difficult or stressful to meet her needs, voicing my to-do lists out loud, or announcing any failures and multiple changes of plan. All she can see is the end result of tasks which relate to her. She wouldn't notice if I was ignoring my own needs or my own plan, because she wouldn't even know about them.

My cognitive journey (stop, start, revise, melt down, or even abort mission), and tasks I do in my personal life, aren't relevant or perceptible to her at all.

In short yeah, I thought the same thing but was assured it's normal (even expected) that the Informant's view will be limited in scope. Even if a parent is used as the informant, they can only measure what they see externally and not the process in their child's thoughts.

Was I hard on myself? No. Those scores weren't all from self-reporting. They were in conjunction with testing I did in the assessment. I was asked to switch between tasks, my short term memory was tested, etc. There's a lot more detail in the report to substantiate these numbers and a lot of it was performance-based tests. The doctor also had my school reports showing missed deadlines, not completing all parts of tasks, etc. There was also this comment to validate that I didn't exaggerate my strengths / weaknesses in the testing.

^
Emotional Control: I think the idea is that poor emotional control will affect our EF, or might be a predictor of it. Perhaps for some people with other mental health diagnoses, it's the only reason for poor EF. In that regard maybe they measure it as a variable to compare against the ASD-related scores?
Yes I suppose the ability to control the emotions is very important for enabling EF, at least EF in the sense that they would define it, and I can well imagine a person who might have very good potential EF skills if only they could bottle their extreme emotions long enough to get anything done. The way they present it suggests, though, that they see it simply as just another EF parameter, but I don't know if they used it that way or not. To me emotional control is more about explaining why EF might be poor than an actual measure of EF. I'm always doing this - finding anomalies in other people's organisation of information, and getting hung up on them.

Sorry, I got mother and daughter mixed up.

Yes I see how another person would naturally assess you differently to how you assess yourself, because they can't know about the failures you see internally. And that jpg showing their personality assessment info shows that they didn't think you were being hard on yourself in your self-assessment.

It's a much more detailed and extensive assessment than mine was, though mine was looking purely at Aspergers, and of course the methodology varies from place to place. I did ask my diagnostician if I might have other things going on, and she said that she'd been aware of that possibility but hadn't seen the slightest evidence of anything but Aspergers during the entire proceedings, and we left it at that.

I have no specific comments on the rest, except that it makes sense.

I think that's what freezes my thinking a lot - if I don't understand it perfectly then it feels like it's a horrible mess that I'm completely failing to get my brain round. Which doesn't encourage me to persevere, though it's surprising how often I do. As for preoccupation with parts, that too. Reminds me of this quote from an 18th century philosophy journal, which has often made me wonder if the guy he's talking about was an Aspie:
"He often attempted too much accuracy in his description of minute things." - John Hunter [Phil. Trans. R. Soc. Lond. 1792 82, p.128]

EF can control actions and expression of emotion. It can also fail to do so.
Emotions can affect EF - and make it hard to control other aspects of EF like attention.
It works both ways.
Also - intense emotions is one way of self-medication. People with ADHD and ASD engage in stimulation seeking and self stimulation behaviors. Panic can be very focusing. Fear and anger can be very un-focusing.
Risk taking behavior (mountain climbing, parachuting, bungie jumping, driving a fast car, picking a fight, leaping before you look) can cause an "rush" that takes away other negative emotions.
EF and emotions affect each other.

For me the key is to be worried enough to take action but not so worried that I shut down (or take random or ineffective / counterproductive action).

Balance is the goal.

https://www.adhdawarenessmonth.org/adhd ... egulation/

Inhibit, Start, Continue, Stop, Choose, Sequence

As far as I am concerned these are like the atoms that all the EF function molecules are made of.
(and it is not hard to think what can go wrong when they happen at the wrong time - or fail to happen at the right time)

Time awareness is another important component, listed as "Deadline" in the diagram from "A Computer Model of Prefrontal Cortex Function".
Goals, and Actions the 5 senses, and rewards - and "thinking ahead - what-if thinking" (called "Simulation") is another.
And Planning and Replanning (when things changes) are other key components.

I also really like the last column which includes a list of what happens if part of the EF DOESN'T work.

Good to know. Yes, that's the problem with specialist papers - if it's not the reader's field, it'll be hard to understand. I think the other problem is that the paper was written from the specialist's point of view (not surprisingly), not ours. They're not talking to us, they're talking to their peers.

I'm sure I have trouble at times forming concepts from info supplied by the outside world - what the exact trouble is I'm not sure, but for whatever reason, the effect I notice is that it take me longer than most people, so if the information is given to me verbally then I can't always keep up.

I'm still having problems acquiring any faith in the system of categorisation the paper presents. It seems to take a long time for me to take in the definitions of the categories, and it all seems very disjointed and irrational to me, rather a mess in fact. I don't know who decides how to categorise things or how they come to think that they've done it in the best way. Maybe it's just that they're lucky to be in a privileged position and they manage to popularise (with their peers) a set of categories that helps THEM to understand it but doesn't help ME very much. Every time I look at the terms I start thinking "I'd have used different categories," though I can't say I know enough about the subject to be able to create my own yet. At school, once the lessons became complex, I was unable to adopt the teacher's way of thinking about a subject. Maybe that's an example of rigid thinking - come to think of it, why isn't rigid thinking an EF category? It seems to impact the matter as much as any of their categories.

Incidentally I had to set up the above quote of your post manually, because when I clicked the Quote button the web page said "You do not have access to this forum" which seems strange because I was already accessing it. I tested the quote button out on another post here and there was no error message. No wonder I spend too much time messing about on the web.

EDIT: The post I quoted seems to have been deleted. I don't know why. I also don't know why the page didn't just say "that post has just been deleted."

Further Reading

https://en.wikipedia.org/wiki/Executive ... ons#Models
https://en.wikipedia.org/wiki/Executive ... l_evidence
https://en.wikipedia.org/wiki/Executive ... Assessment
https://en.wikipedia.org/wiki/Allegory_of_the_cave
https://en.wikipedia.org/wiki/Theory-theory