Closed shnizzedy closed 6 years ago
Appreciate the feedback - thank you. I think we should do both. Your phrasing is exactly what I was hoping to do and is really good for 1 and 2. Let me give this more thought, try to get interesting examples, explore possible structures and get back to you. Would it not be a good thing to not have a uniform fixed structure (I feel some conform to some structures better than the other)
I completely agree that we should "not have a uniform fixed structure [. . . if] some conform to some structures better than the other". We should just be sure to be clear about what structures we are using and when.
Hello Arno,
I have been conceptualizing the system and questionnaire (could explain in detail tomorrow). After huffing and puffing, here is an example https://docs.google.com/spreadsheets/d/1k2RkEpatNZKtA4O3-4NYh7nV9bDdPHxHp9w19iC6rpY/edit?usp=sharing
I could act on your feedback and have something more concrete with more examples for tomorrow.
Additionally, here are some nice visuals we could use for the online questionnaire https://www.16personalities.com/country-profiles/japan
— @anirudh4792
Thanks, Anirudh.
frequency - regular vs. often/good doesn't make sense to me.
Have Jon take a look.
Cheers, @rno
— @binarybottle
You know I'm confused by the format, but leaving that aside:
Thank you. I’ve made some changes. Feels like back to square one – How to generate questions from symptoms and I am having trouble generating them.
One path could be to zoom out a bit, have attention as the neutral behaviour and have several symptoms and questions mapped to it, adding context as a dimension. This approach seems feasible and possibly a little programmatic as we would have around 50-100 core ‘neutral behaviours’.
Right now, we are trying to get 1000 neutral behaviours (from every symptom). I could generate an example for the above.
Note: We are going more levels deeper than E-SWAN which tackles only 1 dimension and is faithful to the DSM vs actually generating behaviours. They also select only a subset of symptoms in ADHD
But I have suggestions on points like one person’s pretty much is another person’s very much and how to overcome these uncertainties.
— @anirudh4792
I really like the idea of abstracting to "around 50–100 core ‘neutral behaviours’."
Thanks Jon! 0) Forget the format for now - it is simply for illustrative purposes 1) E-SWAN is only there just for comparison and to know what they did. The questions from questionnaires were also there just to give you an idea of whats out there in that component. The main aim here is to generate questions, add dimensions to 'neutral behaviours' while keeping in mind how a parent could answer the question (hence, the follow up questions). Subsequently, these questions will be mapped to symptoms, objective measures etc. later in the grand scheme of things 2) is exactly what I was looking for and seems appropriate - just awesome (there is no charge for awesomeness)
2c. cannot < can with difficulty < can < can easily < cannot avoid [doing] OR must [do]
intensity?:
no effect < mild < severe
opposite effect < mildly opposing response < no effect < mild effect < severe
reaction?:
no reaction < mild < severe
severe opposite reaction < mild opposing reaction < none < mild < severe
a. duration: very short < short < typical < long < very long eg: sustaining attention b. frequency: never < rarely < sometimes < often < always eg: running or climbing when inappropriate c. ability: cannot < can with difficulty < can < can easily < cannot avoid [doing] OR must [do] eg: following through on instructions d. opposite effect < mildly opposing response < no effect < mild effect < severe e. REACTION severe opposite reaction < mild opposing reaction < none < mild < severe eg: anticipating social situations, fear of doing something that will lead to embarrassment
However, I feel attention to detail does not fit these (interestingly, one end of the scale for this is considered poor for adhd while other end is considered 'poor/great' for asd/adhd (hyper focus, extreme attention to detail). A very good/bad seemed to fit). There are other examples with similar polarities.
I also tried combining dimensions and it was a mess in my head (can never sustain attention for long etc.)
attention to detail: a. very short to very long attention span b. never attend to anything to always attending to something c. cannot sustain attention to cannot stop attending to things d. attend to everything but what I should (distracted) to cannot attend to anything but center of focus e. (same as d when presented with stimuli or when directed by another)
more examples like attention to detail https://docs.google.com/document/d/15P-Hr2vq_HyJlcZZzWgidzqb9yJmIikwmHYWpdccLs8/edit
I wrote specific comments and suggestions in the document.
Generally, we need to be careful about which part of a complex behavior we are modifying. E.g., for attention to detail are we interested in the duration, frequency, intensity, etc. of the attention or of the detail? For sharing of emotions, the duration, frequency, intensity, etc. of the sharing or of the emotions? For being exposed to scrutiny in social situations, the duration, frequency, intensity, etc. of being in social situations, of being exposed to scrutiny, of the social situations or of the scrutiny?
In that last example, should in social situations be a modifier? I know we've talked about having a vocabulary of contexts to apply to decontextualized behaviors.
The "(same as x)" examples give me pause. Are these dimensions sometimes interchangeable?
@binarybottle, in your attention to detail example above
attention to detail: a. very short to very long attention span b. never attend to anything to always attending to something c. cannot sustain attention to cannot stop attending to things d. attend to everything but what I should (distracted) to cannot attend to anything but center of focus e. (same as d when presented with stimuli or when directed by another)
, I see how these phrases deal with attention but don't see how they relate to detail.
a. duration: very short < short < typical < long < very long eg: sustaining attention b. frequency: never < rarely < sometimes < often < always eg: running or climbing when inappropriate c. ability: cannot < can with difficulty < can < can easily < cannot avoid [doing] OR must [do] d. intensity: opposite effect < mildly opposing response < no effect < mild effect < severe e. consistency f. specificity
attention to detail: a. attends to detail for very short to very long periods For fun, let's include duration of opposite behavior X, where X is "maintains inattention" or "attends to non-detailed generalities":
https://docs.google.com/document/d/1h87oNlDdKiwThbISvX0RcjUgp742GjEM5SqwnB1SBqI/edit?usp=sharing
More examples with different types of consistency/specificity
Will incorporate jon's comments after we discuss them for a to d.
On Thu, Oct 12, 2017 at 4:09 PM, Arno Klein notifications@github.com wrote:
a. duration: very short < short < typical < long < very long eg: sustaining attention b. frequency: never < rarely < sometimes < often < always eg: running or climbing when inappropriate c. ability: cannot < can with difficulty < can < can easily < cannot avoid [doing] OR must [do] d. intensity: opposite effect < mildly opposing response < no effect < mild effect < severe e. consistency f. specificity
attention to detail: a. attends to detail for very short to very long periods For fun, let's include duration of opposite behavior X, where X is "maintains inattention" or "attends to non-detailed generalities":
- X for very long periods vs. attends to detail for very long periods
- X for very long periods vs. X for very short periods
- attends to detail for very short periods vs. X for very short periods b. never attends to detail -> always attends to details c. cannot attend to details -> must attend to details d. attend to everything but details -> attends only to details e: randomly attends to details (or attends to random details) -> attends to all details f1. attends to certain details -> attends to all details (or attends to random details) f2. attends to details under certain conditions -> attends to details under all conditions
— You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub https://github.com/ChildMindInstitute/mhdb/issues/27#issuecomment-336251619, or mute the thread https://github.com/notifications/unsubscribe-auth/AKCstU9dXNKQswP-8cUw9cwBnUOONwJDks5srnIDgaJpZM4PvfmA .
@anirudh4792 -- linking out to google docs ensures that this information will be lost in the future. best to keep these discussions succinct within github until the issue is resolved, and then expand our efforts in other formats/venues...
Yes. But we cant add comments on github More examples with different types of consistency/specificity key: e. consistency f. specificity
1) Being accurate in work or activities e. Is randomly accurate in work -> accurate in all work (40% accurate in manual curation on day 1, 70% on day 2) f1. Accurate in certain types of work -> accurate in all types of work (accurate in manual curation but inaccurate in programming) f2. accurate in work under certain conditions -> accurate in work under all conditions (accurate if slept well, accurate if favorite chocolate is given)
2) Sustaining attention e. randomly sustains attention (or sustains attention randomly) -> always sustains attention (sustains attention always) (day 1 - focusses 75% on the task vs day 2 - focusses 40% on same task) f. Sustains attention under certain conditions -> sustains attention under all conditions (sustains attention if slept well) f2. sustains attention under certain activities -> sustains attention under all activities (does not sustain attention for homework but does so for gaming)
3) Being exposed to scrutiny in social situations e1. reacts when exposed to certain scrutiny -> reacts when exposed to any scrutiny in social situations (baldness in a bar vs stammering while speaking anywhere) e2. Reacts randomly when exposed to scrutiny -> reacts consistently when exposed to scrutiny (laughs it off sometimes vs always starts stammering) f2. exposed to scrutiny under certain social situations-> exposed to scrutiny under all social situations (classes in school vs sports)
4) Sharing of emotions e. shares all emotions at same intensity -> shares different emotions at different intensity (shares happiness and sadness differently) f1. shares certain types of emotions -> shares all types of emotions (shares only happiness) f2. shares emotions under certain conditions -> shares emotions under all conditions (shares emotions only at grandparent's home)
5) Reaction to sensory stimuli e. Always reacts to sensory stimuli at same intensity -> reacts to sensory stimuli at different intensity (reacts angrily one day and reacts terrified the next) f1. Reacts to certain types of sensory stimuli -> Reacts to all types of sensory stimuli f2. Reacts to sensory stimuli under certain conditions -> reacts to sensory stimuli under all conditions (reacts to noises when having a migraine)
6) behaviours performed repeatedly without a seeming purpose <opens and closes doors, switches lights on and of, lines up toys or objects> e. Repeats behaviours with same consistency -> Repeats behaviours with different consistency (eg day 1 - switches lights 25 times, day 2 - 50 times) f. Repeats behaviours under certain conditions -> repeats behaviours under all conditions (eg switches lights on and off only outside home)
Wrapped my head around this too much.
Behaviour = attention to details, dimension = duration Let attention = A (verb) and details = B (noun), let dimension lower limit = X (short) and dimension upper limit = Y (long)
From possibilities of 4! = 12 to 4!/2 = 6
On Jon’s earlier comment, I feel only the opposite behaviour of the verb (not having attention/inattention) should be taken. Opposite nouns are valid only in rare cases and if so, it should be taken as a separate behaviour (eg response to sensory stimuli vs response to non-sensory stimuli)
Why opposite behaviours: In autism, children pay attention to detail to objects of interest and do not even look at other objects (this is not reflected by pays attention to detail only for a short time but by does not pay attention for long time (does not pay attention even for a short time - boo for the english language)
I like what Arno did to ability (cannot < can with difficulty < can < can easily < cannot avoid [doing] OR must [do]). I have not seen this (cannot to can do too much/cannot stop) in any questionnaire.
My preference is for all our dimensions to have this
Possible combinations (A,B); (A, B-opposite); (A-opposite, B); (A-opposite, B-opposite) We select only (A,B); (A-opposite, B); Let (A,B) = N and (A-opposite, B) = N-opposite
The winner (N-opposite, Y) : (N, Y) Having inattention to details for very long vs having attention to detail for very long As it contains within it (N, X) : (N, Y) and (N-opposite, X) : (N-opposite, Y) with (N-opposite, Y) sometimes similar but subtly different to (N, X) and (N, Y) sometimes similar but subtly different to (N-opposite, X)
I was thinking something like a circle rather than a linear scale to account for these subtleties.
2) Consistency and specificity f1. attends to certain details -> attends to all details (or attends to random details) Applying opposite behavior < mildly opposite behavior < mild behavior < good behavior < severe behavior Implies Does not attend to any details (eg in all objects of disinterest) < does not attend to some details (ADHD forgets cc’ing people in emails) < does attend to some details < does attend to all details < does not stop attending to all details (keeps staring at the object of interest)
Coming soon Layer 3 (time) + Layer 2 (consistency, specificity) + Layer 1 (duration, frequency, intensity)
I follow and agree with everything in your previous comment up to the "good behavior", which sounds judgmental. I also don't know why you would need a Layer 3 (time) when consistency directly deals with time. Layers or filters is a reasonable way of framing this problem. Anything in layer/filter/dimension 1 can be filtered/contextualized by layer 2.
What "good behavior" meant in this case was having "good attention to detail" but I agree the wording needs to be changed.
Ive been thinking a lot about this and I feel an 'about average' option should be present (I imagined asking people what their child's attention to detail was and I felt many possible answers could be 'just regular/normal' and does not stand out as inattentive or highly attentive but is not represented by mild behavior/attends to some details).
Also, f1. was imbalanced. Here is a new version for consistency and specificity (I have tried to combine both consistency and specificity which may not be the best option)
Applying opposite behavior < mildly opposite behavior < behavior is uniform in consistency across different situations < mild behavior < severe behavior Implies Does not attend to any details (eg in all objects of disinterest) < does not attend to some details (ADHD forgets cc’ing people in emails) < attends to detail consistently across different situations < does attend to some details < does not stop attending to all details (keeps staring at the object of interest)
I felt Layer 3 (time) was needed to observe trends in different time periods (eg month 1 vs month 2, last 2 months vs last 8 months etc.). Consistency dealt with performing a behavior under different specifications within a specific time period (eg he performed this behavior more on tuesdays and fridays; the behavior was more intense outside home (eg SM).
I think consistency and specificity go hand in hand and I agree that consistency could be taken as a sub-type of Layer 3.
Let's consider the kinds of questions we might ask:
https://docs.google.com/drawings/d/1VhDkEtkBqwFGqUC_eC_6moYhmRZ5MzWDgkBHfF_TtgU/edit?usp=sharing
Awesome. Thank you. Was typing away but will try to communicate by graphical diagrams vs long text henceforth :)
Integrating Arno's flowchart with list of behaviors grouped under 'Attention' (All modifications have been underlined). https://docs.google.com/drawings/d/1hPHY_1uuk4oVPS1u-8Bx100-7JgEYjH61EIIf2ZL4Dw/edit?usp=sharing
I have added 'does not try' as an option under ability (does not try to do homework is different from cannot do homework)
I have added questions under dimensions when applicable (eg How is your attention span or how long can you sustain attention)
For neutral behaviors that could not be added, I have created a table and added them as implications.
Ultimately, I would like for parents to be able to identify and act/keep track of strengths and weaknesses and I have tried to include this concept.
I think that the table we need to create is one that tries to map the different "implications" of attention to the different dimensions (duration, intensity, frequency, context dependency) to see if we can reduce the number of questions that need to be posed.
Also, I would replace 'does not try' with simply 'does not', and I believe this is covered by frequency: 'never'.
The issue of implications has been solved in issue #30 I agree that 'does not try/does not' is covered in frequency and I have removed it (we have other such instances of overlapping dimensions as well)
Updated example for attention https://docs.google.com/drawings/d/1LKlTR9LCxHQciXMyQCPyAr3Ap0DDVCXXkVgAHS2wUXk/edit?usp=sharing
example for non communicative behaviors (google draw was very user unfriendly) https://docs.google.com/drawings/d/1fydATax1Q63THY9Yaew16DBr_rHMsVdSTH09cWnJJs4/edit?usp=sharing
The updated example for attention looks good. This will probably be one of the simplest cases, with a single symptom (attention to detail) for a category (attention), with all other symptoms (limited attention span) subsumed within the category's dimensions (duration).
Regarding non-communicative behaviors:
Agree to both. I have followed this format and order (I shall continue to give more examples using the following code than google draw) Scenario 1) E - subsumed neutral behaviors as examples of core neutral behaviors if applicable as the prefix to the neutral behavior was indicated in the questions eg social use of eye contact - eye contact use and understanding of body postures - body postures use and understanding of gestures <pointing, waving, nodding, shaking head> - gestures Scenario 2) S - subsumed within the category's dimensions Scenario 3) C - symptom (eg attention to detail) for a category (eg attention) get their own dimensions
The three scenarios sound reasonable. Let's just not lose the verbs (use, understand,...).
220 neutral behaviors transformed into questions using the above scenarios, but presented in a more readable and structured format (didnt use the letter code)
Every neutral behavior was included either as an example, dimension or core neutral/parent behavior
Justification of decisions, links to other categories, insights added as comments (in a format understandable to me) to explain during our next meeting)
https://docs.google.com/document/d/19at5dngbUS1Jt4K_Z5laq-Y7fSB8UeNJsPawIQ6BOi4/edit?usp=sharing
I like the direction this is going. A lot of these are questions that make me want to take the questionnaire / play the game!
I scrolled through the core behaviors and took a stab at putting them in a tree:
Letting LucidChart rearrange a hierarchy based on the relationships I defined gives us some interesting levels:
I'm curious if we can draw meaningful groupings among these entities. I'm not sure if any one of these entities, particularly 'controlling urges', is in the right place. I'm curious what you guys think.
Just wow. Thank you very much for the effort.
Overall, I thought the groupings made sense - was very logical to me. Only suggestion - for 'controlling urges', I feel is not an implication of impulsiveness, repetitive behaviors or interests but indulgence in urges (for the sake of neutrality) could be their cause. Also, impulsivity could also be linked to activity
I particularly liked
Understanding nonverbal comm behaviors leads to using them which leads to social communication
How fear has been linked to accepting change, anticipation of social situations and bodily changes
How the 'ADHD' part has been logically presented. Attention and activity leads to implications in hyperactivity and impulsivity.
very cool
Great job, Jon! I really like the organization of this chart! I think that we should put this in a form soon for evaluation by clinicians. Clearly not just show and ask "What do you think?" but more along the lines of pairing different items and asking about the strength of their relationships...
Thank you Arno for the suggestion to look into patterns (among families and across families at the same levels (eg across core parent behaviors or level 1 sub-behaviors) Inherent structure and patterns in dimensions
Insights
1) It is evident that we could merge ability and intensity into one as intensity that reflects ability 2) The parent (underlying cause) can cover all possible dimensions while the sub-behaviors/implications can have specific dimensions as the others are already covered in the underlying cause 3) In some cases where only some options on the scale are needed (eg 0-3 rather than 0-5), redundant ones have been converted into opposite behaviors (see google doc with the comment - Arno opp beh) 4) The dimension 'Duration' is inherently present in the familiy external/internal: activity, interests and attention and their sub-behaviors 5) Jon's structure adds readability by adding/creating parents (these parents have the same structure as their sub-behaviors and we can do the necessary edits to have a format like 2 (eg fear/anticipation has been added as a category while anticipation of social situations becomes a sub-behavior; good move as now we can include/link specific phobias to fear)
Clarification 1) Jon's structure adds readability though activity is the same as hyperactivity (I mislabeled the gdoc; activity encompases hyper and hypo activity) Response to social situation same as response to non-verbal communicative behaviors
Minor doubt that doesnt warrant typing on github 1) Unable to understand the logic of arrows going into sensory stimuli
I removed hyperactivity and added a link from activity to interacting with others.
Re: 1. Unable to understand the logic of arrows going into sensory stimuli, if you're talking about the live LucidChart, I was playing with different organization scheme, hit the free-tier complexity limit, and stopped. I've since reverted to the version in this comment above above.
To be able to deal with more complexity, I translated the LucidChart diagram into TTL.
Annoyingly, I couldn't quickly figure out how to get the chart from LucidChart without manually downloading, so I uploaded the generated ttl to Google Drive as a txt file. There's nothing new in this format besides
Nice work, Jon!
Pulled up an old document from Stan. Please see pages 31 to 38. Do you feel they have categorized symptoms well (not based on disorder but on topics like everyday activities) BASC-3-Rating-Scales-Report-with-Intervention-Recommendations-Sample.pdf
I think they've categorized behaviors rather than symptoms (BASC is Behavior Assessment System for Children), but I think their groupings are reasonable. I particularly like that the BASC solicits self-, parent-, and teacher-perspective assessments.
Sorry. i meant they categorize or group symptoms under core behaviors. @binarybottle see above for Stan's old questionnaire
I'm not sure what to think when I see the following in BASC:
"Autism Probability
This is a manual interpretation report (the user has answered 'never' for behaviors grouped under autism probability @shnizzedy this is what we would not want to do but I get your point with questioning why we should retaining core behaviors in the questionnaire/response analysis vs just keeping the questions that represent them
Also See issue #37 for comparison of different domains
Just to clarify, I've been questioning if these "core behaviors" are behaviors or experiences at all, and, if not, what they are (anything besides abstract concepts based on grouping signs, symptoms, behaviors and/or experiences) and what their utility is or can be.
With the example of Autism Probability
: Seems odd.
:
Seems odd.
is an articulable experience that can be a valuable source of information. Lots of people fret about how they are perceived and with a BASC-like questionnaire, we can see where perceptions align and where they conflict.Autism Probability
are as an entry-point (e.g., someone searching for autism information) and as a means to buy-in (e.g., explaining our questionnaire's coverage). If we were to ask people about the autism probability
of themselves or others, I suspect we'd get something pretty similar to their answers to seems odd
, but less directly (i.e., noisier).My concern here is that we risk enforcing existing models by relying on them to structure our probes. I think a case can be made for some of these "core behaviors", but most of them seem like problematic relics to me.
I strongly agree with @shnizzedy's last paragraph in his previous comment, and I agree that it would be interesting to see how much self vs. other's perceptions dis/agree with each other.
Perhaps we could lump internal experience under "mental experience" or "thought" and replace "behavior" with "thought/behavior" in our discussions?
I'm not sure if "mental experience", "thought" and "behavior" encompass enough. Do pain, attention or irritability fit any of those categories?
I feel pain and irritability come broadly under emotional state and mental experience.
Attention may fall under cognitive level. I like where this is going in that we need precise defined unambiguous classifications or none at all? On Oct 30, 2017 4:04 PM, "Jon Clucas" notifications@github.com wrote:
I'm not sure if "mental experience", "thought" and "behavior" encompass enough. Do pain, attention or irritability fit any of those categories?
— You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub https://github.com/ChildMindInstitute/mhdb/issues/27#issuecomment-340567439, or mute the thread https://github.com/notifications/unsubscribe-auth/AKCstc9laKiF26Id8lmZJymytf6Rmw6Lks5sxivKgaJpZM4PvfmA .
This discussion touches on a variety of interesting topics, but are we satisfied with our use of the term "state" and "neutral state" as they apply to "signs" or "symptoms"? If so, can we close this issue?
I think 'state' is better than 'behavior' and is workable.
I don't dislike the term, I just think it's not inclusive of all the nouns & neutral derivations you're compiling, i.e., I am not convinced that "being alone" or "fixated interests" or "repeated compulsive thoughts" are behaviors.
We should either find a term that fits every item of that type or restrict inclusion or reframe some phrases ("seeking solitude", "fixating on interests", "acting or compulsive thoughts") if we are certain behaviors is the right term for what we want.