andkov commented 7 years ago

the cognitive domains depicted on GitHub should accurately reflect that cognitive domains we are using.

Match the domains on the domain graphs for pulmonary track to the domain structure give in the tables.

@eduggan , you have sent me three excel files :

5 Table 4.xlsx
Grip Covarianes and Correlations table for Sean 2017-06-02.xlsx
Pulmonary Meta-analysis 2017-06-20.xlsx

I presume they all contain the same structure of domains (at least it appears so at a glimpse), I will be using the file 5 Table 4.xlsx for my reference. Please direct me otherwise, if incorrect.

eduggan commented 7 years ago

@andkov Yes, please use "5 Table 4" (it is the same structure as "Pulmonary Meta-analysis," just a little easier to look at).

I know there are a studies and few measures used within grip that we did not use for pulmonary... so anything that is not resolved with "5 Table 4" should be able to be resolved with "Grip Covariances and Correlations table for Sean."

andkov commented 7 years ago

@eduggan , thanks for clarification. I'll then just use Grip Covarianes and Correlations table for Sean 2017-06-02.xlsx to kill two birds with one stone (what a violent idiom, now that I think of it!! why would you want to kill a bird?? )

eduggan commented 7 years ago

@andkov I think the vice versa problem exists too (some things in pulmonary not used in grip)... although I am not sure and only have recently come on to help out this the project. I just know the pulmonary has been fully vetted and put through the ringer (in terms of looking for errors in grouping, ordering, etc). In theory using Grip Covarianes and Correlations table for Sean 2017-06-02.xlsx should be faster... just can't guarantee that ;)

andkov commented 7 years ago

@eduggan , I'm comparing the grip and pulmonary spreadsheets. Do you want to refer to MMSE measure in EAS as "Calculated MMSE" or just "MMSE"? spreadsheets show both use

eduggan commented 7 years ago

@andkov please refer to is as "Calculated MMSE." That correction has not made it's way into grip yet.

andkov commented 7 years ago

@eduggan , should I apply this lable to all MMSE measures in all studies or only in EAS?

eduggan commented 7 years ago

@andkov Only EAS. Thanks!

eduggan commented 7 years ago

EAS used BIMC and converted the scores to a MMSE metric. The most appropriate label for this is calculated MMSE. All other studies with "MMSE" used the actual MMSE.

andkov commented 7 years ago

@eduggan , @ampiccinin , after 6+ hours on this, the lesson I learn is that if any changes are made outside of the reproducible flow - they are major challenge to get back in. I need to think of better ways of creating a way for you to make these edits: the csv I created to do this was not accomodating the change of the Item name. This is a must-do, in the present form this tranfer of information is not sustainable. Not your fault at all, I just didn't have enough time to design an efficient way of doing it.

On the bright side, this will be recyclable for other papers.

eduggan commented 7 years ago

@andkov @ampiccinin Very sorry to do all that work outside of the framework and then hand back all the changes! Just had to do what I could to get the project done. But, yes... hopefully now that we've ironed these structural things out, it will be much easier for the other projects. :)

andkov commented 7 years ago

oh, I get it :) no worries

andkov commented 7 years ago

@eduggan , @ampiccinin @andreazammit In all studies but EAS, we use "Symbol" instead of "Digtit symbol". The underlying name of the item is identical in all studies. Is this intentional? In other words, is the symbol measure somehow different in EAS from other studies that warrants a differnt name?

This is a big deal structurally too: if a measure gets the same lable in this table it gets its separate row in tables like these:

I presume in all of these studies, we are talking about digit symbol substitution test?

andkov commented 7 years ago

@eduggan It seems that a few cognitive tests were omitted, but perhaps intentionally. Here is the complete list of cognitive tests that exist in the physical-cognitive catalog, but have no representation in the pulmonary (or grip, for that matter) spreadsheet. ILSE does not count, because it has no pulmonary or grip physical processes. However, the measure fluency has not been omitted from ELSA and fig_id from SATSA. Please comment if you remember the specifics on these two.

In the meanwhile I will assume these are errors of omissions and will fill in where missing. Aslso, to be thorough, I will assign labels to ILSE as well, so that we, hopefuly, could use the same domain structure for gait-cognitive track. This is how I filled in the structure based on my judgement. Please do let me know if something else fits better.

andkov commented 7 years ago

@eduggan @ampiccinin @andreazammit I believe I have reconstructed the structure of the domains you have imposed in the pulmonary paper. Please take a look to see if you detect any discrepancy.

|domain_b_label             |process_b_label          |
|:--------------------------|:------------------------|

|Attention & Working Memory |Digit Ordering           |
|Attention & Working Memory |Digit Span Backward      |
|Attention & Working Memory |Digit Span Backward Span |
|Attention & Working Memory |Digit Span Forward       |
|Attention & Working Memory |Digit Span Total         |
|Attention & Working Memory |Serial 7s                |
|Attention & Working Memory |Trials B                 |

|Learning & Memory          |Boston Story Delayed     |
|Learning & Memory          |Boston Story Immediate   |
|Learning & Memory          |Figure Memory            |
|Learning & Memory          |Logical Memory Delayed   |
|Learning & Memory          |Logical Memory Immediate |
|Learning & Memory          |Logical Memory Total     |
|Learning & Memory          |Objects Recalled         |
|Learning & Memory          |Objects Recognized       |
|Learning & Memory          |Prose Immediate          |
|Learning & Memory          |Word List Delayed        |
|Learning & Memory          |Word List Immediate      |
|Learning & Memory          |Word Recognition         |

|Mental Status              |Calculated MMSE          |
|Mental Status              |Ideas                    |
|Mental Status              |MMSE                     |

|Perceptual Reasoning       |Block Design             |
|Perceptual Reasoning       |Card Rotation            |
|Perceptual Reasoning       |Figure Copy              |
|Perceptual Reasoning       |Figure Identification    |
|Perceptual Reasoning       |Figure Logic             |
|Perceptual Reasoning       |IPSS Spatial Ability     |
|Perceptual Reasoning       |Line Orientation         |
|Perceptual Reasoning       |Pattern Comparison       |
|Perceptual Reasoning       |Picture Comletion        |
|Perceptual Reasoning       |Progressive Matrices     |

|Processing Speed           |Coding                   |
|Processing Speed           |Number Comparison        |
|Processing Speed           |Perceptual Speed         |
|Processing Speed           |Symbol                   |

|Verbal Ability             |Analogies                |
|Verbal Ability             |Animals                  |
|Verbal Ability             |Boston Naming Test       |
|Verbal Ability             |Categories               |
|Verbal Ability             |FAS                      |
|Verbal Ability             |Fluency                  |
|Verbal Ability             |Information              |
|Verbal Ability             |NART                     |
|Verbal Ability             |Synonyms                 |
|Verbal Ability             |Vocabulary               |

It appears to me that this structure can be used for all three track.

ampiccinin commented 7 years ago

ILSE should have grip

andkov commented 7 years ago

@ampiccinin , you are rigth. ILSE does have grip. And so should I!

andkov commented 7 years ago

@eduggan , @ampiccinin I believe this is issue can be closed. I have re-assembled the data tables to match the domain structure from pulmonary manuscript (based on the spreadsheed that @eduggan sent me). Please refer to https://github.com/IALSA/IALSA-2015-Portland/blob/master/reports/correlation-3/table-data/ as the authoritive source of this information.

I have removed tables broken down by gender, it was a specific request from Sean early on and it seems that it is no longer relevant. Besides, filtering by gender is too simple to justify a separate file.
I have added a new format meta which mimics the structure of the spreadsheet @eduggan sent me
the file you need for pulmary is pulmonary-andro-meta.csv

Requestion for action

please let me know if this is not in the format you need.
please close the issue if all is ok.

eduggan commented 7 years ago

@andkov @ampiccinin

Yes, the use of symbol across studies is an intentional harmonization choice. There are many versions (revisions and adaptations of this measure used across studies); however, they are all identical measuring the same type of performance. EAS requested me to change the name for the sake of accuracy (to which I have temporarily conceded). We will see what we have to do moving forward. The same issue exists for Raven's (Progressive matrices).

eduggan commented 7 years ago

@andkov @ampiccinin @annierobi @dgfinkel

Omitted tests: I know MMSE was accidentally omitted from the LASA analyses.

I am not sure why verbal fluency (verbal abilities) was omitted from ELSA and figure identification (processing speed) was omitted from SATSA. I don't seem to have results of analyses with these measures.

We can follow-up on this with the second review.

eduggan commented 7 years ago

@andkov @ampiccinin

Thanks Andrey! These tables look great and I think that issue is resolved: https://github.com/IALSA/IALSA-2015-Portland/blob/master/reports/correlation-3/table-data/.

However, these figures should be updated to match: https://raw.githubusercontent.com/IALSA/IALSA-2015-Portland/master/reports/domain-map/figure-png/print-domain-map-5.png

andkov commented 7 years ago

@eduggan , yes. That's not a small tweak. I'm working on it.

ampiccinin commented 7 years ago

@andkov @eduggan -

1) did ELSA-verbal fluency and SATSA-figureIdentification show up?

2) Did you receive offline replies from @annierobi or Dgfinkel about these? I see no record of a response here.

3) Is LASA-MMSE now available?

4) which of the issues above is "that issue", which is apparently resolved, and are the remaining ones resolved as well?

5) Are "symbol" and "ravens" naming differences resolved? What needs to happen with this?

6) I am trying to support AZammit in completing the grip paper, so want to ensure we are not having to re-hash all of these. ..and then there is walking...and cog-cog...

Thanks! A

eduggan commented 7 years ago

@andkov @ampiccinin

No. I don't think ELSA-verbal fluency (pulmonary) or SATSA figure identification (pulmonary) have been executed.
No, I have not received off-line replies about these from @annierobi or @dgfinkel.
LASA-MMSE is not available to my knowledge.
The "issue" is correcting the test groupings by domains and the domain titles in the "seed tables". The remaining issue is that visual representation of the tests and their domains on the primary page of each study. That has not been corrected.
Please let me know if you need me to do anything to help support you and @andreazammit to finish the grip paper.

annierobi commented 7 years ago

Sorry I thought that had been discussed in a separate email and must have forgotten to add it here. Verbal fluency in Elsa was not measured at each occasion and so there were too few time points that overlapped with the physical measure (also not available at each measure). Verbal fluency descriptives should still be in the descriptives table. You should be able to see the lack of overlap by looking at the table. I will check tomorrow morning to make sure I am remembering it correctly.

On Fri, Sep 22, 2017 at 9:44 PM Emily Duggan notifications@github.com wrote:

@andkov https://github.com/andkov @ampiccinin https://github.com/ampiccinin

1.

No. I don't think ELSA-verbal fluency (pulmonary) or SATSA figure identification (pulmonary) have been executed. 2.

No, I have not received off-line replies about these from @annierobi https://github.com/annierobi or @dgfinkel https://github.com/dgfinkel. 3.

LASA-MMSE is not available to my knowledge. 4.

The "issue" is correcting the test groupings by domains and the domain titles in the "seed tables". The remaining issue is that visual representation of the tests and their domains on the primary page of each study. That has not been corrected. 5.

Please let me know if you need me to do anything to help support you and @andreazammit https://github.com/andreazammit to finish the grip paper.

— You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub https://github.com/IALSA/IALSA-2015-Portland/issues/179#issuecomment-331600182, or mute the thread https://github.com/notifications/unsubscribe-auth/AKow3ZG5oVqtBtFKWVgF-atVxZ9Janmfks5slGJ4gaJpZM4OBqRp .

IALSA / IALSA-2015-Portland

Cognitive domains structure for Pulmonary #179

Requestion for action