As a MDM user, I want to get a FDZ Question export to reimport in MDM

[kpstephan]

this means

• 1 image per FDZ question and 1 corresponding image JSON per FDZ question
  • the images need to be "plain" which means no white borders, no navigation buttons, no language flags etc.
• 1 corresponding JSON per FDZ question

We need the files to have

• the correct folder structure and
• meaningful file names

[dzhwmeisner]

http://vmdevwiki.dzhw.local/fdz/mdm_ifq-mdm-mod_Update310718.zip send to Robert

[UteH]

@UteH: get feedback from @rbirkelbach and @svietgen about the 3rd zofar export https://github.com/dzhw/Effizienzprojekt_Modul_H/issues/12

[UteH]

Feedback zur 3. Lieferung

Fazit questions

• topic wurde nicht geliefert
• questionNumber anpassen, z. B. 6.1 (not: page6.1)
  • questionNumber, SucessorNumbers, file names
• teilweise
  • engl. Übersetzung nicht vorhanden
  • einzelne JSONs gefehlt
    • according to the questionnaire pdf image and JSON for question number 8.19 are missing (was also the case for the real delivery).
• Ordnerstruktur hatte nicht gepasst
  • teilweise Ordnernamen
    
• Encoding bei englischen Bildern:
  • wir wollen es so abbilden, wie es im Feld war. wenn des encoding damals auch falsch war, wollen wir das Bild ensprechend haben
  • here, here
  • here

Fazit variables

[UteH]

as we have not identified the FDZ-question #87 yet, we will just use this format: Page-ID.#

[UteH]

Hi @dzhwmeisner should I plan to be testing this issue today?

[dzhwmeisner]

Here are current Updates for the IFQ-Survey and the SE21-Survey

http://vmdevwiki.dzhw.local/fdz/MDM_ifq_Update02102018.zip http://vmdevwiki.dzhw.local/fdz/MDM_se21_Update02102018.zip

with the request to test

• topic wurde nicht geliefert ... in work

• questionNumber anpassen (Prozesse mit PFP berücksichtigen!) questionNumber, SucessorNumbers, file names ... set to pageId.#

• engl. Übersetzung nicht vorhanden ... in IFQ-Survey fixed ... SE21-Survey was in parts not multilingual. How to deal with partly translated surveys?

• einzelne JSONs gefehlt (in 3. Lieferung dann ok) need to be checked in current update.

• Ordnerstruktur hatte nicht gepasst need more details to fix

[UteH]

~work in progress, will be further edited~

hidden, because referring to old files. See below

thanks! I have some first remarks, but I need more time to test the exports thoroughly.

Wissenschaftlerbefragung

files

• the files in the mdm are not the files you uploaded earlier (according to the file names) or am I missing something?
• for questions we need to discuss in the FDZ, if we need cancel, index, startpage1, startpage2 (but for now, that's ok)
• I think the sequence is incorrect? When I "click through" the questions, the survey ends the first time wirh https://metadatamanagement-test.cfapps.io/#!/de/questions/que-meisner1-ins1-startpage_b_sw_bb_mca$
  • should the question before (https://metadatamanagement-test.cfapps.io/#!/de/questions/que-meisner1-ins1-page1_b_sc1$?search-result-index=4) have two possible successors?
  • also for https://metadatamanagement-test.cfapps.io/#!/de/questions/que-meisner1-ins1-page1_b_sc1$?search-result-index=4 AND https://metadatamanagement-test.cfapps.io/#!/de/questions/que-meisner1-ins1-page2_b_sc$?search-result-index=6 -> only successor is to leave the survey and no previous question
  • also no successor: https://metadatamanagement-test.cfapps.io/#!/de/questions/que-meisner1-ins1-page4_b_sw_bb_mc$?search-result-index=10
  • I am wondering if startpage1 and 2 were actually on one page? (agree or disagree)
    -> please check the sequence of the questions, something seems messed up here
• topic is always "bla bla" -> you said earlier "in work" -> ok
• all english pages from the beginning until https://metadatamanagement-test.cfapps.io/#!/en/questions/que-meisner1-ins1-startpage_b_mc$ seem incorrect?
• according to the questionnaire image and JSON for question number 8.19 are missing (was also the case for the actual delivery).
  • Is this correct or is the pdf incorrect?
  • should be around page 90.# Please provide feedback
• instrument.title should be "Fra­ge­bo­gen der DZHW-Wis­sen­schaft­ler­be­fra­gung 2016"/"Ques­tion­naire of the DZHW Sci­en­tists Sur­vey 2016" instead of "Erstes Instrument"/"First Instrument"
  • -> we need to deliver the title to you, you need to ask us for that (I looked it up in the MDM)

    file structure

• questions: seems right!
• for the Wissenschaftlerbefragung two folders seem to be named differently:

  file name / question.number

• seems right, I have to check with @anneweber first

  encoding

  • I cannot make it work. even if I do MS Sxcel > Daten > aus Text > csv-Datei the encoding is not correct

Generally, I need more time to test such a complex issue.

[UteH]

@dzhwmeisner is this the right project? https://metadatamanagement-test.cfapps.io/#!/de/studies/stu-meisner1$ ? There seem to be many oddities concerning the sequence of the questions and the filenames of the images do not correspond to the files you uploaded earlier. Am I missing something here? Please provide help

edit: I am waiting for feedback before further testing

[UteH]

Ute replace with new files

[UteH]

I uploaded the files you sent earliert to the mdm (DAP: meisner1)

Wissenschaftlerbefragung

Generally, a lot of details seem to be messed up whereas they had been working in the actual mdm export. Normally, I would have stopped testing because the export has so many bugs. But in terms of time I tried to test the export so that we have a list to work on. We should discuss this.

I often compared the new export with the real export in the mdm with the new test project. Here are some examples of what I found:

questions

additionalQuestionText.en

• positive: dropdown lists are already part of the additionalQuestionText! #85

content (jsons)

• all introductions (de/en) and instructions (de/en) are missing OR they art part of the questionText (whereas in the real MDM they were marked as introduction) - why?

content (images)

I found these errors absolutely randomly, please investigate further

• (almost?) all introductions (de/en) and instructions (de/en) are missing
• differences - why?
  • mdm: questionText and introducion: 10.1, 10.5
  • test: everything in questionText: page107.1, page111.1
• introduction is missing:
  • test: https://metadatamanagement-test.cfapps.io/#!/de/questions/que-meisner1-ins1-page59.1$
  • mdm: https://metadata.fdz.dzhw.eu/#!/en/questions/que-scs2016-ins1-5.9$?search-result-index=1
• answers are missing:
  • test: https://metadatamanagement-test.cfapps.io/#!/en/questions/que-meisner1-ins1-page30a.1$?search-result-index=1
  • mdm: https://metadata.fdz.dzhw.eu/#!/en/questions/que-scs2016-ins1-3.1a$?search-result-index=1
  • also for: https://metadatamanagement-test.cfapps.io/#!/de/questions/que-meisner1-ins1-page30b.1$?search-result-index=1, https://metadatamanagement-test.cfapps.io/#!/de/questions/que-meisner1-ins1-page30c.1$?search-result-index=1
• in the real mdm, we have 128 questions (in the real mdm starting only with 1.1 ..., ), in the test we have 173, please investigate the differences further
  • in the new test version the soft reminders are often exported as a separate question which it should not be

completeness

• according to the questionnaire pdf image and JSON for question number 8.19 are missing (was also the case for the real delivery).
  • Is this correct or is the pdf incorrect/not up to date?
  • should be around page 90.# Please provide feedback
• ~for questions we need to discuss in the FDZ, if we need cancel, index, startpage1, startpage2 (but for now, that's ok)~ it's ok this way (talked to AnneW)

sequence

• the sequence is incorrect. When I "click through" the questions, the survey ends the first time with https://metadatamanagement-test.cfapps.io/#!/de/questions/que-meisner1-ins1-startpage.2$
  • I am wondering if startpage1 and 2 were actually on one page? (agree or disagree)
  • should the question before have two possible successors?
  • after almost every question follows: https://metadatamanagement-test.cfapps.io/#!/de/questions/que-meisner1-ins1-page4.2$?search-result-index=10
  • the following questions do not have a previous question

• often, the following questions are different in then real mdm

  • test: https://metadatamanagement-test.cfapps.io/#!/de/questions/que-meisner1-ins1-page71.1$
  • mdm: https://metadata.fdz.dzhw.eu/#!/de/questions/que-scs2016-ins1-7.6$?search-result-index=1

  -> please check the whole sequence of the questions, almost everything seems messed up

topic

• no topic -> you said earlier "in work" -> ok
• but we still need it

encoding

• for the english texts and images: € is not ecoded correctly, e.g. page26a.1 and page25.1
  • what ist this? page59.1

images

• the quality is fine, (further ideas are an even better resolution (if possible), no white boarders -> but not within this issue!)
• plain white: de, en
  • according to the real delivery, it should be part of the previous question:
  • in MDM: https://metadata.fdz.dzhw.eu/#!/en/questions/que-scs2016-ins1-8.8$
  • in test: https://metadatamanagement-test.cfapps.io/#!/en/questions/que-meisner1-ins1-page80.1$
• it is correct the no buttos etc. are shown
• but: the images are in some way not like we had it for the real Wissenschaftlerbefragung
  • test: page19.1_1
  • mdm: https://metadata.fdz.dzhw.eu/public/files/questions/que-scs2016-ins1-1.20$/images/1.20_2.png -> the soft reminders etc. should be part of the image
  • (then the text should be part of the question-json, right? -> we need to discuss this in the fdz)
• I think, often the answers soft reminders are exported as separate question, but they should be part of the question as displayed to the participant (MDM example)

english version

no claim to completeness! I've been looking for "???" in the images, "zofar" in the questionText and some I detected randomly. Please investigate further

• page87_bewerberauswahl.1 (and bewerberauswahl.2 again)
• page90_1.1
• page90_2.1
• page90_4.1
• question text and image ("()") incorrect page53.1
• little errors, for example
  • english version of startpage.2 incorrect (this question text and error appear frequently)
  • page19.1 -> first category
  • encoding of €: page26a.1,
  • what ist this? page59.1

instrument.title

• instrument.title should be "Fra­ge­bo­gen der DZHW-Wis­sen­schaft­ler­be­fra­gung 2016"/"Ques­tion­naire of the DZHW Sci­en­tists Sur­vey 2016" (de/en) instead of "Erstes Instrument"/"First Instrument"
  • -> we need to deliver the title to you, you need to ask us for that (I looked it up in the MDM)
  • we should put this in the Programmiervorlage

file structure

• import into mdm works

file name / folder name / question.number

• two folders seem to be named differently: (see question.number)
• sorry, there seems to have been a missunderstanding: we want the format to be "26a.1", not "page26a.1"
• bewerberauswahl.1 seem bewerberauswahl.1 seem odd (see file name)

variables

encoding

• I cannot make it work. even if I do MS Sxcel > Daten > aus Text > csv-Datei the encoding is not correct

@dzhwmeisner

Generally, I need more time to test such a complex issue. Also, please do upload the files into the MDM and test the export yourself first. Next time, please wait to build the final version (e. g. with the correct sequence and with topic) so that I can test a more final versions.

[UteH]

~as discussed in the retro we will go back to the last question export that had worked and try to improve from there.~

@dzhwmeisner did edit the new export, so this to do list is valid

In our file system it looks like the import was called "mdm_ifq-mdm-mod_Update240518" -> @dzhwmeisner can you work with this?

[UteH]

~I updated the To Do list.~

@dzhwmeisner did edit the new export, so this to do list is valid

[UteH]

PS: @dzhwmeisner @vdick do you know about our "post validation button"?

with clicking the button, several dependecies are tested, for questions this is (according to the mdm wiki):

• question.instrumentId: there must be an instrument with that id
• question.successors: there must be a question with that id
• Check for a depending question image

[dzhwmeisner]

newest Release is DAP: https://metadatamanagement-test.cfapps.io/#!/de/studies/stu-zofar221018$

Wissenschaftlerbefragung

Generally, a lot of details seem to be messed up whereas they had been working in the actual mdm export. Normally, I would have stopped testing because the export has so many bugs. But in terms of time I tried to test the export so that we have a list to work on. We should discuss this.

I often compared the new export with the real export in the mdm with the new test project. Here are some examples of what I found:

questions

additionalQuestionText.en

• positive: dropdown lists are already part of the additionalQuestionText! #85

(dzhwmeisner) ==> seems to be. will be checked

content (jsons)

• all introductions (de/en) and instructions (de/en) are missing OR they art part of the questionText (whereas in the real MDM they were marked as introduction) - why?

(dzhwmeisner) ==> fixed. JSON Attribute has changed

content (images)

I found these errors absolutely randomly, please investigate further

• (almost?) all introductions (de/en) and instructions (de/en) are missing (dzhwmeisner) ==> fixed. JSON Attribute has changed

• differences - why?

  • mdm: questionText and introducion: 10.1, 10.5
  • test: everything in questionText: page107.1, page111.1

(dzhwmeisner) ==> maybe reworked for mdm by FDZ? Fixed in current Test-Version

• introduction is missing:
  • test: https://metadatamanagement-test.cfapps.io/#!/de/questions/que-meisner1-ins1-page59.1$
  • mdm: https://metadata.fdz.dzhw.eu/#!/en/questions/que-scs2016-ins1-5.9$?search-result-index=1

(dzhwmeisner) ==> maybe reworked for mdm by FDZ? Fixed in current Test-Version

• answers are missing:
  • test: https://metadatamanagement-test.cfapps.io/#!/en/questions/que-meisner1-ins1-page30a.1$?search-result-index=1
  • mdm: https://metadata.fdz.dzhw.eu/#!/en/questions/que-scs2016-ins1-3.1a$?search-result-index=1

(dzhwmeisner) ==> fixed.

• also for: https://metadatamanagement-test.cfapps.io/#!/de/questions/que-meisner1-ins1-page30b.1$?search-result-index=1, https://metadatamanagement-test.cfapps.io/#!/de/questions/que-meisner1-ins1-page30c.1$?search-result-index=1

(dzhwmeisner) ==> fixed.

• in the real mdm, we have 128 questions (in the real mdm starting only with 1.1 ..., ), in the test we have 173, please investigate the differences further

(dzhwmeisner) ==> Soft reminders not separated anymore. Now we have 135 Questions including index,start,end,cancel

• in the new test version the soft reminders are often exported as a separate question which it should not be

(dzhwmeisner) ==> Soft reminders not separated anymore.

completeness

• according to the questionnaire pdf image and JSON for question number 8.19 are missing (was also the case for the real delivery).
  • Is this correct or is the pdf incorrect/not up to date?
  • should be around page 90.# Please provide feedback

(dzhwmeisner) ==> Seems to be87_bewerberauswahl

• for questions we need to discuss in the FDZ, if we need cancel, index, startpage1, startpage2 (but for now, that's ok)

(dzhwmeisner) ==> current Test Version include these questions

sequence

• the sequence is incorrect. When I "click through" the questions, the survey ends the first time with https://metadatamanagement-test.cfapps.io/#!/de/questions/que-meisner1-ins1-startpage.2$
  • I am wondering if startpage1 and 2 were actually on one page? (agree or disagree)
  • should the question before have two possible successors?
  • after almost every question follows: https://metadatamanagement-test.cfapps.io/#!/de/questions/que-meisner1-ins1-page4.2$?search-result-index=10

(dzhwmeisner) ==> fixed. the old Test Version based on Zofar-question perspective. not on on Zofar-page perspective

• the following questions do not have a previous question
  • often, the following questions are different in then real mdm
  • test: https://metadatamanagement-test.cfapps.io/#!/de/questions/que-meisner1-ins1-page71.1$
  • mdm: https://metadata.fdz.dzhw.eu/#!/de/questions/que-scs2016-ins1-7.6$?search-result-index=1

(dzhwmeisner) ==> fixed.

topic

• no topic -> you said earlier "in work" -> ok
• but we still need it

(dzhwmeisner) ==> implemented for Questions on which page an Topic is set. Not for following questions.

encoding

• for the english texts and images: € is not ecoded correctly, e.g. page26a.1 and page25.1

(dzhwmeisner) ==> fixed.

• what ist this? page59.1

(dzhwmeisner) ==> fixed.

images

• the quality is fine, (further ideas are an even better resolution (if possible), no white boarders -> but not within this issue!)
• plain white: de, en

(dzhwmeisner) ==> fixed

• according to the real delivery, it should be part of the previous question:
  • in MDM: https://metadata.fdz.dzhw.eu/#!/en/questions/que-scs2016-ins1-8.8$
  • in test: https://metadatamanagement-test.cfapps.io/#!/en/questions/que-meisner1-ins1-page80.1$

(dzhwmeisner) ==> fixed. the old Test Version based on Zofar-question perspective. not on on Zofar-page perspective

• it is correct the no buttos etc. are shown
• but: the images are in some way not like we had it for the real Wissenschaftlerbefragung
  • test: page19.1_1
  • mdm: https://metadata.fdz.dzhw.eu/public/files/questions/que-scs2016-ins1-1.20$/images/1.20_2.png

(dzhwmeisner) ==> fixed

-> the soft reminders etc. should be part of the image

• (then the text should be part of the question-json, right? -> we need to discuss this in the fdz)
  • I think, often the answers soft reminders are exported as separate question, but they should be part of the question as displayed to the participant (MDM example)

(dzhwmeisner) ==> fixed. the old Test Version based on Zofar-question perspective. not on on Zofar-page perspective

english version

no claim to completeness! I've been looking for "???" in the images, "zofar" in the questionText and some I detected randomly. Please investigate further

• page87_bewerberauswahl.1 (and bewerberauswahl.2 again)

(dzhwmeisner) ==> fixed

• page90_1.1

(dzhwmeisner) ==> fixed

• page90_2.1

(dzhwmeisner) ==> fixed

• page90_4.1

(dzhwmeisner) ==> fixed

• question text and image ("()") incorrect page53.1

(dzhwmeisner) ==> not fixed. this is an placeholder for an answer of a previous question. Questions will not be answered while screenshotting.

• little errors, for example
  • english version of startpage.2 incorrect (this question text and error appear frequently)

(dzhwmeisner) ==> fixed. the old Test Version based on Zofar-question perspective. not on on Zofar-page perspective

• page19.1 -> first category

(dzhwmeisner) ==> fixed

• encoding of €: page26a.1,

(dzhwmeisner) ==> fixed

• what ist this? page59.1

(dzhwmeisner) ==> fixed

instrument.title

• instrument.title should be "Fra­ge­bo­gen der DZHW-Wis­sen­schaft­ler­be­fra­gung 2016"/"Ques­tion­naire of the DZHW Sci­en­tists Sur­vey 2016" (de/en) instead of "Erstes Instrument"/"First Instrument"
  • -> we need to deliver the title to you, you need to ask us for that (I looked it up in the MDM)
  • we should put this in the Programmiervorlage

file structure

• import into mdm works

file name / folder name / question.number

• two folders seem to be named differently: (see question.number)
• sorry, there seems to have been a missunderstanding: we want the format to be "26a.1", not "page26a.1"

(dzhwmeisner) ==> fixed. Pages can named individual. in the current Test-Version only the 'page' will be replaced from name.

• bewerberauswahl.1 seem bewerberauswahl.1 seem odd (see file name)

(dzhwmeisner) ==> not fixed. the schema page# is not required. Pages can named individual. in the current Test-Version only the 'page' will be replaced from name.

variables

encoding

• I cannot make it work. even if I do MS Sxcel > Daten > aus Text > csv-Datei the encoding is not correct

(dzhwmeisner) ==> the CSV-File is UTF-8 encoded. In the current ZIP is an additional SO-8859-1 encoded Version

[dzhwmeisner]

MDM_stu-zofar221018.zip According Archive to stu-zofar221018$

[UteH]

@dzhwmeisner thanks for the new export! I found some inconsitencies again, I think it's easiest, if we talk about it in person. I will contact you via Outlook :calendar:

[UteH]

Info:

• refers to DAP: https://metadatamanagement-test.cfapps.io/#!/de/studies/stu-zofar221018$
• all new remarks are formatted as a checkbox

Wissenschaftlerbefragung

questions

additionalQuestionText.en

• positive: dropdown lists are already part of the additionalQuestionText! #85 (dzhwmeisner) ==> seems to be. will be checked -> Ute: ok for de and en

content (jsons)

• all introductions (de/en) and instructions (de/en) are missing OR they art part of the questionText (whereas in the real MDM they were marked as introduction) - why? (dzhwmeisner) ==> fixed. JSON Attribute has changed ok, see section discussion!

• [ ] string (section) is weird:

  • https://metadatamanagement-test.cfapps.io/#!/de/questions/que-zofar221018-ins1-108.1$?search-result-index=1
  • https://metadata.fdz.dzhw.eu/#!/de/questions/que-scs2016-ins1-10.2$?search-result-index=1
  • I think we need to discuss this, because please remember that we said: please do not put effort in manually beautifying sections
  • so is this an example of the "costs" that come with automation if we decide to not do manual beautifying?
  • Can we avoid this in the future somehow (code conventions?) we need to write this down
  • see section discussion!

• [ ] questionText, introduction and instruction are still inconsistent between mdm and test - section problem, example:

  • test: https://metadatamanagement-test.cfapps.io/#!/de/questions/que-zofar221018-ins1-105.1$?search-result-index=1
  • mdm: https://metadata.fdz.dzhw.eu/#!/de/questions/que-scs2016-ins1-9.1a$?search-result-index=1
  • see section discussion!

• [ ] often: (de and en) "Kein Fra­ge­text vor­han­den" / "No ques­tion­text avail­able"

  • https://metadatamanagement-test.cfapps.io/#!/de/questions/que-zofar221018-ins1-start.1$?search-result-index=2
  • because of the concatenate of all questions of one page
  • if "Kein Fra­ge­text vor­han­den" / "No ques­tion­text avail­able" please delete the attribute

• [ ] in 1.1.json and 37(?) others: more than one number (these are the questionNumbers of the page)

  "number":[
      "1.1",
      "1.2"

  • I talked to René. As the mdm does not use the json attributes questionNumber and InstrumentNumber. (The mdm gets quesionNumber and instrumentNumber from the files and folders.)
  • Thus, we should delete both of these artefacts from the jsons.

• [ ] questionText is incomplete (section according to QML):

  • 25.1
  • 62.1
  • 66.1
  • ...
  • I found these exemplary cases skipping through the QML searching for "section" -> section discussion!

• all introductions (de/en) and instructions (de/en) are missing OR they art part of the questionText (whereas in the real MDM they were marked as introduction) - why?

(dzhwmeisner) ==> fixed. JSON Attribute has changed -> Ute: ok, see section discussion

• [ ] "keine Anleitung vorhanden"

  • https://metadatamanagement-test.cfapps.io/#!/de/questions/que-zofar221018-ins1-110.1$?search-result-index=1
  • if "keine Anleitung vorhanden" / ... please delete this attribute

• [ ] images are ok, but introduction is missing -> how can this be, do we need to do sth. about this?

  • MDM: https://metadata.fdz.dzhw.eu/#!/de/questions/que-scs2016-ins1-8.24$?search-result-index=1
  • test: https://metadatamanagement-test.cfapps.io/#!/de/questions/que-zofar221018-ins1-96.1$?search-result-index=1
  • test: https://metadatamanagement-test.cfapps.io/#!/de/questions/que-zofar221018-ins1-104.1$?search-result-index=1

• [ ] 72.1

  • json: the actual question text (section?) is missing
  • image (de only): zofar code snippet
  • see section discussion!

content (images)

• [ ] content is okay but introductions are bold / not bold -> which one is true?
  • mdm: https://metadata.fdz.dzhw.eu/#!/de/questions/que-scs2016-ins1-10.1$?search-result-index=1
  • test: https://metadatamanagement-test.cfapps.io/#!/de/questions/que-zofar221018-ins1-107.1$?search-result-index=2
  • see section discussion

I found these errors absolutely randomly, please investigate further

• (almost?) all introductions (de/en) and instructions (de/en) are missing

(dzhwmeisner) ==> fixed. JSON Attribute has changed

• differences - why?
  • mdm: questionText and introduction: 10.1, 10.5
  • test: everything in questionText: page107.1, page111.1

(dzhwmeisner) ==> maybe reworked for mdm by FDZ? Fixed in current Test-Version

• introduction is missing:
  • test: https://metadatamanagement-test.cfapps.io/#!/de/questions/que-meisner1-ins1-page59.1$
  • mdm: https://metadata.fdz.dzhw.eu/#!/en/questions/que-scs2016-ins1-5.9$?search-result-index=1

(dzhwmeisner) ==> maybe reworked for mdm by FDZ? Fixed in current Test-Version

• answers are missing:
  • test: https://metadatamanagement-test.cfapps.io/#!/en/questions/que-meisner1-ins1-page30a.1$?search-result-index=1
  • mdm: https://metadata.fdz.dzhw.eu/#!/en/questions/que-scs2016-ins1-3.1a$?search-result-index=1

(dzhwmeisner) ==> fixed. -> Ute: ok. How can that happen? Do we need to do something about this?

• also for: https://metadatamanagement-test.cfapps.io/#!/de/questions/que-meisner1-ins1-page30b.1$?search-result-index=1, https://metadatamanagement-test.cfapps.io/#!/de/questions/que-meisner1-ins1-page30c.1$?search-result-index=1

(dzhwmeisner) ==> fixed. -> Ute: ok. How can that happen? Do we need to do something about this?

• in the real mdm, we have 128 questions (in the real mdm starting only with 1.1 ..., ), in the test we have 173, please investigate the differences further

(dzhwmeisner) ==> Soft reminders not separated anymore. Now we have 135 Questions including index,start,end,cancel -> Ute: ok

• in the new test version the soft reminders are often exported as a separate question which it should not be

(dzhwmeisner) ==> Soft reminders not separated anymore. -> Ute: ok

completeness

• [ ] number of Questions: mdm: 128, test: 135 -> ok, so just FYI:

  • missing in the mdm:
  • ok: index, start, cancel, 123.1
  • not ok: 90_2.1, 90_3.1, 90_4.1 -> you delivered. I think we accidentally deleted them.

• according to the questionnaire pdf image and JSON for question number 8.19 are missing (was also the case for the real delivery).

  • Is this correct or is the pdf incorrect/not up to date?
  • should be around page 90.# Please provide feedback

(dzhwmeisner) ==> Seems to be87_bewerberauswahl -> Ute: ok, seems fine

• for questions we need to discuss in the FDZ, if we need cancel, index, startpage1, startpage2 (but for now, that's ok)

(dzhwmeisner) ==> current Test Version include these questions -> Ute: we will leave it like this!

sequence

• [ ] I think , the sequence is still messed up. These are examples! It occurs often that there are more subsequent questions than in the mdm

  • mdm: 1 subsequent question: 1.1
  • test: 4 subsequent questions: 1.1
  • mdm: 1 subsequent question: https://metadata.fdz.dzhw.eu/#!/de/questions/que-scs2016-ins1-1.2$
  • test: 3 subsequent questions:https://metadatamanagement-test.cfapps.io/#!/de/questions/que-zofar221018-ins1-2.1$?search-result-index=2
  • test: https://metadatamanagement-test.cfapps.io/#!/de/questions/que-zofar221018-ins1-20.1$
  • I think the "exit question" is also not in the right places, e.g. it should be a subsequent question to 25.1, right?

    I cannot test the sequence because I don't know what to compare the question flow to.

• [ ] here, I think the new export (test) is correct and the real one is false

  • test: https://metadatamanagement-test.cfapps.io/#!/de/questions/que-zofar221018-ins1-105.1$
  • mdm: https://metadata.fdz.dzhw.eu/#!/de/questions/que-scs2016-ins1-9.2a$

• the sequence is incorrect. When I "click through" the questions, the survey ends the first time with https://metadatamanagement-test.cfapps.io/#!/de/questions/que-meisner1-ins1-startpage.2$

  • I am wondering if startpage1 and 2 were actually on one page? (agree or disagree)
  • should the question before have two possible successors?
  • after almost every question follows: https://metadatamanagement-test.cfapps.io/#!/de/questions/que-meisner1-ins1-page4.2$?search-result-index=10

(dzhwmeisner) ==> fixed. the old Test Version based on Zofar-question perspective. not on on Zofar-page perspective -> Ute: Ok, sequence still seems odd, see above

• the following questions do not have a previous question
  • often, the following questions are different in then real mdm
  • test: https://metadatamanagement-test.cfapps.io/#!/de/questions/que-meisner1-ins1-page71.1$
  • mdm: https://metadata.fdz.dzhw.eu/#!/de/questions/que-scs2016-ins1-7.6$?search-result-index=1

(dzhwmeisner) ==> fixed. -> Ute: Ok, sequence still seems odd, see above

topic

• no topic -> you said earlier "in work" -> ok
• but we still need it

(dzhwmeisner) ==> implemented for Questions on which page an Topic is set. Not for following questions. -> Ok

• [ ] if topic is "Kein Thema vor­han­den" please delete the attribute from the json

• [ ] no topic in json, but image says so: https://metadatamanagement-test.cfapps.io/#!/de/questions/que-zofar221018-ins1-105.1$

• [ ] https://metadatamanagement-test.cfapps.io/#!/de/questions/que-zofar221018-ins1-cancel.1$?search-result-index=135 is header, not a topic (de/en)

• [ ] https://metadatamanagement-test.cfapps.io/#!/de/questions/que-zofar221018-ins1-start.1$?search-result-index=2 is header, not a topic (de/en)

encoding

• for the english texts and images: € is not ecoded correctly, e.g. page26a.1 and page25.1

(dzhwmeisner) ==> fixed. -> Ute: ok

• what ist this? page59.1

(dzhwmeisner) ==> fixed.

images

• the quality is fine, (further ideas are an even better resolution (if possible), no white boarders -> but not within this issue!)
• plain white: de, en

(dzhwmeisner) ==> fixed

• according to the real delivery, it should be part of the previous question:
  • in MDM: https://metadata.fdz.dzhw.eu/#!/en/questions/que-scs2016-ins1-8.8$
  • in test: https://metadatamanagement-test.cfapps.io/#!/en/questions/que-meisner1-ins1-page80.1$

(dzhwmeisner) ==> fixed. the old Test Version based on Zofar-question perspective. not on on Zofar-page perspective

• it is correct the no buttos etc. are shown
• but: the images are in some way not like we had it for the real Wissenschaftlerbefragung
  • test: page19.1_1
  • mdm: https://metadata.fdz.dzhw.eu/public/files/questions/que-scs2016-ins1-1.20$/images/1.20_2.png

(dzhwmeisner) ==> fixed

-> the soft reminders etc. should be part of the image

• (then the text should be part of the question-json, right? -> we need to discuss this in the fdz)
  • I think, often the answers soft reminders are exported as separate question, but they should be part of the question as displayed to the participant (MDM example)

(dzhwmeisner) ==> fixed. the old Test Version based on Zofar-question perspective. not on on Zofar-page perspective

english version

no claim to completeness! I've been looking for "???" in the images, "zofar" in the questionText and some I detected randomly. Please investigate further

• page87_bewerberauswahl.1 (and bewerberauswahl.2 again)

(dzhwmeisner) ==> fixed

• [ ] https://metadatamanagement-test.cfapps.io/public/files/questions/que-zofar221018-ins1-start.1$/images/start.1_1.png

  (as always: if this was a mistake in the live survey, we want it this way)

• page90_1.1

(dzhwmeisner) ==> fixed

• page90_2.1

(dzhwmeisner) ==> fixed

• page90_4.1

(dzhwmeisner) ==> fixed

• question text and image ("()") incorrect page53.1

(dzhwmeisner) ==> not fixed. this is an placeholder for an answer of a previous question. Questions will not be answered while screenshotting.

• little errors, for example
  • english version of startpage.2 incorrect (this question text and error appear frequently)

(dzhwmeisner) ==> fixed. the old Test Version based on Zofar-question perspective. not on on Zofar-page perspective

• page19.1 -> first category

(dzhwmeisner) ==> fixed

• encoding of €: page26a.1,

(dzhwmeisner) ==> fixed

• what ist this? page59.1

(dzhwmeisner) ==> fixed

instrument.title

• instrument.title should be "Fra­ge­bo­gen der DZHW-Wis­sen­schaft­ler­be­fra­gung 2016"/"Ques­tion­naire of the DZHW Sci­en­tists Sur­vey 2016" (de/en) instead of "Erstes Instrument"/"First Instrument"
  • -> we need to deliver the title to you, you need to ask us for that (I looked it up in the MDM)
  • we should put this in the Programmiervorlage

file structure

• import into mdm works

• [ ] could you delete the folder "json"? so, that it is

  study\MDM\questions
  study\MDM\variables
  (study\MDM\instruments)

  and not

  study\MDM\json\questions

file name / folder name / question.number

• two folders seem to be named differently: (see question.number)
• sorry, there seems to have been a missunderstanding: we want the format to be "26a.1", not "page26a.1"

(dzhwmeisner) ==> fixed. Pages can named individual. in the current Test-Version only the 'page' will be replaced from name. -> Ute: that's fine. We will take the name exactly as it is in zofar. SISO.

• bewerberauswahl.1 seem bewerberauswahl.1 seem odd (see file name)

(dzhwmeisner) ==> not fixed. the schema page# is not required. Pages can named individual. in the current Test-Version only the 'page' will be replaced from name. -> Ute: that's fine. We will take the name exactly as it is in zofar. SISO.

variables

• this is a different issue #93

[UteH]

@andreaschu @dzhwmeisner do you have any kind of question flow diagram (Filterführungsdiagramm) for the Wissenschaftlerbefragung 2016?

I can't work out the "real" question flow

[UteH]

@dzhwmeisner I think I am done for now, thank you. Feel free to make an appointment if you think we should talk about some remarks face to face.

[UteH]

@UteH: compare qml sequence with export, nothing more.

[kastirke]

@UteH : Concerning your mail from 5th November 2018: Question flow within Wissenschaftlerbefragung 2016

dzhwmeisner and I checked the Wissenschaftlerbefragung-Programmieranleitung (last version we own is from 21th October 2015; this is not the final one; there were changes after this date; please ask the project members for the final version). To us, the qml ist right with respect to the variable definition.

For example:

Here you find the whole Anleitung: iFQ_WissBefr_2015_Fragebogen_Kodiert_2015_10_15korr.pdf

[UteH]

@kastirke @dzhwmeisner thank you for these information!

We also talked about how to test the export regarding the question flow of the scs2016. I will only check if the export is correct with respect to the qml and not the Anleitung (because it could contain false information that do not concern the export.)

Thank you for clarifying the situation!

[kastirke]

To us, you will be right by only checking the qml. It defines all variables inluding visible conditions, question flow, instructions, assisting variables and so on. We always advice the projekct teams to update their PV, but we can't quarantee on this.

[dzhwmeisner]

Info:

• refers to DAP: https://metadatamanagement-test.cfapps.io/#!/de/studies/stu-zofar091118$
• all new remarks are formatted as a checkbox

Wissenschaftlerbefragung

questions

additionalQuestionText.en

• positive: dropdown lists are already part of the additionalQuestionText! #85 (dzhwmeisner) ==> seems to be. will be checked -> Ute: ok for de and en

content (jsons)

• all introductions (de/en) and instructions (de/en) are missing OR they art part of the questionText (whereas in the real MDM they were marked as introduction) - why? (dzhwmeisner) ==> fixed. JSON Attribute has changed ok, see section discussion!

• [ ] string (section) is weird:

  • https://metadatamanagement-test.cfapps.io/#!/de/questions/que-zofar221018-ins1-108.1$?search-result-index=1
  • https://metadata.fdz.dzhw.eu/#!/de/questions/que-scs2016-ins1-10.2$?search-result-index=1
  • I think we need to discuss this, because please remember that we said: please do not put effort in manually beautifying sections
  • so is this an example of the "costs" that come with automation if we decide to not do manual beautifying?
  • Can we avoid this in the future somehow (code conventions?) we need to write this down
  • see section discussion!

• [ ] questionText, introduction and instruction are still inconsistent between mdm and test - section problem, example:

  • test: https://metadatamanagement-test.cfapps.io/#!/de/questions/que-zofar221018-ins1-105.1$?search-result-index=1
  • mdm: https://metadata.fdz.dzhw.eu/#!/de/questions/que-scs2016-ins1-9.1a$?search-result-index=1
  • see section discussion!

• [x] often: (de and en) "Kein Fra­ge­text vor­han­den" / "No ques­tion­text avail­able"

  • https://metadatamanagement-test.cfapps.io/#!/de/questions/que-zofar221018-ins1-start.1$?search-result-index=2
  • because of the concatenate of all questions of one page
  • if "Kein Fra­ge­text vor­han­den" / "No ques­tion­text avail­able" please delete the attribute

(dzhwmeisner) ==> fixed

• [x] in 1.1.json and 37(?) others: more than one number (these are the questionNumbers of the page)

  "number":[
      "1.1",
      "1.2"

  • I talked to René. As the mdm does not use the json attributes questionNumber and InstrumentNumber. (The mdm gets quesionNumber and instrumentNumber from the files and folders.)
  • Thus, we should delete both of these artefacts from the jsons.

  (dzhwmeisner) ==> fixed

• [ ] questionText is incomplete (section according to QML):

  • 25.1 (dzhwmeisner) ==> Part of the section problem

  • 62.1 (dzhwmeisner) ==> Part of the section problem

  • 66.1 (dzhwmeisner) ==> Part of the section problem

  • I found these exemplary cases skipping through the QML searching for "section" -> section discussion!

• all introductions (de/en) and instructions (de/en) are missing OR they art part of the questionText (whereas in the real MDM they were marked as introduction) - why?

(dzhwmeisner) ==> fixed. JSON Attribute has changed -> Ute: ok, see section discussion

• [x] "keine Anleitung vorhanden"
  • https://metadatamanagement-test.cfapps.io/#!/de/questions/que-zofar221018-ins1-110.1$?search-result-index=1
  • if "keine Anleitung vorhanden" / ... please delete this attribute

(dzhwmeisner) ==> fixed

• [ ] images are ok, but introduction is missing -> how can this be, do we need to do sth. about this?
  • MDM: https://metadata.fdz.dzhw.eu/#!/de/questions/que-scs2016-ins1-8.24$?search-result-index=1
  • test: https://metadatamanagement-test.cfapps.io/#!/de/questions/que-zofar221018-ins1-96.1$?search-result-index=1
  • test: https://metadatamanagement-test.cfapps.io/#!/de/questions/que-zofar221018-ins1-104.1$?search-result-index=1

(dzhwmeisner) ==> Part of the section problem

• [x] 72.1
  • json: the actual question text (section?) is missing
  • image (de only): zofar code snippet
  • see section discussion!

(dzhwmeisner) ==> Code Conventions needed to make QML-Structure more reliable

content (images)

• [ ] content is okay but introductions are bold / not bold -> which one is true?
  • mdm: https://metadata.fdz.dzhw.eu/#!/de/questions/que-scs2016-ins1-10.1$?search-result-index=1
  • test: https://metadatamanagement-test.cfapps.io/#!/de/questions/que-zofar221018-ins1-107.1$?search-result-index=2
  • see section discussion

I found these errors absolutely randomly, please investigate further

• (almost?) all introductions (de/en) and instructions (de/en) are missing

(dzhwmeisner) ==> fixed. JSON Attribute has changed

• differences - why?
  • mdm: questionText and introduction: 10.1, 10.5
  • test: everything in questionText: page107.1, page111.1

(dzhwmeisner) ==> maybe reworked for mdm by FDZ? Fixed in current Test-Version

• introduction is missing:
  • test: https://metadatamanagement-test.cfapps.io/#!/de/questions/que-meisner1-ins1-page59.1$
  • mdm: https://metadata.fdz.dzhw.eu/#!/en/questions/que-scs2016-ins1-5.9$?search-result-index=1

(dzhwmeisner) ==> maybe reworked for mdm by FDZ? Fixed in current Test-Version

• answers are missing:
  • test: https://metadatamanagement-test.cfapps.io/#!/en/questions/que-meisner1-ins1-page30a.1$?search-result-index=1
  • mdm: https://metadata.fdz.dzhw.eu/#!/en/questions/que-scs2016-ins1-3.1a$?search-result-index=1

(dzhwmeisner) ==> fixed. -> Ute: ok. How can that happen? Do we need to do something about this?

(dzhwmeisner) ==> it was a bug in export logic. No need to do something

• also for: https://metadatamanagement-test.cfapps.io/#!/de/questions/que-meisner1-ins1-page30b.1$?search-result-index=1, https://metadatamanagement-test.cfapps.io/#!/de/questions/que-meisner1-ins1-page30c.1$?search-result-index=1

(dzhwmeisner) ==> fixed. -> Ute: ok. How can that happen? Do we need to do something about this?

(dzhwmeisner) ==> it was a bug in export logic. No need to do something

• in the real mdm, we have 128 questions (in the real mdm starting only with 1.1 ..., ), in the test we have 173, please investigate the differences further

(dzhwmeisner) ==> Soft reminders not separated anymore. Now we have 135 Questions including index,start,end,cancel -> Ute: ok

• in the new test version the soft reminders are often exported as a separate question which it should not be

(dzhwmeisner) ==> Soft reminders not separated anymore. -> Ute: ok

completeness

• [x] number of Questions: mdm: 128, test: 135 -> ok, so just FYI:

  • missing in the mdm:
  • ok: index, start, cancel, 123.1
  • not ok: 90_2.1, 90_3.1, 90_4.1 -> you delivered. I think we accidentally deleted them.

• according to the questionnaire pdf image and JSON for question number 8.19 are missing (was also the case for the real delivery).

  • Is this correct or is the pdf incorrect/not up to date?
  • should be around page 90.# Please provide feedback

(dzhwmeisner) ==> Seems to be87_bewerberauswahl -> Ute: ok, seems fine

• for questions we need to discuss in the FDZ, if we need cancel, index, startpage1, startpage2 (but for now, that's ok)

(dzhwmeisner) ==> current Test Version include these questions -> Ute: we will leave it like this!

sequence

• [x] I think , the sequence is still messed up. These are examples! It occurs often that there are more subsequent questions than in the mdm
  • mdm: 1 subsequent question: 1.1
  • test: 4 subsequent questions: 1.1
  • mdm: 1 subsequent question: https://metadata.fdz.dzhw.eu/#!/de/questions/que-scs2016-ins1-1.2$
  • test: 3 subsequent questions:https://metadatamanagement-test.cfapps.io/#!/de/questions/que-zofar221018-ins1-2.1$?search-result-index=2
  • test: https://metadatamanagement-test.cfapps.io/#!/de/questions/que-zofar221018-ins1-20.1$
  • I think the "exit question" is also not in the right places, e.g. it should be a subsequent question to 25.1, right?

    I cannot test the sequence because I don't know what to compare the question flow to.

(dzhwmeisner) ==> Nadin uploaded a document the implementation of QML based on. The MDM-Version seems to base on a different information source.

• [x] here, I think the new export (test) is correct and the real one is false
  • test: https://metadatamanagement-test.cfapps.io/#!/de/questions/que-zofar221018-ins1-105.1$
  • mdm: https://metadata.fdz.dzhw.eu/#!/de/questions/que-scs2016-ins1-9.2a$

(dzhwmeisner) ==> Nadin uploaded a document the implementation of QML based on. The MDM-Version seems to base on a different information source.

• the sequence is incorrect. When I "click through" the questions, the survey ends the first time with https://metadatamanagement-test.cfapps.io/#!/de/questions/que-meisner1-ins1-startpage.2$
  • I am wondering if startpage1 and 2 were actually on one page? (agree or disagree)
  • should the question before have two possible successors?
  • after almost every question follows: https://metadatamanagement-test.cfapps.io/#!/de/questions/que-meisner1-ins1-page4.2$?search-result-index=10

(dzhwmeisner) ==> fixed. the old Test Version based on Zofar-question perspective. not on on Zofar-page perspective -> Ute: Ok, sequence still seems odd, see above

• the following questions do not have a previous question
  • often, the following questions are different in then real mdm
  • test: https://metadatamanagement-test.cfapps.io/#!/de/questions/que-meisner1-ins1-page71.1$
  • mdm: https://metadata.fdz.dzhw.eu/#!/de/questions/que-scs2016-ins1-7.6$?search-result-index=1

(dzhwmeisner) ==> fixed. -> Ute: Ok, sequence still seems odd, see above

topic

• no topic -> you said earlier "in work" -> ok
• but we still need it

(dzhwmeisner) ==> implemented for Questions on which page an Topic is set. Not for following questions. -> Ok

• [x] if topic is "Kein Thema vor­han­den" please delete the attribute from the json

(dzhwmeisner) ==> fixed.

• [ ] no topic in json, but image says so: https://metadatamanagement-test.cfapps.io/#!/de/questions/que-zofar221018-ins1-105.1$

(dzhwmeisner) ==> already fixed manually. But part of the section problem

• [ ] https://metadatamanagement-test.cfapps.io/#!/de/questions/que-zofar221018-ins1-cancel.1$?search-result-index=135 is header, not a topic (de/en)

(dzhwmeisner) ==> Part of the Code-conventions problem

• [ ] https://metadatamanagement-test.cfapps.io/#!/de/questions/que-zofar221018-ins1-start.1$?search-result-index=2 is header, not a topic (de/en)

(dzhwmeisner) ==> Part of the Code-conventions problem

encoding

• for the english texts and images: € is not ecoded correctly, e.g. page26a.1 and page25.1

(dzhwmeisner) ==> fixed. -> Ute: ok

• what ist this? page59.1

(dzhwmeisner) ==> fixed.

images

• the quality is fine, (further ideas are an even better resolution (if possible), no white boarders -> but not within this issue!)
• plain white: de, en

(dzhwmeisner) ==> fixed

• according to the real delivery, it should be part of the previous question:
  • in MDM: https://metadata.fdz.dzhw.eu/#!/en/questions/que-scs2016-ins1-8.8$
  • in test: https://metadatamanagement-test.cfapps.io/#!/en/questions/que-meisner1-ins1-page80.1$

(dzhwmeisner) ==> fixed. the old Test Version based on Zofar-question perspective. not on on Zofar-page perspective

• it is correct the no buttons etc. are shown
• but: the images are in some way not like we had it for the real Wissenschaftlerbefragung
  • test: page19.1_1
  • mdm: https://metadata.fdz.dzhw.eu/public/files/questions/que-scs2016-ins1-1.20$/images/1.20_2.png

(dzhwmeisner) ==> fixed

-> the soft reminders etc. should be part of the image

• (then the text should be part of the question-json, right? -> we need to discuss this in the fdz)
  • I think, often the answers soft reminders are exported as separate question, but they should be part of the question as displayed to the participant (MDM example)

(dzhwmeisner) ==> fixed. the old Test Version based on Zofar-question perspective. not on on Zofar-page perspective

english version

no claim to completeness! I've been looking for "???" in the images, "zofar" in the questionText and some I detected randomly. Please investigate further

• page87_bewerberauswahl.1 (and bewerberauswahl.2 again)

(dzhwmeisner) ==> fixed

• [x] https://metadatamanagement-test.cfapps.io/public/files/questions/que-zofar221018-ins1-start.1$/images/start.1_1.png

  (as always: if this was a mistake in the live survey, we want it this way)

(dzhwmeisner) ==> Translation is missing in live survey

• page90_1.1

(dzhwmeisner) ==> fixed

• page90_2.1

(dzhwmeisner) ==> fixed

• page90_4.1

(dzhwmeisner) ==> fixed

• question text and image ("()") incorrect page53.1

(dzhwmeisner) ==> not fixed. this is an placeholder for an answer of a previous question. Questions will not be answered while screenshotting.

• little errors, for example
  • english version of startpage.2 incorrect (this question text and error appear frequently)

(dzhwmeisner) ==> fixed. the old Test Version based on Zofar-question perspective. not on on Zofar-page perspective

• page19.1 -> first category

(dzhwmeisner) ==> fixed

• encoding of €: page26a.1,

(dzhwmeisner) ==> fixed

• what ist this? page59.1

(dzhwmeisner) ==> fixed

instrument.title

• instrument.title should be "Fra­ge­bo­gen der DZHW-Wis­sen­schaft­ler­be­fra­gung 2016"/"Ques­tion­naire of the DZHW Sci­en­tists Sur­vey 2016" (de/en) instead of "Erstes Instrument"/"First Instrument"
  • -> we need to deliver the title to you, you need to ask us for that (I looked it up in the MDM)
  • we should put this in the Programmiervorlage

file structure

• import into mdm works

• [ ] could you delete the folder "json"? so, that it is

  study\MDM\questions
  study\MDM\variables
  (study\MDM\instruments)

  and not

  study\MDM\json\questions

  (dzhwmeisner) ==> fixed

file name / folder name / question.number

• two folders seem to be named differently: (see question.number)
• sorry, there seems to have been a missunderstanding: we want the format to be "26a.1", not "page26a.1"

(dzhwmeisner) ==> fixed. Pages can named individual. in the current Test-Version only the 'page' will be replaced from name. -> Ute: that's fine. We will take the name exactly as it is in zofar. SISO.

• bewerberauswahl.1 seem bewerberauswahl.1 seem odd (see file name)

(dzhwmeisner) ==> not fixed. the schema page# is not required. Pages can named individual. in the current Test-Version only the 'page' will be replaced from name. -> Ute: that's fine. We will take the name exactly as it is in zofar. SISO.

variables

• this is a different issue #93

[UteH]

@dzhwmeisner thanks! could you please also provide the new zip?

[UteH]

Info:

• refers to DAP: https://metadatamanagement-test.cfapps.io/#!/de/studies/stu-zofar091118$
• all new remarks are formatted as a checkbox

Wissenschaftlerbefragung

questions

additionalQuestionText.en

• positive: dropdown lists are already part of the additionalQuestionText! #85 (dzhwmeisner) ==> seems to be. will be checked -> Ute: ok for de and en

content (jsons)

• [ ] index

  • "Bevor es losgeht, stimmen Sie bitte den nachfolgenden Informationen zur Datennutzung zu." is missing, because it is part of the header
  • "Anleitung" is truncated (limit: 1000 MB).
  • But the text is actually the questionText (limit: 2048)

• -> code conventions!

• all introductions (de/en) and instructions (de/en) are missing OR they art part of the questionText (whereas in the real MDM they were marked as introduction) - why? (dzhwmeisner) ==> fixed. JSON Attribute has changed ok, see section discussion!

• [ ] string (section) is weird:

  • https://metadatamanagement-test.cfapps.io/#!/de/questions/que-zofar221018-ins1-108.1$?search-result-index=1
  • https://metadata.fdz.dzhw.eu/#!/de/questions/que-scs2016-ins1-10.2$?search-result-index=1
  • I think we need to discuss this, because please remember that we said: please do not put effort in manually beautifying sections
  • so is this an example of the "costs" that come with automation if we decide to not do manual beautifying?
  • Can we avoid this in the future somehow (code conventions?) we need to write this down
  • see section discussion!

• [ ] questionText, introduction and instruction are still inconsistent between mdm and test - section problem, example:

  • test: https://metadatamanagement-test.cfapps.io/#!/de/questions/que-zofar221018-ins1-105.1$?search-result-index=1
  • mdm: https://metadata.fdz.dzhw.eu/#!/de/questions/que-scs2016-ins1-9.1a$?search-result-index=1
  • see section discussion!

• [x] often: (de and en) "Kein Fra­ge­text vor­han­den" / "No ques­tion­text avail­able"

  • https://metadatamanagement-test.cfapps.io/#!/de/questions/que-zofar221018-ins1-start.1$?search-result-index=2
  • because of the concatenate of all questions of one page
  • if "Kein Fra­ge­text vor­han­den" / "No ques­tion­text avail­able" please delete the attribute

(dzhwmeisner) ==> fixed Ute: ok! (de/en)

• [x] in 1.1.json and 37(?) others: more than one number (these are the questionNumbers of the page)

  "number":[
      "1.1",
      "1.2"

  • I talked to René. As the mdm does not use the json attributes questionNumber and InstrumentNumber. (The mdm gets quesionNumber and instrumentNumber from the files and folders.)
  • Thus, we should delete both of these artefacts from the jsons.

  (dzhwmeisner) ==> fixed Ute: I need the jsons

• [ ] questionText is incomplete (section according to QML):

  • 25.1 (dzhwmeisner) ==> Part of the section problem
  • 62.1 (dzhwmeisner) ==> Part of the section problem
  • 66.1 (dzhwmeisner) ==> Part of the section problem

• see section discussion

  • I found these exemplary cases skipping through the QML searching for "section" -> section discussion!

• all introductions (de/en) and instructions (de/en) are missing OR they art part of the questionText (whereas in the real MDM they were marked as introduction) - why?

(dzhwmeisner) ==> fixed. JSON Attribute has changed -> Ute: ok, see section discussion

• [x] "keine Anleitung vorhanden"
  • https://metadatamanagement-test.cfapps.io/#!/de/questions/que-zofar221018-ins1-110.1$?search-result-index=1
  • if "keine Anleitung vorhanden" / ... please delete this attribute

(dzhwmeisner) ==> fixed -> Ute: seems correct, I need the jsons for clarification

• [ ] images are ok, but introduction is missing -> how can this be, do we need to do sth. about this?
  • MDM: https://metadata.fdz.dzhw.eu/#!/de/questions/que-scs2016-ins1-8.24$?search-result-index=1
  • test: https://metadatamanagement-test.cfapps.io/#!/de/questions/que-zofar221018-ins1-96.1$?search-result-index=1
  • test: https://metadatamanagement-test.cfapps.io/#!/de/questions/que-zofar221018-ins1-104.1$?search-result-index=1

(dzhwmeisner) ==> Part of the section problem

• [ ] 72.1
  • json: the actual question text (header?) is missing
  • image (de only): zofar code snippet -> due to manual correction
  • see header discussion! / code convention discussion!

(dzhwmeisner) ==> Code Conventions needed to make QML-Structure more reliable Ute: image (de) is still broken, was this the case for the actual survey? --> @dzhwmeisner please check!

content (images)

• [ ] content is okay but introductions are bold / not bold -> which one is true?
  • mdm: https://metadata.fdz.dzhw.eu/#!/de/questions/que-scs2016-ins1-10.1$?search-result-index=1
  • test: https://metadatamanagement-test.cfapps.io/#!/de/questions/que-zofar221018-ins1-107.1$?search-result-index=2
  • see section discussion

I found these errors absolutely randomly, please investigate further

• (almost?) all introductions (de/en) and instructions (de/en) are missing

(dzhwmeisner) ==> fixed. JSON Attribute has changed

• differences - why?
  • mdm: questionText and introduction: 10.1, 10.5
  • test: everything in questionText: page107.1, page111.1

(dzhwmeisner) ==> maybe reworked for mdm by FDZ? Fixed in current Test-Version

• introduction is missing:
  • test: https://metadatamanagement-test.cfapps.io/#!/de/questions/que-meisner1-ins1-page59.1$
  • mdm: https://metadata.fdz.dzhw.eu/#!/en/questions/que-scs2016-ins1-5.9$?search-result-index=1

(dzhwmeisner) ==> maybe reworked for mdm by FDZ? Fixed in current Test-Version

• answers are missing:
  • test: https://metadatamanagement-test.cfapps.io/#!/en/questions/que-meisner1-ins1-page30a.1$?search-result-index=1
  • mdm: https://metadata.fdz.dzhw.eu/#!/en/questions/que-scs2016-ins1-3.1a$?search-result-index=1

(dzhwmeisner) ==> fixed. -> Ute: ok. How can that happen? Do we need to do something about this?

(dzhwmeisner) ==> it was a bug in export logic. No need to do something

• also for: https://metadatamanagement-test.cfapps.io/#!/de/questions/que-meisner1-ins1-page30b.1$?search-result-index=1, https://metadatamanagement-test.cfapps.io/#!/de/questions/que-meisner1-ins1-page30c.1$?search-result-index=1

(dzhwmeisner) ==> fixed. -> Ute: ok. How can that happen? Do we need to do something about this?

(dzhwmeisner) ==> it was a bug in export logic. No need to do something

• in the real mdm, we have 128 questions (in the real mdm starting only with 1.1 ..., ), in the test we have 173, please investigate the differences further

(dzhwmeisner) ==> Soft reminders not separated anymore. Now we have 135 Questions including index,start,end,cancel -> Ute: ok

• in the new test version the soft reminders are often exported as a separate question which it should not be

(dzhwmeisner) ==> Soft reminders not separated anymore. -> Ute: ok

completeness

• [x] number of Questions: mdm: 128, test: 135 -> ok, so just FYI:

  • missing in the mdm:
  • ok: index, start, cancel, 123.1
  • not ok: 90_2.1, 90_3.1, 90_4.1 -> you delivered. I think we accidentally deleted them.

• according to the questionnaire pdf image and JSON for question number 8.19 are missing (was also the case for the real delivery).

  • Is this correct or is the pdf incorrect/not up to date?
  • should be around page 90.# Please provide feedback

(dzhwmeisner) ==> Seems to be87_bewerberauswahl -> Ute: ok, seems fine

• for questions we need to discuss in the FDZ, if we need cancel, index, startpage1, startpage2 (but for now, that's ok)

(dzhwmeisner) ==> current Test Version include these questions -> Ute: we will leave it like this!

sequence

• [x] I think , the sequence is still messed up. These are examples! It occurs often that there are more subsequent questions than in the mdm
  • mdm: 1 subsequent question: 1.1
  • test: 4 subsequent questions: 1.1
  • mdm: 1 subsequent question: https://metadata.fdz.dzhw.eu/#!/de/questions/que-scs2016-ins1-1.2$
  • test: 3 subsequent questions:https://metadatamanagement-test.cfapps.io/#!/de/questions/que-zofar221018-ins1-2.1$?search-result-index=2
  • test: https://metadatamanagement-test.cfapps.io/#!/de/questions/que-zofar221018-ins1-20.1$
  • I think the "exit question" is also not in the right places, e.g. it should be a subsequent question to 25.1, right? Ute: I cannot test the sequence because I don't know what to compare the question flow to.

(dzhwmeisner) ==> Nadin uploaded a document the implementation of QML based on. The MDM-Version seems to base on a different information source.

-> Ute: I compared the new export randomly with the qml -> seems correct!

• [x] here, I think the new export (test) is correct and the real one is false
  • test: https://metadatamanagement-test.cfapps.io/#!/de/questions/que-zofar221018-ins1-105.1$
  • mdm: https://metadata.fdz.dzhw.eu/#!/de/questions/que-scs2016-ins1-9.2a$

(dzhwmeisner) ==> Nadin uploaded a document the implementation of QML based on. The MDM-Version seems to base on a different information source.

• the sequence is incorrect. When I "click through" the questions, the survey ends the first time with https://metadatamanagement-test.cfapps.io/#!/de/questions/que-meisner1-ins1-startpage.2$
  • I am wondering if startpage1 and 2 were actually on one page? (agree or disagree)
  • should the question before have two possible successors?
  • after almost every question follows: https://metadatamanagement-test.cfapps.io/#!/de/questions/que-meisner1-ins1-page4.2$?search-result-index=10

(dzhwmeisner) ==> fixed. the old Test Version based on Zofar-question perspective. not on on Zofar-page perspective -> Ute: Ok, sequence still seems odd, see above

• the following questions do not have a previous question
  • often, the following questions are different in then real mdm
  • test: https://metadatamanagement-test.cfapps.io/#!/de/questions/que-meisner1-ins1-page71.1$
  • mdm: https://metadata.fdz.dzhw.eu/#!/de/questions/que-scs2016-ins1-7.6$?search-result-index=1

(dzhwmeisner) ==> fixed. -> Ute: Ok, sequence still seems odd, see above

topic

• no topic -> you said earlier "in work" -> ok
• but we still need it

(dzhwmeisner) ==> implemented for Questions on which page an Topic is set. Not for following questions. -> Ok

• [x] if topic is "Kein Thema vor­han­den" please delete the attribute from the json

(dzhwmeisner) ==> fixed. -> seems correct, Ute check jsons randomly

• [ ] no topic in json, but image says so: https://metadatamanagement-test.cfapps.io/#!/de/questions/que-zofar221018-ins1-105.1$

(dzhwmeisner) ==> already fixed manually. But part of the section problem

• [ ] https://metadatamanagement-test.cfapps.io/#!/de/questions/que-zofar221018-ins1-cancel.1$?search-result-index=135 is header, not a topic (de/en)

(dzhwmeisner) ==> Part of the Code-conventions problem

• [ ] https://metadatamanagement-test.cfapps.io/#!/de/questions/que-zofar221018-ins1-start.1$?search-result-index=2 is header, not a topic (de/en)

(dzhwmeisner) ==> Part of the Code-conventions problem

encoding

• for the english texts and images: € is not ecoded correctly, e.g. page26a.1 and page25.1

(dzhwmeisner) ==> fixed. -> Ute: ok

• what ist this? page59.1

(dzhwmeisner) ==> fixed. -> Ute: ok (is this version how it was in the actual survey?)

images

• the quality is fine, (further ideas are an even better resolution (if possible), no white boarders -> but not within this issue!)
• plain white: de, en

(dzhwmeisner) ==> fixed

• according to the real delivery, it should be part of the previous question:
  • in MDM: https://metadata.fdz.dzhw.eu/#!/en/questions/que-scs2016-ins1-8.8$
  • in test: https://metadatamanagement-test.cfapps.io/#!/en/questions/que-meisner1-ins1-page80.1$

(dzhwmeisner) ==> fixed. the old Test Version based on Zofar-question perspective. not on on Zofar-page perspective

• it is correct the no buttons etc. are shown
• but: the images are in some way not like we had it for the real Wissenschaftlerbefragung
  • test: page19.1_1
  • mdm: https://metadata.fdz.dzhw.eu/public/files/questions/que-scs2016-ins1-1.20$/images/1.20_2.png

(dzhwmeisner) ==> fixed

-> the soft reminders etc. should be part of the image

• (then the text should be part of the question-json, right? -> we need to discuss this in the fdz)
  • I think, often the answers soft reminders are exported as separate question, but they should be part of the question as displayed to the participant (MDM example)

(dzhwmeisner) ==> fixed. the old Test Version based on Zofar-question perspective. not on on Zofar-page perspective

english version

no claim to completeness! I've been looking for "???" in the images, "zofar" in the questionText and some I detected randomly. Please investigate further

• page87_bewerberauswahl.1 (and bewerberauswahl.2 again)

(dzhwmeisner) ==> fixed

• [x] https://metadatamanagement-test.cfapps.io/public/files/questions/que-zofar221018-ins1-start.1$/images/start.1_1.png

  (as always: if this was a mistake in the live survey, we want it this way)

(dzhwmeisner) ==> Translation is missing in live survey -> Ute: ok

• page90_1.1

(dzhwmeisner) ==> fixed

• page90_2.1

(dzhwmeisner) ==> fixed

• page90_4.1

(dzhwmeisner) ==> fixed

• question text and image ("()") incorrect page53.1

(dzhwmeisner) ==> not fixed. this is an placeholder for an answer of a previous question. Questions will not be answered while screenshotting.

• little errors, for example
  • english version of startpage.2 incorrect (this question text and error appear frequently)

(dzhwmeisner) ==> fixed. the old Test Version based on Zofar-question perspective. not on on Zofar-page perspective

• page19.1 -> first category

(dzhwmeisner) ==> fixed

• encoding of €: page26a.1,

(dzhwmeisner) ==> fixed

• what ist this? page59.1

(dzhwmeisner) ==> fixed

instrument.title

• instrument.title should be "Fra­ge­bo­gen der DZHW-Wis­sen­schaft­ler­be­fra­gung 2016"/"Ques­tion­naire of the DZHW Sci­en­tists Sur­vey 2016" (de/en) instead of "Erstes Instrument"/"First Instrument"
  • -> we need to deliver the title to you, you need to ask us for that (I looked it up in the MDM)
  • we should put this in the Programmiervorlage

file structure

• import into mdm works

• [x] could you delete the folder "json"? so, that it is

  study\MDM\questions
  study\MDM\variables
  (study\MDM\instruments)

  and not

  study\MDM\json\questions

  (dzhwmeisner) ==> fixed

file name / folder name / question.number

• two folders seem to be named differently: (see question.number)
• sorry, there seems to have been a missunderstanding: we want the format to be "26a.1", not "page26a.1"

(dzhwmeisner) ==> fixed. Pages can named individual. in the current Test-Version only the 'page' will be replaced from name. -> Ute: that's fine. We will take the name exactly as it is in zofar. SISO.

• bewerberauswahl.1 seem bewerberauswahl.1 seem odd (see file name)

(dzhwmeisner) ==> not fixed. the schema page# is not required. Pages can named individual. in the current Test-Version only the 'page' will be replaced from name. -> Ute: that's fine. We will take the name exactly as it is in zofar. SISO.

variables

• this is a different issue #93

[dzhwmeisner]

mdm_ifq-mdm-mod_Update151118.zip

[UteH]

please check question 72.1 @dzhwmeisner