Search the Legacy Support Archive (up to v2020)
  1. Read-only archive (2005–2024). Posting is closed. Need help with MediaLab or DirectRT 2027? Email service@empirisoft.com.

    All words · “exact phrase” · -leave out

    Advanced search

seemingly random writing of data

  1. #1

    seemingly random writing of data

    My group has created a few experiments that link multiple MediaLab questionnaires, and I double-checked to make sure all the data was writing in Excel... all of the data is accounted for, but I've found that ML seems to be writing it in a random order (not the order in which the questionnaires are presented). Ex: I have questionnaires A-B-C linked and the data will write in the order C-B-A. While all of the data is there, we're just worried that something else might be going on to throw it out of the presented order(?).

    Thanks for your consideration!

  2. #2
    jarvis24 Blair Jarvis, Administrator
    Which data folder are you looking at?

  3. #3
    We've been looking at the data in the 'by questionnaire' folder, as apparently this problem came up in a previous study and it was associated w/ viewing data in the 'by variablename' folder. When opening in the VN folder it looks like the data is just writing in alphabetical order of the questionnaires; this is not the case in the QN folder, where the order is neither alphabetical nor the order in which the questionnaires are presented. Interestingly, the first- and shortest- questionnaire presented is a vitals check (we enter blood pressure, heart rate, etc) and that always is written last in the data string...

  4. #4
    jarvis24 Blair Jarvis, Administrator
    Can you zip and post your data folder? I'll take a look at it.

  5. #5
    Here's the master data file; the problem I described earlier is seen in 'base1a', when the presented order of the questionnaires is vitals (sys_bp, co, etc), FTND, sjp, ces, sens, and rfc.

    thanks-

  6. #6
    jarvis24 Blair Jarvis, Administrator
    Just to be clear, is this file an example?

    data/byquestionnaire/base1a4.csv

    If so, I believe it checks out OK. In the byQuestionnaire folder, the data are always written with the questionnaires sorted alphabetically by their names. Note the critical difference here being the names of the questionnaires being sorted verses the names of variables being sorted (as in the case of the byVariableName folder).

    Does that make sense?

    See also: http://www.empirisoft.com/medialab/help/data.htm