Jump to content

cchen

QSR Staff
  • Content count

    194
  • Joined

  • Last visited

  • Days Won

    1

Posts posted by cchen


  1. Thanks, Claire, for your help with the case problem. It worked perfectly.

     

    Regarding sources coded vs coding references. In this case it is a lit review so - apart from a single case where I had to use two sources to form a case - all other publications have been coded as one case each (as a whole as I am looking here at case attributes). But as there is one more source than there are references/cases here I would have expected to see a difference here.

     

    We are glad it worked for you. Regarding sources coded vs. coding references, if 2 sources have been coded at the same case, then sources coded and coding references should still show the same number, since each of those 2 sources is actually coded at the case in one coding reference each, and for those 2 sources together, they are coded at the case in 2 separate coding references. Does it make sense to you?

     

    Regards,

    Claire Chen

    QSR International


  2. Hi,

    I have imported a project from NVivo 8 where I had combined a range of sources into cases. I now want to establish the number of CASES of e.g. reference type & year of publication (analysis of lit review) with a matrix query. The query is no problem. But I don't know how to get the (number of) cases shown in the cells of the resulting matrix. The two options that are the most likely 'Sources coded' and 'Coding references' both don't work (they actually show the same in my case). How can I do this?

    When you open an NVivo 8 project in NVivo 9, a node classification named "Case" will be created and all NVivo 8 case nodes will be assigned to this node classification. If we want to get the number of cases shown in the matrix cells of the result matrix, what you can do is to select Matrix Cell Content option “Nodes Coded”, and then select “Select Classification…” and select node classification “Case”.

     

    The two options that are the most likely 'Sources coded' and 'Coding references' both don't work (they actually show the same in my case).

    'Sources coded' and 'Coding references' might show same numbers, if each source has been coded at a node in one coding reference only. If that is not your case, please provide more detail about your project and matrix and we'll investigate further.

     

    Regards,

    Claire Chen

    QSR International


  3. Have just looked into this issue again. It seems that there are in fact two 'SB' users one 'SB' and one 'S B' - this has happened inadvertently, I don't know how it happened (the SB one looks like a default setting because it is using the name 'Compaq_Owner' which is not something I would have called myself!) and I didn't want 2 users in the project. How can I merge them? And is this the reason for the apparent anomaly I've noticed when coding?

     

    Many thanks.

     

    Hi Susan,

     

    This might be the reason for the inconsistent results on 2 computers that you've noticed. When NVivo launches for the first time, it will ask you to enter user name and initials. However, there will be a default user name displayed, which is your Windows user name on that computer; if you don't choose to change it, it will be your NVivo user name. If you then open a project and do coding etc. in that project, this NVivo user will be recorded as one of the project users in that project. I guess 'Compag_Owner' might be the Windows user name which you have used on one of those computers. To fix your NVivo user name on that computer, please select from main menu Tools -> Application Options, and change User Name on that dialog. However, this will not fix any existing coding in your project.

     

    You cannot merge users in the current version of NVivo 8. However, it might be not that hard to fix the coding manually, if majority of coding is under one of the users. To find out the details about different users' coding, you might find some tools useful, such as Coding Comparison Query, Coding Query, User Coding Stripes, etc. Once owner is identified for each piece of coding, you can choose to uncode for selected user and re-code after your user name on Application Options is fixed.

     

    By the way, in NVivo, users are identified by user name, instead of initials. So if 2 users have 2 different user names, then they will be considered as 2 different users, even if they have same initials.

     

    Does it make sense to you?

     

    Regards,

    Claire Chen

    QSR International


  4. Hi Susan,

     

    Your problem sounds very interesting. Is your project a single user project or multi user project? If you have been using the same project on XP and Vista, do you have same user name for NVivo 8 on those 2 machines? You can open up Application Options to find out the current user name, and turn on User Coding Stripes to find out who has coded this document. Please be aware that overlapping coding by 2 different users will be counted as 2 references in NVivo, while overlapping coding by 1 user (at the same node) will be merged and counted as 1 reference.

     

    However, even if you had multiple users' coding in that document, it still doesn't explain why the number of nodes are different after coding the same selection to the same nodes, if the original documents have exactly the same coding. Susan, may I know if you have used the same copy of project to run this test on XP and Vista? If you have used 2 different copies, then any slight difference in coding or anything else might make a difference.

     

    Regards,

    Claire Chen

    QSR International

×