Jump to content


  • Content count

  • Joined

  • Last visited

  • Days Won


cyjacobs last won the day on January 15 2015

cyjacobs had the most liked content!

Community Reputation

10 Good

About cyjacobs

  • Rank
    Advanced Member

Profile Information

  • Gender
  • Location

Recent Profile Visitors

7,493 profile views
  1. Ernani, An example I often use in workshops is to compare a cluster analysis of a set of interviews by word similarity to one by coding similiarity. Comparing the results, some items will shift (a few of those that had been very similar in word content, may be quite different in coding). In an analysis, this would raise the question - why has this occurred? Double clicking an item in the cluster display will cause it to open with coding stripes displayed for 'Nodes Most Coding.' Opening two items that have discrepant 'similarities' will then allow you to examine the text content and your coding to look for an explanation. Did my coding vary? What is it that's making these two interviews so similar in content but so different in coding? Hope that helps as a way to begin exploring.
  2. Danny, One approach you might try is using a Text Search Query. While the default is to include just your text (and to exclude your Annotations), you can choose to include just your Annotations. Then, you could restrict the query to a few sources at a time, in logical groupings, to present your material to your supervisor by source. If it's important, though, for your supervisor to see the text to which you've attached the annotation (which I would imagine it is!), your best bet really may be to print or export each source, including the annotations (which will then, as you say, appear as end notes). The annotations will also appear (if you choose to include them) if you print or export a node, which would be another approach - to present the material to your supervisor by theme/issue. Hope this is helpful. Cynthia Jacobs, EdD Training and Research Consultancy Regional Manager, Americas QSR International (Americas) Inc. 90 Sherman Street | Cambridge, MA 02140 USA T +1 617 491 1850 F +1 (617) 812 7783 Cell +1 (978) 884 5081 c.jacobs@qsrinternational.com | www.qsrinternational.com
  3. Hello, NumNumNum. Sounds like an interesting project, and here are a couple of thoughts. Rather than nesting nodes, I think I'd create a node for each "actor," with attributes sector (private/public), role (senate/congress/NGO/regulator), committee (Not Applicable for many roles), and so forth. At each of these nodes, I would code all of the testimony from that actor over time. If you're working in NVivo 8 or earlier, these would be Cases. If you're in NVivo 9, you might create a Node folder (called something like "Actors") to contain these nodes. (In NVivo 9, you would then classify these nodes in order to be able to then apply attributes). You mention that you are interested in analyzing over time - are there distinct periods? If so, you could create a node for each of these as well, containing all of the testimony from that period, or create a set for each period. In any case, you are absolutely on the right track in your thinking here. If you think you'd like a bit of consulting as you get started with this, let us know. You can reach me directly at c.jacobs@qsrinternational.com. I'm sure that other folks listening in on the forum will have some other good ideas about project structure. There is never one right way in NVivo... Best, Cindi Jacobs
  4. Devo, I would recommend creating a node in the Cases folder for each student. Then, code all of the sources for each student (each student's drafts and final submission) to their Case node. (You can do this by selecting all sources for a given student in List View, then right click > Code Sources > At Existing Node, then navigate to the Case node for that student). Then, in Classifications, create an attribute for Achievement Level (with values low, medium and high) and Language Coach Training (with values trained, untrained). With this structure, you will then be able to compare coding at several nodes by attribute value using Matrix Coding queries. Hope that helps. Cindi
  5. Tamara, When viewing the number of references in the Sources List View, this is a count of the number of selections within that source that have been coded to any node. The same selection coded to two different nodes would be counted as two references. CJ
  6. Tamara, Memo Links are one-to-one, whole-to-whole links. See Also Links can be created from selected content in any number of sources to any number of memos, so this may help you. The See Also Links are not visible in List View for the memos, but can be viewed from Links. Another approach might to be to code a portion of or the entire memo to the case, depending on the nature of the relationship between the case and the memo. Hope that helps. CJ
  7. LG, Yes. The functionality is identical. The difference is all in the license terms. CJ
  8. cyjacobs

    Sample Project Crashes N9

    Stuart, A few others have had this problem and this approach seems to work: Working with a clean copy of the sample project (which is automatically placed in a folder called NVivo 9 Samples in your Public Documents folder when you install NVivo 9). Then, open NVivo by double clicking that file, rather than opening the sample project from your Welcome screen. Cindi
  9. Crispy, NVivo will not 'allow' you to save if there have been no changes to the project since your last save. If the Save Project option is grayed out on your File menu, this is most likely what's happening. Try adding some text to a memo - the save option should reappear. Hope that helps. Cindi Jacobs