Jump to content

cathorn

Members
  • Content count

    24
  • Joined

  • Last visited

Community Reputation

0 Neutral

About cathorn

  • Rank
    Member
  • Birthday April 7

Contact Methods

  • MSN
    cathorn@hotmail.com
  • Website URL
    http://www.carnegiefoundation.org/who-we-are/staff-directory/chris-thorn/
  • Skype
    chris.thorn

Profile Information

  • Gender
    Male
  • Location
    Stanford, CA USA
  1. is pounding head against new evaluation report template

  2. I have to admit that I don't follow the forum much any more. I have a rule set up to notify me of changes to this thread. I will be interested to see if this feature is added back in Version 8. After 15 years of using QSR products, I pretty much have had to abandon NVivo 7 for my own work. Aside for an inability actually see coding in context - arguably the primary analytical purpose of a qualitative software package, new features in V8 around quantitative notions of intercoder reliability are not particularly relevant to the sort of interdisciplinary work I do either. I would not expect high degree agreement when I have an urban planner and a elementary reading specialist code a set of documents. That's actually the point. I want a variety of insights. I don't do as much training as I used to, but I do a fair amount of training of faculty, scientists, and graduate students here at the University of Wisconsin. While I use NVivo to teach the basics of qualitative analysis with software, that really based on my familiarity with the tool. I can no longer recommend it. I spend a substantial amount of my training time with the CAQDAS framework for choosing a qualitative analysis package. I think link describe a series of analytical practices and discuss how each of the major packages support those practices. We are looking at whether or not to upgrade to the new version. I guess we will treat it like we treat Microsoft software. We never, ever upgrade before service pack 1. With most of the office productivity tools we actually skip generations because the improvements are not worth suffering the pain of upgrading. I'll be interested to see what the sense of the forum is after the new version ships tomorrow. Chris A former Nud*ist and NVivo enthusiast
  3. cathorn

    Macintosh, anyone?

    It seems that Apple has come up with an end run on the development environment problem I described above. It's a wonderful deal for Microsoft, since they don't have to do a thing to sell more Windows OS licenses. Unfortunately for OS X folks, it also reduces the pressure on software firms to deliver OS X native applications. Chris
  4. ziniel, There is no way to merge two NVivo 2 projects without the Merge application . If your team is co-located, the simple way is to share the same project and not have it open at the same time. I know several small teams the manage this quite well. The students keep a shedule of their classes in the office they share and coordinate there work hours to minimize overlap. This works on a shared network across several building on one case. The key is human coordination to make sure folks are not trying to work at the same time. Chris
  5. Ok. That is wildly cool. Great job zfirst. Thanks to Daniel for the SQL install info. Chris
  6. I thought this thread was about the ability to view coding in one's project. Coding reports and coding stripes are both issues we need to have addressed. They are also different issues. Could you address each of them individually? How do we see all of the codes coding any text in our documents and what they code - they actual text. I need to be able to view and assess my coding. I need to be able to walk away from the software and work with other people. Chris
  7. It would be a great help to have a stand-alone version of the volunteering project. Any training changes the file and makes it tough to do follow on training. Could that be put on the training resources page? Chris There are 6 trainers in the room right now who would also appreciate this!
  8. PS to my post above: After working though this with most of the trainers who are attending the sessions I am hosting in Wisconsin, the best we could come up with to get at something close to the old coding detail report was a group find against a set of documents (or a single document) that shows the coding for a set of nodes (or single nodes). I could then do a series of node reports to get the content I need. There is a coding summary report, but it only reports character numbers. What I would really like would be the ability to do the group find to show a subset of coding for a subset of sources and print that all out. Chris
  9. Ooh! Ooh! Good News! I was actually able to encourage a colleague to actually attempt the install of NVivo 7 on a terminal server. Adam & Daniel - I don't have copious details yet, but here is a short version. Windows Terminal Server 2003 64-bit Results: Installation Invalid - "A required registry key for this product was not found." It's not clear to me if the error refers to the application itself or one of the application requirements (.NET 2.0 framework or SQL 2005). 32-bit version of Windows Server 2003 with Citrix Metaframe. Score! We had it running with three open connections and there were no serious resource issues. I tried the Volunteering tutorial and was able to create all of the objects I tried (all of the top level stuff). I also imported a large NVivo 2 project successfully. Two documents did not come in. One had a busted databite in it already and the other had what looked like an indent. Otherwise everything came in intact. I also tried to create new sources, query, model, relationship, and did finds. Everything seemed to work. Not what I would call software testing, but promising enough that we are going to deploy it for a handful of users for a little more testing. Do you have a test plan I could have few student follow to make sure the major features all work? Chris
  10. I'll add my thanks Leonie. Mary....I was not a beta tester, but I did get a chance to work with a pre-release version at a training session. I have to admit that while I did notice and agree that the lack of the ability to print out coding stripes was a problem, I feel completely stupid that I did not notice the inability get more detailed coding reports from documents. I am in the process of shifting an evaluation team from NVivo 2 to 7 in order to take advantage of the embedded graphics and improved query tools in NVivo 7. That is likely to come to a screeching halt on Monday, if I cannot think of a new way to communicate with my team and funders without coding reports. They have always been an important part of my work. I guess the lack of a printer at the training I did contributed to the issue not coming up. Tom and Lyn Richards are actually in town right now teaching at a training I'm hosting. I'll go ask if I am just missing things. There may be a better way to communicate with folks not using the software. I sure hope so. Chris
  11. And another thing..... Take a hard look at the check box for code documents at cases. NVivo 7 handles attributes at case nodes only. This box can be a big help in bringing docs into the case area. If you have many to one relalationships between documents and cases, one can quickly drag and drop case nodes into a tree that makes sense. - that may be a confusing statement, but do take the time to explore this before going to0 far. It is very, very cool Chris
  12. Hi I'm working with a group who get some of my site licenses as part of cooperative agreement on campus. They use a terminal server to provide access to lots of different reseach software to folks all over - no matter what their OS. Serving up any number of stats packages has never been a problem - NVivo 2 and N6 work fine. Their system administrator took one look at the NVivo 7 specs and said, mmmm.....cough......cough.....SQL2005.....NET 2.0......uuuh.....that sound like a system crash waiting to happen. We are not going to trying in on a production system. Has anyone out there tried this. This has been our solution for Mac and Linux users. It means that Atlas 5.0 or NVivo 2 will be the more likely solution in that environment. Chris
  13. Daniel, Thanks very much for this posting. That helps a great deal. Questions about 1 and 4 as follows: 1 - Can the software be run without administrator rights after it has been installed using admin rights? 4 - We are required to use encrypted file systems for some research projects in health, mental health, research on school children, etc. Does that mean that I can have to install the application in an unencrypted location, but can put the data in an encrypted file space? Or are encrypted filesystems not supported at all? Thanks, Chris
  14. cathorn

    Macintosh, anyone?

    Jens et al., The development environment is what rules here. We've been working on cross platfrom tools for years at my place and it is a thankless and wildly difficult task under the best of circumstances. I've had the opportunity to talk with QSR folks at length about the problems of supporting both platforms. There are very few development tools (program environments, windowing systems, media controls, etc.) that work on both platforms - or even more annoying - almost work on both platforms. The choice to move to .NET as a development environment was likely driven by a whole host of features, but the most important were probably: - The ability to hire and/or train talented folks in these tools - The availability of a well tested user interface model from Microsoft (with tools to pull it off) - The object oriented structure available in this environment that allows the development team to bring in new features, fix bugs, etc. with much less labor than other options - Microsoft treats its software development partners pretty well - with early access to new technologies, lots of professional development opportunities, etc. Ok. Here's the big gotcha - no .NET for the Mac. Virtual PC or terminal services (using Citrix or the RDP client) must look very attractive to QSR as an alternative to developing the whole package over again in a completely different technology. My small research group, on the other hand (www.transana.org) went open source with Python as our primary environment and WxPython as the windowing environment. We use open source, but we have 5-6 components that are not at the same level of development and often have ugly dependencies. Normally the Mac version lags the Windows and Linux development effort by 6-12 months - or even longer with multimedia tools. We can give it all away, but we are beholden to other open source teams to deliver many of the features we need. I have no problem doing the math to see that NVivo on the Mac is a no-win solution for QSR. It would cost so much to produce that no one could afford to buy it. Just my $0.02 (USD) at 23:00 CST (GMT -6) Chris
  15. If you have the time, you can come watch Tom Richards spill the beans in person at the TQM 2005 Working Conference in Madison, Wisconsin, USA. Tom will give a plenary on the new features of NVivo 7 and then be around to pester for the next day and a half. Check out the agenda at: http://www.wcer.wisc.edu/tqm/events.html Chris
×