ASA Connect

 View Only

  • 1.  Computational tool for med students

    Posted 09-08-2017 09:15

    I am working with med students. These are mostly pretty clever folks, but not many are computationally sophisticated. For others who work with med students, can you recommend an inexpensive tool for statistical support? Such a tool would not need to do much data management – that's a sphere of statistical effort I would not expect to see done by med students. The tool should be inexpensive. And it cannot be GraphPad Prism. I have problems with that tool.

     

    Paul A. Thompson

    -----------------------------------------------------------------------
    Confidentiality Notice: This e-mail message, including any attachments,
    is for the sole use of the intended recipient(s) and may contain
    privileged and confidential information.  Any unauthorized review, use,
    disclosure or distribution is prohibited.  If you are not the intended
    recipient, please contact the sender by reply e-mail and destroy
    all copies of the original message.



  • 2.  RE: Computational tool for med students

    Posted 09-11-2017 09:03
    My experience with medical residents and fellows suggests starting with t- and chi-squared tests. There are plenty of web sites that will do these  -- find one you like. 

    Again in my experience, the residents/fellows were seldom required to perform the analysis but always had to understand the report. E.g., not required to perform a logistic regression but had to correctly write about odds ratios, confidence intervals, and p-values. So I omitted all discussion of how to perform a logistic regression.



    Original Message


  • 3.  RE: Computational tool for med students

    Posted 09-11-2017 10:46

    Try Epi Info 7 from the CDC.  It's free and pretty well built.

    I don't see most med students needing beyond that and it has the typical things they would need to see

    in the field (Odds, Risk, Basic tests, Sensitivity, Specificity, etc...)  It can handle maps and spacial density

    as well.  If you want them to do complex modeling it's probably the wrong tool, but for the basics it's great.

    Beyond that, students can build instruments, record data, and analyze all in one package without much effort.

    The dashboard is pretty easy to learn and use, plus you can easily filter and subset data for analysis without any code.

    There's also good reporting/report building methods built in.



    ------------------------------
    Joseph Reid
    Associate Prof. of Applied Mathematics and Statistics
    Oregon Institute of Technology
    ------------------------------

    Original Message


  • 4.  RE: Computational tool for med students

    Posted 09-14-2017 19:34
    The NIST has a free statistical program called data plot.  Also has a good e-handbook that covers a lot of material with examples.  

    www.itl.nist.gov/div898/handbook/data plot.htlm

    ------------------------------
    Steven Correia
    Environmental analyst (retired)
    Massachusetts Division of Marine Fisheries
    ------------------------------

    Original Message


  • 5.  RE: Computational tool for med students

    Posted 09-11-2017 13:13

    MedCalc might be a good option.  It has a perpetual license and has a large number of functions.  I think your medical students would like it because it has very simple functions for sensitivity, specificity, etc.  It also has some nice graphing features.

     

    https://www.medcalc.org/


    Mike

     

    Mike Malek-Ahmadi, PhD, GStat

    Research Bioinformatics Analyst

    Banner Alzheimer's Institute

    901 E. Willetta St., Phoenix, AZ 85006

    Phone – 602-839-2806

     




    Original Message


  • 6.  RE: Computational tool for med students

    Posted 09-11-2017 16:00
    JASP might be of interest (https://jasp-stats.org/). It's free, easy to use, and has Bayesian methods implemented.

    ------------------------------
    Stanley E. Lazic, PhD

    Team Leader, Statistics
    AstraZeneca
    Cambridge, UK

    https://stanlazic.github.io/
    ------------------------------

    Original Message


  • 7.  RE: Computational tool for med students

    Posted 09-13-2017 07:40
    ​Thanks for the introduction to JASP Stanley.  It is a really interesting tool especially by introducing the Bayes thinking into common tests.  Very academically oriented too.

    Janet

    ------------------------------
    Janet McDougall
    President
    McDougall Scientific Ltd.
    ------------------------------

    Original Message


  • 8.  RE: Computational tool for med students

    Posted 09-12-2017 09:18
    I use R with the Rstudio interface and mosaic package with my MPH students, many of whom are med students.  They can copy commands and change the relevant parts, which is how many people learn statistical software initially anyhow.

    ------------------------------
    Janet Rosenbaum
    ------------------------------

    Original Message


  • 9.  RE: Computational tool for med students

    Posted 09-13-2017 10:28

    The med school I work at has a site license for SPSS. Because of that, we individual users can get SPSS on our computers for just a small installation fee paid to our IT department. If the med school you work with has a site license for SPSS, then I would recommend SPSS to your med students.

     

    Then there is NCSS. I haven't used NCSS in almost 10 years, but my memory is that it was an excellent software for the price, and was fairly intuitive for people just getting started. Its data tables had a Data View and a Variable View, much like SPSS. I had a couple of biology collaborators whose labs used NCSS in preference to GraphPad.

     

    But having said all that, I also have to say, Don't Overlook Excel. I spend some of my time working with young surgeons in a residency program. They, too, are not computationally sophisticated, but Oh! My! Yes, they are clever. This month, I saw the Excel spreadsheets of three of them. Two of them had figured out on their own, and figured out correctly, how to use the Excel functions for doing t-tests and chi-square tests. In addition, one of them had taught himself how to use Excel's IF function in a manner that was ghastly to look at, but ingenious when you studied it. I found myself thinking, Man, If Only they'd had a little exposure to concepts of how to organize their data, they would be awesome. And I guess that's my point: your med students already have Excel. All they need is a little guidance how to use it effectively.

     

    I know that many people criticize Excel for its supposed lack of accuracy, but I have compared Excel results to SAS results on-and-off for 17 years, and I have yet to find an instance where the two are in disagreement. Excel merely requires close attention to detail, but if they're med students, then the chances are good they have that habit already.  

    Confidentiality Notice: This e-mail message, including any attachments, is for the sole use of the intended recipient(s) and may contain confidential and privileged information. Any unauthorized review, use, disclosure or distribution is prohibited. If you are not the intended recipient, please contact the sender by reply e-mail and destroy all copies of the original message.



    Original Message


  • 10.  RE: Computational tool for med students

    Posted 09-13-2017 12:36
    I would use any package that has a good, user-friendly interface.  Although I recognize the possibilities and flexibility that R and Python have (both have also Bayesian extensions - see for instance https://stats.stackexchange.com/questions/1595/python-as-a-statistics-workbench#1632), using the command line adds a level of complexity for the students.  Interfaces for R such as RStudio still do not have the user friendly feel that applications such as Minitab, JMP, and SPSS have.

    ------------------------------
    Filiep Samyn
    ------------------------------

    Original Message


Related Content