Advertisement
If you have a new account but are having problems posting or verifying your account, please email us on hello@boards.ie for help. Thanks :)
Hello all! Please ensure that you are posting a new thread or question in the appropriate forum. The Feedback forum is overwhelmed with questions that are having to be moved elsewhere. If you need help to verify your account contact hello@boards.ie

Help with T-test

Options
  • 28-10-2008 10:17pm
    #1
    Macha
    Moderators, Science, Health & Environment Moderators Posts: 6,376 Mod ✭✭✭✭


    Hi guys,

    I've 3 weeks to hand in my masters dissertation (9000 words down, 9000 to go..) and after doing a quantitative questionnaire, I've been told by my tutor that I have to do a t-test analysis on my results. He sent me 2 websites and that's pretty much the sum of his help on the topic.

    http://www.socialresearchmethods.net/kb/stat_t.php
    http://changingminds.org/explanations/research/analysis/t-test_table.htm

    I've had a look at these sites but am still a bit confused. Would any of you know a clear book/web resource that explains how to take data from an excel sheet and apply a t-test? Do I first have to get the standard deviation using the excel forumula?

    Thanks in advance

    Taco


Comments

  • Options
    #2
    2Scoops
    Registered Users Posts: 1,845 ✭✭✭2Scoops


    Excel can calculate the t-statistic for you if you go through the Tools menu to 'Data Analysis' (if that option isn't there, go to 'Add-Ins' to insert it).

    The nature of your data will determine whether you should use a paired or unpaired test. Of course, that's just assuming your supervisor is correct in recommending the t-test to begin with. :pac:


  • Options
    #3
    cruiserweight
    Registered Users Posts: 5,307 ✭✭✭cruiserweight


    If you are only testing the statistical significance of your results now I would guess that it is only an add on for your thesis. So as the previous user said Excel has built in functionalty to do what you want to do. But if you want something that is a little better I would suggest using R.


  • Options
    #4
    Macha
    Moderators, Science, Health & Environment Moderators Posts: 6,376 Mod ✭✭✭✭Macha


    2Scoops wrote: »
    Excel can calculate the t-statistic for you if you go through the Tools menu to 'Data Analysis' (if that option isn't there, go to 'Add-Ins' to insert it).

    The nature of your data will determine whether you should use a paired or unpaired test. Of course, that's just assuming your supervisor is correct in recommending the t-test to begin with. :pac:
    Hmm, well he looked over it and that's what he recommended *gulp*
    cruiserweight wrote: »
    If you are only testing the statistical significance of your results now I would guess that it is only an add on for your thesis. So as the previous user said Excel has built in functionalty to do what you want to do. But if you want something that is a little better I would suggest using R.
    Well, unfortunately not...my tutor went AWOL during the summer so I lost 2 months. Plus he's taken a week to get back to me about this..I'm kinda pissed off about it, also because none of this was covered in my course and the implications of doing a quantitative survey were only discussed after I had finished it.

    What is R?


  • Options
    #5
    Red Alert
    Hosted Moderators Posts: 7,484 ✭✭✭Red Alert


    R is technically a programming language, but it's interactive and optimized for statistics. There's plenty of examples and best of all it's free and runs on all computer platforms.


  • Options
    #6
    2Scoops
    Registered Users Posts: 1,845 ✭✭✭2Scoops


    R has a pretty steep learning curve for a beginner and since your analysis looks like it's going to be fairly basic, I suggest you don't worry about R for now. :)


  • Advertisement
  • Options
    #7
    Macha
    Moderators, Science, Health & Environment Moderators Posts: 6,376 Mod ✭✭✭✭Macha


    Thanks guys, I'll stick with the t-test. Ah! panic is setting in!


  • Options
    #8
    IvyTheTerrible
    Closed Accounts Posts: 14,949 ✭✭✭✭IvyTheTerrible


    I would recommend at least reading about t-tests in a basic stats text-book - There is no point in trying to do stats if you don't understand a little bit of what you are doing. Otherwise you'll use the wrong test (Excel has a few options for the t-test) or misinterpret your results.

    I find "Choosing and using Statistics: A biologists guide". Ok it's for biologists but the basic gist of it would be fine for any discipline.


  • Options
    #9
    Macha
    Moderators, Science, Health & Environment Moderators Posts: 6,376 Mod ✭✭✭✭Macha


    Thanks IvyTheTerrible, I do have a statstical analysis book for Excel, which is the application I'm using.

    Also, could anyone shed some light on the best way to present Likert-scale results?

    Thanks in advance


  • Options
    #10
    Victor
    Registered Users Posts: 78,297 ✭✭✭✭Victor


    What is the nature of what you are comparing? Are you comparing before / after type results?


  • Options
    #11
    Macha
    Moderators, Science, Health & Environment Moderators Posts: 6,376 Mod ✭✭✭✭Macha


    Victor wrote: »
    What is the nature of what you are comparing? Are you comparing before / after type results?

    Thanks for the reply Victor, but it's all done at this stage...well except for the last chapter - to be handed in on Monday!

    2Scoups and drango sorted me out and I muddled through. My tutor's voodoo doll is worn out :pac:

    Edit: Actually one last question-is it standard practice to put a "Self reflection" section in the Conclusions & Recommendations chapter? It's in the guidebook I bought but I find it a bit odd..


  • Advertisement
Advertisement