+ Reply to Thread
Page 1 of 6 1 2 3 4 5 6 LastLast
Results 1 to 15 of 78

Thread: Need good, fast stats help, will pay, help on dissertation, inferential analysis ONLY

  1. #1
    Points: 2,067, Level: 27
    Level completed: 45%, Points required for next Level: 83

    Location
    An incredible part of the world
    Posts
    41
    Thanks
    32
    Thanked 1 Time in 1 Post

    Need good, fast stats help, will pay, help on dissertation, inferential analysis ONLY




    I need a stats coach to sort out the inferential stats part of my dissertation - that's all. The SPSS file is done, I have 109 "good" pages of the dissertation written. My colleagues all have writing coaches (which I don't need since I am a decent writer), but they are all good in stats and I stink. This is an EdD program, so not as rigorous in terms of stats, more on practical application of the study in the field. My committee chair said I need to outsource and get some help on stats. This doesn't have to be perfect, just defendable. Thanks! I want to graduate in May. The qualitative chapter is written and approved, the final chapter is halfway done. So graduation is hinging on inferential stats. Sigh.
    Last edited by purplelily; 01-26-2014 at 02:54 AM.

  2. #2
    Devorador de queso
    Points: 95,781, Level: 100
    Level completed: 0%, Points required for next Level: 0
    Awards:
    Posting AwardCommunity AwardDiscussion EnderFrequent Poster
    Dason's Avatar
    Location
    Tampa, FL
    Posts
    12,933
    Thanks
    307
    Thanked 2,629 Times in 2,245 Posts

    Re: Need good, fast stats help, will pay, help on dissertation, inferential analysis

    What kind of data/models are you looking at?
    I don't have emotions and sometimes that makes me very sad.

  3. The Following User Says Thank You to Dason For This Useful Post:

    purplelily (01-22-2014)

  4. #3
    Points: 2,067, Level: 27
    Level completed: 45%, Points required for next Level: 83

    Location
    An incredible part of the world
    Posts
    41
    Thanks
    32
    Thanked 1 Time in 1 Post

    Re: Need good, fast stats help, will pay, help on dissertation, inferential analysis

    Oh thank God someone is awake, it's 8AM here! My survey had the typical demographic questions, one Likert scale question and a bunch of yes/no questions - it was elite surveying with an audience that would stop answering if it got too difficult (per my testing of the original instrument). Yes/no questions - "no" is the answer that supports my hypothesis. Everything I read says that yes/no questions have no weight associated (even though I gave them points to analyze). Someone else said to use logistic regression. I don't care if there is no correlation or my hypothesis is disproven, I am not trying to pound a round peg into a square hole, if that makes sense.

    I guess the bottom line is that I have no idea which tests to run. I know how to use SPSS but not sure which tests. I've built all of the tables linking the questions to the specific hypothesis and research thread. Anyhow, thank you for listening!
    Last edited by purplelily; 01-26-2014 at 02:54 AM.

  5. #4
    Points: 3,631, Level: 37
    Level completed: 88%, Points required for next Level: 19
    staassis's Avatar
    Location
    New York
    Posts
    226
    Thanks
    2
    Thanked 41 Times in 39 Posts

    Re: Need good, fast stats help, will pay, help on dissertation, inferential analysis

    purplelily, how many hours of coaching are we taking about? Do you have just a couple of questions or do you need somebody to coach you through whole data analysis + write-up stage?

  6. The Following User Says Thank You to staassis For This Useful Post:

    purplelily (01-22-2014)

  7. #5
    Points: 2,067, Level: 27
    Level completed: 45%, Points required for next Level: 83

    Location
    An incredible part of the world
    Posts
    41
    Thanks
    32
    Thanked 1 Time in 1 Post

    Re: Need good, fast stats help, will pay, help on dissertation, inferential analysis

    Thanks for the note- shouldn't be long at all - I just need guidance on which tests I should run to best address my hypothesis based on how I asked the survey questions. I understand SPSS and can analyze the results and do my own write up about statistical significance and whatever else.

  8. #6
    Points: 3,631, Level: 37
    Level completed: 88%, Points required for next Level: 19
    staassis's Avatar
    Location
    New York
    Posts
    226
    Thanks
    2
    Thanked 41 Times in 39 Posts

    Re: Need good, fast stats help, will pay, help on dissertation, inferential analysis

    Ok, then please describe in 2-3 paragraphs your data and the research questions.

  9. The Following User Says Thank You to staassis For This Useful Post:

    purplelily (01-22-2014)

  10. #7
    Pirate
    Points: 15,159, Level: 79
    Level completed: 62%, Points required for next Level: 191
    victorxstc's Avatar
    Posts
    875
    Thanks
    229
    Thanked 332 Times in 297 Posts

    Re: Need good, fast stats help, will pay, help on dissertation, inferential analysis

    purplelilly, instead of paying, I suggest you doing two FREE things:

    1. Click the thanks button underneath each comment you receive.

    2. When you got more familiar with stats, or even right now (if you know it at some levels), try to help other question askers, and this way kindly contribute to the community.

    Thanks

    tmp pst
    "victor is the reviewer from hell" -Jake
    "victor is a machine! a publication machine!" -Vinux

  11. The Following User Says Thank You to victorxstc For This Useful Post:

    purplelily (01-22-2014)

  12. #8
    Points: 2,067, Level: 27
    Level completed: 45%, Points required for next Level: 83

    Location
    An incredible part of the world
    Posts
    41
    Thanks
    32
    Thanked 1 Time in 1 Post

    Re: Need good, fast stats help, will pay, help on dissertation, inferential analysis

    Of course! I am not seeing a "thanks" button, though, I will keep looking. Of course I am thankful for any help, and will give back 100%.
    Last edited by purplelily; 01-26-2014 at 02:55 AM.

  13. #9
    Pirate
    Points: 15,159, Level: 79
    Level completed: 62%, Points required for next Level: 191
    victorxstc's Avatar
    Posts
    875
    Thanks
    229
    Thanked 332 Times in 297 Posts

    Re: Need good, fast stats help, will pay, help on dissertation, inferential analysis

    Thanks a lot dear purplelily and welcome to the community
    Could you elaborate on your question? What is your study about? What are the variables? What are the goals of study? What do you want to prove? How many groups are there in your study? How many specimens are there in each group? Do you have subgroups too? How many? and How many samples within each subgroup? With this information, it would be much easier to figure it out.
    "victor is the reviewer from hell" -Jake
    "victor is a machine! a publication machine!" -Vinux

  14. The Following User Says Thank You to victorxstc For This Useful Post:

    purplelily (01-23-2014)

  15. #10
    Points: 2,067, Level: 27
    Level completed: 45%, Points required for next Level: 83

    Location
    An incredible part of the world
    Posts
    41
    Thanks
    32
    Thanked 1 Time in 1 Post

    Re: Need good, fast stats help, will pay, help on dissertation, inferential analysis

    Hello again! Thank you so much, just talking this out is making me feel better!

    Ok, I will try to make this simple. The study is about leader development in higher ed (or lack thereof, my hypothesis). I am loosely using the military model as the framework. Although some elements of development may be present informally, this is about a formal structure for equitable, consistent development across the masses. (there is a crisis, so this is a hot topic) I have a ton of yes/no questions.

    In SPSS, I gave Yes 2 points and No 1 point. No always supports my hypothesis.

    Thanks for any insight you may have - anything is helpful at this point!
    Last edited by purplelily; 01-26-2014 at 02:55 AM.

  16. #11
    Pirate
    Points: 15,159, Level: 79
    Level completed: 62%, Points required for next Level: 191
    victorxstc's Avatar
    Posts
    875
    Thanks
    229
    Thanked 332 Times in 297 Posts

    Re: Need good, fast stats help, will pay, help on dissertation, inferential analysis

    Thanks. Lets reword your informative response a little bit.

    So if I understood correctly, you have a sample of 21 college presidents. You have interviewed 14 of them more in depth, but what you are caring about right now is the group of those 21 presidents, not those 14.

    Then each of those 21 has been asked questions about their past. And you want to assess whether their past activities has had any impact on their presidency position.

    Is "presidency position" a matter of Yes/No? Or for example the age at which they have elected as college president matters?

    I am asking this because you need to compare two groups, in order to be able to assess the effects of your independent variables (which are the past activities of those people). You should either go find a new "control" group consisting of academic people who are at the same age and similar demographics with those presidents, but have not been elected as president. Then by comparing the variables of those groups, you can assess your H0 hypothesis (I think you mean H0, by saying "H1", right?).

    However, if you are not going to find a new Control group, you need to find some ranks within your very group. By rank I mean some superiority of some presidents versus the lack of that superiority in the other ones. For example a very competent person can become a president at the age of 35 years, while another one can become president at 55... This can be used to qualify them. If you had such data to sort your 21 subjects according to some academic qualifications, tell me so that we can move forward to the statistics necessary for assessment of the variables.

    However, if you cannot find any source of contrast in your 21 subjects, and also cannot find a control group, I think (according to my current understanding of the situation) we cannot determine the role of the past activities (or as you put "Deliberate leader development activities").
    "victor is the reviewer from hell" -Jake
    "victor is a machine! a publication machine!" -Vinux

  17. The Following 2 Users Say Thank You to victorxstc For This Useful Post:

    Madonna (01-27-2014), purplelily (01-23-2014)

  18. #12
    Pirate
    Points: 15,159, Level: 79
    Level completed: 62%, Points required for next Level: 191
    victorxstc's Avatar
    Posts
    875
    Thanks
    229
    Thanked 332 Times in 297 Posts

    Re: Need good, fast stats help, will pay, help on dissertation, inferential analysis

    About coding "Yes" and "No", I would suggest coding them as 1 and 0 respectively (as in On and Off). However, 2 and 1 is mathematically correct to be assigned for Yes and No. But that is not common and also can be confused with ranked variables.
    "victor is the reviewer from hell" -Jake
    "victor is a machine! a publication machine!" -Vinux

  19. The Following User Says Thank You to victorxstc For This Useful Post:

    purplelily (01-23-2014)

  20. #13
    Fortran must die
    Points: 58,790, Level: 100
    Level completed: 0%, Points required for next Level: 0
    noetsi's Avatar
    Posts
    6,532
    Thanks
    692
    Thanked 915 Times in 874 Posts

    Re: Need good, fast stats help, will pay, help on dissertation, inferential analysis

    With so few cases I would not try any statistics (other than basic descriptives where you are testing nothing). I would use qualitative coding. There is simply no way with so few cases that you can argue you have a representative sample (from a quantitative approach which assumes random sampling in any case). It is unlikely that any statistical method is going to work with just 14-21 cases anyhow (note that others disagree with this, it is based on my own experience and comments in regression text for example).

    Qualitative methods do not assume a minimum sample size nor is power a issue. Instead they are tied to two approaches. How reasonable is it you have enough people to respresent some phenomenon. Or alternatively have you reached saturation where you are not learning anything new (note that many qualitative researchers would argue that there is no need to generalize at all so sample size is meaningless, but your committee probably won't agree especially if they are quantiatively trained). I think 14 cases is too few to analyze any complex phenomenon, but you might get lucky with your committee. You simply have to ask them.

    You need to get a good, practical not theoretical, book on qualitative methods. Stay away from people like Lincoln and Guba who are interested in the theory of qualitative methods as compared to how to actually do it I did a qualitative dissertation so it is possible. But it is a lot of hard work.
    "Very few theories have been abandoned because they were found to be invalid on the basis of empirical evidence...." Spanos, 1995

  21. The Following User Says Thank You to noetsi For This Useful Post:

    purplelily (01-23-2014)

  22. #14
    Pirate
    Points: 15,159, Level: 79
    Level completed: 62%, Points required for next Level: 191
    victorxstc's Avatar
    Posts
    875
    Thanks
    229
    Thanked 332 Times in 297 Posts

    Re: Need good, fast stats help, will pay, help on dissertation, inferential analysis

    Quote Originally Posted by noetsi View Post
    With so few cases I would not try any statistics (other than basic descriptives where you are testing nothing). I would use qualitative coding. There is simply no way with so few cases that you can argue you have a representative sample (from a quantitative approach which assumes random sampling in any case). It is unlikely that any statistical method is going to work with just 14-21 cases anyhow (note that others disagree with this, it is based on my own experience and comments in regression text for example).
    I think it is not that black and white noetsi. As you suggested sometimes ago, this is a matter of "more and less", not necessarily yes or no. So I think although n = 24 is not high and that it exerts Limitations over the results, it is still not by any means useless. Of course it can be used and even detect some quite significant results... If it could not detect significant results, again there will be useful information for future studies with larger samples... So I would of course use qualitative methods, but would not discard quantitative methods.
    "victor is the reviewer from hell" -Jake
    "victor is a machine! a publication machine!" -Vinux

  23. #15
    Points: 2,067, Level: 27
    Level completed: 45%, Points required for next Level: 83

    Location
    An incredible part of the world
    Posts
    41
    Thanks
    32
    Thanked 1 Time in 1 Post

    Re: Need good, fast stats help, will pay, help on dissertation, inferential analysis


    Thank you so much to both of you! This is great stuff and makes so much sense. So nice to chat with someone about this, my colleagues are all bogged down in their own studies and I don't want to overload them. I am digesting all of the advice right now and coming to better grips about what I need to do in the inferential stats section. I am going back into SPSS to change the yes/no to 1/0, I loved the on and off analogy. For me, "off" supports the hypothesis, so I just need to remember that working through the data. Or should I make it 0/1?

    Drinking a Red Bull and digging back in.
    Last edited by purplelily; 01-26-2014 at 02:56 AM.

+ Reply to Thread
Page 1 of 6 1 2 3 4 5 6 LastLast

           




Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts






Advertise on Talk Stats