+ Reply to Thread
Results 1 to 3 of 3

Thread: How to do a post hoc analysis on non-parametric data in SAS

  1. #1
    Points: 5, Level: 1
    Level completed: 9%, Points required for next Level: 45

    Posts
    1
    Thanks
    0
    Thanked 0 Times in 0 Posts

    How to do a post hoc analysis on non-parametric data in SAS




    Good morning!

    I know there is a statistically significant difference between groups for my non-parametric data-set but I'm not sure how to find where those differences lie between groups..

    I'm comparing race/ethnicity (Asian Pacific Islander, Hispanic, etc) by academic discipline (dental, nursing, medicine, etc), and am able to tell that there is a difference, but not between which groups. I'm running this in SAS, and the code I used to find the difference is:

    PROC GLM data = local.analysisfile;
    CLASS school;
    MODEL race = school;
    MEANS school /LSD;
    run;

    Here's an example of the data:

    Code: 
    <style type="text/css">
    	table.tableizer-table {
    	border: 1px solid #CCC; font-family: Arial, Helvetica, sans-serif;
    	font-size: 10px;
    } 
    .tableizer-table td {
    	padding: 4px;
    	margin: 3px;
    	border: 1px solid #ccc;
    }
    .tableizer-table th {
    	background-color: #104E8B; 
    	color: #FFF;
    	font-weight: bold;
    }
    </style><table class="tableizer-table">
    <tr class="tableizer-firstrow"><th></th><th>Dental</th><th>Law</th><th>Medicine</th><th>Nursing</th><th>Pharmacy</th><th>Social Work</th><th>Total</th></tr>
     <tr><td>&nbsp;</td><td>n = 50</td><td>n = 78</td><td>n = 117</td><td>n = 72</td><td>n = 67</td><td>n = 131</td><td>&nbsp;</td></tr>
     <tr><td>Race/Ethnicity</td><td>&nbsp;</td><td>&nbsp;</td><td>&nbsp;</td><td>&nbsp;</td><td>&nbsp;</td><td>&nbsp;</td><td>&nbsp;</td></tr>
     <tr><td>Asian/Pacific Islander</td><td>18</td><td>13</td><td>25</td><td>3</td><td>29</td><td>5</td><td>93</td></tr>
     <tr><td>&nbsp;</td><td>(36%)</td><td>(16.7%)</td><td>(21.4%)</td><td>(4.2%)</td><td>(43.3%)</td><td>(3.8%)</td><td>(18.1%)</td></tr>
     <tr><td>Black/African American</td><td>3</td><td>14</td><td>3</td><td>11</td><td>6</td><td>23</td><td>60</td></tr>
     <tr><td>&nbsp;</td><td>(6.0%)</td><td>(18%)</td><td>(2.6%)</td><td>(15.3%)</td><td>(9.0%)</td><td>(17.6%)</td><td>(11.7%)</td></tr>
     <tr><td>Hispanic</td><td>4</td><td>5</td><td>3</td><td>6</td><td>1</td><td>5</td><td>24</td></tr>
     <tr><td>&nbsp;</td><td>(8.0%)</td><td>(6.4%)</td><td>(2.6%)</td><td>(8.3%)</td><td>(1.5%)</td><td>(3.8%)</td><td>(4.7%)</td></tr>
     <tr><td>White/Caucasian</td><td>24</td><td>46</td><td>82</td><td>52</td><td>29</td><td>95</td><td>328</td></tr>
     <tr><td>&nbsp;</td><td>(48.0%)</td><td>(59.0%)</td><td>(70.1%)</td><td>(72.2%)</td><td>(43.3%)</td><td>(72.5%)</td><td>(63.7%)</td></tr>
     <tr><td>Other</td><td>1</td><td>0</td><td>4</td><td>0</td><td>2</td><td>3</td><td>10</td></tr>
     <tr><td>&nbsp;</td><td>(2.0%)</td><td>(0.0%)</td><td>(3.4%)</td><td>(0.0%)</td><td>(3.0%)</td><td>(2.3%)</td><td>(1.9%)</td></tr>
    </table>
    Any suggestions? I can't find any consensus on the best next step and am new to stats.

    Thanks,
    Jenny

  2. #2
    Omega Contributor
    Points: 38,396, Level: 100
    Level completed: 0%, Points required for next Level: 0
    hlsmith's Avatar
    Location
    Not Ames, IA
    Posts
    7,001
    Thanks
    398
    Thanked 1,186 Times in 1,147 Posts

    Re: How to do a post hoc analysis on non-parametric data in SAS

    You code box is unbelievably jumbled up.


    Race is a class variable, how can your treat it as a dependent variable in this model? Do you have it coded as 1, 2, 3,...,etc. If so, that makes no sense. What is your sample size?

  3. #3
    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: How to do a post hoc analysis on non-parametric data in SAS


    Fischer's LSD test appears to require parametric data, you are doing a series of t test. So I don't think it makes much sense to use that.

    https://en.wikipedia.org/wiki/Post_hoc_analysis

    This tends to support the view that there are no formal ad hoc test for non-parametric data but offers some solutions [if you can read it]
    http://www.researchgate.net/post/Can...riedmans_ANOVA
    "Very few theories have been abandoned because they were found to be invalid on the basis of empirical evidence...." Spanos, 1995

+ Reply to Thread

           




Tags for this Thread

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