P-value: Stored Results of Ranksum

[Priscilla Fung]

Hi,

I was wondering if there is a way to obtain the p-value from stored results of ranksum tests. Below are the options I found through "help ranksum", none of which match.
Any other options showing correlation between a categorical and continuous variable would also work!

Stored results

ranksum stores the following in r():

Scalars
r(N_1) sample size n_1
r(N_2) sample size n_2
r(z) z statistic
r(Var_a) adjusted variance
r(group1) value of variable for first group
r(sum_obs) actual sum of ranks for first group
r(sum_exp) expected sum of ranks for first group
r(porder) probability that draw from first population is larger than
draw from second population

Thank you!
Priscilla

[Carlo Lazzaro]

Priscilla:
you may want to consider:

.  webuse fuel2

. ranksum mpg, by(treat)

Two-sample Wilcoxon rank-sum (Mann–Whitney) test

       treat |      Obs    Rank sum    Expected
-------------+---------------------------------
   Untreated |       12         128         150
     Treated |       12         172         150
-------------+---------------------------------
    Combined |       24         300         300

Unadjusted variance      300.00
Adjustment for ties       -4.04
                     ----------
Adjusted variance        295.96

H0: mpg(treat==Untreated) = mpg(treat==Treated)
         z = -1.279
Prob > |z| = 0.2010
Exact prob = 0.2117

. return list

scalars:
                  r(N) =  24
                r(N_1) =  12
                r(N_2) =  12
             r(group1) =  0
            r(sum_obs) =  128
            r(sum_exp) =  150
              r(Var_a) =  295.9565217391304
                  r(z) =  -1.278817949868369
                  r(p) =  .2009611714248259
                r(p_l) =  .1004805857124129
                r(p_u) =  .8995194142875871
            r(p_exact) =  .211683793390618
          r(p_l_exact) =  .105841896695309
          r(p_u_exact) =  .9006281442342823

. g wanted=r(p_exact)

. list in 1/10

     +----------------------------+
     | mpg       treat     wanted |
     |----------------------------|
  1. |  20   Untreated   .2116838 |
  2. |  23   Untreated   .2116838 |
  3. |  21   Untreated   .2116838 |
  4. |  25   Untreated   .2116838 |
  5. |  18   Untreated   .2116838 |
     |----------------------------|
  6. |  17   Untreated   .2116838 |
  7. |  18   Untreated   .2116838 |
  8. |  24   Untreated   .2116838 |
  9. |  20   Untreated   .2116838 |
 10. |  24   Untreated   .2116838 |
     +----------------------------+

.

[Priscilla Fung]

Hi,

Thank you so much for the response Carlo!

Unfortunately, I am using Stata 15, so only the options below are shown for me when I use "ranksum" followed by "return list"

Ho: age_a~rt(ncigr3..==0) = age_a~rt(ncigr3..==1)
z = 1.870
Prob > |z| = 0.0615

. return list

scalars:
r(N_2) = 19
r(N_1) = 91
r(Var_a) = 15993.25
r(z) = 1.870091180546235
r(sum_exp) = 5050.5
r(sum_obs) = 5287
r(group1) = 0

However, I found Nick's archived post (http://www.stata.com/statalist/archi.../msg00622.html) and the below code seems to work

ranksum age_atchemostart if cond1!=.& cond2!=. & ncigr3var!=., by (ncigr3var)
scalar pvalue = 2*normprob(-abs(`r(z)'))
display pvalue // to check that it matches
putexcel D6 = pvalue, nformat(number_d3)

Thanks again!
Priscilla

[Carlo Lazzaro]

Priscilla:
happy with reading that you found a solution.
In your future posts, please mention the Stata release that you're using (as per FAQ). Otherwise, interested listers assume that you are worworking with Stata 18 and reply accordinglyl (and unhelpfully, if your Stata release is older than the last one).Thanks.