Comparing paired income data on Stata

[Ciaran DK]

Hey all,

I am trying to find a statistical test to use on Stata which will compare changes in income bracket before and after an intervention.

The data is ordinal with 10 income brackets (1 = lowest, 10 = highest). The data is also paired as it is involves the same 66 subjects before and after an intervention.

I looked into Ordered Logistic Regression but I don't think this is quite the right test to use.

Is/are there a test(s) I can use to establish whether there has 1) been a statistically significant change in incomes groups following the intervention 2) to determine the overall direction of change?

I have attached my dataset below.

Many thanks for your help,

Ciaran

Pre-intervention income group	Post-intervention income group
2	2
4	4
4	1
4	4
1	1
1	1
1	1
2	1
5	6
1	1
5	5
6	6
1	1
2	3
2	2
1	1
2	3
4	1
2	2
1	1
1	1
6	6
2	2
1	1
1	1
2	3
10	10
3	3
2	4
5	5
1	1
4	5
1	4
1	1
5	1
4	4
6	7
7	1
4	6
1	1
7	6
5	2
1	1
4	5
3	1
3	4
1	1
1	1
1	1
1	1
2	2
2	2
1	1
1	1
2	2
2	2
1	5
3	3
1	1
1	2
7	9
1	1
3	3
3	4
3	3
2	2

[Carlo Lazzaro]

Ciaran:
this seems a job for -signrank- (please note that your original variables are abbreviated as -V_1- and -V-2-):

. input V_1 V_2

           V_1        V_2
  1. 2 2
  2. 4 4
  3. 4 1
  4. 4 4
  5. 1 1
  6. 1 1
  7. 1 1
  8. 2 1
  9. 5 6
 10. 1 1
 11. 5 5
 12. 6 6
 13. 1 1
 14. 2 3
 15. 2 2
 16. 1 1
 17. 2 3
 18. 4 1
 19. 2 2
 20. 1 1
 21. 1 1
 22. 6 6
 23. 2 2
 24. 1 1
 25. 1 1
 26. 2 3
 27. 10 10
 28. 3 3
 29. 2 4
 30. 5 5
 31. 1 1
 32. 4 5
 33. 1 4
 34. 1 1
 35. 5 1
 36. 4 4
 37. 6 7
 38. 7 1
 39. 4 6
 40. 1 1
 41. 7 6
 42. 5 2
 43. 1 1
 44. 4 5
 45. 3 1
 46. 3 4
 47. 1 1
 48. 1 1
 49. 1 1
 50. 1 1
 51. 2 2
 52. 2 2
 53. 1 1
 54. 1 1
 55. 2 2
 56. 2 2
 57. 1 5
 58. 3 3
 59. 1 1
 60. 1 2
 61. 7 9
 62. 1 1
 63. 3 3
 64. 3 4
 65. 3 3
 66. 2 2
 67. end

. signrank V_1=V_2

Wilcoxon signed-rank test

        sign |      obs   sum ranks    expected
-------------+---------------------------------
    positive |        8       471.5       632.5
    negative |       15       793.5       632.5
        zero |       43         946         946
-------------+---------------------------------
         all |       66        2211        2211

unadjusted variance    24505.25
adjustment for ties      -38.38
adjustment for zeros   -6858.50
                     ----------
adjusted variance      17608.38

Ho: V_1 = V_2
             z =  -1.213
    Prob > |z| =   0.2250

.

[Ciaran DK]

Thank you Carlo!