ROC   -   Test out for DTM


                   ROCKIT (Windows95 version 0.9.1 BETA):                      

             Maximum Likelihood Estimation of a Binormal ROC Curve


                  From CONTINUOUSLY-Distributed Test Results




          -----------------------------------------------------
             Original input of   44 Actually-NEGATIVE cases
          -----------------------------------------------------

       .908         .598         .640         .640         .640
       .640         .640         .598         .598         .640
       .598         .640         .640         .640         .908
       .640         .640         .640         .908         .640
       .640         .640         .598         .598         .598
       .598         .640         .598         .640         .640
       .908         .598         .640         .640         .640
       .640         .598         .598         .598         .598
       .640         .640         .598         .640
   



          ----------------------------------------------------
             Original input of   56 Actually-POSITIVE cases
          ----------------------------------------------------

       .908         .908         .938         .938         .908
       .640         .938         .908         .640         .640
       .640         .938         .908         .908         .908
       .640         .908         .598         .908         .938
       .938         .908         .908         .938         .640
       .908         .908         .640         .908         .640
       .908         .908         .640         .640         .640
       .640         .908         .598         .908         .938
       .908         .640         .908         .908         .640
       .938         .640         .640         .908         .908
       .938         .908         .938         .908         .640
       .938
   


                                                      


      ROCKIT (Windows95 version 0.9.1 BETA):                      
          ROC data by Philip                                          

             Maximum Likelihood Estimation of the Parameters
                      a Single Binormal ROC Curve

      Name of Input File being used: Clinical DT test in.



      Condition 1: ROC         

      Total number of actually-negative cases =   44.
      Total number of actually-positive cases =   56.

      Data collected on a nominally continuous scale.
      Larger values of the test result represent stronger evidence that the
      case is actually-positive (e.g., that the patient is actually abnormal)


    Operating Points Corresponding to the Input Data Categorized by the LABROC5 Scheme:

    FPF:  .000  .000  .091  .659 1.000
    TPF:  .000  .214  .661  .964 1.000



             -----------------------------------------------------
               Initial Estimates of the Binormal ROC Parameters: 
             -----------------------------------------------------

                  a =  1.4836
                  b =   .9440

      z(k) =   .392 -1.200 -2.446

                Procedure Converges after   5 Iterations


               =====================================================
                  Final Estimates of the Binormal ROC Parameters    
               =====================================================


 Binormal Parameters and Area Under the Estimated ROC :
        a           =        1.4641
        b           =         .7672
        Area (Az)   =         .8773
        Area (Wilc) =         .8362

     1: z(k) =  -.414  1.355  2.948

 Estimated Standard Errors and Correlation of these Values:
        Std. Err. (a)  =      .2829
        Std. Err. (b)  =      .1980
        Corr(a,b)      =      .5705
        Std. Err. (Az) =      .0376
        Std. Err.(Wilc)=      .0397

 Symmetric 95% Confidence Intervals
        For a :        (  .9096, 2.0187)
        For b :        (  .3790, 1.1553)

 Asymmetric 95% Confidence Interval
        For Az:        (  .7878,  .9363)


                Variance-Covariance Matrix:
                ===========================

                a       b   z( 1)   z( 2)   z( 3)
     a       .0801
     b       .0320   .0392
     z( 1)   .0239   .0102   .0378
     z( 2)   .0222  -.0202   .0125   .0674
     z( 3)  -.0272  -.1025  -.0077   .1051   .3814


                Correlation Matrix:
                ===================

                a       b   z( 1)   z( 2)   z( 3)
     a      1.0000
     b       .5705  1.0000
     z( 1)   .4337   .2642  1.0000
     z( 2)   .3028  -.3930   .2484  1.0000
     z( 3)   .0000   .0000   .0000   .0000   .0000

        Estimated Binormal ROC curve, with Lower and Upper
        Bounds of the Asymmetric Point-wise 95% Confidence
        Interval for True-Positive Fraction at a Variety
        of False-Positive Fractions:


            FPF      TPF        (Lower Bound, Upper Bound)

            .005     .3042       (    .0911  ,    .6214   )
            .010     .3741       (    .1437  ,    .6634   )
            .020     .4555       (    .2194  ,    .7091   )
            .030     .5083       (    .2761  ,    .7377   )
            .040     .5481       (    .3221  ,    .7591   )
            .050     .5800       (    .3609  ,    .7763   )
            .060     .6068       (    .3946  ,    .7909   )
            .070     .6300       (    .4243  ,    .8036   )
            .080     .6503       (    .4507  ,    .8148   )
            .090     .6684       (    .4746  ,    .8250   )
            .100     .6847       (    .4963  ,    .8342   )
            .110     .6995       (    .5161  ,    .8427   )
            .120     .7132       (    .5344  ,    .8506   )
            .130     .7257       (    .5512  ,    .8579   )
            .140     .7374       (    .5669  ,    .8648   )
            .150     .7483       (    .5814  ,    .8713   )
            .200     .7935       (    .6417  ,    .8987   )
            .250     .8282       (    .6873  ,    .9201   )
            .300     .8559       (    .7234  ,    .9372   )
            .400     .8980       (    .7779  ,    .9621   )
            .500     .9284       (    .8185  ,    .9782   )
            .600     .9514       (    .8514  ,    .9885   )
            .700     .9690       (    .8800  ,    .9947   )
            .800     .9826       (    .9069  ,    .9981   )
            .900     .9928       (    .9351  ,    .9996   )
            .950     .9968       (    .9524  ,    .9999   )


      Estimated Relationship between the Critical Test-Result Value
      (which separates 'positive' results form 'negative' results)
      and the Corresponding Operating Point on the Fitted Binormal
      ROC Curve:

 ***********************************************************************

          Critical Test           (  FPF ,   TPF )
          Result Value


                 .9230            (  .002,   .213)
                 .7740            (  .088,   .664)
                 .6190            (  .661,   .963)