1          OPTIONS NONOTES NOSTIMER NOSOURCE NOSYNTAXCHECK;
 55         
 56         /* MathReg1.sas  */
 57         %include '/folders/myfolders/441s16/Lecture/readmath2b.sas';
 NOTE: Format YNFMT is already on the library WORK.FORMATS.
 NOTE: Format YNFMT has been output.
 NOTE: Format CRSFMT is already on the library WORK.FORMATS.
 NOTE: Format CRSFMT has been output.
 NOTE: Format NFMT is already on the library WORK.FORMATS.
 NOTE: Format NFMT has been output.
 
 NOTE: PROCEDURE FORMAT used (Total process time):
       real time           0.00 seconds
       cpu time            0.00 seconds
       
 
 159        /* readmath2b has dummy variable definitions
 160                e1-e4,e6 for ethnic (Reference category is East Indian)
 161                gender=1 for Female
 162                mtongue=1 for English
 163                c1-c3: c1 = 'Catch-up' c2 = 'Mainstream' c3 = 'Elite'   */
 164        
 165        title2 'Variable Selection for Predicting Grade';
 166        
 
 NOTE: The infile '/folders/myfolders/exploremath.data.txt' is:
       Filename=/folders/myfolders/exploremath.data.txt,
       Owner Name=root,Group Name=vboxsf,
       Access Permission=-rwxrwx---,
       Last Modified=18Jan2016:17:34:49,
       File Size (bytes)=44583
 
 NOTE: 579 records were read from the infile '/folders/myfolders/exploremath.data.txt'.
       The minimum record length was 75.
       The maximum record length was 75.
 NOTE: Missing values were generated as a result of performing an operation on missing values.
       Each place is given by: (Number of times) at (Line):(Column).
       99 at 80:24    99 at 117:13   
 NOTE: The data set WORK.MATHEX has 579 observations and 34 variables.
 NOTE: DATA statement used (Total process time):
       real time           0.02 seconds
       cpu time            0.02 seconds
       
 167        proc freq;
 
 168             title3 'Check dummy variables';
 169             tables sex*gender / norow nocol nopercent missing;
 170             tables tongue*mtongue / norow nocol nopercent missing;
 171             tables (e1-e4 e6) * ethnic / norow nocol nopercent missing;
 172             tables (c1-c3) * course / norow nocol nopercent missing;
 173        
 
 NOTE: There were 579 observations read from the data set WORK.MATHEX.
 NOTE: PROCEDURE FREQ used (Total process time):
       real time           0.24 seconds
       cpu time            0.23 seconds
       
 174        proc reg plots = none; /* Suppress diagnostic plots for now*/
 
 175             title3 'Model A: Predict University Calculus Grade from HS Information';
 176             model grade = hsgpa hscalc hsengl;
 177        
 178        /* It is very interesting to know what proportion of the remaining
 179           variation is explained by each variable, controlling for the other two.
 180           F = t-squared, and
 181                               a = sF/(n-p + sF)
 182        */
 183        
 
 NOTE: PROCEDURE REG used (Total process time):
       real time           0.06 seconds
       cpu time            0.06 seconds
       
 184        proc iml;
 
 NOTE: IML Ready
 185             title3 'Proportion of remaining variation for HS information';
 186             n = 323;
 186      !               p = 4;
 186      !                      s = 1;
 187             print "hsgpa controlling for hscalc and hsengl";
 188             t = 8.00;
 188      !                F = t**2;
 188      !                           a = s*F/(n-p + s*F);
 189             print a;
 190        
 191             print "hscalc controlling for hsgpa and hsengl";
 192             t = 3.14;
 192      !                F = t**2;
 192      !                           a = s*F/(n-p + s*F);
 193             print a;
 194        
 195             print "hsengl controlling for hsgpa and hscalc";
 196             t = -3.26;
 196      !                 F = t**2;
 196      !                            a = s*F/(n-p + s*F);
 197             print a;
 198        
 NOTE: Exiting IML.
 NOTE: PROCEDURE IML used (Total process time):
       real time           0.03 seconds
       cpu time            0.04 seconds
       
 199        proc reg plots = none;
 
 200             title3 'Model B: Predict University Calculus Grade from Diagnostic Test';
 201             model grade = precalc calc;
 202        
 
 NOTE: PROCEDURE REG used (Total process time):
       real time           0.05 seconds
       cpu time            0.05 seconds
       
 203        proc reg plots = none;
 
 204             title3 'Model C: Do the diagnostic test and HS info both contribute?';
 205             model grade = hsgpa hscalc hsengl precalc calc;
 206             Diagnostic_Test:  test  precalc=calc=0;
 207             HS_Information:   test  hsgpa=hscalc=hsengl=0;
 208        
 
 NOTE: PROCEDURE REG used (Total process time):
       real time           0.09 seconds
       cpu time            0.08 seconds
       
 209        proc iml;
 
 NOTE: IML Ready
 210             title3 'Proportion of remaining variation explained by diagnostic test';
 211             print "Precalc and calc controlling for hsgpa hscalc hsengl";
 212             n = 289;
 212      !               p = 6;
 212      !                      s = 2;
 212      !                             F = 8.28;
 213             a = s*F/(n-p + s*F);
 213      !                           print a;
 214        
 NOTE: Exiting IML.
 NOTE: PROCEDURE IML used (Total process time):
       real time           0.02 seconds
       cpu time            0.02 seconds
       
 215        proc reg plots = none;
 
 216             title3 'Model D: See if Course makes a contribution';
 217             model grade = hsgpa hscalc hsengl precalc calc c1 c3;
 218             Course: test c1=c3=0;
 219             Diagnostic_Test:  test  precalc=calc=0;
 220        
 
 NOTE: PROCEDURE REG used (Total process time):
       real time           0.08 seconds
       cpu time            0.07 seconds
       
 221        proc glm;
 
 222             title3 'Model D again with proc glm';
 223             class course;
 224             model grade =  hsgpa hscalc hsengl precalc calc course;
 225             contrast 'Replicate Test of Course' course 1 -1  0,
 226                                                 course 0  1 -1;
 227             contrast 'Diagnostic Test F = 9.06' precalc 1, calc 1;
 228        
 
 NOTE: PROCEDURE GLM used (Total process time):
       real time           0.09 seconds
       cpu time            0.08 seconds
       
 229        proc reg plots = none;
 
 230             title3 'Model E: Include Language, Sex and Ethnic Background';
 231             model grade = hsgpa hscalc hsengl precalc calc
 232                           mtongue gender e1-e4 e6;
 233             TroubleVars: test mtongue=gender=e1=e2=e3=e4=e6=0;
 234             Nationality: test e1=e2=e3=e4=e6=0;
 235        
 
 NOTE: PROCEDURE REG used (Total process time):
       real time           0.10 seconds
       cpu time            0.09 seconds
       
 236        proc reg plots = none;
 
 237             title3 'Model F: Discarding Gender and Nationality';
 238             model grade = hsgpa hscalc hsengl precalc calc mtongue;
 239             EnglishTongue: test hsengl=mtongue=0;
 240        
 
 NOTE: PROCEDURE REG used (Total process time):
       real time           0.07 seconds
       cpu time            0.07 seconds
       
 241        proc iml;
 
 NOTE: IML Ready
 242             title3 'Proportion of remaining variation explained by mother tongue';
 243             print "Mtongue controlling for hsgpa hscalc hsengl precalc calc";
 244             n = 287;
 244      !               p = 7;
 244      !                      s = 1;
 244      !                             t = -2.23 ;
 244      !                                         F = t**2;
 245             a = s*F/(n-p + s*F);
 245      !                           print a;
 246        
 NOTE: Exiting IML.
 NOTE: PROCEDURE IML used (Total process time):
       real time           0.02 seconds
       cpu time            0.01 seconds
       
 247        proc reg plots = none;
 
 248             title3 'Model G:  Drop mtongue and calc';
 249             title4 'Compare  R-Square = 0.4556, Adj R-Sq = 0.4460 From Model 3';
 250             model grade = hsgpa hscalc hsengl precalc;
 251        
 
 NOTE: PROCEDURE REG used (Total process time):
       real time           0.06 seconds
       cpu time            0.05 seconds
       
 252        proc iml;
 
 NOTE: IML Ready
 253             title3 'Proportion of remaining variation explained by Pre-calculus';
 254             print "precalc controlling for hsgpa hscalc hsengl";
 255             n = 289;
 255      !               p = 5;
 255      !                      s = 1;
 255      !                             t = 3.63 ;
 255      !                                        F = t**2;
 256             a = s*F/(n-p + s*F);
 256      !                           print a;
 257        
 NOTE: Exiting IML.
 NOTE: PROCEDURE IML used (Total process time):
       real time           0.02 seconds
       cpu time            0.02 seconds
       
 258        proc reg plots(only) = ResidualPlot ;
 
 259             title3 'Model H: Combine precalc and calc instead of dropping calc';
 260             title4 'Compare  R-Square = 0.4492 from Model 7';
 261             model grade = hsgpa hscalc hsengl totscore;
 262        
 
 NOTE: PROCEDURE REG used (Total process time):
       real time           0.50 seconds
       cpu time            0.25 seconds
       
 263        proc iml;
 
 NOTE: IML Ready
 264             title3 'Proportion of remaining variation explained by Pre-calculus';
 265             print "totscore controlling for hsgpa hscalc hsengl";
 266             n = 289;
 266      !               p = 5;
 266      !                      s = 1;
 266      !                             t = 3.92 ;
 266      !                                        F = t**2;
 267             a = s*F/(n-p + s*F);
 267      !                           print a;
 268             print "For prediction, I am happy with Model 8: hsgpa hscalc hsengl totscore";
 269        
 NOTE: Exiting IML.
 NOTE: PROCEDURE IML used (Total process time):
       real time           0.05 seconds
       cpu time            0.05 seconds
       
 270        proc reg plots = none;
 
 271             title3 'Model I: Same as Model H but including Mother Tongue';
 272             model grade = hsgpa hscalc hsengl totscore mtongue;
 273        
 
 NOTE: PROCEDURE REG used (Total process time):
       real time           0.10 seconds
       cpu time            0.06 seconds
       
 274        proc reg plots = none;
 
 275             title3 'Try automatic (stepwise) selection';
 276             model grade = hsgpa hscalc hsengl precalc calc
 277                           mtongue gender e1-e4 e6
 278                           / selection = stepwise slentry = 0.05 slstay = 0.05 ;
 279                           /* Default slentry = slstay = 0.15 */
 280        
 281        
 282        OPTIONS NONOTES NOSTIMER NOSOURCE NOSYNTAXCHECK;
 294