/*Oct 2002*/
/*Based on comments at the ASA meeting, I am examining item level more closeyl by doing factor analysis*/

/*specifying library*/
%let direc=C:\AAA PROJECTS\ECBW\N2P\;
%let WD=C:\AAA PROJECTS\ECBW\N2P\;
libname here "&direc" ;
libname here "&WD" ;
libname there "&direc.data4" ;/*data storage*/

data base;set here.all;
keep STDID SEX
/*F1s12A  F1s12B  F1s12C  F1s12D  F1s12E  F1s12F  F1s12G  F1s12H  F1s12I
F1s12J  F1s12K  F1s12L  F1s12M  F1s12N  F1s12O  F1s12P  F1s12Q  F1s12R */
/*the values of these are already reverse coded in nelsrunprog4.5sas*/
late cut skip cheat copy fight gang racist sexist steal destroy smoke drink drug weapon abuse talkback disobey
F_abuse F_destroy F_fight F_copy
;
/*these are already reverse conded in nelsrunprogrog4.5.sas*/
late=F1S12A;
cut=F1S12B;
skip=F1S12C;
cheat=F1S12D;
copy=F1S12E;
fight=F1S12F;
gang=F1S12G;
racist=F1S12H;
sexist=F1S12I;
steal=F1S12J;
destroy=F1S12K;
smoke=F1S12L;
drink=F1S12M;
drug=F1S12N;
weapon=F1S12O;
abuse=F1S12P;
talkback=F1S12Q;
disobey=F1S12R;

/*these are items that behave differently across gender*/
/*
if sex=2 then do;
F_abuse=abuse;
abuse=.;
F_destroy=destroy;
destroy=.;
F_fight=fight;
fight=.;
F_copy=copy;
copy=.;
end;
*/
run;
proc sort;by sex;run;



/*Here to create an excel table deviance_frequency_table.xls*/
/*Here to create an excel table deviance_frequency_table.xls*/
/*Here to create an excel table deviance_frequency_table.xls*/
*****************************************************************************;
*****************************************************************************;
*****************************************************************************;
*****************************************************************************;
proc freq;
by sex;
tables
late cut skip cheat copy fight gang racist sexist steal destroy smoke drink drug weapon abuse talkback disobey
/*F_abuse F_destroy F_fight F_copy*/;
ods output  OneWayFreqs=freqresult;
run;

data freqresult2;set freqresult;
jun=_n_;
new=table||sex;
run;

proc sort;by new;run;

data freqresult3;set freqresult2;
by new;
if first.new then count=0;
count+1;
if count=3 then code="Sometimes";
if count=4 then code="Often";
if count=2 then code="Rarely";
if count=1 then code="Never";

run; 
proc sort;by jun;run;


data boy;set freqresult3;
keep table code Malefreqency Malepercent;
if sex=1;
Malecode=code;
Malefreqency= Frequency;
Malepercent=percent/100;
run;

data girl;set freqresult3;
keep table  Femalefreqency  Femalepercent;
if sex=2;
Femalecode=code;
Femalefreqency= Frequency;
Femalepercent=percent/100;
run;

data bothsex;merge boy girl;
length desc $ 40;
if table="late" then desc="To be late for school";
if table="copy" then desc="To copy someone's homework";
if table="talkback	" then desc="To talk back to teachers";
if table="disobey	" then desc="To disobey school rules";
if table="skip" then desc="To skip school a whole day";
if table="cut" then desc="To cut a couple of classes	";
if table="fight" then desc="To get into physical fights";
if table="cheat" then desc="To cheat on tests";
if table="sexist" then desc="To make sexist remarks";
if table="smoke" then desc="To smoke on school grounds";
if table="racist" then desc="To make, racist remarks";
if table="gang" then desc="To belong to gangs";
if table="weapon" then desc="To bring weapons to school";
if table="destroy" then desc="To destroy, school property";
if table="drink" then desc="To drink alcohol at school";
if table="steal" then desc="To steal belongings frm school";
if table="abuse" then desc="To abuse teachers";
if table="drug" then desc="To use drugs at school";
run;

proc export data=bothsex outfile="C:\My Documents\AERA\deviance_frequency_table.xls"
dbms=excel2000 replace;run;

*****************************************************************************;
*****************************************************************************;
*****************************************************************************;
*****************************************************************************;

option noxwait xsync;
%let win= C:\WINSTEPS\winsteps;
%let WD=&direc;

%let scale=F1dev;
%let ni =22;/*n of items*/
%let codes=1234;

data F1dev;set base;
/*women have DIF for these items, so treat them separately*/
/*
if sex=2 then do;
F_abuse=abuse;
abuse=.;
F_destroy=destroy;
destroy=.;
F_fight=fight;
fight=.;
F_copy=copy;
copy=.;
end;
*/


blank=' ';
file "&WD.F1dev.dat";
put 
(STDID) (5.0)
(blank)(1.0)
(sex) (1.0)
( 
late cut skip cheat copy fight gang racist sexist steal destroy smoke drink drug weapon abuse talkback disobey
/*F_abuse F_destroy F_fight F_copy*/
)(1.0);
run;


/*
late cut skip cheat copy fight gang racist 
sexist steal destroy smoke drink drug weapon abuse talkback disobey
*/
/*these are lables for variables, which will come, in the WINSTEP control file, in between END and END NAMES*/

filename new "C:\WINSTEPS\F1dev.con";
data _null_;
file new;
put '&inst	';
put "	item1=8";
put "	name1=1";
put "	ni=&ni";
put "	HLINES=n";
put "	;target=y";
put ";table=";
put "	person=student";
put "	item=question";
put "	codes=&codes";
put "	udecim=6";
put ";	pfile=F1dev.pfile";
put ";    ifile=F1dev.ifile";
put "	csv=n";
put "	data=&WD.F1dev.dat";
put '	&end';
put "late";
put "cut";
put "skip";
put "cheat";
put "copy";
put "fight";
put "gang";
put "racist";
put "sexist";
put "steal";
put "destroy";
put "smoke";
put "drink";
put "drug";
put "weapon";
put "abuse";
put "talkback";
put "disobey";
put "F_abuse";
put "F_destroy";
put "F_fight";
put "F_copy";
put "	END NAMES";
run;


filename idfile "C:\WINSTEPS\idfile2.txt";
data _null_;
file idfile;
put "12";
put "7";
put "14";
put "16";
put "13";
put "10";
put "11";/*up to here May 2003*/
put "15";
put "8";
put "9";
run;

x  "start &win  F1dev.con F1dev_whole.out  ifile=F1dev_whole.ifile pfile=F1dev_whole.pfile ";

x  "start &win  &WDF1dev.con F1dev._whole.out  ifile=F1dev._whole.ifile pfile=F1dev._whole.pfile ";
x  "start &win  &WDF1dev.con F1dev._boys.out  pselect=??????1*  ifile=F1dev._boys.ifile pfile=F1dev._boys.pfile";
x  "start &win  &WDF1dev.con F1dev._boys.out  pselect=??????1*  ifile=F1dev._boys.ifile idfile=idfile.txt  pfile=F1dev._boys2.pfile";

/*without mistfit May 2003*/
x  "start &win  &WDF1dev.con F1dev._boys.out  pselect=??????1*  ifile=F1dev._boys.ifile idfile=idfile2.txt  pfile=F1dev._boys3.pfile";
x  "start &win  &WDF1dev.con F1dev._girls.out  pselect=??????2*  ifile=F1dev._girls.ifile pfile=F1dev._girls.pfile";
*x  "start &win  &WDF1dev.con F1dev._09.out  pselect=???????????????????????9*  ifile=F1dev._09.ifile";
*x  "start &win  &WDF1dev.con F1dev._10.out  pselect=??????????????????????10*  ifile=F1dev._10.ifile";
*x  "start &win  &WDF1dev.con F1dev._11.out  pselect=??????????????????????11*  ifile=F1dev._11.ifile";
*x  "start &win  &WDF1dev.con F1dev._12.out  pselect=??????????????????????12*  ifile=F1dev._12.ifile";

/*CHECKING DIF*/
data  blank;
set _null_;
 ENTRY=.;
 MEASURE=. ;
 STTS=.; 
 COUNT=.; 
 SCORE=.; 
 STDER=.; 
 INMSQ=.; 
 INZSTD=.; 
 OUTMS =.; 
 OUTZSTD=.; 
 DISPL =.; 
 CORR =.; 
 WEIGHT=.; 
 G =.; 
 M ="        "; 
 NAME="        "; 
 MFIT =.; 
subgroup="        "; 


%macro edward (suji);
FILENAME raw2 "&WDF1dev._&suji..ifile";
data outdata;
infile raw2 missover;
input
ENTRY MEASURE STTS COUNT    SCORE  STDER INMSQ INZSTD OUTMS OUTZSTD DISPL CORR WEIGHT G M  $ NAME $;
if stTS  < = -2 then measure=.;
MFIT=0;
if OUTMS > 2 then MFIT=1;
if entry > -999;
subgroup="&suji";
proc sort;
by name;
run;

proc append base=blank  data=outdata;
run;

%mend edward;
%edward (suji=whole);
%edward (suji=boys);
%edward (suji=girls);
*%edward (suji=09);
*%edward (suji=10);
*%edward (suji=11);
*%edward (suji=12);


*%macro dummy;
data here.F1dev._idif;
set blank;
run;
proc sort;by subgroup;run;


ods html file="&WDF1dev.body.htm" 
contents="&WDF1dev.contents.htm" 
page="&WDF1dev.page.htm" 
frame="&WDF1dev.index.htm";
filename outfiles "&WD";

goptions gsfname=outfiles dev=gif gunit=pct cback=white htitle=6 htext=3 ftext=swissb ctext=black vsize=8  hsize=8;
title1 font=duplex "Item difficulty varying across grades";
footnote1 "";
symbol interpol=spline /*join*/ value=dot height=1;
proc gplot data=here.F1dev._idif;
plot measure*subgroup=name / /*overlay*/;
run;
quit;

ods html close;
proc print data=here.F1dev._idif;run;
*%mend dummy;

/*Person measures*/
/*Person measures*/
/*Person measures*/
/*Person measures*/
/*Person measures*/

*%macro ken (scale=);
*%let scale=whole;
/*whole*/
FILENAME raw1 "&direc.F1dev._whole.pfile";
data F1dev;infile raw1 missover;
*length TchID $ 8;
input ENTRY  measure STTS  COUNT   SCORE 
STDER  INMSQ  INZSTD   OUTMS 
OUTZSTD  DISPL   CORR WEIGHT   
STDID ;
if stTS  < = -2 then measure=.;
MFIT=0;
if OUTMS > 2 then MFIT=1;
if entry > -999;
F1dev.R=measure;
F1dev.ERR=STDER;
F1dev.MFIT=MFIT;

F1dev.RLOG=log(measure+ 10);
run;
proc sort;by STDID;run;

/*boys*/
FILENAME raw1 "&direc.F1dev._boys2.pfile";
data BF1dev;infile raw1 missover;
*length TchID $ 8;
input BENTRY  Bmeasure BSTTS  BCOUNT   BSCORE 
BSTDER  BINMSQ  BINZSTD   BOUTMS 
BOUTZSTD  BDISPL   BCORR BWEIGHT   
STDID ;
if BstTS  < = -2 then Bmeasure=.;
BMFIT=0;
if BOUTMS > 2 then BMFIT=1;
if Bentry > -999;

BF1dev.R=Bmeasure;
BF1dev.ERR=BSTDER;
BF1dev.MFIT=BMFIT;
BF1dev.RLOG=log(Bmeasure+ 10);
run;
proc sort;by STDID;run;



/*boys*/
/*I further reduced the number of items following Nick's advice*/
FILENAME raw1 "&direc.F1dev._boys3.pfile";
data CF1dev;infile raw1 missover;
*length TchID $ 8;
input CENTRY  Cmeasure CSTTS  CCOUNT   CSCORE 
CSTDER  CINMSQ  CINZSTD   COUTMS 
COUTZSTD  CDISPL   CCORR CWEIGHT   
STDID ;
if CstTS  < = -2 then Cmeasure=.;
CMFIT=0;
if COUTMS > 2 then CMFIT=1;
if Centry > -999;

CF1dev.R=Cmeasure;
CF1dev.ERR=CSTDER;
CF1dev.MFIT=CMFIT;
CF1dev.RLOG=log(Cmeasure+ 10);
run;
proc sort;by STDID;run;


data bothscales;
merge F1dev BF1dev CF1dev;
by STDID;
run;

proc univariate  data=bothscales  plot;
var F1devR  F1devRLOG
BF1devR  BF1devRLOG
CF1devR  CF1devRLOG;
run;

proc standard data=bothscales  out=bothscales mean=0 std=1;
var F1devR F1devRLOG
BF1devR  BF1devRLOG
CF1devR  CF1devRLOG
;
run;


data here.F1dev;set bothscales;
keep STDID F1devR F1devRlog
BF1devR BF1devRlog
BMFIT MFIT
CF1devR 
CF1devRlog CMFIT;

*if F1devR > 2.5 then F1devR=2.5;
*if F1devR < -2.5 then F1devR=-2.5;
*if F1devRLOG > 2.5 then F1devRLOG=2.5;
*if F1devRLOG < -2.5 then F1devRLOG=-2.5;

proc univariate  data=here.F1dev  plot;
var F1devR  F1devRLOG
 BF1devR  BF1devRLOG
 CF1devR  CF1devRLOG;
run;

proc corr data=here.F1dev;
var  F1devR BF1devR CF1devR;run;

proc freq data=here.F1dev;
where MFIT > -999;
tables
BMFIT MFIT
CMFIT;
run;

proc univariate data=here.F1dev plot;
class BMfit;
var BF1devRlog
CF1devRlog

;run;