Follow-up to previous question

Francios — Thu, 09 Mar 2017 17:26:56 GMT

I have been implementing several techniques to address this issue, but I need some help. I am trying to predict students success based on the number of years in high school. Please see below clarification of my initial post. Number of years is dichotomized 1 = 3years and 2= 4 years. The response variable is ordered based on the 1 to 5, with 5 being the highest level of success. I also have type of school attended which is dichotomized top tier = 1 and lower tier = 2. I also have gender, male and female. The sample size is 56000 representing 24 schools, which randomly selected from 400 schools. So this is what I have:

Predictor = number of years (coded 1 and 2)

Outcome = Admissibility to university (coded 1 – 5)

Gender = (female and Male, coded 1 and 2)

School Type (A and B, coded 1 and 2)

Run Ordinal Logistics see SAS code and output below:

ODS LISTING CLOSE;

ods graphics on;

ODS RTF FILE = '\\Client\C$\SHS_DATA_CURRENT\LOGIT_YEARS.RTF';

proc logistic data=SHS plots(only)=(effect(polybar)oddsratio(range=clip))DESCENDING;

class YEARS(param=ref ref= "4YEARS");

WHERE YEARS NE ('3N4YEARS');

model ACCEPT=YEARS / SCALE=NONE AGGREGATE covb;

oddsratio YEARS ;

ODDSRATIO ACCEPT;

OUTPUT OUT=PREDICTED2 PRED=PRED;

title PREDICTING STUDENT ADMISSIBILITY TO UNIV. BASED ON YEARS IN HIGH SCHOOL;

run;

ods rtf close;

ods graphics off;

ods listing;

The Problem I am having is the PO assumption is not held ---please see my print out.

I have also tried other suggested techniques such as empirical test of parallelism of my variables – see sample code below and out on document

proc freq data=SHS;

table ACCEPT*YEARS / out=os;

WHERE NOT MISSING(ACCEPT);

run;

PROC SORT DATA = OS;

BY YEARS;

RUN;

proc transpose data=os(WHERE=(YEARS NE '3N4YEARS')) out=tran;

by YEARS; var count;

run;

data a; set tran;

const=0;

c1=log((sum(of col1-col1)+const)/(sum(of col2-col5)+const));

c2=log((sum(of col1-col2)+const)/(sum(of col3-col5)+const));

c3=log((sum(of col1-col3)+const)/(sum(of col4-col5)+const));

c4=log((sum(of col1-col4)+const)/(sum(of col5-col5)+const));

run;

ODS RTF FILE = '\\Client\C$\SHS_DATA_CURRENT\LOGIT_YEARS.RTF';

TITLE 'EMPIRACAL PLOTS OF ACCEPT ON YEARS';

proc sgplot;

series y=c1 x=YEARS;

series y=c2 x=YEARS;

series y=c3 x=YEARS;

series y=c4 x=YEARS;

yaxis values=(-6 to 6);

xaxis integer;

run;

My question what do I do next. Is this the end of my analysis. Should I continue interpreting my results based on the fact that the lines are parallel?

Any help will be appreciated.

Re: Follow-up to previous question

mkeintz — Sun, 12 Mar 2017 20:26:04 GMT

Instead of "follow-up to previous question", how about making a subject line that actually describes the question?

topic Follow-up to previous question in Statistical Procedures

Follow-up to previous question

Re: Follow-up to previous question