Classification Table

[Jason Schoeneberger]

Hi All,

 

I don’t think I’m missing something, but correct me if I’m wrong.

 

Let’s assume I want to classify students into a Pass-Fail category on an assessment using a second assessment with a continuous scale score.  Does anyone know of a way to generate a classification table if I have data for both assessments?  Maybe someone has a macro written to accomplish this.

 

What would be ideal is a table that looks like such:

 

2nd Score	Fail	Pass	Below Score - Fail	% Below Who Fail	Above Score - Fail	% Above Who Fail	Below Score - Pass	% Below Who Pass	Above Score - Pass	% Above Who Pass	Accuracy
3	3	0	5	100.0%	6407	64.2%	0	0.0%	3577	35.8%	35.9%
4	1	0	8	100.0%	6404	64.2%	0	0.0%	3577	35.8%	35.9%
5	4	0	9	100.0%	6403	64.2%	0	0.0%	3577	35.8%	35.9%
6	6	0	13	100.0%	6399	64.1%	0	0.0%	3577	35.9%	35.9%

 

 

From this, you can look to see what score can be used as an ideal cut-off to result in the most accurate classification.  If someone has something already, that’d be great.

 

Jason

[Maguin, Eugene]

Jason,

 

Perhaps somebody on the list is instantly familiar with what you are doing but i'm not. If you haven't gotten anything helpful from somebody else, would you explain your colum headings and how they are calculated or determined. And, would you explain the relationship between the first and second assessments? This might help the process along.

 

Gene Maguin

 

 >> I don’t think I’m missing something, but correct me if I’m wrong.

 

Let’s assume I want to classify students into a Pass-Fail category on an assessment using a second assessment with a continuous scale score.  Does anyone know of a way to generate a classification table if I have data for both assessments?  Maybe someone has a macro written to accomplish this.

 

What would be ideal is a table that looks like such:

 

2nd Score	Fail	Pass	Below Score - Fail	% Below Who Fail	Above Score - Fail	% Above Who Fail	Below Score - Pass	% Below Who Pass	Above Score - Pass	% Above Who Pass	Accuracy
3	3	0	5	100.0%	6407	64.2%	0	0.0%	3577	35.8%	35.9%
4	1	0	8	100.0%	6404	64.2%	0	0.0%	3577	35.8%	35.9%
5	4	0	9	100.0%	6403	64.2%	0	0.0%	3577	35.8%	35.9%
6	6	0	13	100.0%	6399	64.1%	0	0.0%	3577	35.9%	35.9%

 

From this, you can look to see what score can be used as an ideal cut-off to result in the most accurate classification.  If someone has something already, that’d be great.

 

Jason