KR-20 help needed

[jhogan1]

I'm hoping someone can help me. We are using a program called Parscore with our students. It calculates a KR-20 score; but, I want to verify it.

I'm unfortunately not to fluent in statistics and the formulas I find while googling aren't to helpful to me.

I'd like to be able to plug in the student responses to a database/spreadsheet and have a formula that can verify the KR-20 for me.

Is this something anyone can help me with?

 

[Masteras]

The questions must be in yes or no answer/wrong or right. Try this page: http://www.eflclub.com/elvin/publications/2003/Elvinsoutput.xls and here:
http://www.gifted.uconn.edu/Siegle/...ility and Validity/reliabilitycalculator2.xls

 

[jhogan1]

The questions must be in yes or no answer/wrong or right. Try this page: http://www.eflclub.com/elvin/publications/2003/Elvinsoutput.xls and here:
http://www.gifted.uconn.edu/Siegle/...ility and Validity/reliabilitycalculator2.xls

Wow....thanks. That's what I've been looking for.

 

[jhogan1]

this is proving to be truly interesting to me.

We gave 3 versions of the test.

Version's B and C have .69 and .70 KR20 scores. Not too uncommon. Version "A" came out with a .00 KR20.

All 3 combined came up with a .55

So my problem apparently is with Version A.

I just don't know if that is an anomaly enough in and of itself to be concerned.