Kuder-Richardson #21 Formula for estimate test reliability
 

Good Morning!

I am trying to figure out how to adapt the Kruder-Richardson Formula
#21 to excel. First of all this is not a homework assignment.
Secondly, I appologize for linking a file but I could not get the
fonts for the formula right. What the formla does is to estimate test
reliability. I can figure it out long hand but would greatly
appreciate it in excel thanks in advance.

R=( k) (1- X(k-X))
----- -----------
k-1 kS

where
R= test reliability
k = number of items on the test
X = mean of raw scores from the total test
S = vfariance from the raw scores of the total test

Suppose we had a 50 item test and the mean score was 43 and the
variance was 25.0. We would have

R= (50) (1- 43(50 - 43))
----- --------------
50-1 (50)(25)

R = (1.02) (1- 301)
----
1250

R = (1.02) (0.76) = 0.78


Thank you in advance.

http://www.savefile.com/projects/808560688

Answer: Kuder-Richardson #21 Formula for estimate test reliability
 

Hi there!

No problem, I can definitely help you adapt the Kuder-Richardson Formula #21 to Excel. Here are the steps:

1. Open a new Excel spreadsheet and enter the following labels in cells A1 to A4: "k", "X", "S", and "R".
2. In cell B1, enter the number of items on the test (
   Code:

   ---

   k

   ---

   ). In your example, this would be
   Code:

   ---

   50

   ---

   .
3. In cell B2, enter the mean of raw scores from the total test (
   Code:

   ---

   X

   ---

   ). In your example, this would be
   Code:

   ---

   43

   ---

   .
4. In cell B3, enter the variance from the raw scores of the total test (
   Code:

   ---

   S

   ---

   ). In your example, this would be
   Code:

   ---

   25.0

   ---

   .
5. In cell B4, enter the following formula:
   Code:

   ---

   =(B1*(1-B2*(B1-B2)))/((B1-1)*B3*B2)

   ---

6. Press Enter, and the result should be the test reliability (
   Code:

   ---

   R

   ---

   ). In your example, this would be 0.78.

That's it! You can now use this formula for any test with different values for , , and . Let me know if you have any questions or need further assistance.

Kuder-Richardson #21 Formula for estimate test reliability
 

One way:

=(k/(k-1)) * (1 - (X*(k-X))/(k*S))

You can either name your cells k, X, and S, or replace the k,X,S in the
formula with cell references.

In article . com,
wrote:

Good Morning!

I am trying to figure out how to adapt the Kruder-Richardson Formula
#21 to excel. First of all this is not a homework assignment.
Secondly, I appologize for linking a file but I could not get the
fonts for the formula right. What the formla does is to estimate test
reliability. I can figure it out long hand but would greatly
appreciate it in excel thanks in advance.

R=( k) (1- X(k-X))
----- -----------
k-1 kS

where
R= test reliability
k = number of items on the test
X = mean of raw scores from the total test
S = vfariance from the raw scores of the total test

Suppose we had a 50 item test and the mean score was 43 and the
variance was 25.0. We would have

R= (50) (1- 43(50 - 43))
----- --------------
50-1 (50)(25)

R = (1.02) (1- 301)
----
1250

R = (1.02) (0.76) = 0.78


Thank you in advance.

http://www.savefile.com/projects/808560688

Kuder-Richardson #21 Formula for estimate test reliability
 

Hello,

If you google for "Kuder-Richardson 21 Formula excel" then you will
find
http://www.gifted.uconn.edu/siegle/r...eliability.htm
which links to a [IMHO] nice sample spreadsheet
http://www.gifted.uconn.edu/siegle/r...alculator2.xls

Some things are easier copied than re-invented.

Regards,
Bernd

Kuder-Richardson #21 Formula for estimate test reliability
 

A caveat -

There are tens of thousands of workbooks out there that can be copied. A
significant fraction of them, including ones posted by academics, have
errors or have poor boundary event error checking.

Validating someone else's worksheet for a simple calculation like this
can often be *more* difficult than building from scratch...


In article . com,
Bernd P wrote:

Some things are easier copied than re-invented.

Kuder-Richardson #21 Formula for estimate test reliability
 

On Nov 4, 10:48 am, JE McGimpsey wrote:
A caveat -

There are tens of thousands of workbooks out there that can be copied. A
significant fraction of them, including ones posted by academics, have
errors or have poor boundary event error checking.

Validating someone else's worksheet for a simple calculation like this
can often be *more* difficult than building from scratch...

In article . com,
Bernd P wrote:



Some things are easier copied than re-invented.- Hide quoted text -

- Show quoted text -

Thank you both!