Welcome Guest (Log in)

- Author
- Rating
- Type
- Revision
- Downloaded
- Updated

- Alistair Campbell
- Stack
- 6
- 898 Times
- 05 December 2009

Calculate z-score or Q(1-P)

2009-12-05

Fixed a global declaration problem which made the whole thing not work. Should be all fine now.

----------------------------------------------------------------------------------------------

This stack is an adaptation of a set of javascript functions developed by John Walker (http://www.fourmilab.ch/rpkp/experiments/analysis/zCalc.html). They enable the calculation of a z-score for a standard normal distribution given the probability (area under the normal curve) or calculation of a probability (Q) given just the z-score.

I am developing an application to calculate whether differences between subtests on psychological tests (neuropsych mainly) are significant (a la John Crawford's work on the TEA). These functions are necessary to generate tables of differences for different levels of significance.

It's fairly specialised but relevant to people who might be doing programming of statistical routines.

Fixed a global declaration problem which made the whole thing not work. Should be all fine now.

----------------------------------------------------------------------------------------------

This stack is an adaptation of a set of javascript functions developed by John Walker (http://www.fourmilab.ch/rpkp/experiments/analysis/zCalc.html). They enable the calculation of a z-score for a standard normal distribution given the probability (area under the normal curve) or calculation of a probability (Q) given just the z-score.

I am developing an application to calculate whether differences between subtests on psychological tests (neuropsych mainly) are significant (a la John Crawford's work on the TEA). These functions are necessary to generate tables of differences for different levels of significance.

It's fairly specialised but relevant to people who might be doing programming of statistical routines.

Your Comment

Click here to log in and post a comment