Jan-paul Van Wingerden

12-22-1998, 06:26 PM

Dear BIOMCH-l readers,

Yesterday I posted a request concerning a formula to calculate

percentiles. The replies were very helpfull although the main reply is:

it cannot be done simply.

That gets me to a more basic question:

Why convert to percentiles at all?!

When calculating z, by using the formula:

z = (x-mu) / rho

you get the number of standard deviations above or below the mean.

Using the cumulative probability table you wille get the percentiles,

which is, in my opinion the same value but only in another shape.

So, why using percentiles (using tables or difficult calulation) while

the z score already gives the answer.

The only reason I can think of is that percentiles adress more to our

imagination (we have some idea to what a percentile means, while the

z-score remains more abstract)

But maybe there are more legitimate reasons for using percentiles.

Do you have any ideas on this topic?

The main reason why I'm having these questions is that I'm comparing

individual scores with populations and want to see at what level they

score (strenght, mobility, endurance etc).

By the way If anyone is interested in the replies I got please let me

know. I dont think it will be a good idea to dump all the formula's on

the discussion list right away.

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

To unsubscribe send SIGNOFF BIOMCH-L to LISTSERV@nic.surfnet.nl

For information and archives: http://isb.ri.ccf.org/biomch-l

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

Yesterday I posted a request concerning a formula to calculate

percentiles. The replies were very helpfull although the main reply is:

it cannot be done simply.

That gets me to a more basic question:

Why convert to percentiles at all?!

When calculating z, by using the formula:

z = (x-mu) / rho

you get the number of standard deviations above or below the mean.

Using the cumulative probability table you wille get the percentiles,

which is, in my opinion the same value but only in another shape.

So, why using percentiles (using tables or difficult calulation) while

the z score already gives the answer.

The only reason I can think of is that percentiles adress more to our

imagination (we have some idea to what a percentile means, while the

z-score remains more abstract)

But maybe there are more legitimate reasons for using percentiles.

Do you have any ideas on this topic?

The main reason why I'm having these questions is that I'm comparing

individual scores with populations and want to see at what level they

score (strenght, mobility, endurance etc).

By the way If anyone is interested in the replies I got please let me

know. I dont think it will be a good idea to dump all the formula's on

the discussion list right away.

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

To unsubscribe send SIGNOFF BIOMCH-L to LISTSERV@nic.surfnet.nl

For information and archives: http://isb.ri.ccf.org/biomch-l

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