Monday, January 30, 2012

How to calculate percentage of standard deviation between 2 values?

how to calculate percentage of standard deviation between 2 values for example, z=0 and z=1How to calculate percentage of standard deviation between 2 values?
Ah! I think what you are asking is how to calculate the probabilities under the normal curve between two z-scores.



Well, there are two fundamental ways:

1) calculator

2) table



I like the table approach myself.



1) First, with a calculator, I've used TI calculators. They have a function that calculates this.

On my TI-83, just to the left of the [Clear] button, there is the [Vars] button, and above that in yellow is DISTR



Type [2nd] -%26gt; [Vars] (which accesses DISTR), select normalcdf(



Then use the command with the following format:

normalcdf(z1, z2), where z1 is the smaller z-value.

For example:

normalcdf(0,1)

Or, normalcdf(-2.11,2.04)

etc.



2) Using a table.

There are commonly two types of z tables:

i) CUMULATIVE z tables

ii) z tables showing probability from the center (z = 0)



i) To use a cumulative z-table, look up the probability of the right-most z-value, and subtract the probability of the left-most z-value.

So for, P(0 %26lt; z %26lt; 1), look up 1 and 0.

P(z %26lt; 1) = .8413

P(z %26lt; 0) = .5000

Subtract to get .3413.



ii) To use a standard z table, to find the probability between two z-values, if the first one is 0, then simply look up the other value in the table.

P(0 %26lt; z %26lt; 1) = .3413



Suppose you wanted P(-1.96 %26lt; z %26lt; 1)

You would look up the probabilities from both numbers and ADD.



For both numbers being positive, P(.55 %26lt; z %26lt; 1.22), look up both numbers and subtract the smaller probability from the larger one.



If both numbers are positive, this is analogous to the previous situation.



For a one-tailed lookup, e.g. P(z %26lt; -2)

look up the probability and subtract it from .5.



----------



A more targeted and complete answer would have been possible if you told me what tools you had at your disposal.

No comments:

Post a Comment