Re: Using standard deviation to make predictions
From: David Jones (dajxxx_at_ceh.ac.uk)
Date: 01/19/05
- Next message: Paula: "Which test to use"
- Previous message: Vimal: "Re: Using standard deviation to make predictions"
- In reply to:(deleted message) snarly: "Using standard deviation to make predictions"
- Messages sorted by: [ date ] [ thread ]
Date: Wed, 19 Jan 2005 12:03:44 -0000
snarly wrote:
> Hi group,
>
> I understand and I can calculate mean, median, variance and standard
> deviation. I've read that one can make predictions about data using
> standard deviation.
>
> How would I do this?
>
It depends on how technical you want to be.
The mean and/or median provide an indication of the "typical size" of
as-yet-unobserved values. The standard deviation and/or
mean-absolute-deviation provide an indication of the sizes of the
likely departures of as-yet-unobserved values from such a "typical
size". Thus this is a prediction about the future of a different kind
from that provided by the mean and/or median.
Given more assumptions, you could be led to statements that certain
percentages of future observations will lie within a given range about
a predicted central value.
One of the major basic assumptions here is of course that "past" is
representative of the "future".
- Next message: Paula: "Which test to use"
- Previous message: Vimal: "Re: Using standard deviation to make predictions"
- In reply to:(deleted message) snarly: "Using standard deviation to make predictions"
- Messages sorted by: [ date ] [ thread ]
Relevant Pages
|
