 
| Content-Transfer-Encoding: |
7bit |
| Sender: |
|
| Subject: |
|
| From: |
|
| Date: |
Wed, 30 Oct 2002 11:14:08 EST |
| Content-Type: |
text/plain; charset="US-ASCII" |
| MIME-Version: |
1.0 |
| Reply-To: |
|
| Parts/Attachments: |
|
|
In a message dated 10/30/02 6:54:35 AM, [log in to unmask] writes:
>But you have a 5% chance of being wrong. I'm gearing up for my comps this
>Friday, and of course one of the questions will be on Biostatistics.
>
>beth >t
Only if you are using inferential statistics. Most stats Kyle was talking
about are descriptive stats. And, (this is a technical point for Beth the OT
for her Biostat question) the 5% relates to the level of chance. For example,
if I find the mean salary difference between groups are statistically
significant, that indicates only that the difference in the two salaries
could be found by chance only 5% of the time. An infinitesimal distinction,
but important, nonetheless.
Another example is the almost perfect correlation between ministers
salary in New York State and wine production in California.
Bobby
|
|
|
