 |
|
| Science Forum Index » Space - Consult Forum » probability question... |
|
Page 1 of 1 |
|
| Author |
Message |
| Nikolay... |
 Posted: Wed Sep 30, 2009 4:55 am |
|
|
|
Guest
|
Hello,
Here is my problem :
If a man has 10% chance to be laid off during the year 2010, what is
the chance he will be laid off from 1st to 31st March 2009?
I make the hypothesis than the rate of 10% is constant during the
months of the year. However, when using the binomial distribution I
have the following results for each month:
Pourcentage Month
10% Jan
9% Feb
8% Mar
7% Apr
7% May
6% Jun
5% Jul
5% Aug
4% Sep
4% Oct
3% Nov
3% Dec
28% Still employed
This does not correspond to 10% for the entire year, it is much more
(1-0.2 . Moreover, the percentages are obviously not equal month
compared to another. I do not thing the problem should be treated as a
coin toss. Should I divide 10% by 12 to have equal probability?
Thanks in advance for you help.
Nik |
|
|
| Back to top |
|
|
|
| Rich Ulrich... |
| Posted: Wed Sep 30, 2009 12:43 pm |
|
|
|
Guest
|
On Wed, 30 Sep 2009 07:55:46 -0700 (PDT), Nikolay
<nikolay at (no spam) nikolov-consulting.ch> wrote:
[quote:0ceb7b6484]Hello,
Here is my problem :
If a man has 10% chance to be laid off during the year 2010, what is
the chance he will be laid off from 1st to 31st March 2009?
I make the hypothesis than the rate of 10% is constant during the
months of the year. However, when using the binomial distribution I
have the following results for each month:
Pourcentage Month
10% Jan
9% Feb
8% Mar
7% Apr
7% May
6% Jun
5% Jul
5% Aug
4% Sep
4% Oct
3% Nov
3% Dec
28% Still employed
This does not correspond to 10% for the entire year, it is much more
(1-0.2.
[/quote:0ceb7b6484]
It makes plenty of sense to me, the fact that 10% attrition
per month does not equal 10% attrition for a year.
Consider: Monthly Survivorship (S)
You seem to want (1-S), where
0.90= S^12
[quote:0ceb7b6484]Moreover, the percentages are obviously not equal month
compared to another. I do not thing the problem should be treated as a
coin toss. Should I divide 10% by 12 to have equal probability?
[/quote:0ceb7b6484]
Well, if there is *constant* attrition as a rate applied to
a sample, there will have to be a *decreasing* number (N)
that are attrited, when the sample becomes smaller.
Yeah, 12 comes in there somewhere, if you start with Months.
--
Rich Ulrich |
|
|
| Back to top |
|
|
|
| Nikolay... |
| Posted: Wed Sep 30, 2009 9:40 pm |
|
|
|
Guest
|
Hi Rich,
thanks for your reply. The formula "0.90= S^12" is exactly what I was
looking for.
Have a nice day.
Regards,
N
On Sep 30, 8:43 pm, Rich Ulrich <rich.ulr... at (no spam) comcast.net> wrote:
[quote:79e836ab53]On Wed, 30 Sep 2009 07:55:46 -0700 (PDT), Nikolay
niko... at (no spam) nikolov-consulting.ch> wrote:
Hello,
Here is my problem :
If a man has 10% chance to be laid off during the year 2010, what is
the chance he will be laid off from 1st to 31st March 2009?
I make the hypothesis than the rate of 10% is constant during the
months of the year. However, when using the binomial distribution I
have the following results for each month:
Pourcentage Month
10% Jan
9% Feb
8% Mar
7% Apr
7% May
6% Jun
5% Jul
5% Aug
4% Sep
4% Oct
3% Nov
3% Dec
28% Still employed
This does not correspond to 10% for the entire year, it is much more
(1-0.2.
It makes plenty of sense to me, the fact that 10% attrition
per month does not equal 10% attrition for a year.
Consider: Monthly Survivorship (S)
You seem to want (1-S), where
0.90= S^12
Moreover, the percentages are obviously not equal month
compared to another. I do not thing the problem should be treated as a
coin toss. Should I divide 10% by 12 to have equal probability?
Well, if there is *constant* attrition as a rate applied to
a sample, there will have to be a *decreasing* number (N)
that are attrited, when the sample becomes smaller.
Yeah, 12 comes in there somewhere, if you start with Months.
--
Rich Ulrich- Hide quoted text -
- Show quoted text -[/quote:79e836ab53] |
|
|
| Back to top |
|
|
|
|
|
All times are GMT - 5 Hours
The time now is Wed Dec 09, 2009 5:21 pm
|
|