Would you like to inspect the original subtitles? These are the user uploaded subtitles that are being translated:
1
00:00:00,004 --> 00:00:03,001
- [Instructor] When you analyze your data using statistics,
2
00:00:03,001 --> 00:00:06,007
you should know what aspect of your data you want to test.
3
00:00:06,007 --> 00:00:08,004
The statement that embodies your test
4
00:00:08,004 --> 00:00:11,003
is called a hypothesis.
5
00:00:11,003 --> 00:00:13,006
Your hypothesis is an educated guess
6
00:00:13,006 --> 00:00:15,008
about the characteristics of a data set
7
00:00:15,008 --> 00:00:18,000
and the circumstance it describes.
8
00:00:18,000 --> 00:00:21,002
For example, you could create a hypothesis that says
9
00:00:21,002 --> 00:00:22,008
the state a customer lives in
10
00:00:22,008 --> 00:00:24,008
is related to the amount of olive oil
11
00:00:24,008 --> 00:00:27,000
they order from your company.
12
00:00:27,000 --> 00:00:29,003
There are two parts of a hypothesis
13
00:00:29,003 --> 00:00:32,001
used for hypothesis testing.
14
00:00:32,001 --> 00:00:34,001
The first is the null hypothesis,
15
00:00:34,001 --> 00:00:38,000
which simply says that factor A has no effect on B.
16
00:00:38,000 --> 00:00:39,006
In the example I gave earlier,
17
00:00:39,006 --> 00:00:42,003
it would say that the state a customer lives in
18
00:00:42,003 --> 00:00:46,001
has no effect on how much they spend at my store.
19
00:00:46,001 --> 00:00:50,004
The alternative hypothesis says factor A affects B.
20
00:00:50,004 --> 00:00:52,008
So the state that a customer lives in
21
00:00:52,008 --> 00:00:56,001
does affect how much they spend at my store.
22
00:00:56,001 --> 00:00:57,003
An alternative hypothesis
23
00:00:57,003 --> 00:01:00,002
can be directional or non-directional.
24
00:01:00,002 --> 00:01:03,009
A directional hypothesis looks at either a value
25
00:01:03,009 --> 00:01:07,002
that is greater than or less than a target.
26
00:01:07,002 --> 00:01:10,000
For example, customers who were part
27
00:01:10,000 --> 00:01:12,001
of a specific marketing group
28
00:01:12,001 --> 00:01:14,007
might spend more than the average
29
00:01:14,007 --> 00:01:19,000
based on the effect of that advertising,
30
00:01:19,000 --> 00:01:21,007
whereas customers who are previous customers
31
00:01:21,007 --> 00:01:25,002
but had not ordered for a while would spend less.
32
00:01:25,002 --> 00:01:28,005
A non-directional alternative hypothesis
33
00:01:28,005 --> 00:01:32,005
looks for a comparison to a specific value.
34
00:01:32,005 --> 00:01:35,005
You might say that 30% of customers who ordered one product
35
00:01:35,005 --> 00:01:37,000
would order another.
36
00:01:37,000 --> 00:01:41,007
And you're looking for a variation either above or below.
37
00:01:41,007 --> 00:01:45,000
So it can be both instead of just one.
38
00:01:45,000 --> 00:01:48,006
So, how do you create an effective alternative hypothesis?
39
00:01:48,006 --> 00:01:49,008
Well, first, you need to state
40
00:01:49,008 --> 00:01:52,006
that there is a relationship between the variables
41
00:01:52,006 --> 00:01:54,002
and then base your alternative
42
00:01:54,002 --> 00:01:56,001
on your knowledge of the world.
43
00:01:56,001 --> 00:01:59,004
For example, if you believe that your advertising campaign
44
00:01:59,004 --> 00:02:00,009
will make a difference in sales,
45
00:02:00,009 --> 00:02:02,009
you need to state that.
46
00:02:02,009 --> 00:02:05,007
You should express your hypothesis simply and briefly
47
00:02:05,007 --> 00:02:09,004
and also make sure you can test your alternative hypothesis.
48
00:02:09,004 --> 00:02:11,006
Your hypothesis needs to be based on data
49
00:02:11,006 --> 00:02:12,009
that you are able to collect
50
00:02:12,009 --> 00:02:15,000
and that you believe should be reliable,
51
00:02:15,000 --> 00:02:18,000
so that you can perform an accurate test.
4035
Can't find what you're looking for?
Get subtitles in any language from opensubtitles.com, and translate them here.