Hi Allan,
I don't find either it is a sensible idea to try to make inference on
correlation based only on four points. I simply showed using simulation
that Paaveen's results are not a contradiction.
Daniel
On Sun, Jul 29, 2018 at 6:49 AM, Allan Reese <[log in to unmask]>
wrote:
> This is a great example of Mathsworld. That's the mindset that goes: "I'm
> in a maths class/here are some numbers/I am *required* to do a sum/there is
> a single right answer/any other answer is *wrong*."
>
> It is literally (in the literal sense) not sensible to correlate and infer
> from four data points, and a disservice to Paaveen to do so.
>
> Allan
>
> Date: Sat, 28 Jul 2018 11:40:21 -0300
> From: Daniel Molinari <[log in to unmask]>
>
>> Subject: Re: correlation coefficient
>>
>> If the four points are randomly selected such that 0 <= X, Y <= 10, a
>> simulation run 10000000 times yields
>> P(r > 0.89) = 0.056 and P(abs(r) > 0.89) = 0.111. These outcomes agree
>> with Paaveen's results.
>>
>> Daniel
>>
>>
>>
>> On Tue, Jul 24, 2018 at 1:57 PM, paaveen jeyaganth <[log in to unmask]>
>> wrote:
>> i have 4 data point i did a pearson correlation end up with
>> r= 0.8919
>> p= 0.1081
>> why is that it's not significant since it's high correlation 0.89
>> because of sample size??
>>
>> Thanks
>> Paaveen
>>
>>
> You may leave the list at any time by sending the command
>
> SIGNOFF allstat
>
> to [log in to unmask], leaving the subject line blank.
>
You may leave the list at any time by sending the command
SIGNOFF allstat
to [log in to unmask], leaving the subject line blank.
|