[R] Is there a R command for testing the difference of two liear regressions?

[David Winsemius]

On Jan 27, 2012, at 4:10 PM, Michael wrote:

> I changed the notation for data from x to z...
>
> That's it. Should be very clear now... Thanks!
>
> Data: z1, z2, ..., z_{n+1}
>
> y1 = z_1,z_2,.........  z_n
> y2 = z_2, z_3,......... z_{n+1}
>
> x1 = 1, ..., n
> x2 = 1, ..., n
>
> y1 = A1+ x1 * B1 + epsilon_1
> y2 = A2 + x2 * B2 + epsilon_2
>
> H0: B1 and B2 are statistically significally different...

So in hopes of clarifying, ....So you want to test whether estimated  
slopes are different after you slide a data-window one unit to the  
right on the y-scale. Are you willing to say anything else about the  
mathematical properties of Y? is it for instance measured at equal  
time intervals?

-- 


>
>
>
> On Fri, Jan 27, 2012 at 2:41 PM, Mark Leeds <markleeds2 at gmail.com>  
> wrote:
>
>> now i'm confused because you first use y_1, y_2 and then use y  
>> later. I
>> would take
>> a look at that earlier paper i mentioned. I think it's along the  
>> lines of
>> what you want. Unfortunately. I don't have a computer copy of it. I  
>> got it
>> from a library service where I once worked.
>>
>>
>> mark
>>
>>
>> On Fri, Jan 27, 2012 at 3:35 PM, Michael <comtech.usa at gmail.com>  
>> wrote:
>>
>>> Thanks all.
>>>
>>> Here are a more clear statement of my question:
>>>
>>> Data: z1, z2, ..., z_{n+1}
>>>
>>> y1 = z_1,z_2,.........  z_n
>>> y2 = z_2, z_3,......... z_{n+1}
>>>
>>> x1 = 1, ..., n
>>> x2 = 1, ..., n
>>>
>>> y = A1+ x1 * B1 + epsilon_1
>>> y = A2 + x2 * B2 + epsilon_2
>>>
>>> H0: B1 and B2 are statistically significally different...
>>>
>>> Any more thoughts?
>>>
>>> Thanks  a lot!
>>>
>>> On Fri, Jan 27, 2012 at 1:39 PM, Mark Leeds <markleeds2 at gmail.com>  
>>> wrote:
>>>
>>>> Hi Richard: I read michael's question as meaning that he says two
>>>> univariate no intercept
>>>> regression model where the predictor data is different in each  
>>>> model so
>>>> that
>>>>
>>>> x1 = x_11,x_12,.........  x_1n
>>>> x2 = x_21, x_22,......... x_2n
>>>> y = y_1, .....y_n
>>>>
>>>> y = x1 * B1 + epsilon_1
>>>> y = x2 * B2 + epsilon_2
>>>>
>>>> and he wants to see which coefficient ( B1 or B2 ) "works"  
>>>> better. But I
>>>> could be wrong
>>>> which I only realized after reading your recommendation. michael:  
>>>> if i'm
>>>> wrong, then disregard the paper reference that I sent earlier.
>>>>
>>>>
>>>> Mark
>>>>
>>>>
>>>>
>>>>  On Fri, Jan 27, 2012 at 2:29 PM, Richard M. Heiberger <rmh at temple.edu 
>>>> >wrote:
>>>>
>>>>> It looks like you might be asking for the anova() on two models.
>>>>>
>>>>> M1 <- lm(y ~ x1 + x2 + x3, data=something)
>>>>> M2 <- lm(y ~         x2 + x3, data=something)
>>>>> anova(M1, M2)
>>>>>
>>>>> Please send a reproducible example to the list if more detail is  
>>>>> needed.
>>>>>
>>>>> Rich
>>>>>
>>>>> On Thu, Jan 26, 2012 at 11:59 PM, Michael <comtech.usa at gmail.com>
>>>>> wrote:
>>>>>
>>>>>> Hi al,
>>>>>>
>>>>>> I am looking for a R command to test the difference of two linear
>>>>>> regressoon betas.
>>>>>>
>>>>>> Lets say I have data x1, x2...x(n＋1).
>>>>>> beta1 is obtained from regressing x1 to xn onto 1 to n.
>>>>>>
>>>>>> beta2 is obtained from regressing x2 to x(n＋1) onto 1 to n.
>>>>>>
>>>>>> Is there a way in R to test whether beta1 and beta2 are  
>>>>>> statistically
>>>>>> different?
>>>>>>
>>>>>> Thanks a lot!
>>>>>>
>>>>>>       [[alternative HTML version deleted]]
>>>>>> .

David Winsemius, MD
West Hartford, CT