correlation between independent samples

[Tom Salvitti]

Hi to everybody.
I have a dubt.
I have two indipendent sample...Is it correct, to see if there is a correlation, to use the command below




pwcorr [varlist] [if] [in] [weight] [, pwcorr_options]



Thanks a million to everybody

[Carlo Lazzaro]

Tom:
correlation is between variables.
What are you searching for, exactly?

[Tom Salvitti]

A parameter was measured on two independent samples. To see if there is a relationship between these two groups (children and adults) is it correct to use the command described above? Thanks a lot

[Tom Salvitti]

Is it correct ???

[Carlo Lazzaro]

Tom:
if the parameter is continuous, you my want to consider:

. use "C:\Program Files\Stata17\ado\base\a\auto.dta"
(1978 automobile data)

. correlate price foreign
(obs=74)

             |    price  foreign
-------------+------------------
       price |   1.0000
     foreign |   0.0487   1.0000


. regress price i.foreign

      Source |       SS           df       MS      Number of obs   =        74
-------------+----------------------------------   F(1, 72)        =      0.17
       Model |  1507382.66         1  1507382.66   Prob > F        =    0.6802
    Residual |   633558013        72  8799416.85   R-squared       =    0.0024
-------------+----------------------------------   Adj R-squared   =   -0.0115
       Total |   635065396        73  8699525.97   Root MSE        =    2966.4

------------------------------------------------------------------------------
       price | Coefficient  Std. err.      t    P>|t|     [95% conf. interval]
-------------+----------------------------------------------------------------
     foreign |
    Foreign  |   312.2587   754.4488     0.41   0.680    -1191.708    1816.225
       _cons |   6072.423    411.363    14.76   0.000     5252.386     6892.46
------------------------------------------------------------------------------

. test 1.foreign

 ( 1)  1.foreign = 0

       F(  1,    72) =    0.17
            Prob > F =    0.6802





.

[Tom Salvitti]

Dear, thank you very much. You are very kind. I'll explain the problem better. I have two samples (mice). The first sample (children) consists of three types of children (Typical, HR, LR). The second sample (adult) is made up of only one group. Vocalizations (continuous variable) were measured on these groups. To see the type of relationship between these two groups (children, adult), is it correct to use the commands I wrote earlier? IS IT CORRECT to correlate (Children Typical and adult), (Children HR and adult) and (children LR and adult)? or what do you recommend? Many thanks in advance.

[Tom Salvitti]

I don't know Carlo Lazzaro if my explanation was clear...I Hope so... Thanks a million in advance Tommaso

[Tom Salvitti]

What do you think??

[Carlo Lazzaro]

Tommaso:
thanks for claryfing.
-pwcorr- does not allow categorical variables.
I would feel more comfortable with -regress-:

. use "C:\Program Files\Stata17\ado\base\a\auto.dta"
(1978 automobile data)

. g vocalization=price

. g patients=rep78
(5 missing values generated)

. label define patients 2 "children_typical" 3 "children_HR" 4 "children_LR" 5 "adults"

. label val patients patients

. regress vocalization i.patients

      Source |       SS           df       MS      Number of obs   =        69
-------------+----------------------------------   F(4, 64)        =      0.24
       Model |  8360542.63         4  2090135.66   Prob > F        =    0.9174
    Residual |   568436416        64     8881819   R-squared       =    0.0145
-------------+----------------------------------   Adj R-squared   =   -0.0471
       Total |   576796959        68  8482308.22   Root MSE        =    2980.2

-----------------------------------------------------------------------------------
     vocalization | Coefficient  Std. err.      t    P>|t|     [95% conf. interval]
------------------+----------------------------------------------------------------
         patients |
children_typical  |   1403.125   2356.085     0.60   0.554    -3303.696    6109.946
     children_HR  |   1864.733   2176.458     0.86   0.395    -2483.242    6212.708
     children_LR  |       1507   2221.338     0.68   0.500    -2930.633    5944.633
          adults  |     1348.5   2290.927     0.59   0.558    -3228.153    5925.153
                  |
            _cons |     4564.5   2107.347     2.17   0.034     354.5913    8774.409
-----------------------------------------------------------------------------------

. mat list e(b)

e(b)[1,6]
           1b.         2.         3.         4.         5.          
     patients   patients   patients   patients   patients      _cons
y1          0   1403.125  1864.7333       1507     1348.5     4564.5

. test 2.patients=5.patients

 ( 1)  2.patients - 5.patients = 0

       F(  1,    64) =    0.00
            Prob > F =    0.9687

[Tom Salvitti]

I probably haven't explained myself. Vocalization was measured on these two groups (group 1 adult; group 2 HR_CHILDREN, LR_CHILDREN, TYPICAL_CHILDREN). IS IT CORRECT to correlate first (Children Typical vocalization and adult vocalization) then (Children HR vocalization and adult vocalization) and at the end (children LR vocalization and adult vocalization) using pwcorr? The variables vocalization is numeric.. Many thanks in advance Carlo Lazzaro

[Tom Salvitti]

Now, it's clearer Carlo Lazzaro
thanks a million in advance

Why do you say it's categorical???

Best Tommaso

[Carlo Lazzaro]

Tommaso:
you have one categorical predictor (i.e., different groups of patients).

[Tom Salvitti]

Thanks a lot @CarloLazzaro

[Tom Salvitti]

Dear Carlo Lazzaro thanks a lot...I ve stata 16...Is it the same???
How can we interprate the output?
Using your example, what can we say about the relationship between these different groups of patients?
Thanks a million in advance...

[Carlo Lazzaro]

Tommaso:
my previous -test- example tells you that we cannot reject the null that there's no difference in -vocalization- when -children_typical- and -adults- patients are compared.
You can go on comparing the other groups of patients in the very same way.
See also -contrast-, -margin- and -marginsplot-.
As an unsolicited advice (that you can safely discard), please note that these issues are covered in any decent textbook on (bio)statistics and fully explained in Stata .pdf manual.