matlab - use watts in random signal generation

Question

i would like to understand what is a different from electrical point of view in generation of two kind of random noise,one let say white noise and second Gaussian random signal,let us consider following two code

 k=randn(10,1)

k =

   -1.3499
    3.0349
    0.7254
   -0.0631
    0.7147
   -0.2050
   -0.1241
    1.4897
    1.4090
    1.4172



 mean(k)

ans =

    0.7049

and std is

 std(k)

ans =

    1.2191

first of all i have tried to convert it to white noise,so did following thing

k=k-mean(k)

k =

   -2.0548
    2.3300
    0.0205
   -0.7679
    0.0098
   -0.9099
   -0.8290
    0.7848
    0.7041
    0.7123

>> mean(k)

ans =

     0

and also

k=k./std(k);

 std(k)

ans =

    1.0000

>> mean(k)

ans =

 -1.1102e-017

i have got one white noise,now i am concerning to the following noise

y1 = wgn(10,1,2);



 y1 = wgn(10,1,2)

y1 =

   -2.0118
   -0.7443
   -0.6304
   -0.9785
   -1.1116
   -0.3348
   -0.4021
   -2.6500
    3.3511
   -1.1340


       y1=y1-mean(y1);
>> y1=y1./std(y1);
>> [mean(y1) std(y1)]

ans =

    0.0000    1.0000
   i got another  noise,but  it says that For example:

the command below generates a column vector of length 10 containing real white Gaussian noise whose power is 2 dBW. The function assumes that the load impedance is 1 ohm. i have used this source and change 50 with 10

http://matlabtutorial.blogspot.com/2010/12/matlab-tutorial-for-generating-white.html what is a role of impendance and watt here?how circuit works when we are giving him to such command?i have used plot to plot these two noise and get following picture

first one is white noise,second one Gaussian noise.please help me