Hello, 

I have a question and I would appreciate to know your insight.  I am new to this area and your help will be appreciated.  In DSP, we can look at the frequency response of a Moving average easily and derive the transfer function and look at its power spectrum.  Now, sometimes we use the moving average (sliding window) and then further we calculate the Pearson Correlation for example.  My question is how I can look at the frequency response of a correlation based sliding window and thus further see how it affects my signal of interest in the frequency domain?  

What if instead of the correlation in the moving average I have some other function in the time domain, and I calculate it within the sliding window.  What is the general approach and how could I possibly implement it in MATLAB.

Best regards

Just take the FFT of your function.  The convolution of two functions in the time domain is the same as the multiplication of two transforms in the frequency domain.  Look up "convolution theorem" to get the details.