Hi mantow,
Every time you change the phase offset you should see a discontinuity in the output when viewed with an oscilloscope. If you change the phase offset by 180 degrees you should see the output signal get inverted. Any other phase offset change I think will be hard to determine without synchronizing the phase changes so they occur at a the peaks or zero-crossings in the output. The RAM has a feature that can synchronize the changes if you want to test that instead. See page 35 in the datasheet, 'RAM Direct Switch Mode with Zero Crossing'.
If your oscilloscope can capture data and store to a CSV file you could do a fair bit of math to determine the phase changes. Also, if you can synchronize two DDS parts that would also make it much easier to tell the phase difference between the two.
