Slope question

[PaulVincent1]

I am having a problem with understanding some slope data.  I downloaded ned 10 and ned 30 and created a 10m and 30m mosaic of my study area using mosaic to new raster.  I then used those mosaics to derive the slope of my study area.  I then use the Extract by Mask tool to find the slope around two points of interest for various distances up to 25 kilometers.  So far, so good the output looks fine.  However, as the mask distances get larger and as the areas begin to overlap it seems that the average of the values of slope from the ned 30 and ned 10 should trend toward an average.  Instead, they get wildly different.  That has me perplexed, so I hope someone might have a suggestion or could point me to a resource to help me understand what could be causing this variation. Many thanks.

[DanPatterson_Retired]

Is one dataset in % slope and the other in slope angle (in degrees)?

[PaulVincent1]

Both are in degrees.

[DanPatterson_Retired]

some ideas of the numbers might help and when you said overlap, there is a discrepancy.  The extract by mask will have no influence on the values obtained, and if you mosaicked the data sets together prior to running slope, then the slopes should be consistent within each grouping (ie 10 and within 20), however, the slopes at a particular spot need not be the same as you change the resolution particularly in areas of greater topographic variations.  I would examine the affect that the scale of the input files has had upon the slopes themselves and not try to combine their values between samples