It's hard to tell whether it's an accelerometer from that dataset, since you appear to be testing two orientations with a two axis device.

If you want to try the following six orientations, you'll be able to tell if it's an accelerometer:

Front and back
Left edge and right edge
Bottom edge and top edge

You should notice the following patterns:
One pair of orientations will always show small values for x and y
Another pair of orientations will show a large value for x, and small value for y. The sign of x will be positive in one of the orientations, and negative in the other orientation.
Another pair of orientations will show a large value for y, and small value for x. The sign of y will be positive in one of the orientations, and negative in the other orientation.

It is possible to simplify the measuements, but this is fairly robust.