After 5 s of presenting this map view, the computer would switch to the FFOV and provide a beep, prompting a response of either "same" or "different." The objects in the second view could differ from objects in the first view in their relative position, shape, or size.
Seven elevation angles of the FFOV were manipulated as the third factor (15 [degrees], 30 [degrees], 37 [degrees], 45 [degrees], 52 [degrees], 60 [degrees], or 75 [degrees]).
Although the experimental design using map and FFOV angles was fully crossed, the statistical design using an elevation angle disparity variable cannot be a full-factorial; because of real-world constraints, only a few map angle/FFOV angle combinations can produce large foreshortening deviations.
A subtle but critical implication of the foreshortening disparity result is the desired flight path approach angle, or the angle of the FFOV. At low elevation angles, small angular disparities result in much larger foreshortening disparities than at high elevation angles.
In a within-subjects design, three factors were manipulated: map elevation angle (45 [degrees], 60 [degrees], or 90 [degrees]), azimuth angle disparity (0 [degrees], 22 [degrees], 45 [degrees], 67 [degrees], or 90 [degrees], all rotated in the same direction from the referent), and FFOV elevation angle (30 [degrees], 45 [degrees], 60 [degrees], or 75 [degrees]).
That is, the map might not need to rotate until a critical amount of azimuth disparity between the FFOV and the map is reached.