Entry test for MENSA I had about 5 years ago didn't have any words or shapes of any real-world objects. There was only one kind of questions. You had to pick image that would fit in the right bottom corner of (usually) 3x3 image grid. It was all about picking apart those objects and noticing how their elements change as you move from row to row and from column to column. Me being programmer given me some advantage I think because I perceived some transitions as bitwise operations and frame by frame animations.