Mazeh, Holczer, and Shporer (2015) have presented an approach that can, in principle, use the derived transit timing variation (TTV) of some transiting planets observed by the $Kepler$ mission to distinguish between prograde and retrograde motion of their orbits with respect to their parent stars rotation. The approach utilizes TTVs induced by spot-crossing events that occur when the planet moves across a spot on the stellar surface, looking for a correlation between the derived TTVs and the stellar brightness derivatives at the corresponding transits. This can work even in data that cannot temporally resolve the spot-crossing events themselves. Here we apply this approach to the $Kepler$ KOIs, identifying nine systems where the photometric spot modulation is large enough and the transit timing accurate enough to allow detection of a TTV-brightness-derivatives correlation. Of those systems five show highly significant prograde motion (Kepler-17b, Kepler-71b, KOI-883.01, KOI-895.01, and KOI-1074.01), while no system displays retrograde motion, consistent with the suggestion that planets orbiting cool stars have prograde motion. All five systems have impact parameter $0.2lesssim blesssim0.5$, and all systems within that impact parameter range show significant correlation, except HAT-P-11b where the lack of a correlation follows its large stellar obliquity. Our search suffers from an observational bias against detection of high impact parameter cases, and the detected sample is extremely small. Nevertheless, our findings may suggest that stellar spots, or at least the larger ones, tend to be located at a low stellar latitude, but not along the stellar equator, similar to the Sun.