Three input modalities for manipulation techniques in Mobile Augmented Reality have been compared. The first one employs only multi-touch input. The second modality uses the movements of the device. Finally, the third one is a hybrid approach based on a combination of the two previous modalities. A user evaluation (N=12) on a 6 DOF docking task suggests that combining multi-touch input and device movement offers the best results in terms of task completion time and efficiency. Nonetheless, using solely the device is more intuitive and performs worse only in large rotations. Given that mobile devices are increasingly supporting movement tracking, the presented results encourage the addition of device movement as an input modality.