It has been shown in previous studies that haptic guidance improves the learning outcomes of handwriting motor skills. In this paper, we present a comparison between full guidance and partial guidance using a haptic learning tool which supports these two modes. The full guidance mode leads the user along a pre-recorded trajectory, whereas the partial guidance mode allows the user a free movement and provides corrective forces if they deviate significantly from the desired path. Experimental results with 22 participants demonstrated that there is no significant difference between partial haptic guidance and full haptic guidance for improving learning outcomes, while both have significantly improved the learner's performance compared to no haptics feedback. However, when the two modes are combined, partial guidance followed by full guidance yielded better overall performance. We conclude the paper by summarizing our findings and providing perspectives for future work.