There are few studies which have explored the way paralinguistic information varies across different languages, specifically languages of tone versus languages of stress. It has been shown that speakers of tone use less F0 related cues in the production of verbal expressions of emotions than speakers of stress do. This study proposes a new method of teaching English intonation patterns, specifically those that aid in the expression of emotion, to speakers of tone languages. A total of 12 participants were tested, all who spoke English as their second language, 8 whose first language was tone based. One portion of the participants reacted to and mimicked an exaggerated group of recordings while the other portion reacted to the non-exaggerated version of these same recordings. 2 scoring methods were used during the analysis portion of our experiment. The first was a visual scoring process. Each of the participants' pitch contours were compared directly against the model’s using the Praat software. The second method was a perceptual one. During this scoring process we had 2 scorers listen to the recordings of all the participants. Both scorers indicated the emotion they perceived for statement. Our results show that the participants who responded to the exaggerated version of the model's recordings received more accurate and consistent perceptual ratings. Anger was the most difficult emotion for the participants to imitate and for both scorers to recognize. Identifying effective language learning strategies will have important implications in language teaching.