In teaching phonetics, I have always required students to submit a final project. This was my experience as a student studying phonetics (as an undergraduate and as a graduate student) after all. The project is a phonetic description of a language that the student is unfamiliar with. Students work with a speaker, practice their transcription skills, analyze their data, and examine some of the acoustic properties of the language.
I do phonetic description as part of my research, so I like the project idea. Yet I realize that this type of project isn't for everyone. Students often struggle with it and every semester that I teach phonetics, I get both good projects and ones which miss the mark. Among the problems that I encounter are the following:
a. Students do not understand that one must establish contrasts before you analyze the phonetic properties of the language.
Establishing contrasts requires that students have a little background in phonology, but typical phonetics courses do not require much in the way of phonology. One solution here might be to require more background before taking phonetics, but at a major public university where enrollment is a concern in higher-level courses, being more selective is sometimes not an option.
b. Students do not understand the point of spectrograms.
Students will include pages of spectrograms in a final paper with no explanation of what the images are supposed to reflect at all. I think this is a specific case of a more general issue that I will call "the instagramification of prose." The image does not speak for itself. You must guide the reader through it. Otherwise, it just occupies space. One solution to this might be to devote more time in the semester to reading the literature and writing.
c. With vowels, anything goes.
Students will produce a cursory description of the vowel system because consonants are easier for them. They might even plot an acoustic vowel space that looks extremely odd but will forge ahead and ignore the fact that it does not match their transcriptions. I don't know immediately how to solve this.
d. Bad ears. I hate to say it.
I want to encourage students to pursue projects where they analyze the phonetics of Xhosa or Danish or Zapotec. However, some students just struggle to hear phonetic contrasts. They can hear an aspirated/unaspirated contrast among stops but might not distinguish between different back vowels, e.g. [o] vs. [ɔ] or [ʊ] vs. [ɯ]. Then they choose a tough language for their project. Do you lead such students away from more phonetically difficult languages because you feel they will struggle too much or does doing so discourage such students? If you include more listening exercises in the semester and the students still do poorly on them, does this help them or hurt them?