I am looking for studies or meta-studies in which it has been systematically explored how two presented pieces of music (two rhythms, two melodies, two chords) may deviate such that test persons do perceive them as essentially the same, i.e. "recognize" them (without having to be able to name them)?

The deviations may be simple and intra-modal:

 - minor vs. major intervals (in melodies and chords)
 - otherwise slightly deviating intervals
 - transposition (of melodies and chords)
 - speed (of melodies and rhythms)
 - single vs. double dotting (in rhythms)

Or the deviations may be more complex and cross-modal:

 - How does deviation in rhythm affect the recognition of melodies?
 - How does deviation in chords or accompaniment affect the recognition of melodies?
 - How does deviation in melody affect the recognition of rhythms?

In both cases the studies might ask: Which kinds of deviations at which crucial locations (inside the piece of music) or with which frequency of occurrence affect their recognition (by individual test persons)?

My question is a reference request: **Where do I find such a study or meta-study?** Are there "classical" studies?