Is your research not flawed already with a definition of "punch". What do you categorise as this term because I don't think I could actually describe in words what I personally attribute to "punch"?

The definition of punch was actually proposed in research before my own. Part of what we're trying to test here is the definition that was proposed in that earlier work.

In the original research, tests were done in which people were asked to manipulate kick drum envelopes, and judge music, or pink noise bursts (there were a few different tests) to determine the parameters of punch. During this research the experiment participants weren't given a definition for what punch was, but regardless of this, their results tended tended to be similar. The idea being that 'punch' is just name we give to describe a particular perceptual phenomenon, part of what we're trying to do is uncover, in terms of technical features of the signal, what controls that particular aspect of our perception. The use of the term punch comes from various other research before us defining semantics commonly used to describe audio.

It may be that when we get the results back from this current online test, where we're trying to test more people and more audio, that we see people have wildly different ideas of what punch is.

Which would render the validity of your research rather pointless if you collect data indescriminately wouldn't it? The dangers of asking questions that are not fully understandable are pretty obvious in the present political climate - we had a referendum and nobody seems to understand what the question is. Your situation seems rather similar? I'd do your survey, but as an educationalist, would need the paramaters a lot more tighter than answering questions on my assumption of what your criteria actually are? I believed PhD research to be far more quantative and qualititive in preciseness? Fair enough I'm old, but my university would never have allowed inverted commas in a proposal. The first question from my supervisor would have been - define what you mean by punch. My understanding of it is surely irrelevant, as your understanding could be very different, which would compromise the data? Like the medical ones where you'd never get away with asking if somebody felt 'unwell' - we all know what unwell means - but a tighter definition needs to be provided.

I don't mean to criticise but if the data is to be valid, it must be accurate. I see no way to align multiple examples of 'punch' into a cohesive, scientifically established component to be used in comparison.

There are several instances in literature where consensus has been seen as to the definition of punch, all of which point towards the same/similar signal features. These show when people are given the term punch as the discriminator, the judgement they make accordingly is aligned. If we provide a definition based on the signal features reported that have been found to correlate well with punch, we'd be skewing the results towards these parameters. Which, although they have been shown to be related, may not be the be-all end-all.

Thank you for your questions, these are certainly points to consider.