My Psychology Project

[avaserfi]

I would suggest you combine the two suggestions you have in the beginning
of your thread. It will be a lil' bit more complex but will achieve better picture of a person's perception regarding audio compressions in relation to loudness. It would be advisable to use 3 different audio compressions with 3 different level of loudness. I would think it is best to subject the volunteers to listen to the lowest audio compressed then the highest and lastly the mid-audio compressed track. Give them each 3 seconds to rest their ears (i don't know why i pick 3). With simple statistics analysis this can be done pretty easily.

That would be an interesting topic, but I believe starting with dynamic music vs dynamically compressed would be ideal and perhaps expending the research later would be more ideal. Also, there has been more than enough research regarding perception of loudness as well as dynamics vs loudness in my honest opinion.

Likely the plan will be have two 25 second intervals that was originally a dynamic source one being the lossless copy and the other being dynamically compressed. Rotating which is A and B between subjects, so the lossless copy isn't always heard first further eliminating bias. I have yet to determine timing, but will likely have the participant listen to the first clip wait ten seconds then listen to the second (keeping it in one file so the timing is not done by hand). I will allow each participant to listen to the full file as many times as they would like. Once they feel they have had enough time with the piece I will give them a questionnaire which will ask various questions to help determine which is liked more and factors involved with that choice.

If anyone sees any problems with this, has questions suggestions or any comments at all let me know .

In a research like this, there should be several assumptions need to be made, such as: little or no interference from outside noise, the headphone used has linear frequency response and the audio track covers lowest and highest audible frequency and don't forget, the samples (persons who are picked to be the victims of your research) are not tone deaf...

1) Outside noise will be controlled for by using over the ear headphones and use of a quiet room
2) The headphones I am looking to purchase have an almost perfectly flat response and a calibration file will be used to ensure complete linearity
3) The tracks don't actually don't need the full range of human hearing. The project is comparing if the subjects prefer a dynamic source vs a dynamically compressed source. As long as there is a quality recording that is very dynamic especially when compared to the compressed source (think jazz or an orchestra).

PS: May be take a sample of equal number of sexes (male/female). Then you can see how different the two sexes are in this research.

I do plan on having both sexes in my study, but there is no need for the samples to be equal. If the groups are each large enough, regardless of size actual size, statistical analysis can still be done successfully.


Lastly, I am trying to come up with some good dynamic music sources to use for the study. I am thinking something that many people haven't heard will be ideal so the subjects will not have a preconceived idea of how the specific song should sound. Does anyone have suggestions? Right now I am thinking that I will use a track from Chesky records or Jazz in the Pawn Shop.

 

[zhimbo]

these suggestions you make may seem logical, but in reality add great bias to the experiment which would make it completely ungeneralizable and thus useless.

Bias? I'm not sure about that. Add noise to the data, yes, for sure, but it would only be bias if people consistently turn one sort of sample up higher than the other for some reason.

Which may in fact be true, and if the research question is really a perception question, that would be bad. But if it's more of a question of consumer preference, then I think it would even be preferable to match real world conditions - and people will adjust volume until it "sounds right" (whatever that means) in the real world.

But I suppose this is a Psychology class, and not a Market Research class.

I haven't looked at the link yet that looked at matching "perceived loudness" -seems like a tall order to me. I'm no auditory expert, but I have some experience at designing psychophysical studies (visual).

 

[avaserfi]

Bias? I'm not sure about that. Add noise to the data, yes, for sure, but it would only be bias if people consistently turn one sort of sample up higher than the other for some reason.

Which may in fact be true, and if the research question is really a perception question, that would be bad. But if it's more of a question of consumer preference, then I think it would even be preferable to match real world conditions - and people will adjust volume until it "sounds right" (whatever that means) in the real world.

But I suppose this is a Psychology class, and not a Market Research class.

I haven't looked at the link yet that looked at matching "perceived loudness" -seems like a tall order to me. I'm no auditory expert, but I have some experience at designing psychophysical studies (visual).

I understand what you mean, but the perception that louder is better is well established along side that it has been proven with strong significance that dynamic sources are perceived to be louder than less dynamic sources (done with both music and pink noise). With this knowledge it would be necessary to control for these situations as allowing a subject to change volume freely could easily add bias towards one sample versus the other as subjects could possibly turn up the dynamic source during quieter situations and possibly turn it down at others while just setting the compressed to another and just prefer the compressed due to its ease.

This is my only source I can actually provide a link to:

http://links.jstor.org/sici?sici=0022-4294(199521)43:1<22:CLJODI>2.0.CO;2-D

 

[Highlander]

I am trying to come up with a good research question.

It seems to me that the more correlation you show between variables the greater your final mark will be, albeit at the expense of an increase in project size, though the relationship between the two need not necessarily be proportionate.

It also seems to me that instead of looking for correlation and drawing conclusions only from a given sample, that by repeating the experiment at one or more times you could additionally determine if the correlations and conclusions drawn are their self consistant over time, thereby lending credibility to any original findings.

A couple of ideas for experiments:

1. Determining a relationship between extraneous noise and the level at which it consistantly impedes judgement in a comparison of compressed/uncompressed sound through open/closed headphones. Additional sampling by you of average ambient noise on, say, a trafficked street, a bus or train/tube would allow you to determine the level of compression that could be used by people in practice with little to no audible loss of sound quality.

2. Determining if and by how much visual stimuli impedes people's judgement in a comparison of compressed/uncompressed sound through headphones. Assuming you use your laptop you could run one set of tests with the candidate told to look only at a blank screen, another with a static picture on the screen (a flower, a car, a naked woman (or man - whatever floats your boat ), whatever) and yet another with the candidate required to continuously play tic-tac-toe etc. This test would doubtless highlight differences in multi-tasking ability of the sexes.

Food for thought.

 

[masak_aer]

2. (a flower, a car, a naked woman (or man - whatever floats your boat ), whatever) and yet another with the candidate required to continuously play tic-tac-toe etc. This test would doubtless highlight differences in multi-tasking ability of the sexes.

Food for thought.

Mmmm..if you were one of the appointee for the research and you were shown a naked pic of a man...and you somehow with your in depth experience tells you which compression's good/bad with good sound quality or whatnot....and it just so happens that the acceptable and outstanding quality of the recording played during that time..so you have to say outstanding...Won't that be another interesting question to whether you are referring to the recording or the pic you are looking at?

Not that there's anything wrong with that...i'm just saying...

PS: Now i am confused...

 

[avaserfi]

Well, everything is said and done and the results are in!

Everything was completed with 45 participants - 27 males and 18 females between the ages of 18-56. In the end 86.7% of listeners preferred the uncompressed file (Average RMS 19dBfs) to the compressed file (Average RMS 11dBfs) this ends us up with a 95% confidence interval saying that between 76.8% to 96.6% of participants would prefer the uncompressed to the compressed in the general population.

Just to see if there were any variables involved with preference I compared age, sex, music ability, and music interest to see if any trends were apparent and there were none which shows that this preference was universal.

I also have a independent study next semester so I can dedicate far more time to expanding this experiment so more results will come.

Here is an image of the files the people listened to the first part is clearly uncompressed and the second is compressed. Thanks to Chris (WmAx) for setting the files up for me.

 

[Tomorrow]

Andrew,

Doesn't the federal gov't. give multi-million dollar grants for less?! Now, if you could apply audio preference to political speeches....whoa....billions!!

 

[avaserfi]

Andrew,

Doesn't the federal gov't. give multi-million dollar grants for less?! Now, if you could apply audio preference to political speeches....whoa....billions!!

Man, I wish I could get a grant for some research on these things. I would be doing some stuff far more complicated than this simple perceptual research like this.

The follow up paper I write will likely be submitted to some journals though unsure of which yet, but thats the plan. Maybe that will help me get a job or some grants .