Why Do Participants Perceive Lower Frequency Sounds as Higher Pitched?

[cng99]

A research on perception of pitch

Homework Statement

I've been working on a project. It's regarding the human perception of sound. I basically tested around 100 human subjects and made a report of it. The data of 60 individuals is in "google_science_fair..zip" attached with the post.

Here's what I exactly did as part of the test.

Step 1) I asked them to listen to 10 sounds (tones actually). Each tone had a number assigned. The ten tones are of
1.)100 Hz
2.)150 Hz
3.)200 Hz
4.)300 Hz
5.)400 Hz
6.)450 Hz
7.)500 Hz
8.)550 Hz
9) 650 Hz
10)700 Hz

(You can find all the sounds in an attachment called "named.zip")

Step 2) After this I took a test of them. They were made to listen to 5 sounds and they were supposed to identify what Sound number is being played. The participants were clearly told that it doesn't matter how accurate they are. The motive was to associate a higher number for sounds of higher frequency.

Step 3) I also asked their age after this to note down.

Now for every test of 5 sounds I conducted on any individual, the First sound and the Third sound were always same (500 Hz), however the subjects always gave a different rating to the two sounds. And the difference in the rating seemed to depend on the 2nd sound.

Apart from this a consistent observation is also that for some reason participants tend to rate a larger number to 300 Hz sound than to a 500 Hz. Even after being very clear that a higher frequency sound means a larger number, the 300 Hz sound seemed to some as higher pitched sound than the 500 Hz sound.

Why did this happen? And is there any other conclusion that could be made from the data?

Please analyze the data carefully. I've worked hard for this.

Homework Equations

No equations involved.

The Attempt at a Solution

My two conclusions

1.) Even if two sounds of same frequency are played with another sound in between, subjects may perceive it differently depending on the sound played in between.

2.) Sometimes an subjects fail to identify if the pitch has increased or not. there seems to be some relation of this with regard to what sounds they're accompanied with.

Anything I'm missing? Any suggestions would be accepted. Also, I'd love to be told where I went wrong.

 

[chemisttree]

Hook them up to electrodes and tell them that incorrect answers will result in shocks of increasing intensity. Data will look very different.

 

[cng99]

Hook them up to electrodes and tell them that incorrect answers will result in shocks of increasing intensity. Data will look very different.

Are you serious? I've worked really hard for this and if you can't respect that at least don't make fun of it.

 

[chemisttree]

Have it your way. I don't think many people will download a zipped file, unzip it, carefully go over the data and interpret it for you.
Why did you choose the data from 60 subjects out of 'around 100'? Sampling bias?
Did you make the subjects listen to 10 tones and rank them? Or were they comparing only two tones at a time? Could there be an design bias at work here?

 

[cng99]

Why did you choose the data from 60 subjects out of 'around 100'? Sampling bias?
Did you make the subjects listen to 10 tones and rank them? Or were they comparing only two tones at a time? Could there be an design bias at work here?

I took 60 because the other 40 were a part of a case study which involved people of similar age group all studied with a single type of test sequence.

The subjects were made to listen to 10 tones and after they're done listening a test was conducted for 5 tones which they had to identify. Hardly anyone got them right but the idea was to associate a higher number to a higher sound. Hence even if someone's giving the wrong answers, he would assign larger numbers to shriller sounds, which was the motive.

 

[cng99]

So no one's interested after all.
Alright, I understand.

 

[Ophiolite]

As described your experiment does not seem to be well designed. That may be a reflection of the lack of detail in describing the experiment, rather than in the methodology.

But you then ask us to devote time to analysing data for an experiment that may be poorly designed when that is a task that is really yours. That may be a faulty percpetion, but that is kind of how you have come across. I think you can see that might account for the poor response.

However, I thought I'd take a look. Sorry - your database layout does not lend itself to pivot table analysis, which would have allowed a serious look in minimal time. I am not prepared to take the time to convert your data table to usable form. If you will do so, then I'll be pleased to look at analysing your data.