Why do people cite surveys?

borisp · #1 10-18-2007, 06:55 AM

It is well known that the human animal is notoriously incompetent with regard to responding to surveys.

For example, in one survey, 60% of respondents claimed they were self conscious about their dandruff.

The catch? Dandruff only affects 35% of the population. Now, my precise numbers may be off, but you get the idea.

My point is that folks on this forum need to understand that most "surveys" are nonsense. This is directly in line with Borodog's point that the "what if you were..." threads are meaningless.

iggymcfly · #2 10-18-2007, 07:10 AM

People often use "dandruff" as an umbrella term to describe not only dandruff but dry scalp and other similar conditions as well. I don't see how this semantic argument proves anything.

onesandzeros · #3 10-18-2007, 07:34 AM

[ QUOTE ]
Why do people cite surveys?

[/ QUOTE ]

Because surveys always lie.

tame_deuces · #4 10-18-2007, 07:44 AM

The fun thing is....

That you somehow take some singular pieces of anecdotal evidence about surveys as truth that they don't work is actually an extremely good argument FOR survey research.

Noodles. · #5 10-18-2007, 09:42 AM

You know what is even more funny? That you cite a survey proving your point that surveys are not reliable to prove a point.

DblBarrelJ · #6 10-18-2007, 09:59 AM

[ QUOTE ]
You know what is even more funny? That you cite a survey proving your point that surveys are not reliable to prove a point.

[/ QUOTE ]

You know, statistically, 85% of statistics are made up on the spot.

Seriously though, and I admit this, even though I'm guilty as well. Surveys are cited because, if you really look hard enough, you can find a survey to prove anything. Take religion for example. If I were to stand outside of a church at around 12:45 on Sunday afternoon conducting a survey, I could probably get a very skewed reading on the question "Do You believe in the Genesis theory of creation?"

I interview 2000 people coming out of several different churches on several consecutive Sundays, and I have a survey that I can cite that proves that "97% of Americans believe in the Genesis theory of creation" with a sample size of 2000 interviews.

LondonBroil · #7 10-18-2007, 10:17 AM

I have dandruff

tame_deuces · #8 10-18-2007, 10:20 AM

[ QUOTE ]
[ QUOTE ]
You know what is even more funny? That you cite a survey proving your point that surveys are not reliable to prove a point.

[/ QUOTE ]

You know, statistically, 85% of statistics are made up on the spot.

Seriously though, and I admit this, even though I'm guilty as well. Surveys are cited because, if you really look hard enough, you can find a survey to prove anything. Take religion for example. If I were to stand outside of a church at around 12:45 on Sunday afternoon conducting a survey, I could probably get a very skewed reading on the question "Do You believe in the Genesis theory of creation?"

I interview 2000 people coming out of several different churches on several consecutive Sundays, and I have a survey that I can cite that proves that "97% of Americans believe in the Genesis theory of creation" with a sample size of 2000 interviews.

[/ QUOTE ]

This is why survey research is almost a discipline of its own, and why academic openness (as always) is vital in surveys, so you can know what means have been done not to get skewered results, and what means may give skewered results in your sample etc.

hitch1978 · #9 10-18-2007, 12:16 PM

[ QUOTE ]

For example, in one survey, 60% of respondents claimed they were self conscious about their dandruff.

[/ QUOTE ]

98% of students at my middle school were very conscious of my science teacher's dandruff.

Hello Mr. Beard.

GoRedBirds · #10 10-18-2007, 01:23 PM

[ QUOTE ]
This is why survey research is almost a discipline of its own, and why academic openness (as always) is vital in surveys, so you can know what means have been done not to get skewered results , and what means may give skewered results in your sample etc.

[/ QUOTE ]
Photoshop material? Also, QFT. There are good surveys and bad surveys depending on the actual makeup of any representative sample, sample size, survey phrasing, order of questions, yada yada... It's important to know how a survey is conducted before it's used or accepted as any form of support.