Quote from: ducrider45 on April 23, 2012, 10:52:46 AM
Done, and no liberal votes here

same here... i just always assumed I was the only conservative person who rode a ducati.... not that I'm trying to talk religion or politics (the two no's)
...but it just jumped out to me as a survey on safe motorcycle riders vs their political views

I'm so sorry to piss in your thread and be negative, but I have to - after extensive studies and work within the field of methodology and statistics.

There is no way to know what characterizes the respondents in such a survey - based on voluntary participation
Respondents should be picked randomly - best way of reaching a representative selection.

I can explain further by PM if you read this and want to improve the study.

Quote from: stopintime on April 23, 2012, 01:37:03 PM
I'm so sorry to piss in your thread and be negative, but I have to - after extensive studies and work within the field of methodology and statistics.

There is no way to know what characterizes the respondents in such a survey - based on voluntary participation
Respondents should be picked randomly - best way of reaching a representative selection.

I can explain further by PM if you read this and want to improve the study.

So sorry to piss on your post but he can characterize the respondents to the survey, what he cannot do is make a valid inference to a larger population. 

smells like ammonia around here.

Quote from: j v on April 23, 2012, 02:20:17 PM
smells like ammonia around here.

Done   ...  And the population he picks is entirely in relation to what thesis he is trying to prove.

If he is comparing characteristics of people who hang around in chat rooms with say, another population canvassed in a bar and another canvassed in an art museum, he could come up with some interesting conclusions.

Or maybe he is just comparing Ducati owners who participate in chat rooms and who respond to online questionnaires appreciation of art to that of similar HD owners.

Population bias is only a problem if you don't build it into your experimental design or into your analysis.

He is innocent until proved guilty !

Quote from: crazykyy on April 22, 2012, 06:37:06 PM
Anyways here she is, my motorcyclist survey. It is pretty straight-forward; nothing is designed to trick you.

The motorcyclist questions were a smokescreen to trick us into answering the psychobabble questions...  and the research is actually a qualitative project being done on the various forums.  stopintime and akmnstr fell into the trap!

Done!  Do I get a cookie?

we are ALL part of the experiment 

The point is just that there is no way of knowing who responds and what they have, or don't have, in common with a representative (usually random) selection.

The only info that could come out of this is that a bad selection of x-brand forum'ers answered "this" compared to - a bad selection of y-brand forum'ers said "that".

It won't be a lie, but not particularly valuable...

Quote from: stopintime on April 23, 2012, 02:59:31 PM
It won't be a lie, but not particularly valuable...

Quote from: stopintime on April 23, 2012, 02:59:31 PM
The point is just that there is no way of knowing who responds and what they have, or don't have, in common with a representative (usually random) selection.

The only info that could come out of this is that a bad selection of x-brand forum'ers answered "this" compared to - a bad selection of y-brand forum'ers said "that".

It won't be a lie, but not particularly valuable...

It looks like he is using the exact same link for all the forums so it is actually a self-selected selection of people who read motorcycle forums (not necessarily motorcyclists, just people who play one on the internet) answered "this."  How else are you going to be able to draw conclusions on the motorcycling public?  Were you to ask the same survey questions at a large gathering of motorcyclists, you still would have survey results tainted by any number of qualifiers (IBA meeting? HOG meeting? DOC gathering? Vintage motorcycles?).

Even were he to mail the survey to every single licensed motorcycle, he would still miss a representative sample due to those who don't register their motorcycle for whatever reason.  And not all motorcyclists hold a valid operator's permit, so that selection criteria is less than useful anyway.

The results of this survey are likely to be no less valid to today's motorcyclists than the Hurt report... from better than 3 decades ago!

dammit, he sucked me in, too

Ugh.  I'd like to change my answer on the "Rarely get irritated." question.  If there had been one more question.  I would have lost it. 

You should see how many questions are on the Minnesota Miltiphasic Personality Inventory....
http://en.wikipedia.org/wiki/Minnesota_Multiphasic_Personality_Inventory

Quote from: NAKID on April 23, 2012, 05:10:01 PM
You should see how many questions are on the Minnesota Miltiphasic Personality Inventory....
http://en.wikipedia.org/wiki/Minnesota_Multiphasic_Personality_Inventory

There is another one you take...in conjunction with the MMPI..for police jobs. Around 1000 questions. Drives you bonkers. I figured em out though and can finish in about 40 minutes.