A survey on Railroader Demographics


EGardner

New Member
An acquaintance of mine is doing a survey of Model Railroader demographics, and they are hoping to collect enough responses to provide some meaningful information to people about trends in our hobby. If you are interested in taking part, here is the link:
Model Railroader Demographics Survey

Here is the description and info that they sent me:
I am doing an informal survey of Model Railroader demographics. Basically, I have observed a trend in the local railroading groups, and from articles I have read this trend may be fairly widespread. I am hoping to get as many people to complete this survey as possible to get a good picture of the demographics of our hobby.
If I get a significant number of responses and it results in useful information, I may be able to post some analysis and information about it.
Disclaimers & notes:
  • This is for informational and/or academic purposes only. The info will not be sold to anyone.
  • The results (once complete) will be available freely upon request
  • I reserve the right to edit comments to remove offensive material or personally identifiable information.
  • The results are intended to be anonymous, but based on comments it may be possible to identify individuals. The more people who take part, the less risk of any one person being identified.
  • SurveyMonkey's IP tracking has been disabled.
  • Please complete the survey only once.

If anyone is interested in commenting on the subject, I can pass along their contact info. Or one can post on their thread in Reddit.
 
Only poor question was about the age of the modelers in your area, that will be pure speculation. Rest of the questions are pretty reasonable.
 
to get a good picture of the demographics of our hobby.
How does this survey intend to achieve that when it is purely voluntary? So 30000 people, world wide, complete the survey - what exactly does that tell anyone other than 30000 people world wide completed the survey? It certainly doesn't offer anything conclusive (or remotely close to it) about the demographics or health of the hobby. In fact, you wont know where the demographic apply as there are no questions to that effect.

Consider this ... say there are 3,000,000 people in the world who are actively engaged in this hobby, yet only 30,000 do the survey that's only 1% of the hobby population so the results would be meaningless as a "good picture" of the demographics let alone anything else.

Why does someone want to know the demographic of the hobby, the age (basically) of the participants? What purpose does it serve? You (or your friend?) would be better of sending out an email to every person in the world asking them if they are involved in the hobby and asking for their age and country only. At least that might offer something that could be regarded as credible ... IF anyone responded that is.

Sorry, but I see this as a futile attempt that will yield nothing of substance, as honorable and well intentioned as it may be.
 
Last edited:
Consider this ... say there are 3,000,000 people in the world who are actively engaged in this hobby, yet only 30,000 do the survey that's only 1% of the hobby population so the results would be meaningless as a "good picture" of the demographics let alone anything else.

Actually you can get a representative sample that is much smaller than the population and get pretty good accuracy. Pollsters do it all the time. I used to know the math for it, we used it at work to determine how many employee survey responses we needed to get back to determine if the survey was valid. That's why pollsters can sample 1000 people and get a pretty good idea of what the population is.

How does this survey intend to achieve that when it is purely voluntary?

That is really the problem, to be really valid, it requires a "random" sample. That is suspect if the only source is from this forum. However if the survey was being put forward on several different forums, then it might be more representative.

The questions about age and years in the hobby are not necessarily independent of each other, since one will have less years in the hobby than age (your mom reading MR while you are "in utero" doesn't count.)

In the words of Mark Twain, "....lies, damn lies, and statistics."
 
Dave is correct. The sample size required for a statistically-significant survey is much smaller than you might think. The key is to be sure your pool of respondents is representative of the group you are researching.

If you were to survey only members of model railroad forums, your results might be somewhat skewed. But if you survey members of forums, members of clubs, NMRA members, attendees at model railroad show and conventions, etc., your results would be more representative of the total population of model railroaders.

If you assume the total number of model railroaders is 3,000,000, a 1% sample size (3.000) would be more than enough -- provided all important groups of model railroaders are included.

- Jeff
 
If you assume the total number of model railroaders is 3,000,000, a 1% sample size (3.000) would be more than enough
And most likely why surveys (of any type) are notoriously inaccurate - regardless of what the survey company states or the company requisitioning the survey says.
 
And most likely why surveys (of any type) are notoriously inaccurate - regardless of what the survey company states or the company requisitioning the survey says.
I totally disagree there. A well written and balanced survey properly distributed in a proper population can be very accurate. Most surveys I've seen do not fall within this category. In fact, I would say most surveys (especially the political one) are written intentionally to be self serving. On top of that Survey Monkey is not a robust enough product to provide a truly good survey.
 



Back
Top