Gamer Survey - please respond

[benkyo]

Greetings all.

I am currently taking a Statistics class and the final assignment is doing a case study on a chosen topic. We're tasked with developing a survey, gather data, and then running statistical analysis on that data. We must then present our findings to the class.

The topic I chose is Tabletop Role Playing Games. Specifically, how do people get into them and why do they play. I love tabletop RPGs; it is my favorite past time.

I have developed a survey that gathers various demographic statistics and then goes into my core questions of how, why, and what do you play. I tried to be as encompassing as possible, so please do not be offended if I excluded answer choices specific to the flavors of tabletop RPGs you enjoy (most of my experience is with D&D and Pathfinder). The survey is completely anonymous; I never ask for your name or email address, and I only see the raw answers to the questions asked.

If you can spare a few minutes, please take the survey, you'll be helping me out with my project.

Thank you in advance.

https://www.surveymonkey.com/s/QJDJMP3

[Talakeal]

Ok done. That was easy.

Almost too easy; the questionnaire seems really simple and straightforward. Might I ask what you are setting out to learn / prove?

[lytokk]

Quick and simple, like surveys should be.

[Esprit15]

Always nice to help out a fellow student. Hope your study goes well.

[benkyo]

I am coming into this project with nothing but curiosity, I have no hypothesis as to what I think the results will look like. However, after running statistical analysis, I am hoping patterns emerge; i.e. most people like x campaign style with y rule set. I am also curious to see how the majority of people got into gaming. I am at the collecting data point right now, and will probably not run any analysis until I get around 200 responses (that's the minimum my professor wants for a sample size). If there's interest, I can definitely share the results.

[lytokk]

Would be interesting to see. Did a lot of work with demographic and census data in college, and even today I use my spatial statistics books at work.

[warty goblin]

Done. As a statistician I feel compelled to point out that without randomization, statistical methodology for inference on the broader population is invalid.

[Anxe]

I'd be interested in seeing the results as well.

warty goblin is right though. This is only going to tell you the habits of OOTS forum goers right now. You should try posting it somewhere else as well. It'd also be good if you handed slips of paper with the link to people at local gaming clubs or shops. If you only ask online, then you'll only get answers from people who go on the internet. That might skew your data.

[benkyo]

Valid points.

I have posted this to a couple other sites and am getting data from them as well. I think there is only one gaming store in my town, so the internet is definitely the best way to gather results. I am going to specify, when I present, where the data came from so there won't be an assumptions or misstatements that my data is a representation of the population. This is also a niche topic and I am only interested in responses from people who play.

[warty goblin]

I'd be interested in seeing the results as well.

warty goblin is right though. This is only going to tell you the habits of OOTS forum goers right now. You should try posting it somewhere else as well. It'd also be good if you handed slips of paper with the link to people at local gaming clubs or shops. If you only ask online, then you'll only get answers from people who go on the internet. That might skew your data.

This won't actually solve the problem either. Unless the method by which you choose people to survey can justifiably be considered random (which, unfortunately, isn't true of a volunteer or convenience sample), there's really no justification for statistical methods. I actually work in survey statistics, and the underlying design and the randomized design is absolutely critical for correctly analyzing the results. If you don't have randomization, or some understanding of the selection method, you have a very hard time justifying how the people in your actual survey relate to the entire population.

Now if you have demographic information about the population of interest, you *may* be able to justify considering your sample random based on how your sample demographics match up to the population demographics, particularly if you employ appropriate weights to the results. This is difficult to do - I spent all last summer working on this exact problem - and frankly the results are perforce shaky at best. Methodologically speaking, it's far better to implement a real sampling design from a sample frame than taking a convenience sample of volunteers and hoping for the best.

(This is, incidentally, why I'm leery of the Big Data people's constant miracle claims. Basically nobody understands the selection process for most of those problems, so it's very unclear how to relate your scads of data to the actual universe at large. I'm not saying the problem isn't solvable, but having a couple petabytes of data isn't necessarily as helpful as it at first seems for inference outside of the context of the people who generated it. Then of course you've got the mother of all Bayesian setups.)

[Pluto!]

The survey's response options make me think its writer hasn't had much exposure to games beside D&D.

[benkyo]

this is totally a sample of convenience, which is ok for my project. i do not have to do anything but gather data and then perform statistical analysis techniques learned in class.

Pluto!: you are correct. most of my exposure has been d&d and pathfinder. i have gotten several comments on other forums about this. it's a fault of my survey (i did seek advice when creating my survey, but most of my friends are also mostly d&d/pathfinder fans)

[Anxe]

I've always been puzzled by that part of statistical analysis of surveys though. If you pick the participants randomly, they are still free to decline the survey. Doesn't that still mean its volunteer based?

[Doorhandle]

The survey's response options make me think its writer hasn't had much exposure to games beside D&D.

My concern exactly.

[Amphetryon]

I've always been puzzled by that part of statistical analysis of surveys though. If you pick the participants randomly, they are still free to decline the survey. Doesn't that still mean its volunteer based?

Even if you were to compensate the survey-takers in some way, they'd still be free to decline.

Note that, depending on how rigorously one defines 'random,' it becomes impossible to call a sample size truly random.

[Brother Oni]

Only comments I have are for clarity (Asian means someone from East Asia, not South Asia?) and there's no option to indicate that someone no longer plays except for a '0' in the typical week hours played question.

[Eldan]

What should I do about the money I earn? Because I make 3500 dollars (about) per month, but that's from a low-paid internship and below standard minimum wage.

Also, question 13 should really offer multiple choice.

Edit: and a lot of the questions that follow it, too.

[Lorsa]

I guess we aren't here to analyze how well written the questions where. As I understood it, it's just analyzing the data using some techniques so for all intents and purposes we could answer completely fake and it would still be a valid project. Which leads me to be a bit confused about why you even construct this thing and ask people to answer. Couldn't you simply writen in a couple of hundred answers yourself and skip this step entirely?


On the subject of how statistics can be troublesome; there are other ways that surveys are often misinterpreted as well. For example, in this survey, we answer what we like most or why we continue playing. That says nothing about why people actually play, just why they think they play. They might not have experienced all the various ways of playing yet, or had GMs that weren't very good at one thing or for that matter not really know themselves very well.

It's the same as if you ask people if they are social, outgoing people and 75% of them ask yes. From that you can't really conclude that 75% of people are social and outgoing, only that 75% think that they are. Not even that but only that 75% of the people answered that they are social and outgoing on this survey since for all we know they could be lying. Which really doesn't say much...

[BWR]

Done.

Is the OP interested in feedback on the questionnaire?

[warty goblin]

I've always been puzzled by that part of statistical analysis of surveys though. If you pick the participants randomly, they are still free to decline the survey. Doesn't that still mean its volunteer based?

This is a major problem, to which there are a variety of solutions. If you believe that your missing information is Missing At Random - that there's nothing particular about the people not answering - you can treat the answers you do get as random. Recalculating the weights however is not necessarily trivial.

The other option is to impute the missing values. Very often in the survey context you do have complete demographic information, even about the people who don't respond. If there is strong correlation between the demographics and the people who do answer, you can essentially make up answers for the people who don't answer without biasing your results. The difficulty becomes dealing with the extra variability you get from imputation since your actual sample size is smaller than the sample size you end up using. Imputation is particularly useful for dealing with cases where people answer most of a survey, but leave a few items blank.

But missing values remain one of the major problems in human surveys. Interestingly, survey givers have gotten very good at getting a hold of people over the years, to the point where the non-contact rate is often pretty negligible, but the nonresponse rate of people who are contacted has gone through the roof. I worked on a survey this summer that had an 8% response rate this summer.

[Anxe]

This is a major problem, to which there are a variety of solutions. If you believe that your missing information is Missing At Random - that there's nothing particular about the people not answering - you can treat the answers you do get as random. Recalculating the weights however is not necessarily trivial.

The other option is to impute the missing values. Very often in the survey context you do have complete demographic information, even about the people who don't respond. If there is strong correlation between the demographics and the people who do answer, you can essentially make up answers for the people who don't answer without biasing your results. The difficulty becomes dealing with the extra variability you get from imputation since your actual sample size is smaller than the sample size you end up using. Imputation is particularly useful for dealing with cases where people answer most of a survey, but leave a few items blank.

But missing values remain one of the major problems in human surveys. Interestingly, survey givers have gotten very good at getting a hold of people over the years, to the point where the non-contact rate is often pretty negligible, but the nonresponse rate of people who are contacted has gone through the roof. I worked on a survey this summer that had an 8% response rate this summer.

So it is still a problem, but there are workarounds. Cool!

How would you do this survey to get random results?

[warty goblin]

So it is still a problem, but there are workarounds. Cool!

How would you do this survey to get random results?

General survey sampling methodology* is to start with a sampling frame, which is essentially a big list in some form or other of everybody you're interested in. Note that although everybody you're interested in should be on the list, you don't actually have to have a list that enumerates the entire population explicitly. For instance if you're interested in the population of people who live in houses, a list of houses over the area of interest is sufficient.

That's probably the hardest part of this context, because there's no real list of gamers anywhere, like there is of households, etc. You can however list off popular gamer hangouts like gaming stores and websites. This suggests a two-stage cluster design, at least to me. Enumerate the gamer hangouts, draw a random sample of some sort**. Then camp out the gamer hangouts for some reasonable period of time and collect contact information for as many people who appear as possible. Then randomly sample and contact a subset of people from each gamer hangout to collect the actual information.

There's still some problems with this design. Cluster designs (when your frame only lists groupings of people, not individuals) assumes the clusters are mutually exclusive. They clearly are not in this design as presented, since I can show up in multiple hangouts. This makes it very hard to calculate the probability of sampling an individual gamer. If one is comfortable assuming exclusivity (and the data itself should provide evidence of this by cross-referencing contact information) then the inference is pretty straight-forwards. Otherwise, somebody's getting a Masters out of the deal, at minimum.

Under that assumption however, inference is fairly straightforwards. You have your probability of hangout i being in the sample, say p1i. You have the probability of gamer j from hangout i being sampled, say p2j|i. Calculate the overall probability of gamer j being in the sample as pj = (p1i)(p2j|i), weight all responses for gamer j by wj = 1/pj, add up, and you're good to go. Note that this gives an estimate for population total, an estimate for population mean is harder to get, because in this case we need to estimate the total population size as well. That gets into ratio estimators and a whole nother ball of rather complicated wax. Also, the variances won't be the familiar sample variance, but a rather different deal that also accounts for finite population size and the different sampling weights.


*Although I mostly work in the context of land use surveys, which changes things a bit. When we have nonresponse, it's because we design the survey that way.

**It's worth noting that it's extremely rare to use equal-weight sampling - the kind that is tacitly assumed by most general statistical methodologies - in surveys. This is because intelligently choosing unequal weights can lead to a more precise estimate. Mind, most statistical methodologies are also explicitly model-based, whereas pure survey inference uses only the design of the survey itself, with either no or very limited model assumptions. It's also quite common to build models of various sorts that use the results of a survey as response variables. I'm working on such a project right now.

[benkyo]

Done.

Is the OP interested in feedback on the questionnaire?

Absolutely.
I have gotten a ton of feedback on the questions I have asked (most concerning the obvious D&D/pathfinder slant...which is due to my inexperience with other systems). Although I will not modify my survey based on said feedback for this project (it would cause me to wipe all data and I am under a time constraint), I am finding this process interesting enough that I might try to run a similar survey again, and that survey will be a bit more polished and more encompassing.

As for the survey not asking why you stopped playing...I really am looking for people who do still play. I had a choice in my rough draft for "I quit playing" but I felt that it wasn't necessary to what I was aiming to collect.

As always, thank you all for helping me with this project! :smallsmile:

[benkyo]

What should I do about the money I earn? Because I make 3500 dollars (about) per month, but that's from a low-paid internship and below standard minimum wage.

Also, question 13 should really offer multiple choice.

Edit: and a lot of the questions that follow it, too.

I would go with whatever your best guess is. The only thing I am going to do with that question is say xx% of tabletop gamers make $yyyyy/year.

I get that feedback a lot, that more questions should have allow for multiple choices. I was trying to force you to pick one (which, from other feedback I have gotten, lead a lot of people to selecting "Other"). That is a design flaw with my survey.

[Terraoblivion]

There is no entry for being in a relationship but not yet cohabitating. It's kinda important for me, given that the reason I'm not isn't due to a lack of desire or it not being that serious, but because family reunification is kinda tricky and time consuming.

Also, the section on what systems people have played seems highly arbitrary in its selection beyond D&D. I'm pretty sure Rifts never even got close to being as popular as any of the main World of Darkness games, much less World of Darkness as a whole, for example. Other than World of Darkness, Warhammer licenses are curiously absent too. Those two really seem like ones that should be on a list of systems like that, given how big they have been in both Europe and the US.

Question 13 is also a bit confusing, I think it's meant to refer to your favorite system and not a favorite playstyle, but I'm not sure.

And like others have said, this is heavily biased towards D&D. To the point that a lot of questions don't have meaningful answers for different, more story-oriented systems. In general, I think that reading up on available literature and connecting with roleplaying culture in general before doing something like in the future would be a good idea, including informally asking people to have a better informed basis for creating questions in the future. I find myself answering "other" a lot and so would basically everyone I play with, so there is at least a segment of some size that isn't being represented by these questions.

Not just that, a lot of the questions are based around what makes intuitive sense to you. How is a fighter and a martial artist distinct? I know how it makes sense in D&D, but in wuxia everyone is a martial artist, even if they wear heavy armor and carry a broadsword, and when you get down to it, any set of formalized techniques for beating up or killing others constitute martial arts. I also fail to see how pilot and hacker are common archetypes, but socialites, engineers or medics aren't. I know that I've seen far more examples of any of those three than I've seen of pilots and hackers combined. Really, it takes knowledge to formulate the questions for a good survey, so there's a considerable amount of studying to do to create a theoretical framework to base your study around, otherwise you're liable to end up with something as fruitless as phrenology.

It also helps to make your questions and the options clear. What is "techno-epic"? What does "alignment is for the birds" mean in this context? Unclear questions and options provide a lot of noise in the data because people interpret them highly differently.

Also, I have to ask...Why isn't there a campaign setting style for science fiction other than Post-Apocalyptic and whatever Techno-Epic refers to? It seems like kind of an obvious blind spot. Just like historical settings, pulp adventure or military settings. This just strikes me as really bizarre.

[Pluto!]

The survey's response options make me think its writer hasn't had much exposure to games beside D&D.
I want to elaborate on this so it's not read just a dig at the OP's experience.

Spoilered because this discussion probably isn't appropriate for people who intend to but haven't yet participated in the survey:

The reason I say that the quiz seems framed by D&D-only experience is that many of the "Why?" responses it allows participants to select are D&D-specific design goals, while the design goals and features of other games aren't accounted for.

My apologies if my examples of the quiz are a little hazy. I submitted the quiz and am pretty much writing off memory.

"Character Balance" is a notable design goal of modern D&D that either isn't prioritized or isn't relevant in other RPGs. Any system with randomized stats is by definition an imbalanced game, and many systems like Mutants and Masterminds are popular despite lacking any reasonable pretense of balance. And on the other end of the equation, there are tons of games where it just isn't an applicable question - games like Dust Devils, Dogs in the Vineyard, Fiasco or, to a lesser extent, FATE, work based on qualitative rulesets that don't really have room for one character to be mechanically superior to another in any way.

And equally conspicuous, the survey didn't really account for the design goals and "hooks" of other games, like the adaptability of the system, the speed/ease of use, the way the rules reflect the subject matter (most games have that make the conflicts that are most important to the game's fiction fun to resolve - Fiasco with its storytelling votes, SpyCraft with its mini-games for Chases, Interrogations, Seduction and other Spy movie-like interactions, Dogs in the Vineyard with its escalating-argument rules; and in another way some games like Dust Devils and Dread have mechanics that uniquely reflect their stories' conflicts in ways that build their stories' motifs or tension), the game's subject matter (games like Continuum, Everyone Is John or Don't Rest Your Head just present wildly different sorts of stories than other systems can be used to reflect).

The question about the "type of character" a person tends to play was clearly applicable for class-based games, which are few and far between these days. Like the setting question, it also seemed extremely narrow and arbitrary in its available options, but it had a bigger problem in that the idea of identifying a character as a "Thief" versus a "Hacker" versus a "Pilot" either isn't relevant (as in a character could reasonably be one or more) or meaningful (in that many systems wouldn't care about the distinction).

The quiz also betrays some priorities that seem specific to the survey creator. Specifically, I'm looking at the "lack of variance" options. Variability is a feature that's been deliberately added to many games and many houserules - most clearly in the forms of critical hits, exploding dice and the ever-popular critical fumble D&D houserules. "Balance" also fits into this, because even though I don't think a lot of people would explicitly say "I like playing imbalanced games more," I think people enjoy some of the facets of games that emerge from imbalance - see the enthusiasm of D&D's character optimization fans or the appeal of stat rolls that kept some groups (and the game designers' default assumptions) using them for years after point buy made an appearance in the DMG. So treating absence of variability or presence of balance as reasons for choosing a game without accounting for their opposites seems like the survey is encouraging responses to the effect that "balance is favorable, variability is not," which makes any conclusions in those directions suspect.

[Socksy]

Completeeeed~!>w<
Aren't the questions "How old are you", "How long have you been gaming for", and "When did you start gaming" slightly redundant if all three are in there?
What do my ethnicity/income have to do with anything?

[benkyo]

Terraoblivion:
I understand your concerns. In my rough draft of my survey I had included World of Darkness, but somehow missed it when I was creating the online survey. For your other points, which many others have also made, the only way I could fix the survey would be to wipe all data and start over. I am really loath to do that at this point since I have gotten so many responses and I'd hate to ask everyone to retake the survey.

Part of the issue is definitely my lack of knowledge of the other systems. The other error I made when crafting my rough draft was that I did not seek advice from people outside my current gaming groups. However, if these flaws in the survey cause you to select "Other" for most questions, that does provide me with relevant data. If a good percentage of respondents were similarly forced to answer in that fashion, then I will present my findings to my class what the issues with my survey were.

[warty goblin]

Completeeeed~!>w<
Aren't the questions "How old are you", "How long have you been gaming for", and "When did you start gaming" slightly redundant if all three are in there?
What do my ethnicity/income have to do with anything?

They aren't redundant if there are times periods of a person's life where they haven't been gaming. If I'm 24, started gaming at 18, and have been gaming for 4 years, that implies I wasn't gaming for two of those years.

Ethnicity, income, etc are all useful for post-stratification. If you're interested in how many people of a particular ethnic category game, but don't have the ability to group respondents based on ethnicity before you give the survey (because you don't know their ethnicity yet), there are techniques for producing estimates of favorite game type among ethnicity ___, to give a contextual example.

[Yogibear41]

Done, had to go with other on alignments because I have never played the same alignment more than twice. I'd be interested in seeing the results of class/alignment particularly. I once saw a study for world of warcraft I believe it was that related income to class selection, with the results being people of higher income seemed to be drawn to "less magical" classes such as the warrior, while lower income were more likely to play things like mages, I thought that was interesting, the guy wrote up something which essentially equated to: rich people don't feel the need to play "flashy magical classes because their lives are already above average" or some silly nonsense like that. One of my friends at the time said thats BS, because I play a warrior and I'm broke as S**T (I lol'd) Would be interesting to see if your survey showed a similar trend or blew the previous study out of the water. (hoping it would blow it out of the water)

[benkyo]

Completeeeed~!>w<
Aren't the questions "How old are you", "How long have you been gaming for", and "When did you start gaming" slightly redundant if all three are in there?
What do my ethnicity/income have to do with anything?

All three age questions are there simply so I do not have to edit the raw data and make a separate column before running some stat tests.

Ethnicity/income are just standard demographic questions. They really don't have anything to do with anything other than a one line statement in my presentation that xx% of respondents make this much money and are of a certain ethnicity. I believe the general stereotype for tabletop gamers is that they are middle-aged white men living in their parents basement (I still get weird looks when I tell people I'm play D&D). Perhaps the data from my survey will show that that is not the case.

[BWR]

My questions and comments will probably be a bit harsh but I'm not trying to be a jerk, I'm merely trying to point out what I see as problems. Badly worded questionnaires are a pet peeve of mine (https://www.youtube.com/watch?v=G0ZZJXw4MTA), since they are practically meaningless. My gf for instance was at one point asked which of 3 politicians she preferred, without the option of answering 'none of the above'.

In general, the questions showed a certain bias and starting point. While they might be valid questions depending on which subset of the population of the world you are interested in finding out about, it was not good for a world-wide survey of roleplaying games.


Question 3. This really only works if you are American. Otherwise the categories are too vague as to be meaningless. Black or African American, Hispanic American, Native American, Asian - these don't work well outside of the US. In the case of Asian, for instance, does Middle Eastern count as Asian? Pakistani/Indian? Isn't that a group that includes a wider variety of skin tone, languages, ethnicities and cultures than 'Hispanic American'? And what is the point of the question? finding out what skin color they have? Is it tied to certain socio-economic elements that may be particular to the US but not the rest of the world?

Q 4. Single but cohabiting with SO - if you have an SO you aren't exactly single, are you? "Single, never married" Why is this interesting? Why is it interesting to know the difference?

Q.s 8-9 why is exact age important and not age category? (e.g. younger than 10, 10-15, 15-20, 21-30 etc.)

Q 11 -12. As others have noted, quite a limited selection. Why were these in particular chosen and not other big names like various White Wolf games? If you don't know about much more than the ones you've listed, maybe you should research a bit more the variety of RPGs out there before making the list. If people can list exact ages, why not allow them to list exact RPGs? 'Other' contains quite a lot of games, some of which may be more popular than some of the ones you've listed there. And for most people, Pathfinder and D&D count as the same thing. And D&D can be anything from 0D&D to 4e, which individually are more different than Pathfinder and 3.x,

Q 13. I can't quite put my finger on what bugs me about this question. The options are a bit...sloppy? What exactly does 'world is more creative' mean? Simply that you like the setting? But what if a setting is not particularly unique or innovative, like the Forgotten Realms (ignore how it's unique and see how because it set the standards for much of high fantasy RPG settings that it's common). And what if more than one choice applies to your reason for liking a setting?

Q. 14, see comments for 8-9

Q. 16, Again, what about multiple options? What precisely is the difference between 'variety of what you can play' and 'the ability to completely become someone else'? Again, somewhat fuzzy phrasing. And I'm pretty sure most gamers do a combination of RPGs, movies, video/computer games. And read too. Asking why someone does something instead of something else when they do both is meaningless. Simply asking 'what is it about roleplaying that interests you' would be better.

Q 17. Is this one any different than 16 in any meaningful sense?

Q 18. Again, the questionnaire shows a heavy bias towards D&D style games. The term 'adventure path' is basically the same as a campaign, isn't it? Or possible a 'short story arc'. And as a specific type of campaign written by a a company and published in 6 installments, it's rather exclusive. And what if you don't really care what you play?

Q 19. Again, the options don't really seem thought out. Some of them are not mutually contradictory and can be equal or near equal factors in why you enjoy something.

Q 20. What's a "Techno epic"? What about other options like "urban fantasy" or "space opera" or "comedy" or "horror" or combinations of various categories?

Q 21. you have RAW. you have house rules. You have a mix. What's left?

Q 22. How is a martial artist different from a fighter-type? Why are pilots different from fighter types? Why is 'hacker' a common enough category but 'courtier' isn't? Wouldn't a hacker count as a rogue/skill monkey type?

Q 23. D&D specific. Doesn't really apply to a most other games. They might or might not have something similar. Why is this an interesting question? "Alignments are for birds" isn't a terribly professional way of putting it.

[Eldan]

I showed this questionaire to six buddies of mine who all play RPGs.

Interestingly, they all agreed on one question. Or rather, one answer to it.
"What's an "alignment"?"

Because none of them have ever played D&D. Or any of the games on that list.

[benkyo]

I have thought this over and I am going to close the current survey. I will go back through it, improve the questions and answer choices based on recommendations, and then repost the new improved survey. I thank everyone who participated, and I hope everyone will be willing to take the improved version.

Thanks again for everyone's feedback and help with this project.

[Lorsa]

I thought your assignment was just to make A survey, not necessarily a good one. :smallsmile:

But if you are going to improve the questions, I found it curious how the first question asked about your gender but the answers related to your sex. As far as I know those are two separate things.

[Talakeal]

One thing about alignments, I don't normally play with them, but when I do I normally play neutral good. What should I have chosen?

[benkyo]

I have revamped the survey based on feedback from many of you.

So, if you can help me out again and can spare a few minutes, please click on the link below.
https://www.surveymonkey.com/s/QJDJMP3

Thank you in advance for your participation.

[Dorian Gray]

I did it!
It's better than the last one, I think.

[benkyo]

I did it!
It's better than the last one, I think.

Thanks.
And I hope it is better than the first one:smallwink:

[Anxe]

Done. That one is better. Only change I'd suggest now is drop down menus for the years categories.

[benkyo]

Sorry to be pest, but I still need some more respondents for my survey.
Your help is greatly appreciated.

[Amphetryon]

Done, again.

[Rhynn]

Okay, I answered all the questions, but before I submit - some of the questions (especially #21*) make me think this is intended for US residents - is this correct? I wouldn't want to mess up your statistics if that's your assumption.

* Basically, your results here would be worthless for inferring anything about actual relative wealth, since the same income converted into USD in Norway and in South Africa would mean vastly different buying power and economic class.


#4 is hilarious. You asked for it, I guess.

#7 is going to give you weird math, because the number you're looking for is unclear. Are you looking for a mean unit rate or a mode or what? On most weeks of a year, I spend 0 hours actually tabletop gaming, so is that a 0? Should I estimate the hours-per-year and divide by 52? I just went with 0, since that seemede the answer the question was asking for.

#10 is missing the answer option "both" (or "either").

#12 is completely missing what is probably the main contention between old-school and modern D&D, which is "adjudication-based" (that is, the rules are guidelines and the GM ultimately adjudicates all situations).

Also, I guess this is a US cultural thing, but how is "Middle Eastern" a "race/ethnicity" but e.g. white Europeans and Americans are both lumped under "White/Caucasian" (which, for instance, ethnic Arabs and ethnic Spaniards are) ?

I already commented on #21.


Note that, depending on how rigorously one defines 'random,' it becomes impossible to call a sample size truly random.

Fortunately, statistics has confidence intervals based on sample size. :smallcool:

[benkyo]

Okay, I answered all the questions, but before I submit - some of the questions (especially #21*) make me think this is intended for US residents - is this correct? I wouldn't want to mess up your statistics if that's your assumption.

* Basically, your results here would be worthless for inferring anything about actual relative wealth, since the same income converted into USD in Norway and in South Africa would mean vastly different buying power and economic class.


#4 is hilarious. You asked for it, I guess.

#7 is going to give you weird math, because the number you're looking for is unclear. Are you looking for a mean unit rate or a mode or what? On most weeks of a year, I spend 0 hours actually tabletop gaming, so is that a 0? Should I estimate the hours-per-year and divide by 52? I just went with 0, since that seemede the answer the question was asking for.

#10 is missing the answer option "both" (or "either").

#12 is completely missing what is probably the main contention between old-school and modern D&D, which is "adjudication-based" (that is, the rules are guidelines and the GM ultimately adjudicates all situations).

Also, I guess this is a US cultural thing, but how is "Middle Eastern" a "race/ethnicity" but e.g. white Europeans and Americans are both lumped under "White/Caucasian" (which, for instance, ethnic Arabs and ethnic Spaniards are) ?

I already commented on #21.



Fortunately, statistics has confidence intervals based on sample size. :smallcool:
It is kind of intended for US residents, but the power of the internets allows anyone to fill out the survey. If you are worried about actual spending power compared to US currency, then I would use a conversion based off of the current exchange rate (One million Turkish Lira did used to equal $1.76 US).

For #4: i really did ask for it. Was the only way to make everyone happy (my lists were quite lacking).
For #7: if there are breaks in your campaigns, I would estimate your yearly time and then divide by 52. I did the same thing when I filled this out (life happens)
For #10: I am trying to force you to pick one or the other, which one you prefer the most.
For #12: i would recommend "Based on RAW but extensively rewritten"
For #17: i am not sure it is strictly a US cultural thing (my initial survey was very american-centric). I searched for more complete demographic examples, and that's how the majority broke them down. I did ask myself if I should break out European descent a bit more, but decided against it...which definitely could have been the wrong call. If none of the choices work for you, I would select Other.

[Rhynn]

Thanks for the reply! I've submitted my answers.

[benkyo]

I want to thank everyone for their help and participation in my survey.
However, i do need more respondents.
So, if you have a few minutes, and haven't done so already, please fill out my survey. The updated link is in the first post.

[Airk]

Filled out.

[benkyo]

Filled out.

Thank you! :smallsmile:

[Rephath]

Done! Would like to see your final data.

[TriForce]

i found it hard to make a accurate guess of my education, since the a lot european education systems are not compatible with US ones, other then that, all good!

[Krazzman]

i found it hard to make a accurate guess of my education, since the a lot european education systems are not compatible with US ones, other then that, all good!

This exactly. Took me 2 minutes to get an rough equivalent.
Also just skipped the money question due to lazyness. (thanks for the not to list option)

[benkyo]

This exactly. Took me 2 minutes to get an rough equivalent.
Also just skipped the money question due to lazyness. (thanks for the not to list option)

Thank you for participating.
Sorry about the education question being US specific.

[Soarel]

Took the survey. I'm interested in seeing the results to see whether it confirms or refutes the stereotypes about tabletop gamers.

[benkyo]

Took the survey. I'm interested in seeing the results to see whether it confirms or refutes the stereotypes about tabletop gamers.

Thank you.

[WrathMage]

Done, was interestin.

[Asmayus]

Done! The only question that was a little weird was "how many hours do you play a week?" ...for me at least, the answer is 0. My group tends to get together only a few times a year. :smalltongue:

[banthesun]

Okay, done. It could have done with a question about whether people mainly play or GM, though, since that could potentially turn up some interesting results.

[benkyo]

Okay, done. It could have done with a question about whether people mainly play or GM, though, since that could potentially turn up some interesting results.
Thank you.
Adding a GM question is a good idea. I'll keep that in mind if I do another survey like this in the future.


Done! The only question that was a little weird was "how many hours do you play a week?" ...for me at least, the answer is 0. My group tends to get together only a few times a year. :smalltongue:
Thank you.
The question probably could have been phrased better like "On average, how many hours per week do you play?"
That is generally what I am after.
If one week you play an hour, but the next week 3, then that averages out to 2.
So, if you only meet up a few times a year, I would add those hours up then divide by 52 to get your weekly number (and round up if it is close).

[Exediron]

Done, although being a statistical outlier and a member of the radical role-playing fringe I doubt I'll contribute much to the greater correlations being made. Like others, I'll be interested to see if the data reveals any strong connections.

PS: On monetary conversion, it isn't as simple as just changing currencies - a job that pays 40,000 GBP in the UK doesn't have the same buying power as a job that pays 67,000 USD in the USA despite the two being theoretically the same amount of money, so it's hard to draw correlations about status or relative wealth level from only that data point. Some countries cost more to live in than others, for example.

[Dorian Gray]

Done, although being a statistical outlier and a member of the radical role-playing fringe I doubt I'll contribute much to the greater correlations being made. Like others, I'll be interested to see if the data reveals any strong connections.

PS: On monetary conversion, it isn't as simple as just changing currencies - a job that pays 40,000 GBP in the UK doesn't have the same buying power as a job that pays 67,000 USD in the USA despite the two being theoretically the same amount of money, so it's hard to draw correlations about status or relative wealth level from only that data point. Some countries cost more to live in than others, for example.

Although technically that goes for, say, NYC vs Cleveland as well as the US vs other nations. Money is never going to be static.

[INDYSTAR188]

Done OP. Interested to see your results.

[veti]

Done.

Question 4 is... too open ended, I suspect some of my answers will be meaningless to you 'cuz they're systems you've probably never heard of (and likely never will again). I'm also wondering what the input limit is on that field (how many of the characters I typed will have made it through to your results). Next time, I suggest you pick 10-15 different systems (D&D, including Pathfinder, is *one* system for this purpose, different editions don't count) and ask people to tick those they've played. Include an "Others" box allowing people to list any system they've played extensively (i.e. spent more than 10% of their total lifetime gaming time on it).

[Coidzor]

Done, assumed gaming was specifically table top RPG gaming given the context. Probably just clicked through the first screen too quickly.

Could've used another option or two on the Homebrew vs. published settings question.

[GungHo]

Using "gaming" as shorthand for "tabletop gaming" is a little confusing when you're asking for years, hours spent, etc.

[benkyo]

Done.

Question 4 is... too open ended, I suspect some of my answers will be meaningless to you 'cuz they're systems you've probably never heard of (and likely never will again). I'm also wondering what the input limit is on that field (how many of the characters I typed will have made it through to your results). Next time, I suggest you pick 10-15 different systems (D&D, including Pathfinder, is *one* system for this purpose, different editions don't count) and ask people to tick those they've played. Include an "Others" box allowing people to list any system they've played extensively (i.e. spent more than 10% of their total lifetime gaming time on it).

Thank you for taking the survey.
I had choices originally and several people made the same comment that the list wasn't inclusive enough. I made it open ended to satisfy a large percentage of respondents. I tested the text box and it was able to handle over 600 words, so you should be fine. Going through some of the answers I have received thus far, there are numerous systems I have never heard of (which is why people were, rightfully so, complaining about my list of choices).

[benkyo]

I want to thank everyone who participated in my survey. :smallsmile:
I have enough responses now for my project so I am going to close down the survey.

Again, a big Thank You to everyone who participated.