How much integrity is there to online sex surveys?

[odissius2b]

How much do you believe the sex surveys on line? A lot? Some? Not at all? Here's why I ask. The Internet is anonymous. You don't know who I am and I don't know who you are. Theoretically, your responses to honest questions should be valid. But we both know this is not true. There are individuals who post a screen name age as 100..which is obviously a lie, and if they lie about that...what else will they lie about? On the other hand, there are statements made which we have not reason to doubt. In other words, an anonymous survey may be much more valid than one done by telephone or vis-a-vis.

Here is an example of what I think is a valid Internet survey on 213 respondents on dick.net. :
"Have you ever seen a handsome man and thought 'I would love to suck his cock' "?
The respondents were mostly self described as "gay" or "bisexual" and here were the results:
Yes 93%
No 1%

Here is another question from the same site of 217 respondents: "How old were you when you first tasted someone else's cum?"
10-14 45%
15-19 28%
20-40 18%

Now, these questions are obviously, being answered by a subset of the usual "population" ...but, i think they represent an interesting survey which has never been done before.

Your comments?

[tenni]

I don't know if these questions have been asked before.

I wonder about control factors and statistical analysis. There are statistical analysis systems to report the reliability and validity of the results.
ie http://en.wikipedia.org/wiki/Reliability_(psychometrics)

http://en.wikipedia.org/wiki/Validity_(psychometric)

What I notice first is that the percentages do not add up to 100%. The cum tasting question doesn't report the "never" or "not yet" percentage. That makes the results questionable unless an answer is provided. The phrasing of the questions and categories to respond to can impact the validity of the survey.

You refer to the sexuality of the respondents and that might be one control factor (sexuality) but only a category of MSM (men who have sex with men rather than bisexuals or gay). The information gathered would have some validity but not as much as one done with professional standards of control and analysis. You mention them being a sub group and all the factors that make up this sub group should be reported.

Another control factor might be the year of the survey as the results may change over time. Are there any control factors to prevent repeated answers from the same individual? etc. I think that this site's surveys control repeated responses but someone questioned a political survey on this site with questions of control(repeat voting)

I would think that a random telephone survey would probably not ask these questions due to the personal nature of the questions. A sample group of say university students of a specific age might give more validity to the survey but again the control factors would restrict the generalization of the responses to the population at large. The cum tasting question would lose out on the upper end of the age groups though if the sample was restricted to college attending men. Education would then become a control factor rather than a variable.

So, these questions are interesting and more surveys might affect the reliability of the answers.

[ExSailor]

There's going to be very little since people love to lie in general and they will lie on supposedly anon. surveys. Although I guess it depends on what the questions are and who is giving the survey.

[Gearbox]

It's VERY common for men to lie about their age, build, name and relationship status on hookup sites. So I dought they'd give the Gospel truth to some non-entity they wont ever meet in a survey.But maybe they only fib when there's some 'prize'? Who knows?

[odissius2b]

It's VERY common for men to lie about their age, build, name and relationship status on hookup sites. So I dought they'd give the Gospel truth to some non-entity they wont ever meet in a survey.But maybe they only fib when there's some 'prize'? Who knows?

Good points, Gearbox. But, you and I both know, that someone out there...has the ability to identify both you and me (i.e. name, DOB, address, etc). That's not the point. Those, people: FBI, CIA, etc, have neither the time nor resources to identify people on trivial questions posted above. And you are absolutely correct about people lying, and I can't think of any better example than Bill Clinton lying about Monica Lewinski. But, the only reason he did that was because he didn't think he would get caught. Correct me if you think I'm wrong...but wouldn't you be MORE LIKELY to tell the truth on an anonymous survey, such as the one listed above...than say a telephone survey?

[odissius2b]

There's going to be very little since people love to lie in general and they will lie on supposedly anon. surveys. Although I guess it depends on what the questions are and who is giving the survey.

You've just raised an age-old philosophical paradox. And it goes something like this:
1. All Cretans are liers
2. I am a Cretan
3. Therefore, I'm telling a lie

[Gearbox]

Yes I think anon 'box ticking' would be less intimidating than a phone survey, and more likely to yield honest answers. But even so, I think a layout of comforting, non-judgemental and understanding would help any kind of questionare.For example - I've been asked my sexuality in person 4 times by questionares. The first twice I lied and stated hetro, because I didn't feel comfy telling those particular people the truth. I had the attitude of ''Fek you! It's none of your buisness!''. But the last 2 I gave the truth, coz they were nice.lol (Yes! I'm on a government database due to that)Written questionares could use some sugar too IMO.

[12voltman59]

When it comes to the sorts of sex surveys that appear on sites like this and at others--you just have to take them with "a grain of salt"--they are only meant as a fun exercise and nothing more since there is no valid statistical model being used to create them, the ways the questions are asked are not designed to gather any actual data---they are just seeking responses from people who care to respond, they are a 'self-selected" group and that is not what actual surveys want--they target the "populations" of those who they wish to take the survey--not to stack them towards some end---but to actually vary the population of those who would take the survey so they get a better result--on a site like this----since we all have some common sexual interests as us being self-identified as being "bisexual"---we are too much of a closed group to get any accurate results and then, its only a small number who respond anyhow.

When it comes to actual, academic, research sorts of surveys----if they are done in proper fashion in all the parameters that go into making up a true survey---then the results can at least be trusted to some degree---but even just one such survey is only so good, because in order for the results to be more strongly trusted----surveys like that need to be done on a periodic basis---with many of the sex researchers who do these sorts of things--they will take such surveys over many years-perhaps decades, selecting various study populations to target and such--and then when they have done all of that--then and only then will they publish the results and give their take on the results and what they indicate.

Here is a primer on how they are done: http://en.wikipedia.org/wiki/Survey_methodology

[Annika L]

Integrity? Oh please. The rule of thumb is: none required? none assumed.

[Roxas]

I could see some problems with inacuret reporting on the second survey because some people may not want to admit that they didn't have the specific sexuel experiance till later in life