A claim I heard last night from a "chastity expert" was that about 50% of teens who self reported as homosexual ceased self reporting as homosexual sometime in their twenties.

Standing and saying "citation needed" to her would have been satisfying, but socially inappropriate. Does anyone recognize this as an actual study?

  • 2
    Do you have a name of the expert? It may be possible to find something she has written on the subject so we can understand where she is making the claim from. And I assume by stop you are saying they no longer engage in homosexual activities, and instead choose heterosexual partners.
    – Chad
    Dec 5, 2011 at 21:40
  • 1
    I too would like the name of the expert, but I think the claim is clear enough (except maybe for "sometime in their twenties"). People who self-reported one thing self-report another, later.
    – user792
    Dec 5, 2011 at 22:00
  • 20
    Just the phrase "chastity expert" sets my skeptic senses tingling. I suspect that there is an agenda which has a long history of using plain old lies to promote themselves... Dec 6, 2011 at 3:33
  • 1
    @LarianLeQuella - there are plenty of people who are "chastity experts". Ranging from specific subsets of BDSM community to assorted spiritual people to nerds. Suspecting an agenda doesn't make one NOT capable of knowing what one is talking about. Anecdotally, there are plenty of ex-lesbian ex-Dommes who eventually went hetero-switch route. Doesn't mean they are enough to sway the statistics to 50%, but the demographis is not empty.
    – user5341
    Dec 6, 2011 at 11:39
  • 2
    @DVK I know, however if you read the rest of the question, the context starts to become very clear. This is most likely a religious group/expert with the anti-homosexuality agenda. I think Oddthinking's answer already skewers the 50% claim. Dec 6, 2011 at 11:42

1 Answer 1


Partial answer only, to help others give a better answer.

One question that is raised here is whether teens give accurate answers in surveys. One possible measure of inaccuracy is "test/retest" inconsistency. That is, if you ask the same person the same question again, do you get the same answer?

Test–Retest Reliability of Self-Reported Sexual Behavior, Sexual Orientation, and Psychosexual Milestones Among Gay, Lesbian, and Bisexual Youths Eric W. Schrimshaw, Margaret Rosario, Heino F. L. Meyer-Bahlburg and Alice A. Scharf-Matlick, ARCHIVES OF SEXUAL BEHAVIOR, Volume 35, Number 2, 220-229, 2006. DOI: 10.1007/s10508-005-9006-2

They asked 64 youths (aged 14-21) recruited from GLB-focussed organisations about their sexual identity twice, two weeks apart, and compared the answers.

They found sexual identity had Cohen's Kappa value of 0.89, which is a strong indication that they got very similar results - but not identical ones.

  • 8
    Translating to English - "Earth shattering news!!!: Teenagers change their minds!". This should qualify for IgNoble ;) +1
    – user5341
    Dec 6, 2011 at 11:43
  • 1
    @DVK: Sort of right. If I could show that 50% of self-identifying GLB teenagers "changed their mind" after two weeks (or even if it was 25%), it would undermine the power of the original claim that they change their answers after 10 years. However, this study shows that a vast majority of the teenagers aren't as fickle with their answers about sexuality as one might expect. Nonetheless, some do (and we could probably come up with a dozen ideas about why.)
    – Oddthinking
    Dec 6, 2011 at 11:54
  • 5
    11 % change their mind in two weeks? Seems like quite a lot for a trait which is considered to change in years, if at all. Was there any check done to see if those teenagers take the survey seriously at all? It is long since I was a teenager, but if I remember it correctly, I was quite able to fill a questionary with crazy funny answers just for the heck of it. :)
    – Suma
    Dec 6, 2011 at 12:22
  • 4
    Unfortunately, saying 11% changed their minds is too strong a statement to read from κ=0.89. This value isn't that simple a measure of agreement. From Wikipedia: "It is generally thought to be a more robust measure than simple percent agreement calculation since κ takes into account the agreement occurring by chance".
    – Oddthinking
    Dec 6, 2011 at 12:49
  • 7
    This would actually be a decent answer, all by itself, if it explained what a value of 0.89 in the Cohen's Kappa scale was. And gave some benchmark examples to put it in context.
    – Jonathon
    Oct 22, 2015 at 22:38

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .