Wikipedia:Reference desk/Archives/Humanities/2017 October 17

= October 17 =

Does Dubai have more Sunnis or Shias?
Any definition: city, emirate, metro area, urban or high-rise area. Has this changed recently? Does the plurality change depending on how long someone's been there? I.e. Does counting everyone there at one moment give a different answer than excluding people just passing through without touristing or the previous plus tourists or the previous plus short-term residents or the previous plus non-citizens or the previous plus naturalized citizens or the previous plus people with recent ancestors who weren't born there? Are citizens or residents more likely to be one or the other the richer or more powerful they are? Sagittarian Milky Way (talk) 18:05, 17 October 2017 (UTC)
 * Demographics of Dubai mentions that about 15% are Shia. That makes it in line with what is known of global Shia Islam demographics.  The only majority Shia countries in the world are Iran, Iraq, Azerbaijan and Bahrain.  Most other Islamic majority countries hover around 10-15%.  -- Jayron 32 18:45, 17 October 2017 (UTC)
 * This article says most UAE Shiites live in Dubai or the second largest city in the Dubai metro area and only 30% of the UAE lives in Dubai so it doesn't seem one could assume more Sunnis than Shias. Sagittarian Milky Way (talk) 22:45, 17 October 2017 (UTC)
 * If 15% of all UAE residents are Shia, and 30% of UAE residents live in Dubai, then we're all but assured that at least 50% of Dubai residents are NOT Shia, because if more than 50% of Dubai residents were Shia, that would only work out if essentially zero Shia lived anywhere else in the UAE, which is technically possible, but probably unlikely. -- Jayron 32 11:08, 18 October 2017 (UTC)
 * "A majority of Shi’ites — both citizens and noncitizens — live in Dubai, Abu Dhabi, or Sharjah" (so they don't ALL live in Dubai).  Alansplodge (talk) 15:52, 18 October 2017 (UTC)
 * Only 76% of the UAE and 60% of Dubai (2002 though) are Muslim so only they're available to be Sunni. I suppose the best description is that it's a diverse place. Sagittarian Milky Way (talk) 01:27, 19 October 2017 (UTC)

How do researchers get participants to tell the truth [about academic misconduct, political convictions, sexual histories]?
Researchers often do topics like these. But people can lie. What can researchers do to counter the lying and get more accurate reporting of these figures? And do they ever know it's accurate, or do they recognize it as a report? 140.254.70.33 (talk) 18:35, 17 October 2017 (UTC)
 * Yes, yes, nothing, no, yes. See, , , , . -- Jayron 32 18:48, 17 October 2017 (UTC)


 * Even without a participant deliberately lying to a researcher, Demand characteristics can affect the results. Some participants will want to help the researcher, and will seek clues as to what he wants. If a sex researcher seems to be aaffiliated with a conservative organization and looks pleased or more interested when the participant says he has been faithful to his wife and lived a vanilla sex life then the results may be different than when the researcher seems  bored with a boring sex life, but pays more attention, or raises an eyebrow, makes eye contact, or smiles and nods when sexual escapades or perversion is told of.  Demand characteristics or observer expectancy has been pointed out as causing children to make false accusations that adults have molested them, in the Day-care sex-abuse hysteria of a few years back, but children are not te only ones affected. As the cited article says, other participants in any experiment may have a "screw you" attitude and seek to "ruin" the experiment. When people self-select by volunteering for a survey, there is a greater likelihood of getting braggarts or persons spinning a yarn. Kinsey sometimes would get an entire club or organization to encourage its members to participate to broaden the sampling. A college researcher might disguise the purpose of a study by calling it by a random name like "Orange" instead of "Sex survey." The survey might be collected via a terminal rather than an interviewer, with assurance of anonymity, rather than person to person.  Edison (talk) 13:07, 18 October 2017 (UTC)
 * There's also issues with sampling matters, this article describes the crux of the problem: The people who choose to participate in studies come from a limited sampling of socioeconomic backgrounds, and are not representative of an entire population. Even IF all participants responses could be trusted, you still have the problem of assuming your sample represents the population at large... -- Jayron 32 13:12, 18 October 2017 (UTC)


 * The original Kinsey report had big problems with skewed samples, which included people who were chosen because they would say colorful and vivid things, rather than because they were average in any way. You can read  this interview for claims that analyzing anonymous Google data can be more useful in some contexts than relying on questionnaires or surveys... AnonMoos (talk) 09:30, 19 October 2017 (UTC)


 * There is a statistical method I've heard of, and maybe someone here remembers the name of it and can explain it in more detail. The researcher tells the subject, "Flip a coin, and don't tell me whether it comes up heads or tails. If it comes up heads, write your answer the following question: [an innocuous yes or no question]. If it comes up tails, write your answer the following question: [a question about whether the subject has ever engaged in a particular illegal, controversial, or embarrassing behavior]." The idea is that the subject will free to answer the latter question honestly because the researcher won't know which question they answered, and the researcher will be able to estimate the prevalence of the illegal, controversial, or embarrassing behavior by statistical methods, although I forget how that exactly works. --Metropolitan90 (talk) 16:08, 22 October 2017 (UTC)