Wikipedia:Reference desk/Archives/Miscellaneous/2013 January 30

Welcome to the Wikipedia Miscellaneous Reference Desk Archives
The page you are currently viewing is an archive page. While you can leave answers for any questions shown below, please ask new questions on one of the current reference desk pages.

January 30

Wikipedia pictures

Hello, I was wondering where Wikipedia gets your pictures from that are in your articles? Do you buy them from people or they all donated? — Preceding unsigned comment added by 108.222.6.140 (talk) 02:43, 30 January 2013 (UTC)

They are all cost-free pics. Some we take ourselves, others are public domain, etc. StuRat (talk) 03:16, 30 January 2013 (UTC)

Ignore StuRat's answer — it's impressively incorrect and not even a good simplification. He should really know better. I suspect he does. They are not "all free pics", at all, and such is a dangerous misconception.

Correct answer: the images on Wikipedia fall into roughly two categories. The first are images that are available under a free license. This means they may have terms attached to how they are used, but those terms are sufficiently broad to allow us to use them here without any possible copyright violation. Many of them are in the public domain, but a huge number of them are not. Many of them have been uploaded and licensed by users of Wikipedia, but a lot of them are taken from other webpages, Flickr, and places where people have indicated the licenses are sufficiently free for Wikipedia. To get an idea of the kinds of terms that can be attached to these, see our article on copyleft.

The second category are images that are in fact copyrighted by others. They are used under a clause of American copyright law known as fair use. It is complicated but basically it says that it isn't a copyright violation to use other people's images without their permission under certain conditions (such as scholarly analysis), and whomever uploaded the image to Wikipedia thinks it satisfies those conditions. Not everyone agrees about this and often such images are removed. The conditions are complicated by our article explains them as best it can. Our policy page at Wikipedia:Non-free content explains how this works in the context of Wikipedia.

The bottom line is that Wikipedia pays no one for its images, but that doesn't necessarily mean that they are being used with permission, or even free for other people to use. You will have to view the individual license pages for any images on here (just click on it) to see what the terms of its use on Wikipedia are, and what implications that has for its copyright status. --Mr.98 (talk) 03:25, 30 January 2013 (UTC)

You apparently misunderstood what I meant by "free". In the context of a Q about whether we pay for them, this means no, we don't: "They are cost-free". You must have thought "free" meant something else. StuRat (talk) 04:09, 30 January 2013 (UTC)

The personal attack portion of 98's comments could justifiably be zapped. ←Baseball Bugs ^{What's up, Doc?} carrots→ 04:17, 30 January 2013 (UTC)

That is still completely incorrect. They are not all offered up "free". We take them and use them and don't pay anyone but that doesn't make them free. Under that definition, anything is "free" if you didn't pay for it — whether you got it legitimately or not. It's not really "free" if the reason you didn't pay for something is because you figure you probably won't get sued over it,a and if you did, you might be able to claim that it its into a legal loophole. (There is a reason that the relevant policy page is called Wikipedia:Non-free content.) Bugs, personal attacks? Where? "He should really know better. I suspect he does." Do you really think that's a personal attack? Again, Bugs — grow up. Some of us are trying to help people get correct answers, here. You tell me how your contribution above does that. Assuming that a long-time user like StuRat would have at least a passing familiarity with how images work here on Wikipedia is not a personal attack. --Mr.98 (talk) 18:34, 30 January 2013 (UTC)

Saying "that is incorrect" would have sufficed. Attacking the other user's integrity in front of the OP is not appropriate, and you should know better than to do that. ←Baseball Bugs ^{What's up, Doc?} carrots→ 18:55, 30 January 2013 (UTC)

All Wikipedia pics are free as in free beer. Some are free as in free speech. See Gratis versus libre. --Jayron32 04:35, 30 January 2013 (UTC)

Not all are free as in free beer. Some are stolen as in stolen beer. Things people steal may not cost them any money, but that doesn't make them free. The flowers in my garden are pickable by any passing stranger, but that doesn't mean they have the right to just take them without my permission. They might be freely accessible, like many of the photos we're using, but they are not free. -- Jack of Oz ^[Talk] 07:50, 30 January 2013 (UTC)

True, but no money exchanged hands to cause the pictures to be displayed here. Stolen beer still doesn't require any outlay of funds. Stolen is merely a subset of free. --Jayron32 13:59, 30 January 2013 (UTC)

No, it's not. Free means the thing is morally available to you, whether you take it or not. Stolen means it may be physically available, but not morally available - but you take it anyway. -- Jack of Oz ^[Talk] 14:14, 30 January 2013 (UTC)

That would be free as in speech, free as in freedom as in liberty. Free as in beer only refers to outlay of money. --Jayron32 20:10, 30 January 2013 (UTC)

I think we can all agree that calling fair use images "free" confuses things more than it clarifies them, even under a lay (and not legal) definition of "free." --Mr.98 (talk) 18:40, 30 January 2013 (UTC)

I thought the word was clear enough, based on the context of the Q, but apparently not. I've now clarified my response. Are you going to remove your personal attack, which was based on your misunderstanding of what I said ? StuRat (talk) 18:47, 30 January 2013 (UTC)

I have to back up StuRat on this one. The questioner asked whether Wikipedia paid for pictures. Stu responded that, no, Wikipedia does not pay for pictures, which as far as I know is perfectly true (though I suppose there is no absolute bar on the Foundation paying a photographer to create a work under an acceptable license, if it would be of sufficient value to the encyclopedia). I'm somewhat at a loss to explain why the reaction has been so harsh, unless it's coming from the anti-fair-use POV. --Trovatore (talk) 04:50, 31 January 2013 (UTC)

• All entertainment at wikipedia such as that above ��is free. μηδείς (talk) 12:16, 30 January 2013 (UTC)

There are many sources of images within Wikipedia.

• Pictures donated by the people who took them under a variety of licenses that allow free use by anyone.
• Pictures that are so old that copyright on them has expired.
• Pictures from sources such as the US government that are in the public domain.
• Pictures that ARE copyrighted - but which Wikipedia uses under the tricky "fair use" provisions of copyright law. Generally we can only do this if there is no other photograph available, and there is no reasonable possibility of getting a replacement, and we make no profit on it, and we don't compete against the owner of the copyright for business, and if we use a reduced resolution version, and if we only use it as an essential part of an article about the subject (ie, not just as decoration). The precise details of what allows us to use such images can be tricky.
• Pictures that really shouldn't be here - people upload photos without the legal right to do so all the time. We try to get rid of them as fast as we can - but if people lie about the status of an image, then it can be very hard for us to find and remove them.

When you re-use a picture that you find on Wikipedia, you should be especially aware of the last two possibilities. Just because Wikipedia can claim excemption under the "fair use" provision of the copyright law, doesn't mean that you can! You can find out what rights we have to the image by clicking on it and reading the information on that page.

If you need pictures for your own use that are NOT "fair use" - I'd suggest you use Wiki Commons which is where Wikipedians put pictures that are believed to be truly free for anyone to use.

SteveBaker (talk) 16:40, 30 January 2013 (UTC)

Steve, that's a pretty good summary, but I think the bit about "make no profit on it" is not quite right. Of course the Foundation doesn't make a "profit" per se on anything in Wikipedia, but there is no general prohibition against claiming fair use in a context where you do make a profit. That a use is noncommercial can be one aspect of a fair-use defense, but it is not in general required. (Otherwise, for example, you couldn't quote from another book, in a book you're planning to sell.) --Trovatore (talk) 04:02, 31 January 2013 (UTC)

Oh, also, you're somewhat conflating fair-use law with the internal requirements English Wikipedia has set up for the use of fair-use media. The Wikipedia requirements, in my non-expert view, appear to be rather conservative, probably because there is a large contingent here that is hostile to allowing them at all. I do not think, for example, that the "essential part of an article" thingie corresponds to anything in the law, though I'm willing to be corrected by anyone who actually has expertise in the area. --Trovatore (talk) 04:17, 31 January 2013 (UTC)

While you're correct we go beyond the fair use requirements for a number of reasons, a well acknowledged fact, I think the example you chose isn't great. Steve specific wording included the phrase i.e. not just for decoration so it's clear they're not thinking of a middle ground. While as with most things in law, you should never say never the fair use requirements make it far more difficult to sustain a claim of fair use if your use is solely decorative as in that case there's no good reason why you have to use the copyrighted work and it's harder to be 'transformative'. You don't have to take my word for it [1] [2] basically say he same thing. We do generally require a far greater importance the is likely to be needed for a fair use claim, but as I said Steve didn't seem to be thinking of a middle ground. A better example would be excluding fair use images from anywhere besides the encyclopaedia proper, including here and somewhat controversially the main page. In some of these and other cases, a fair use case could be easily sustained. We also generally disallow images of living people although I

think the use of fair use is complicated, in real life it often doesn't come to it becuase people often se publicity photos and similar where they have limited permission. Nil Einne (talk) 18:53, 31 January 2013 (UTC)

Yes, fair use is a tricky matter. The problem is that the law doesn't define specific steps you have to take in order to take advantage of the fair use provision. You're left with a pile of semi-requirements that are taken together by a judge, weighed and balanced, and an opinion formed. For a site like Wikipedia, that's horrible! In cases where we're on the margins, we rely on editors and admins to weigh that same evidence and (hopefully) come to the same decision that a judge would. That's really unlikely to produce good results! Hence we over-compensate. If you want to see how messy and complicated this can be, check out the image being discussed in our Lenna article and the fuss that this created on the talk page, and on the talk page of the image itself. The image was uploaded here in 2005, discussion over whether we can use the image under fair use started in 2007 and flared up again several times later - culminating in a huge debate in 2011 - and still ongoing as contributory evidence for the possible de-sysopping of one of the Wikipedia admins!

Given that level of difficulty, you can see why there is considerable call for the banning of all such "fair use" images on English-language Wikipedia. The practice of using "fair use" images has already been banned from German-language Wikipedia (the http://de.wikipedia.org/wiki/Lena_%28Testbild%29 article has no picture of "Lenna"). On the other hand, the result of that blanket ban can be some fairly ridiculous problems. Take, for example, the Andy Warhol article. The English-language version includes a low-resolution photograph of Warhol's famous soup-can painting...under "fair use", we are allowed to show this picture as an aid to discussion of Warhol's art. It's not there anymore - but for a long time the German article had a photograph of an actual can of soup instead! I believe that that image was removed because the actual soup can label is also copyrighted and a close-up photograph of it would have to be used under fair use too! However, Wikipedia articles about copyrighted works of art are considerably diminished by not being able to show the reader a picture of the artwork being discussed! The German article about the "Lenna" image is a fairly useless stub because of that.

Hence this remains an open battleground within Wikipedia with good arguments on both sides. The idea that failing to display images that we're legally allowed to display makes for a much worse encyclopedia is in direct conflict with the idea that Wikipedia may be legally copied and mirrored, quoted and re-used by other sites - where those same images might not be legal as fair use. The balance between Wikipedia's "openness" and it's usefulness as an encyclopedia is a difficult one.

That's why our OP has to be careful when re-using images from English-language Wikipedia - they are not all freely usable for just any purpose. If you took the picture of the soup can from Andy Warhol, printed it out, framed it, and sold it in a store - I'm fairly sure that you'd be sued for copyright infringement - and you'd be unable to claim "fair use" as Wikipedia could if it were to be sued for displaying it in that article. On the other hand, the photograph of Warhol's childhood home from the same article can be printed, framed and sold perfectly legally.

SteveBaker (talk) 13:57, 1 February 2013 (UTC)

Clinical trials funded by pharmaceutical companies

Are clinical trials funded by pharmaceutical companies but conducted by reputed universities reliable? I mean is it possible for a company to manipulate the result in their favor? What are the measures taken to prevent this? --PlanetEditor (talk) 03:09, 30 January 2013 (UTC)

It's probably not easy for them to overtly influence the outcome, as that could get them in major trouble. On the other hand, they might tend to select universities, which, in the past, have provided results favorable to their company. Knowing this, universities might tend to subtly alter their results to accommodate the companies. StuRat (talk) 03:23, 30 January 2013 (UTC)

They can exert pressure, most easily by digging up flaws in the methodology if the result comes out in a way they don't like. That happened in the lab where I was a graduate student, and it was a pretty unpleasant experience for my advisor (who didn't give in to the pressure, by the way). Looie496 (talk) 03:33, 30 January 2013 (UTC)

Ah, a fine example of a StuRat-makes-stuff-up answer. References, who needs 'em? --Mr.98 (talk) 03:34, 30 January 2013 (UTC)

"Another really famous case was where an investigator did a clinical study that was sponsored by a manufacturer of a thyroid hormone, a drug called Synthroid, and the university didn't have publication rights. When the investigator submitted to a medical journal, the sponsor made him pull it." [3]. (It doesn't sound like the paper submitted was sufficiently positive.) StuRat (talk) 04:23, 30 January 2013 (UTC)

The full quote is "When the investigator submitted to a medical journal, the sponsor made him pull it. I think that was a case where the university really learned the importance of protecting the academic freedom of investigators." It's not talking about submitting faulty information to the FDA. To be clear, I'm not defending the actions made in the allegation... but I'm just saying, that snippet from a TV interview is out of context given the OPs question. Shadowjams (talk) 08:20, 30 January 2013 (UTC)

I disagree. This is one example of how the relationship between a university and pharma led to the suppression of a study by the uni. The question wasn't limited to the FDA, it didn't even mention it. One way to "manipulate results in your favor" is to publish favorable studies and suppress the rest. Also see Richard Avery's comment below. StuRat (talk) 16:55, 30 January 2013 (UTC)

"They found that there was a statistically significant relationship between funding source and qualitative conclusion. Unfavourable conclusions were reached by only 5% (1/20) of drug company-sponsored studies, compared with 38% (9/24) of non-profit sponsored studies." [4]. StuRat (talk) 04:35, 30 January 2013 (UTC)

Hmm, this makes sense. --PlanetEditor (talk) 06:54, 30 January 2013 (UTC)

Add to that the fact that hundreds of trials are undertaken each year the results of which are never published because the results are unfavourable to the pharmaceutical companies. This hidden information is an international disgrace, Ref: Bad Pharma. Richard Avery (talk) 08:02, 30 January 2013 (UTC)

The FDA requires extensive testing in what are known as clinical trials, which have distinct phases. The drug companies themselves pay for those trials. The vast majority of them end in failure. We have at least 300,000 articles on soccer, but we don't apparently have an article on FDA drug trials, which we should, which would go into detail about the specific steps. New Drug Application appears to be the closest we have to that procedure.

Long story short, clinical trials are necessarily funded by the drug companies, and no, those, under the supervision of the FDA show very little evidence of being biased. The penalties for falsifying data are high, and the benefits are slim (because it won't take long to find out). You might look at Vioxx to see an extreme example... one in which the process wasn't perfect. The FDA gets as much criticism btw for not approving drugs that are approved in other areas (mainly Europe). Shadowjams (talk) 08:06, 30 January 2013 (UTC)

Not that I disagree with the gist of your statement, but I think there's been evidence of a problem in the way clinical trials were done, namely while drug companies may not have much influence over the outcome of the trials, they frequently have had the choice whether and when to publish them, even in cases where it's an approved drug. This has been an issue in the Vioxx case but is also generally considered a wider issue. [5] [6] [7] [8] [9] [10]. Some of the data from the trials may be provided to the FDA (and other regulatory agencies), but this still makes it difficult for others to evaluate the available data such as in a systematic review, for example when deciding which drug is the best choice. Things have changed a lot in the past few years so things aren't quite so bad although many still think we have some way to go. Nil Einne (talk) 11:52, 30 January 2013 (UTC)

The publication bias is an under-recognized (until recently) phenomena and it creates a serious problem in all kinds of research. But... and I don't know enough to answer this... does that apply to FDA trials? My gut instinct is that they have to disclose (maybe not publish) everything they find. I know that's true for adverse effects. Vioxx got pulled I believe because of post-approval testing like that. So I guess I question if publication bias has bearing directly on FDA trials. Shadowjams (talk) 16:37, 30 January 2013 (UTC)

First, this question isn't about FDA trials, where did you get that from ? If doctors become aware that an FDA approved drug has problems, due to published negative results, they will be less likely to prescribe it. This is why suppression of those studies is so important. And, even if we limit the discussion to the FDA approval process, which for some reason you want to do, such studies could affect that, too. That is, even if they aren't submitted to the FDA, if the FDA becomes aware that there are studies out there showing alarming results, this might make them ask for additional studies and slow down the approval process. StuRat (talk) 17:05, 30 January 2013 (UTC)

It is about clinical trials, as opposed to clinical studies. The term "trials" has a particular meaning, and I use the FDA as an example. Perhaps China and Europe have other approval processes that are worth mentioning, but I'm unfamiliar with those. Approved drugs often (if not always) have a post-approval process that requires monitoring of them, so publication is hardly an issue. If the drug starts having serious side effects (or even unserious ones) they have to be reported. Failing to do so sends people to prison. Shadowjams (talk) 17:24, 30 January 2013 (UTC)

Trials doesn't automatically mean FDA trials. We just had another Q on how side-effects often go unreported. As was stated there, doctors who find possible side effects are not required by law to report them, and often don't. StuRat (talk) 18:09, 30 January 2013 (UTC)

One of the things I tried to emphasise, but perhaps failed to, is that while I'm not saying that the FDA, or any of the major regulatory bodies do bad work, most people recognise it's a bad idea to just rely on the FDA. An important part of science is the ability for random other people to review the results, where they may for example catch things that may have missed, or simply disagree with the conclusions. The same for the FDA. And in fact, it's not just about disagreeing but having the best information available. As I did mention, if there are multiple drugs you have to decide somehow which one to use. I suspect often the decision isn't actually that well considered, but when it is, it's difficult to make a good decision if for one drug you're considering there are a bunch of studies which showed it didn't do much, but these were never published. (I don't think relying on the fact most drugs have some studies which were never published so it balances out works.) The conclusions of any meta-review is far more likely to be somewhat flawed if it's only reviewing half or less of the data. You'd note that most of the sources concentrate on negative results but the safety aspect can come in to it as well. If the studies suggesting side effects are common for one drug were never published, the decision process on which drug to use is also likely to be flawed. While the FDA (and other such bodies) do require safety information in particular to be distributed and also have some role in gauging effectiveness and of course in making sure claims by manufacturers are accurate, it's best if professionals can actually see the data. Just for emphasis, this doesn't have to mean the FDA or whoever made a mistake in the approval or post approval process (although there was some suggestion they were too slow in the Vioxx case). It's possible the drug should be on the market even if far fewer people are likely to choose it. P.S. In case it isn't obvious, I'm primarily referring to post approval. P.P.S. I mentioned there being multiple candidates but of course in some cases the decision might be a drug vs surgery, or treatment with a drug vs no treatment (or perhaps just palliative). Nil Einne (talk) 12:50, 31 January 2013 (UTC)

Or the case that bothers me most, where they prescribe a drug with potentially lethal side-effects, rather than getting the patient to eat right and exercise, say to lower their blood pressure. StuRat (talk) 07:03, 2 February 2013 (UTC)

I think two things are being mixed up here. Big pharma can fund pre-clinincal research in universities, where bias and pressure are most definitely a risk and have happened, as StuRat has correctly cited. Then there are Phase 1 and further clinical trials, which are done under the authority of the local drug regulation agency, eg FDA. Those are generally very well done, although this being science, something can always go wrong. And Shadowjam, why would we need an FDA specific trial article when we have the perfectly fine clinical trial article? — Preceding unsigned comment added by Fgf10 (talk • contribs) 09:27, 30 January 2013 (UTC)

Because the largest pharmaceutical producing nation in the world has a very specific procedure for drug approval that involves literally billions of dollars on important drugs, let alone in aggregate. For all the articles we have on miscellaneous topics, this one seems rather important. But it's a lot harder to write than copy pasting in box scores. Shadowjams (talk) 16:39, 30 January 2013 (UTC)

I wouldn't say the FDA trials are necessarily done well: a company wants to get approval as soon as possible, so they go for short trials that would suffice the FDA criteria. That's why in the case of SSRI's for example, the difference with placebos is small, but enough for the FDA. Others have used these as proof that SSRIs don't work,, which I think is a misrepresentation. If the FDA had set higher standards, the companies would have chosen trials of longer duration. Four weeks for an antidepressant like prozac, that has such a long half-life it takes weeks to reach a steady state is too short for good results. Ssscienccce (talk) 17:06, 30 January 2013 (UTC)

That's a bold statement to base on one case, a case that our article on the subject dispels within the lead ("For patients with very severe depression, the benefit of medications over placebo is substantial"). There's way more than one 4 week study done on an approved drug. To characterize prozac's approval as a 4 week study is incorrect. Shadowjams (talk) 17:20, 30 January 2013 (UTC)

That's not what I meant, prozac was just an example. Of course there was more than one trial, and also trials longer than four weeks; but because the FDA disregards negative results (or did so in the 80s), there's no harm in trying to get a positive result with a short trial, one can always get lucky. My point was that if you want to know how good or bad SSRIs work, these FDA approval trials won't give you that information because they were not designed for that, all they had to do was show a statistically significant effect. That meta-study mentioned in the lead did exactly that, use the data to measure something that the tests weren't designed for. Ssscienccce (talk) 18:23, 30 January 2013 (UTC)

Ok I see your point... I think I had a different conception of what you meant by "done well." I think they're done quite well as far as accomplishing their stated goals, like safety, and having at least some efficacy. But you're probably right in a broader picture about them falling short of some other goals. Our publication bias article's pretty good--pre-registering studies for instance--but my ultimate point in this discussion is that drug-approval clinical trials are designed to avoid this as best possible by requiring all collected data to be revealed. Shadowjams (talk) 19:06, 30 January 2013 (UTC)

Ben Goldacre has written Bad Pharma, in which he pretty clearly dislikes this apparent conflict of interest. You might want to give it a read if you after some references (I'm assuming he has them in there at least!) 80.254.147.164 (talk) 12:38, 30 January 2013 (UTC)

I did a graduate level study last year on dental implants sponsored by an implant company -- we put over a hundred implants into 18 dogs and those dogs were special purpose-bred (about $700) and had to be maintained in an veterinary OR kennel under appropriate supervision for 18-37 weeks, like special diet, oral hygiene regimen, etc. I think the cost was somewhere in the ballpark of $200K, which did not include the write-off for all the implants/components donated by the company. This information I give you just so you can get an idea of the fees that can be associated with clinical research, and this is just dogs. Primates can run $5K each and the housing and care is likely more expensive. So this gives research, as a whole, the strong motivation to obtain corporate sponsorship -- much like the Insurance Institute for Highway Safety will run crash and safety tests on any vehicle type as long as the vehicle company provides 2 or 3 free specimens for running the various tests. I'd speculate that the best way companies can interfere with research is with some form of promoting publication bias in that only a favorable result will be published. Now, this didn't happen with my study, and one can't say for sure when it does occur, but if publication is delayed or altogether averted post-hoc because of bad findings, only the good news gets out. DRosenbach ^{(Talk | Contribs)} 16:44, 30 January 2013 (UTC)

That's a good point. I think an important point that's easy to gloss over is that "publication" in a journal is not the same as a study, and a lot of things are reported to the FDA (or whatever governing body) outside of just "publications," which take months if not more to get reviewed and published. Because you're familiar with the process, do you think that's a valid distinction? Shadowjams (talk) 17:27, 30 January 2013 (UTC)

I can't speak for drug studies, because that's not my field, but for dental implants and periodontal surgery -- which are the leading disciplines that contribute the corpus of dental literature -- unpublished papers are nearly meaningless. In fact, even published papers in unrespected or sub-par journals garner sniggers. I mean, if you have some important new findings on cloning cell metabolism, you don't publish in Popular Mechanics, and if you do, academics wonder about your methods, your data or your findings. So, too, if there's some great new finding (or even a mediocre new finding) that's important, why publish in the Australian Dental Journal if you could have bee published in Journal or Clinical Periodontology. DRosenbach ^{(Talk | Contribs)} 17:59, 30 January 2013 (UTC)

You might all be interested in the graph at the bottom of this page (as it's a British Medical Journal site I feel confident in its figures!). --TammyMoet (talk) 19:07, 31 January 2013 (UTC)

Turbo tax ownership

who owns turbo tax — Preceding unsigned comment added by Hopper6713 (talk • contribs) 19:26, 30 January 2013 (UTC)

Intuit Consumer Tax Group. In the future check the Turbo Tax article first. Shadowjams (talk) 19:54, 30 January 2013 (UTC)