How To Find Scientific Information On The Internet Without Getting Duped

Actual scientific data is your best defence against misinformation, but locating real facts amongst selectively-quoted studies and outright lies online can be difficult. We've told you before how to tell if something controversial is actually true, but what if you want to read up on a subject without stumbling into half-truths and pseudoscience? Here's how to use the internet as a powerful research tool without being led astray.

The internet is full of useful, well-documented information, and all of it is right at our fingertips. The problem is that the signal-to-noise ratio can be low. Search engines attempt to separate the real science from unsourced opinions and so-called "experts" only interested in selling books or pushing a particular viewpoint, but those automated services don't always get it right. With these tips, you'll learn how to quickly cut through the weeds and get to the good stuff.

Recognise Your Two Biggest Research Enemies

Before you hit Google and start researching, you have two big obstacles to watch out for:

  • Your own confirmation bias. Confirmation bias is your natural tendency to find, believe and source information that agrees with (or confirms) your already-held opinions about a topic. It's a problem even for highly-educated scientists and experts in their field, and it's something you'll need to be ready to battle when you're looking into a topic that's new to you. You may be presented with information that will challenge your preconceived notions and beliefs. That's OK -- that really just means you need to keep an open mind and seek to understand and find evidence for all sides of an argument (especially the ones you disagree with.) For more on confirmation bias, read this excellent article on the topic by journalist and author David McRaney. Photo by Yi Chen.
  • Questionable sources of information. Possibly the only thing worse than confirmation bias is unsourced, poorly-cited articles that draw conclusions without backing them up. Even good scientists sometimes cite a study that doesn't support their conclusions or report a study's conclusions blindly without examining its validity.

In combination, these two tendencies can convince people to believe even the most tenuous claims when researching a new subject. Keep them in mind and watch out for anything that seems too good to be true, or doesn't pass the almighty sniff test.

Fire Up Your Critical Thinking Skills

If your goal is to read up on current research in quantum physics or understand a specific type of cancer, the first thing to know is that true understanding doesn't come in 10 minutes. If you really want to dig into a topic, you'll have to do a fair bit of hunting and reading. Through it all, you'll also have to make sure what you find is corroborated elsewhere. (Hey, it's called research for a reason).

Open your favourite search engine, and start looking for information on the topic you'd like to learn more about using a handful of relevant keywords. These search engines won't be the end of your search, but they're fine for getting your feet wet.

You'll probably find yourself wading through thousands of results from hundreds of sources (we'll get to how to differentiate the good ones from the bad ones in a moment), but this initial step is a good way to understand the depth of information available for the subject you're researching. It's also a great way to get an overview of what's available so you can refine your searches to get right to what you want to know.

In the initial research phase, you're likely to encounter many pages from Wikipedia. This is definitely a case where checking the references counts. Wikipedia is a great place to read up on the basics, and a good jumping-off point for more reading. Its real power is in its own citations -- even if an article is out of date or inaccurate, its sources are worth checking. It's inaccurate to assume Wikipedia is always accurate or always out-of-date; some articles are well-maintained, others end up ignored or swamped with irrelevant material.

Next, try some of these more scholarly search engines. They'll lead you to more credible, cited information, along with journal articles and reference material that can help:

  • Google Scholar cuts out a lot of material and searches directly for articles and research papers in well-regarded journals.
  • Scirus restricts its indexing to journal articles, papers, scientists' websites and homepages, courseware, government research institutions, patents, and other legal documents.
  • PLOS, or the Public Library of Science, is a non-profit public advocacy group and publisher with the aim to make as many scientific journal articles and paper open and accessible to the public as possible. PLOS's journal publications (including PLOS One, which you may already know,) offer open public access to some of the latest research in a number of fields.

Try to avoid second or third party articles that write about studies or research unless they link to or quote the study in question. If they do, go and read that instead. Relying on third party articles is like playing 'Chinese whispers' -- the actual information from the first party will end up be distorted by the time it gets to you. Often the conclusions of a third-party article or press release can be wildly different from the conclusions in the study itself. Even the most well-intentioned outlets and highly trained journalists suffer from mistaken conclusions.

Learn To Identify Good Sources

Remember those "questionable sources" we talked about earlier? When you start your search, you'll need to learn to separate them from the good stuff. Scammers and hoaxsters go out of their way to disguise marketing copy as "journal articles" and "clinical testing". Here are a couple of ways to tell whether the reference you're reading is legitimate:

  • Is the paper from a real, actively published scholarly journal? Hit Google (or any of the above-mentioned scientific search engines) and look up the journal name. See when the latest edition was published, and if there's any way to read the latest articles (or even abstracts) from it. Look at the journal's own website (it should have one, even if it's just a page by its sponsoring organisation) and see if it's peer reviewed. You might be surprised how often a company will string together impressive-sounding words to make up a journal name and then claim "results" from testing of its product appeared there. One scientist even made up a journal so she would have a place to publish her own "studies". Verification is your friend.
  • Is the paper from a known lab, institution, university and/or author(s)? Beyond the journal that accepted and reviewed the paper, search for the name of the organisation or research institution that sponsored the paper. If it's a university, this will be easy -- but if you find a company masquerading as a research institution, be wary. Similarly, look up the names of the lead authors on the paper. They (or the labs in which they work) should be easy to find, and you should be able to verify their credentials. They may even discuss their recent work on their own web sites.
  • Can you find references to the paper and its authors in multiple places? If you Google the name of the paper, you should be able to find it in multiple places, not just a handful of sycophantic blogs. This is particularly a problem with health news: one study parroted on opinionated blogs can become a "Google fact" even if the study is fake. With most legitimate studies, you should be able to find coverage and discussion in a number of places that both support and criticise it.
  • Is the paper itself well-researched? Can you actually find the citations it uses? Before a researcher paper is published, it must cite relevant evidence. All previous work that led to current research, peripheral work being done by other labs, and any basic theories taken for granted in the paper all have to be noted. You can usually find these citations throughout a paper and at the end for further reading. If the paper you're reading is heavy on ramble and light on citations and references, that should be a red flag. Look up those citations and see if they are what the author claims they are. With luck, you'll have even more to read and more data to use in your own research.
  • Can you easily see who funded the study the paper is based on? Having a study funded by a private entity doesn't make it a bad study in itself. It can and should raise questions over the nature and the conclusions of the research, but if the methodology is sound, there's no reason not to accept its results. However, studies paid for and commissioned by private entities are often used in marketing, so look for independent research on the same topic. Even the lack of independent research can say a lot about the paper you're reading.
  • Can you read the whole paper? This isn't always a deal breaker, but see if you can get your hands on the full text of the paper. We'll get to some creative ways to do so in a moment, but look for journal articles and studies that offer at the very least the abstract, citations, and conclusions. Be wary if you get a webpage with nothing more than a conclusions section, especially if there's no link to the rest of the paper (because it doesn't exist.)

These tips don't just apply to journal articles; they go for articles and entire web sites as well. Granted, third party reporting isn't held to the same standard as peer-reviewed journals, but most well-intentioned publications at least try to cite their sources, offer additional reading, and avoid drawing unfounded conclusions from individual studies.

Learn To Understand Journal Articles

Now that you have a pile of journal articles, it's important to understand exactly what they are and how to read them. First of all, avoid the temptation to assume a journal article represents indisputable fact. Researchers use journal articles to share new research, discuss theory among experts, and as a forum to share knowledge. They are not an end-product, and no lab says "Well, we published a paper, we're done studying cancer!"

Reading a paper can often be daunting. Depending on the field you're reading about, you may be awash in language you're not familiar with, or mathematical calculations you're not familiar with. That's OK -- this is where you get to cheat a little bit. Read the abstract and introduction first and see if the topic is relevant to what you're researching. Read the conclusions to see how the methodology panned out and what the results of the research were, along with suggested avenues of future study. Then go through the data and the methodology if you can, and try to understand how the study was conducted. What was the sample size? How was the study controlled, if a control was necessary? In good papers, all of these things are discussed openly.

The video above does a great job explaining in layman's terms what scholarly journals and academic papers are and how to understand them. Remember, almost every journal article, even if it obtains broad consensus, is a jumping-off point for additional study. It's almost never the last word on a topic, and never should be. The beauty of scientific research is that it's always testing new theories and methods until old patterns and assumptions are either well-explained and duplicated or break down entirely.

Reach Out For More Information

Not all journals are available online. If sitting at home isn't cutting it for you, and you need more or easier access to quality data, you have some options: Photo by Timothy Vollmer.

  • Visit your library. State and university libraries are particularly useful in this context. You'll often have free digital access to scholarly journals of all types, and staff who can advise you where to find specific information. You can usually print or save the articles for future reference.
  • Reach out to scientists and science advocates. Sometimes the best way to get your the full text of a study or to ask a pointed question about someone's research is to ask the person who did it. In many cases, researchers will be happy to provide a copy of their paper to someone who's interested, especially if they're trying to educate themselves. If you have a question, many researchers are happy to entertain them to further understanding of their work. Reach out -- the worst you can hear is nothing.

The idea that you can't trust the Internet is misplaced. The key is to find well-sourced, well-cited information on a given topic before you can trust it. It's not even that difficult to do -- it's just not something that aligns with the fast-paced world of blog comments and Twitter arguments. Becoming literate in a subject area or even remotely knowledgeable requires time, patience and a good bit of actual open-minded research. It's not something you can do in time to respond wittily to someone who made you angry on Facebook.

That said, it's worth doing every time. You'll benefit with better understanding of a topic, and you'll be more prepared to have an informed dialogue, make better health and wellness decisions, or get excited about new research in the future.


Comments

    I don't think you should trust an abstract without reading the article, these often overstate what the research found, after all, it's the scientific equivalent of advertising, if it was boring no one would read the paper.

    This is good stuff to help me do research for my literature review. But what if I have defined my topic and searched for and wide but can not find the correct information i need to support my claims, What do i do now? make a new topic? i am really stuck in a dead end and need some help? please

    Last edited 18/03/13 10:13 pm

Join the discussion!

Trending Stories Right Now