What words come to mind when you hear the word "accuracy?" Do you think of an archer nocking an arrow to a bow and then hitting the center target? Or perhaps you think of a diagnostic medical test's reliability. When conducting research, you should consult resources that are "accurate." But how can we determine accurate?
The Oxford English Dictionary defines accuracy as: "The extend to which something is accurate; precision, exactness," or, "The Closeness of a measurement, calculation, or specification to the correct value."
Stemming from these definitions, in research, we should seek resources that provide information that is reliable, truthful, free from explicit bias, and free from errors.
Accuracy matters. Bad information can have bad consequences.
Accurate resources are written by field experts, cite scholarly sources, describe the research methodology employed, and transparently provide data that was collected or used in the study.
Guiding questions:
- Can I verify this information in another scholarly source? Have the study's results been replicated?
- If the information is accurate, similar findings should be corroborated by other authors and/ or authorities. This is discussed more fully in the "Lateral Reading" tab.
- Take for example the room-temperature semi-conductor. While this made a splash in science news channels, the work could not be replicated in other labs.
- If the information is accurate, similar findings should be corroborated by other authors and/ or authorities. This is discussed more fully in the "Lateral Reading" tab.
- Are citations provided in the work?
- Extraordinary claims require extraordinary evidence. Accurate sources will reference verifiable literature.
- Is the resource peer-reviewed (referred) or edited?
- Peer-review is a rigorous process where an article is reviewed by a panel of researchers before the work is published in a scholarly journal. See more about peer-review below.
- Is there any obvious bias or conflict of interest?
- Bias comes in many forms-- political, commercial, systematic, and personal, among others. We each have our own worldview that informs how we engage with the world around us, which is not necessarily a bad thing. What we do need to be on the lookout for are biases or conflicts of interest that skew information, intentionally mislead, or misinform.
- For example, should we trust a pharmaceutical company's analysis of their own drug, or should we look to third party researchers to confirm the drug's efficacy? We should look to additional research teams!
- Bias comes in many forms-- political, commercial, systematic, and personal, among others. We each have our own worldview that informs how we engage with the world around us, which is not necessarily a bad thing. What we do need to be on the lookout for are biases or conflicts of interest that skew information, intentionally mislead, or misinform.
There is a lot to consider when evaluating accuracy. Start slow by incorporating the guiding questions one at a time into your work, or choose the questions that best fit your information need. As a researcher, ask bold questions about the resources that you want to consult in your work.
Ethics in Peer-review
Peer-review is a process where a source is evaluated before being published in an academic journal. Peer-review teams are usually comprised of an editorial board, with each member contributing their discipline expertise. There are different types of peer review, including open, single blind, double blind, and triple blind, among others.
The peer-review process is intended to provide articles with greater accuracy, as the source is evaluated by other experts. Peer-review is often very rigorous, and authors usually complete several rounds of review before their work is published. This provides authors with opportunities to make changes to their work, to create a stronger final research product.
The intent of peer-review is noble and good-- but it can also easily go awry. 
Factors that complicate peer-review:
- Bias: Reviewers my have conscious or unconscious bias related to the author, author's institution, and/or research topic. This can lead to unfair article acceptances or rejections. Additionally, reviewers may be subject to confirmation bias, where they only accept research that aligns with their own knowledge and theories.
- Lack of reviewer expertise: While the goal is to have editorial boards comprised of experts, when highly specialized research is presented, not all board members may have rigorous experience with the topic.
- Conflicts of interest: Reviewers might have personal, professional, or financial ties to other authors, institutions, or funding agencies.
- Unethical research practices: Peer review does not always catch unethical practices such as plagiarism, data fabrication or manipulation, and other forms of research misconduct. Take for example Redaction Watch, a blog devoted to highlighting academic articles, often published, that are now redacted.
What should I do now?
When consulting an article, do some background research about the journal. Many databases provide hyperlinks to journal information. Or, you could preform a Google search to find the journal's publisher, which should also have publication information. Look outside of the journal itself to find resources that point to the resource's creditability.
Even though an article may be peer-reviewed, you still need to keep your eyes peeled, which is why source evaluation is so critical!
Cartoon made by Hilda Bastian and is available under a CC BY-NC-ND license. More information is accessed here.