Quality of Scientific Evidence

To enhance our science communication, it’s crucial to comprehend the varying quality of scientific evidence. Not all evidence holds the same weight.

Expert knowledge ranks as the least reliable justification in science. While appealing to authority can be valid, it only works when no reasonable counterargument exists—which is often not the case.

Case reports and case series provide slightly stronger evidence. They describe individual cases or a collection of similar cases, such as documenting a patient’s fever and recovery after taking aspirin.

Following that are animal studies, which serve as useful models but can be flawed due to significant biological differences between humans and animals. While ethical considerations favor animal studies, their applicability to human health can be limited.

Cross-sectional studies gather population data and can reveal correlations but fail to establish causation.

Case-control studies improve upon this by identifying individuals with and without a condition and investigating prior exposures. Though less costly, these studies are retrospective and may lack robustness.

Cohort studies are more reliable as they are prospective, tracking individuals over time based on exposure to a potential cause, thus minimizing random chance effects.

Randomized controlled trials are the gold standard, offering the most reliable primary studies by minimizing confounding variables, though they can be costly and ethically challenging.

At the top of the hierarchy are systematic reviews and meta-analyses, which evaluate multiple studies rather than individual phenomena. They provide a comprehensive overview of research quality.

Other ways to evaluate studies can include: - Systematic reviews are more reliable than individual studies. - Human studies are more credible than animal studies. - Experimental research surpasses observational studies. - Blinded studies yield stronger results than non-blinded ones. - Larger studies provide more robust data than smaller studies. - High statistical significance is preferable to low significance.

Scientific justification can range from anecdotal to robust, with the latter often indicating a well-supported theory. The ideal scenario involves a systematic review indicating strong, consistent high-quality studies backing a theory.

Though this analysis focuses on medical research, similar principles apply across other scientific fields. Systematic reviews can aggregate findings in disciplines like physics and chemistry, allowing for greater control over variables in experimental designs.

Proper Word Choice

While discussing evidence quality is vital, it’s equally important to understand the nature of science itself. Scientific theories cannot be confirmed or proven true; this is known as “the problem of induction,” a long-standing issue in the philosophy of science. Therefore, terms like "confirmed," "proven," and "fact" are inappropriate in science communication. Even beyond this philosophical dilemma, uncertainty persists, underscoring the need to scrutinize evidence quality.

Sensationalism can be mitigated by avoiding misleading language and recognizing the evidence hierarchy. Headlines proclaiming "the cure for cancer" or similar claims often misrepresent research due to inappropriate word choice.

Citations

An article I encountered discussed mercury levels in freshwater fish. Wouldn't it be beneficial to read that article directly? It’s frustrating when sources lack citations or provide references to other articles without proper attribution.

This lack of citation parallels the issues of sensationalism in science communication. Proper citation allows readers to verify original studies and assess their accuracy.

With hyperlinks, it’s simple to link back to the original research. While not mandatory, including a link early in the discussion and a full citation at the article’s conclusion enhances the reader's experience. I utilize various tools to simplify citation creation.

Citation Machine

Citation Machine is an excellent tool for generating citations in various formats, including APA, MLA, and Chicago. It can pull data from websites, scholarly publications, and books. Both free and paid versions are available.

doi2bib

doi2bib allows for easy citation generation in BibTeX format from digital object identifiers (DOIs). It’s particularly useful for those using LaTeX, as it facilitates easy organization of potential sources.

BibTeX entry from URL

This extension generates a BibTeX reference from the active URL, streamlining the citation process.

BibTeX to APA

BibTeX to APA converts Google Scholar BibTeX references into APA format. While primarily in French, it’s a handy tool for quickly transforming citations.

While I sometimes take a relaxed approach to citation in blog writing, it’s exasperating when articles claim to reference studies yet fail to provide accessible citations. This poor practice is prevalent in various media, including prominent outlets like The New York Times. Linking to original studies or including citations at the article’s end is straightforward, and it’s essential to avoid sensationalism while accurately interpreting study results.

Summary

Proper citation is crucial in articles. Always link to scientific papers when referenced—don’t bury citations midway through the discussion. Avoid sensational language like "proven" and "fact," and always consider the quality of the evidence being presented.

Originally published at the Freelance Journalism Alliance of the Guild Association

This article serves as a concise overview of the necessary approaches to science communication and journalism. The Freelance Journalism Alliance intends to use this information, along with additional insights from various sources, to create a series of MOOCs.