Educate Yourself
As we know, information is more accessible than ever. Unfortunately, increased accessibility comes with consequences. Most anyone has access to platforms that reach a vast number of people. There are no reliable measures in place to filter the truth from half-truths or even blatant attempts at misinformation. Instead the truth must contend with a collective lack of research literacy; proliferation of fallacies; and even lapses in judgement by those, who have been trained to decipher information.
Without any training or knowledge of research or any particular topic, a powerful tool to equip in the pursuit of objective knowledge is skepticism. With skepticism, one is able to more readily discard bias and scrutinize information for errors or intentional manipulation. What is there to gain from this research? Who benefits from this research? Who conducted this research? Who approved this study? What studies contradict each other and why? What are the study limitations? How recent is the research? Is their an authoritative study for this topic? There are a multitude of questions, which can and should be asked before information or research is accepted as concrete. Once information is acquired and deemed valid, there is much more to consider.
Think about something like cholesterol. Over the years we have seen a number of conflicting trends, which were based on science and even promoted by authoritative bodies. For instance, a well accepted meta-analysis titled: “Blood cholesterol and vascular mortality by age, sex, and blood pressure: a meta-analysis of individual data from 61 prospective studies with 55,000 vascular deaths,” determined that high cholesterol is a predictor of higher mortality rates in those 40 and above (Lewington et al., 2007). This premise was and is widely accepted amongst the general public and scientific community with seemingly convincing support. More recently, others have opposed this premise. Dubroff (2017) asserted that correlation does not always mean causation and that this case is an example (DuBroff, 2017). Ravnskov et al. (2016) conducted a review, which showed inadequate and even inverse findings, which opposed low density lipoprotein cholesterol (LDL-C) as a direct causal agent of increased mortality. How many people have been exposed to this information? How well will these studies hold up over scrutiny and time? Which studies are correct?
Here is, perhaps, a more familiar example. Growing up, some may have hear the age old advice that “breakfast is the most important meal of the day.” In contrast, now studies are finding that intermittent fasting (IF) and fasting can be valuable tools towards dis-ease reversal and modulating insulin sensitivity. A 60 second search on google.scholar.com produces an onslaught of supportive data. For example, Malinowski et al. (2019) discusses how IF creates an adaptation, which increases insulin sensitivity and moderates blood glucose levels for long periods of time.
Clearly, as a whole, we are in a constant state of learning and reevaluation. We are continuously being presented with new ideas, which is immeasurably vital towards growth. However, it is crucial that – as a society – we learn how to select and assimilate the most reliable information possible through critical thinking and personal reflection in the absence of propaganda and marketing campaigns. In the spirit of deeper analysis, below is a glossary containing important research terminology.
References
Dubroff, R. (2017). Cholesterol paradox: a correlate does not a surrogate make. BMJ Evidence Based Medicine, 2017(22), 15-19. https:doi.org/10.1136/ebmed-2016-110602
Link: http://cardiacos.net/wp-content/uploads/ArticulosMedicos/20180920/2016-Cholesterolparadox_-acorrelate-does-not-a-surrogate-make.pdf
Related Link: http://ebm.bmj.com/content/22/1/15#ref-list-1
Lewington et al. (2007) Blood cholesterol and vascular mortality by age, sex, and blood pressure: a meta-analysis of individual data from 61 prospective studies with 55,000 vascular deaths. The Lancet, 370(9602), 1829 – 1839. https://doi.org/10.1016/s0140-6736(07)61778-4
Link: https://pubmed.ncbi.nlm.nih.gov/18061058/
Ravnskov et al. (2019). Lack of an association or an inverse association between low-density-lipoprotein cholesterol and mortality in the elderly: a systematic review. BMJ Open, 2016(6). https://doi.org/10.1136/bmjopen-2015-010401
Link: https://bmjopen.bmj.com/content/6/6/e010401.citation-tools
Factors of Research Validity
Evidence-Based: Conjectures, which are accompanied by significant amounts of supportive data.
Peer-Reviewed: Research, which is presented to be subject to the rigors of thorough analysis by qualified professionals of their respected field.
Credentials: Professional designations bestowed by recognized institutions, which give a degree of credibility to individuals.
Conflicting Interests: Research performed with the intent to support a particular end, which benefits the researcher or is to the detriment of the researchers competition.
Methods: Modalities – that should [but may not] follow scientific protocols – which are utilized to conduct valid research.
Confounding Variables: Factors that influence or mask a relationship between other variables.
Missing Variables: Variables, which influence research outcomes, but are not accounted for during data analysis.
False Assumptions & Conclusions: example. All people, who consume water, die. Drinking water is the cause of death.
Control Group: During research a selected number of individuals or animals, who do not receive any intervention. Data from this group is then compared to the group, which received the intervention to determine the significance of the intervention.
Double Blind: Both testers and subject volunteers are unaware of the effectual elements of any interventions or implementations, which eliminates the likeliness for any parties to influence outcomes.
Placebo: Outcomes may be realized through strong belief, which may produce results – even in the absence of a particular intervention. It is imperative that placebo results are compared to the actual intervention being analyzed.
No-cebo: Negative expectations, which produce negative results without intervention.
Study Size: The number of participants within the study. A larger cohort decreases the likeliness of anomalies within the study and is likely to present a more reliable sample of the population.
Bias: The development or design of research to support a predisposition or guide outcomes towards a desired end.
Algorithm Bias: https://guides.lib.fsu.edu/algorithm
Randomization: Selection of subjects are indiscriminate.
Study Length: The subjects of research studied over longer time presents a more complete outcome. Without longitudinal studies, projected outcomes are only educated speculations.
Resources
https://www.ebsco.com/products/research-databases/cinahl-complete
https://about.proquest.com/en/products-services/publichealth/
https://ovidsp.ovid.com/
https://jamanetwork.com
https://www.nejm.org/
https://scholar.google.com/
https://www.healthdata.gov/
https://www.himss.org/
https://aapsonline.org/
https://vashiva.com