Close
Is all research created equal? Certainly not! But, with the new age of information at the fingertips, it becomes challenging to discern what is reliable from the not-so-reliable that often slicks our screens.
Scientific research is divided into 6 categories that range from low to high impact. The diagram below illustrates the divisions of research in order from least to most reliable:
Starting from the bottom, descriptive quantitative research is case studies, observational and ‘anecdotal’ evidence. A researcher watches and interviews subjects, then documents the findings. Evidently, this research is ranked low based on quality and efficacy because of its capacity for human bias and subjectivity. For example, “my close friend Joe lost 89kg by cutting carbohydrates…” – Joe’s story is portrayed as fact without any verification…very questionable indeed.
Descriptive quantitative research can be ranked and measured in a structured manner. It is ideal for descriptions of a particular group or population, as results are reported statistically in numbers and figures.
Correlational regression analysis…you’d do well to casually drop this into conversation! This research analyses the strength of relationships between two or more variables. For example, is carbohydrate intake associated with onset of Diabetes? From any results using this research method we could never say that carbohydrates cause Diabetes but only that the two are perhaps correlated or linked.
Quasi-experimental (another good name drop!) compares one group that receives intervention with another group that is matched on all other areas, except receives no intervention. Real life example: does eating a low carbohydrate diet increase risk of vitamin deficiency? Two matched groups eating a similar diet are designated, but one group eats a markedly lower carbohydrate content than the other. It is imperative that all other contributing factors such as exercise and total energy intake are controlled. Other elements improving the quality of this research: longer study length, greater group similarity and greater participation numbers.
Finally the good stuff! Experimental design uses randomisation and is known fondly as RCT evidence (randomized control trial). Subjects are assigned randomly (with neither the researcher nor the participant knowing) to either the control group (no intervention) or the intervention group. This is the gold standard of scientific research because with little to no bias and a sound study design, one can deduce causation ie. (For arguments sake!) a low carbohydrate diet causes risk of vitamin deficiency.
We can go a step further and group results from a number of similar studies to determine the impact across a large number of participants. This is called a meta-analysis and it packs quite a punch in research circles. Large meta-analyses of RCT trials is considered the crème de la crème of all research and certainly trumps a collection of anecdotes like our friend Joe’s, above.
Statements made by Prof. Time Noakes such as “90% of all nutrition research is association based”, is inaccurate because any guideline from a Nutritionist or Registered Dietitian, is substantiated by sufficient, conclusive RCT evidence. Don’t be fooled by research that sounds impressive but has low validity ranking based on the scientific evidence. Broaden your mind, yes! But be sure to verify all that you read by pursuing your own search for the reliable research.
References:
admin@nutritionalsolutions.co.za
Bryanston: (011) 463 5502
Waverley: (011) 023 8051/2
Pretoria: (011) 463 5502/074 808 2130
Paarl: (011) 463 5502 / 084 571 8009
JHB South: (010) 035 4800
