“Only that which can be studied can become evidence based.”
Evidence based training methods
In this essay we will be taking a look at the characteristics of evidence based training methods and some of its possible consequences. As the name implies, evidence based are all the forms of practice that have an empirical basis for their claims. Someone, at some time, has set up hypothesis, research questions, gathered the data and analyzed it thoroughly.
Good scientific research has a couple of necessary components, derived from basic philosophy of science. For one, a scientific claim needs to be falsifiable. You need to be transparent enough so that others may repeat your experiment and answer the same questions with the same tools so your claim can be verified or falsified, respectively. The mechanism you are claiming something about must be distinguishable from chance or from other circumstances.
People then write scientific papers about what they studied and these writings get reviewed by other knowledgeable people in the field. All this leads to a rather large body of studies that then get picked up by practitioners who base their methodologies on that knowledge. The following critique is not about that knowledge might be wrong, or that science is not useful. Rather, it is a laying out of to what kind of practice this will almost inevitably lead:
Short term goals with repetitive tools.
This may seem obvious to those more practiced in philosophy of science, but a common error about the scientific method and its resulting truth claims is the confusion of these two propositions:
a: Scientific methods lead to true claims
b: There are no true claims that are not scientific
Given a perfect method, for arguments sake, statement A may be correct, but that does not mean that B is also true. Especially when it comes to training, where most of the time people are looking to study how physical attributes of humans change in response to activity. Because of the limitations regarding humans and what they like to do with their lives, it is not possible to fully study all the activities people are doing and definitely not across all time scales. There will always be truths too difficult to study in practice, that does not negate their existence, but simply put, limits the scope of truths available for description by the scientific method in its current paradigm. We simply do not have Laplace’s Daemon to solve it for us. We will always work with limited information, reduced to fit the confines of analysis, with conclusions that have a transfer problem.
First of all, it will always be easier to study the rapid adaptive process in a body then long term adaptations. Finding experimental subjects and using the space and computing power to process the data, and to keep as many variables as equal as possible it is obviously easier to study a group of people doing a specific weightlifting regime for two months then for two years. Let alone five years. Ten years. Imagine setting up a longitudinal design where you study the adaptive change of one person across their lifespan. The experiment would not only span the life of the subject, but also multiple researchers careers. The inevitable conclusion of this must be the following:
Short term results will be more evidence based.
Secondly, if we wanted to study the impact of a certain movement on specific adaptation, you need to isolate the variables and repeat the activity in exactly the same way. Say you want to study the impact of a lunge on leg strength you will need a precise description of the movement that will be trained, and a way of measuring the results, say vertical jump height, amount of reps to failure or output on a leg extension machine. The lunge will need to be of exactly a certain depth, a certain amount of repetitions with a set speed. Imagine you want to study 20 people for this condition, training in total 8 times over 2 months, doing one movement at one speed with one set amount of repetitions(1). If it takes an hour for every session, that would be four weeks of full time work for one researcher. This is doable.
Imagine now, you vary the amounts of repetitions (to 25/50/100), speed (slow, medium, fast) interval between days (decreasing, spread out or increasing) , depth of executions of the movement (deep, middle, high) and for each of those combinations condition you need at least 20 people(2). If every research sessions took an hour, that would be 6,2 years of full time work. Because it needs to be done within two months, you would need forty researchers working on it. Not so doable anymore. The more variables you have, and the more combinations you make the scale of the research balloons out of control. This will lead us to expect the following:
Exactly repeated movements will be more evidence based.
There is a second reason why the statement above will be enforced, namely, replicability. A study that cannot be replicated fails to meet the terms of falsifiability. Only studies that can be imitated can be falsifiable.
In other words, most evidence based methods, whether it be about strength training, endurance, or flexibility will gravitate to imitation of simple, repetitive movements with short term effects.
Instead of the experiments mirroring normal activity of humans, we have been living with scientific influence over training methods for so long, that our baseline has become methods that imitate laboratory conditions, rather than real life. It is precisely because science cannot get a perfect grasp on reality that training methods imitating its process leads to situations that recreate artificially static conditions that do not really translate well into application for daily life, because life is messy. There are good reasons to believe that our bodies do not move well when it mostly does repetitive imitations. Variation and adaptability and grace are better hallmarks of good movement.
Luckily there is a growing body of work that takes this messiness into account. These are the scientific works based on the Dynamical Systems Approach, nested in chaos theory, and differential learning and ecological psychology. It will take some time to expand on all of these and the consequences for training, which we will do later and elsewhere.
Xander Barel
-
This leads to Trainings sessions * amount of respondents * Repetitions * Movements * Speed * Interval days * Months. One condition for every variable gives: = 8x20x1x1x1x1 = 160 data collections for analysis. -
This leads to 8 days, for 20 people, in each of 3x3x3x3 conditions: 160*81= 12960 data collections.