Training with Purpose: Correlation and Causation

Introduction

The problem with correlation and causation is that they, on one hand, are similar, but in reality they are two different things. This subject is something I had touched on in my blog post titled “A does not equal B…” However, it was more of a rant, and it never really described the factors at hand. The purpose here is to expand and discuss, looking at the variables that may actually correlate or cause certain effects. Through understanding these concepts, it can provide a new perspective on training and programming.

Correlation and Causation Defined

Correlation is a statistical relationship between two variables that display dependence. This would mean that there is a measurable statistical relationship between the two variables, along with two sets of data. Correlations between variables can be positive (when one variable increases, so does the other) or negative (when one variable increases, the other decreases). In the realm of powerlifting, a positive correlation would be a bench press increasing as a result of doing more volume in a special exercise, such as a close-grip bench. In a sport such as track and field, a 100-meter sprinter may have set a PR due to perfecting his start technique, which means that a drill such as block starts showed a positive correlation. Causation means that one thing will directly cause the other. The end product, aka the effect, is a direct consequence of the first variable, the cause. In some ways this may be true. In a very general sense, if you train, you will get a training effect of some sort. Notice I didn’t say stronger, faster, or any other general term because it really depends on what exactly you do and what the desired training effect is. Every effect has a cause, but the problem with cause and effect is that it only works in perfectly clear-cut situations. Example: Effect = Burns on the hand. Cause = Placing your hand in an open flame. Unfortunately, these don’t exist in the realm of training. Let’s look at two different variables: A and B. One relationship, if we are to discuss causality, would be that A may be the cause of B. Another causality could be that B, in fact, causes A. Of course, we could argue that A and B are the effects of a common cause (a third variable: C), but they did not directly cause each other. To confuse this even more is the fact that B can cause A at the same time that A causes B, which contradicts A causing B and vice versa. A final possibility is that it is merely a coincidence, or that the correlations are so complex or inconclusive that it is easier to label it as a coincidence. As we can see, there are many reasons why it isn’t as easy as saying A causes B.

Photo: © Igorabond |

Dreamstime Stock Photos &

Stock Free Images

How this relates to training…

At this point I may have either confused a lot of people or pissed someone off, but if you are still with me, I will finally start to get to the point of all of this. A lot of people seem to have the notion that A causes B. However, when presented with the notion of B causing A, they may refute it. Here is a classic example not related to training:

• A fire that is observed to be large will have a large number of firefighters fighting the fire.
• Reverse causation would mean that more firemen responding to fight a fire would cause an increase in the size of the fire (B causes A).

As we can see, this isn’t true. In training, a lot of people don’t think this way necessarily. For shits and giggles, let’s apply a common fallacy from training and see if some of these still hold true:

• Elite Olympic lifters are observed to display a high level of explosive strength in various tests. Their training includes cleans, snatches, and their variations.
• By including more cleans, snatches, and their variations in training, athletes can display the same amount of explosive strength as Elite Olympic lifters in various tests.

By looking at this example, we need to examine what the missing part of the equation is. The observation here is that the Olympic lifters are explosive and they perform Olympic lifting variations in their training. So, by the rule of reverse causation, if we simply include these lifts in anybody and everybody’s training, then they too will be able to display a high level of explosive strength. However, this is not necessarily true. In a simply theoretical sense, let’s say we went and taught some average citizen off the street how to do Olympic lifts with at least adequate form. We then clean, snatch, and jerk them non-stop for an extended period of time. After this period, we then assess their explosive strength in a barrage of tests and compare it to the Olympic lifter. How many people think that this will actually bring them to the same level?



The answer should be obvious—it won’t. Has it ever crossed some coaches' and trainers' minds that the explosive strength that these high-level lifters display is what allows them to succeed at the Olympic lifts but is not directly caused by performing the lifts themselves? Usually, athletes gravitate toward sports at which they will succeed, and this could definitely be an example of this. If we compare it to the fire example, the amount of explosive strength is the fire and the Olympic variations are the firemen. Just like in that case, the more cleans, snatches, and jerks performed will not necessarily make the level of explosive strength go up. Another example of a fallacy would be bidirectional causation, which means that A causes B and B causes A. This would mean that an increase in A causes an increase in B; therefore, an increase in B would cause an increase in A. Let’s look at the following example:

• A world record holding lifter uses a high volume, high frequency program.
• High volume, high frequency programs cause world records.

The issue at hand here is that the lifter is at an advanced level and also uses high frequency and high volume training. However, what needs to be considered is the following question: Is he a world record-caliber lifter because he uses this much volume and frequency? Or is it the fact that he now has exhausted other options and must raise his volume and frequency to continue to progress or maintain his current level? The fact that he has used this type of programming and reached the level he has may be true in this instance. Still, on premise, it isn’t logical to say that all anyone would have to do is mimic his program down to the sets, reps, and exercises in order to be able to produce the same world record results. This will depend on the current qualification of the lifter and what fits his or her needs. Another consideration is the third factor, which could be described as C. This lurking variable could influence the outcome of two events that appear to be related, yet it is really the third variable that contributes to the cause. This variable may be unbeknownst to most, or maybe it is something that is overlooked. I am going to give an example here, but before I do, I will throw in a disclaimer. I am not opposed to lifters competing in gear or untested federations. Most of my competing is done with single or multi-ply gear. Where someone chooses to compete is his or her own business. However, bench shirts, squat/deadlift suits, and performance enhancers can skew the variables of assessing a training program quite a bit. An alteration to gear or taking more ergogenic aids may be an underlying cause for an effect. Even changes in one's regular daily routine, diet, etc. can have an effect. Now, if you aren’t too busy getting your panties in a twist over it, let’s look at an example.

• Lifter X set a PR in the bench press and used DE bench with chains and bands over a 12-week training cycle from his last meet.
• Therefore, using DE bench with chains and bands for 12 weeks causes PRs.

While this isn’t too far-fetched, what Lifter X may not have told you is that he had his bench shirt taken in, so it was giving him more support. Therefore, we would have to consider how much of a PR it was. Was this PR only attributed to the bench shirt, or did the training play a part? Or maybe in addition to the alteration, he worked his technique in his shirt and learned how to set and use it in order to give him this PR. Of course, he could have also changed his supplementation, his sleeping schedule, or his diet in a way that contributed to this performance. Without knowing all of the lurking variables, it is too simple to say that only 12 weeks of DE benching causes a PR. This isn’t limited to DE benching and PRs. We can sub any programming scheme (Block, Cube, Sheiko, etc.) or any exercise for that matter and argue this. It can also be made for any sport. For instance, we could say that a Top 25 NCAA football team uses the HIT method on machines only, and then try to say that this type of off-season work causes their success. Without considering the other variables (recruiting, overall depth at each position, technical and tactical preparation, etc.), any attempt to pinpoint a single cause is futile.

Conclusion

After examining the difference between causes and correlations, it is easy to see that in something like training it isn’t as simple as thinking one variable is the clear-cut cause for any outcome. The human body is an ever-adapting entity that responds differently to various factors. In addition, when observing those who are exceptional at a certain discipline, it has to be considered that they are good in their field because of pre-existing factors. Their discipline did not necessarily create these abilities that they display. In the future, I plan to elaborate more on this issue and view the differences between correlations in varying qualifications of athletes. Until then, always consider all factors and do not jump to conclusions.