What Are Biomotor Abilities?
Do you know what are biomotor abilities? How did they ’emerge’? The purpose of this video is to explain you the ontology of biomotor abilities, the certain flaws of their use (The Root problem, buckets, periodization based on those qualities) and also provide statistical analysis in RStudio with simulated data using Factor Analysis and Hierarchical Clustering.
Click here to download PDF slides used in the video.
This is members only content so if you would like to watch the video and to download R script which is used to produce the analysis, please click on the button below and become a member.
I forgot to mention in the video that this ‘dimension reduction’ and search for underlying latent factors is very common in psychology, especially tests of personality (e.g. Myers-Briggs test) and intelligence. This field is usually called psychometrics. I might spend some time reading vignette on psych package by William Revelle (vignette, website, chapter “Constructs, Components and Factor Analysis“).
I even stumbled on a book from 1950 called “The Factorial Analysis of Human Ability” by Sir Godfrey Thompson. It is amazing what we can learn by checking the history. I have glanced over some introductory chapters (free on that link above) and similar problems emerge between psychometrics and ‘theory of human bio-motor abilities’.
My message, and I believe is the same one as Verkhoshansky’s, is that biomotor abilities are not some fixed constructs/entities. We need to be familiar with techniques that yielded those constructs (and some of them are simple human opinion through observations, which is of course biased, and factor analysis techniques and other psychometric methods that also have flaws) and how do we use those constructs and buckets. Whole periodization and training theory is based on those constructs – but we need to understand them and see how some methods are flawed and some discussions useless. For example, discussing strength as quality is different to powerlifter and to a sprinter.
Yes, we need certain models to understand perceptual~movement vagueness and to be able to function, but eventually we need to remember the George P.E. Box statement that “All models are wrong… some are useful”.
I will finish this addendum with new way to visualize correlation matrix (see the video) using corrplot package in R.