How to Analyze Movement Screen Tests? [Addendum]
Continuing on the previous post I had an idea to make further analysis. Please note that this is only a playbook.
The idea is to transpose the data, and instead of clustering the athletes, we cluster the metrics or tests. What we are looking for is to find tests/metrics that are similar (within our group of athletes) or that might assess the same factor/quality. Not sure how much is this similar to PCA/EFA, so someone please correct me if I am being stupid here.
By the way, all tests used in movement screen are available here »
We need to load and munge the data
# Load the Movement Screen Data
movementScreenData <- read.table("http://complementarytraining.net/wp-content/uploads/2013/10/Functional-Screening.csv",
header = TRUE, sep = ",")
# Remove first 8 columns
movementScreenData <- movementScreenData[, -1:-8]
# Scale the data
movementScreenData <- scale(movementScreenData)
movementScreenData <- as.data.frame(movementScreenData)
# transpose the data
movementScreenData.matrix <- as.matrix(movementScreenData)
movementScreenData.matrix <- t(movementScreenData.matrix)
We are going do Hierachical Clustering and mark 10 clusters (No clue why I used this number) using both ward and complete method
HCWard <- hclust(d = dist(movementScreenData.matrix), method = "ward")
plot(HCWard, xlab = "", ylab = "", yaxt = "n", main = "Ward Method")
rect.hclust(HCWard, k = 10)
HCComplete <- hclust(d = dist(movementScreenData.matrix), method = "complete")
plot(HCComplete, xlab = "", ylab = "", yaxt = "n", main = "Complete Method")
rect.hclust(HCComplete, k = 10)
What I miss now is how to combine Testing clusters with Athlete clusters? Anyone having a clude let me know.
Responses