exploratory vs confirmatory analysis

Question 1

exploratory analysis

Answer 1

conducted when you have a hypothesis but no clear plan or strategy, or when you have lots of variables tat are maybe associated with outcome but not sure which ones or in which way (no clear predictions)
- provides tools for hypothesis generation particularly via data visualisation

Question 2

confirmatory analysis

Answer 2

conducted when you have a specific research question to test
- does the data support or not support your argument

Question 3

steps in exploratory analysis

Answer 3

1. check the coding of data and visualise variables of interest
2. compare models of interest to find if variables of interest are good predictors or outcome variable
3. compute k fold cross validation MSE for each model
4. identify the best fitting models

Question 4

overfitting (exploratory and confirmatory analysis)

Answer 4

tendency for statistical models to mistakenly fit sample specific noise as if it were signal