When data is highly skewed and we want to look at correlation, then we can apply a log transformation.
Some other transformations like inverse, log and square root to our data to find correlation.
Correlation doesn't imply causation!
Longitudinal vs cross-sectional study
cross-sectional: data on participants from a single snapshot of time. Effect of age / generation on height, confound each other. Cheaper/faster/convenient.
longitudinal: same participants over a period of time are followed over. Effect of age on height is not confounded by generation. NOT.
