8 Oct - Processing (Luca)

Question 1

What is data transformation?

Answer 1

All that has to do with cleaning, reshaping, transforming, augmenting data

Question 2

What can frequency distributions show us?

Answer 2

1. We can spot outliers, we can see the 2. range, 3. we can see the shape of the curve.

Question 3

What is correlations?

Answer 3

Measure the strength and direction of the association (covariance) between
random variables. A correlation is not necessarily linear! Also the strength of the slope does not equal strength in correlation.

Question 4

What defines a strong and weak correlation?

Answer 4

trick question: there is no minimum value that defines a strong correlation, it depends on the problem at han.

Question 5

What is the pearson correlation test?

Answer 5

It measures the linear association between two variables.The goodness of fit (R^2) is exactly the pearson correlation.

Question 6

What is the bonferroni correction?

Answer 6

The more tests are performed, the higher the chance finding a false
positive result. It is very conservative.

Question 7

What are the pearson correlations assumptions?

Answer 7

• Continuous variables
• Paired samples (x and y values are available for all samples)
• Independent observations
• Linear relationship between the two variables
• Roughly Gaussian-distributed variables
• Absence of outliers

Question 8

What is the spearman correlation?

Answer 8

Measures the monotonic association between two ranked variables
and tests its statistical significance

Question 9

What are the spearman assumptions?

Answer 9

• Ordinal variables
• Paired samples (x and y values are available for all samples)
• Independent observations
• Monotonic relationship between the two variables