Data Collection & Preparation Flashcards

A dataset's 'income' attribute is missing more often for high earners who decline to answer. What type of missingness is this?

Answer: Missing not at random (MNAR)

When the probability of missingness depends on the unobserved value itself, the data are missing not at random.

Which imputation method best preserves relationships between variables when filling missing values?

Answer: Multiple imputation using predictive models

Multiple imputation models each missing value from other attributes and reflects uncertainty across several imputed datasets.

An analyst rescales 'age' to the range [0,1] using (x − min)/(max − min). What technique is this?

Answer: Min-max normalization

Min-max normalization linearly maps values into a fixed range, typically [0,1].

Why is min-max normalization sensitive to outliers?

Answer: A single extreme value stretches the range and compresses all other values

Because min and max define the scale, one extreme point squeezes the remaining values into a narrow interval.

A retailer merges two customer tables where one stores 'CustID' and the other 'Customer_Number' for the same field. Which data integration problem is this?

Answer: Entity identification problem

The entity identification problem involves matching equivalent attributes or entities across different sources.

Which statistic is commonly used to detect redundancy between two numeric attributes during integration?

Answer: Pearson correlation coefficient

A high absolute Pearson correlation indicates that one numeric attribute may be largely redundant given the other.

Stratified sampling is most appropriate when:

Answer: The population has important subgroups with very different sizes

Stratified sampling draws from each subgroup so that small but important strata are represented.