DA0-001 · Question #3
A customer list from a financial services company is shown below: A data analyst wants to create a likely-to-buy score on a scale from 0 to 100, based on an average of the three numerical variables…
The correct answer is D. Normalize the variables. Normalization rescales each variable to a common range (typically 0 to 1 or 0 to 100), ensuring that no single variable dominates the average simply because it has a larger numerical scale. For example, income in dollars (e.g., $50,000) would vastly outweigh age (e.g., 35) and…
Question
A customer list from a financial services company is shown below:
A data analyst wants to create a likely-to-buy score on a scale from 0 to 100, based on an average of the three numerical variables: number of credit cards, age, and income. Which of the following should the analyst do to the variables to ensure they all have the same weight in the score calculation?
Exhibit
Options
- ARecode the variables.
- BCalculate the percentiles of the variables.
- CCalculate the standard deviations of the variables.
- DNormalize the variables.
How the community answered
(38 responses)- A8% (3)
- B18% (7)
- C3% (1)
- D71% (27)
Explanation
Normalization rescales each variable to a common range (typically 0 to 1 or 0 to 100), ensuring that no single variable dominates the average simply because it has a larger numerical scale. For example, income in dollars (e.g., $50,000) would vastly outweigh age (e.g., 35) and number of credit cards (e.g., 4) in a straight average - making income the near-sole determinant of the score, which is not intended. After normalization, each variable contributes proportionally and equally. Recoding is used to transform categorical values into numeric ones. Calculating percentiles creates a ranking-based transformation, which is a form of normalization but is not the standard term used here. Calculating standard deviations measures variability but does not by itself rescale the data for equal-weight averaging.
Topics
Community Discussion
No community discussion yet for this question.