Outliers and data cleaning Flashcards
Studying this course? Keep your progress in one place.

Flip each card. Be honest: the ones you mark Not yet come back first.
13 cards
Tap a card to flip it. Swipe right if you knew it, left if not.
See all 13 cards
- What is an outlier?
- A value unusually far from the rest of the data, as defined by a stated rule.
- State the quartile rule for outliers.
- More than above , or more than below .
- State the standard deviation rule for outliers.
- More than standard deviations from the mean (e.g. 2 or 3).
- Upper boundary when and ?
- How are outliers shown on a box plot?
- As separate points beyond the whiskers.
- How do you deal with missing data?
- Exclude it from the calculation and note it; do not treat it as 0.
- How do you deal with an obvious error such as for a spend?
- Correct it if the true value is known, otherwise remove it.
- What do you do with a genuine outlier?
- Keep it, comment on its effect, and perhaps give results with and without it.
- Which average is least affected by an outlier?
- The median.
- Which measure of spread is least affected by an outlier?
- The interquartile range.
- Why is a value exactly on the boundary not an outlier?
- The rule says 'more than', so it must be beyond the boundary.
- What should you check first about a suspected outlier?
- Whether it is an error, by checking it against the source.
- A wrong value should be . How does the total change?
- Subtract and add .