Deciding what to do with an outlier

Name:

Date:

Class:

Quick Questions

Shade in the A or B circle for each question.

1. A class survey on hours of sleep shows 8, 7, 9, 8, 70. What is the first thing to do with 70?

A Delete it straight away because it is big

B Check whether it is a data entry error

2. A student typed 1.65 m, 1.70 m, 1.68 m, 165 m for heights. The 165 was meant to be 1.65 m. What should you do?

A Correct it to 1.65 and note the change

B Keep 165 because it is real data

3. In a dataset of 100m sprint times, one runner is much faster than the others, and the timing was checked and confirmed. What should you do?

A Keep it because it is a genuine extreme value

B Remove it because it is far from the rest

4. Which statement is true about outliers?

A An outlier is always a mistake that must be deleted

B An outlier is a value far from the rest, and it may be an error or a real value

5. A weather log shows a max temperature of 240 degrees for a Perth day, and there is no way to find the real value. What is a sensible choice?

A Remove it, because it is an impossible error you cannot correct

B Keep it so the dataset has no gaps

Match the Pairs

Draw a line connecting each item on the left to its match on the right.

Left column		Right column
1=MAX(B2:B9)		AChanging a real value to look normal. This makes the data inauthentic. Keep a genuine extreme value and do not alter it.
2=IF(B2>100,"Check","OK")		BFlags a value that is out of the possible range so you can check if it is a data entry error
3A typing slip such as 810 instead of 81 is a data entry error, so correct it once you confirm the true value from the source		CB9: 810 changed to 81
4A genuine extreme value, such as a real 98 in a set of lower marks, is kept because it is true data		DLeo Science 98: verified, keep
5Deleted the highest mark without checking it		ERemoving an outlier without checking whether it is an error or real. A genuine value may be lost and results become biased.
6Changed 98 to 80 so it fits the pattern		FFinds the biggest value in a column so you can spot a possible outlier that is far above the rest

Definitions

outlier

Fill in the Blanks

Complete each line of code by filling in the missing parts.

=___(B2:B9)

=IF(B2_100,"Error","OK")

=______(B2:B9)

IF isError THEN correct(value) ____ keep(value)

=_______(B2:B9,">100")

Spot the Error

Shade in the letter of the line that contains an error.

Question 1

A IF age < 120 THEN

B check the original form

C IF form says 15 THEN

D correct the value to 15

Question 2

A value ← 175

B IF value < 250 THEN

C mark as possible error

D ELSE keep the value

Question 3

A Value: 9.58 s in 100 m sprint

B Source: official world record

C Decision ← delete the value

D Log: genuine extreme value

Question 4

A outlier ← 450 in Qty column

B Checked the original receipt

C Receipt says 45, a typo

D Log: kept 450 as genuine

Question 5

A Max ← MAX(B2:B9)

B Min ← MAX(B2:B9)

C IF Max > 100 THEN check it

Answer Key

Deciding what to do with an outlier

Section A: Quick Questions

B — Check whether it is a data entry error

A — Correct it to 1.65 and note the change

A — Keep it because it is a genuine extreme value

B — An outlier is a value far from the rest, and it may be an error or a real value

A — Remove it, because it is an impossible error you cannot correct

Section B: Match the Pairs

=MAX(B2:B9) → Finds the biggest value in a column so you can spot a possible outlier that is far above the rest

=IF(B2>100,"Check","OK") → Flags a value that is out of the possible range so you can check if it is a data entry error

A typing slip such as 810 instead of 81 is a data entry error, so correct it once you confirm the true value from the source → B9: 810 changed to 81

A genuine extreme value, such as a real 98 in a set of lower marks, is kept because it is true data → Leo Science 98: verified, keep

Deleted the highest mark without checking it → Removing an outlier without checking whether it is an error or real. A genuine value may be lost and results become biased.

Changed 98 to 80 so it fits the pattern → Changing a real value to look normal. This makes the data inauthentic. Keep a genuine extreme value and do not alter it.

Section C: Definitions

outlier — A value far from the rest. Check if it is a data entry error or a real extreme value, then correct, remove or keep it, and record your decision.

Section D: Fill in the Blanks

=MAX(B2:B9)

=IF(B2>100,"Error","OK")

=MEDIAN(B2:B9)

IF isError THEN
  correct(value)
ELSE keep(value)

=COUNTIF(B2:B9,">100")

Section E: Spot the Error

A — An age above 120 is the suspicious outlier, so the test should be age > 120. Using < would flag normal ages instead.

B — Values far above the rest should be flagged as possible errors, so the test should be value > 250. With < the normal value 175 would be flagged.

C — The source confirms the time is a real world record, so it is a genuine extreme value. It should be kept, not deleted.

D — The receipt shows 450 was a data entry error and the real value is 45. The log must record that it was corrected from 450 to 45.

B — MAX gives the largest value. To find the smallest value, and spot a low outlier, use MIN(B2:B9).