1332.0.55.002 - Statistical Language!, 2008  
ARCHIVED ISSUE Released at 11:30 AM (CANBERRA TIME) 27/06/2008  First Issue
   Page tools: Print Print Page Print all pages in this productPrint All  
Contents >> Mode >> When is the mode not a good measure of central tendency?

WHEN IS THE MODE NOT A GOOD MEASURE OF CENTRAL TENDENCY?

There are a number of reasons when the mode would not be the best measure of central tendency. Some are illustrated below.


Example

If there is more than one mode these modes will not be meaningful if a single measure of central tendency is required.

Employment details of 'The Widget Company'

Observation
Position in company
Salary ($)

1
CEO
100 000
2
Manager
100 000
3
Manager
59 000
4
Factory Worker
52 000
5
Factory Worker
51 000
6
Factory Worker
50 000
7
Factory Worker
49 000
8
Trainee
40 000
9
Trainee
40 000


In the above table there are two modes, 100,000 and 40,000, therefore it is impossible to determine one measure of central tendency using mode in this case.

Diagram: mode- multiple modes


Example

If there is no mode in a data set (as all data items are different) then a different measure of central tendency is required.

Employment details of 'The Widget Company'

Observation
Position in company
Salary ($)

1
CEO
100 000
2
Manager
98 000
3
Manager
59 000
4
Factory Worker
52 000
5
Factory Worker
51 000
6
Factory Worker
50 000
7
Factory Worker
49 000
8
Trainee
45 000
9
Trainee
40 000


In the above table there is no mode as all salaries are different.

Diagram: mode- no modes


Example

The mode may not fall near the centre of the data.

Employment details of 'The Widget Company'

Observation
Position in company
Salary ($)

1
CEO
100 000
2
Manager
98 000
3
Manager
59 000
4
Factory Worker
52 000
5
Factory Worker
51 000
6
Factory Worker
50 000
7
Factory Worker
49 000
8
Trainee
40 000
9
Trainee
40 000


In the above table the mode is now $40,000, which is not a good measure of central tendency.

Diagram: mode- extreme outliers




Previous PageNext Page