Technique 2.89 Weighted Factor Analysis
Introduction
This involves converting each factor to a common scale, ie from 1 to 100 (best is 100). This is sometimes called normalising data. Some variables have positive and others negative values.
Sometimes this approach is criticised as it is used to justify your preference or biases, ie the weighted is too subjective!!!
For example, when determining the best town to live in, some factors are weight and included in the table below
| Factor |
Weight (%)
|
Town A
|
Town B
|
Town C
|
Town D
|
| average annual rainfall | 2 |
||||
| # sunny days | 3 |
||||
| snowfall (inches) | 1 |
||||
| watershed quality | 2 |
||||
| air quality | 2 |
||||
| comfort index | 8 |
||||
| time to coast | 7 |
||||
| availability of private schools | 15 |
||||
| achievement index | 10 |
||||
| presence of friends/family | 8 |
||||
| University/College town | 5 |
||||
| Art Centre | 4 |
||||
| quality of town centre | 4 |
||||
| physical beauty of area | 5 |
||||
| existing small companies | 4 |
||||
| size | 4 |
||||
| recreation facilities | 6 |
||||
| shopping facilities | 5 |
||||
| public transport | 5 |
||||
| Total | 100 |
"...where to live case illustrates how you can start with a simple list of issues and elements that are related to your problem statement, disaggregate the elements further into indicator variables, then finally add concrete measures and weights. The rest is straightforward arithmetic based on a considered ranking of features..."
(sources: Charles Conn et al 2018)