Analytic Hierarchy Process (AHP)

Priority Vector and Inconsistency Ratios in Shiny

Wahsabii Neanderthal (wahsabii@gmail.com)

Purpose

To calculate an analytic hierarchy process (AHP) priority vector with its inconsistency ratios.

Scope

  • In scope:

    • Input: a square pairwise comparison matrix in the form of a delimited text file accessible to the user's local computer.
    • Process: a R Shiny application using an AHP function from the pmr package.
    • Output: an AHP priority vector with Saaty and Koczkodaj inconsistency ratios to the screen.
  • Out of scope:

    • Anything not using a CSV pairwise comparison matrix text file as an input with an output a computed priority vector with inconsistency ratios to the screen using a R Shiny application. The number of simulations for inconsistency ratios are also not considered.

The Fundamental Scale for Pairwise Comparisons

  • In an AHP matrix, the row entry is compared to the column entry.
  • Each pair is compared twice: once above and below the diagonal with reciprocal values.
  • As in the example below, the element A1 in the first row has a strong importance, 5, over element A2 in the second column. Also, row A2 element is 1/5 the value of the A1 element. The fundamental scale is as follows:
Value Definition
1 Equal importance
3 Moderate importance
5 Strong importance
7 Very strong or demonstrated importance
9 Extreme importance

Note: 2, 4, 6, and 8 are for compromises between the above values.

Process Steps with Examples


Result is the following priority vector; please compare with Process #5 example output above:

## [1] 0.61 0.24 0.10 0.04

Results of the AHP Shiny App