My team and I are in the middle of a prioritization problem that involves 350 alternatives (see figure for context about alternatives) or so. I have used the AHP to support the decision-making process in the past with only 7 or 8 alternatives and it has worked perfectly.
I would like to know if the AHP has a limit on the number of alternatives, because consistency may become a problem as Dr. Saaty's method provides Random consistency Indexes for matrix sizes of up to 10.
I was thinking in distributing the 350 alternatives in groups of 10, according to an attribute or classification criteria, to be able to use the RI chart proposed by Dr. Saaty.
If there are other more adecuate multi-criteria analysis tools, or different approaches to calculate the RI for larger matrices, please let me know.
Greetings and thank you,
Dear Jose
I believe that the limit is your patience in making pair-wise comparisons
In my opinion you can't partitionate the number of alternatives because the whole is a system where everything is related with everything.
Your alternatives are independent or are there relationships among them?
Are you sure that you criteria are independent? It seems to me that casting cost is related with casting complexity and manufacturing capacity is independent of cost and time
If any of these are related you can't use AHP
Dear Jose,
I share the same opinion of the colleague who preceded me, however I suggest you take a look at the links and files attached this can be profitable for you.
https://mpra.ub.uni-muenchen.de/45348/1/MPRA_paper_45348.pdf
Article Criticisms of the Analytic Hierarchy Process: Why they often...
Best regards
Good day,
Use TOPSIS and ranking the alternatives and using a cut off value take the required factors for AHP analysis
Hi,
Due to the Miller's rule of 7 minus plus 2, a person has a limited capacity of processing information.
Hence it is recomended to include 7 minus plus 2 alternatives.
Best regards
Dear Baris
Then you are solving just an exercise in solving an inexistent problem, because the original has 350 alternatives
This is one of the main drawbacks of AHP; you are subordinating your real problem to what a psychological theory says
The only limit is in the arising complexity and hence in the risk to get inconsistency in the pair-wise comparison.
Dear Epicoco
In my opinion using AHP or ANP you are not solving a problem but a simplification of it
I think you need to use a clustering technique and segregate your decision variables into similar groups. If I`m right there are no limits for amount of alternatives.
- Badges
- Science topic
- Similar topics
- Mathematics
- Algebra
- Matrix