All of the measures are in AHP are quantitative. You would need open-ned questions or set-structured interviews to produce qualitative dat.a

I agree with David L Morgan but have worked on some AHP projects. Other fields use the word "qualitative" differently than social sciences. In AHP, qualitative means assessing the quality of items of interest by scoring them; the scores are not necessarily hard measurements. Hence, the meaning of qualitative can vary between fields. Here is but one such example:

Wedley, W. C. (1990). Combining qualitative and quantitative factors—an analytic hierarchy approach. Socio-Economic Planning Sciences, 24(1), 57-64. https://d1wqtxts1xzle7.cloudfront.net/100775187/0038-012128902990028-620230406-1-qm78wr-libre.pdf?1680799661=&response-content-disposition=inline%3B+filename%3DCombining_qualitative_and_quantitative_f.pdf&Expires=1743448761&Signature=cVAhnzhseynojOw83wpvZYqJJxAQaAGcXM8BE7kuIUrFhCD6FGgTcPkXLe6Jv~rtRsSh54LZ1OQ6dMEwQ6htuCnCe2l2MDj5jHV4q9lomfz~YY726Yo9e1je4IKVo4FhicU~iEGpiRSwr2xVn7kfNq2yCiHMNBzfL2rDouSm~27~TSCAYmtSi2PhkFy1ZCaiCW2lew6JTjS1BXO0IvvUDmXpP4R0T4O3hELlJZnrXYEGqEii5ZQUZJUeYSymsv30jMFKktz-nPyR9AVetybtsfAfUaUaX7E8HfPiIoehrNoi84giLLjuZxIGpYor74jVIGZhl7f~JOotb~DcYf076Q__&Key-Pair-Id=APKAJLOHF5GGSLRBV4ZA

The Analytic Hierarchy Process (AHP) approach is a quantitative research method that, depending on its application, can also include qualitative research method.

The Analytic Hierarchy Process (AHP) method starts with qualitative judgments, as decision-makers provide their opinions on the relative importance of criteria and alternatives. These judgments, though qualitative in nature, are converted into numerical values for comparison through pairwise comparisons. In this way, AHP combines qualitative inputs with quantitative analysis.

AHP thus combines both qualitative and quantitative research methods, making it a "mixed-methods approach."

You cannot do "pairwise comparisons" on qualitative data, which consist of textual responses. As David C. Coker said, there seem to be multiple meanings of "qualitative" here, but as far as I can tell AHP consists entirely of ratings, some of which are ordinal, but none of which produce any kind of text.

Although it is not possible to perform "pairwise comparison" on qualitative data directly in AHP, it is possible to do so indirectly. Because qualitative data usually consists of subjective descriptions, opinions, or concepts that don't easily lend themselves to numerical comparison, conducting "pairwise comparisons" on qualitative data—like textual responses—can be challenging. It is possible to accomplish this, nevertheless, by converting qualitative data into an organized format that may be utilized for the Analytic Hierarchy Process (AHP). Prior to doing an AHP analysis on qualitative data, the qualitative data should be encoded into numerical values (1-9 relative importance scale). The results of the AHP analysis can then be simply decoded into qualitative terms.

As an illustration, consider evaluating several products using three criteria: cost, ease of use, and quality. It is a qualitative decision-making procedure to start. Furthermore, it is a qualitative assessment if an expert states that "cost is more important than ease of use." The pairwise comparison itself relies on subjective, qualitative perceptions to determine which aspect is more important than another (e.g., cost vs. ease of use). You can use scales, like the 1–9 relative importance scale in AHP, to translate your qualitative conclusions—such as that cost is more important than ease of use—into numerical values. This is where quantitative techniques come in.

(That was how we can do "Pairwise Comparison" on qualitative data.)

Normalizing pairwise comparison matrices, calculating eigenvalues, determining weighted scores, and ranking the products according to these weighted criteria are examples of quantitative mathematical operations that are part of the process. The weighted scores for each product are examples of the numerical results from AHP, and you rank the products quantitatively before decoding these values to qualitative terms.

In conclusion, AHP is a combined method that involves both qualitative aspects (expert judgment, pairwise comparison, and intuitive input) and quantitative aspects (numerical representation means encoding the qualitative data to numerical values according to a 1-9 scale, mathematical processing, and final ranking). The final output is quantitative — a ranking of alternatives based on weighted scores, which then can be decoded to qualitative data.

In its nature, AHP is a qualitative approach, but the involvement of a quantitative procedure makes it a combined approach.

Attaulah Darzar's answer proves David Morgan's point: AHP is a different qualitative than social science research or mixed methods. Why? Pairwise is not qualitative, ever, in social science research. There's also qualitative in engineering, an entirely different endeavor. None is inherently good or bad, just each serves a different purpose. So, yes, AHP is definitely qualitative and quantitative, but be careful to differentiate from the most popular and common definition of qualitative research. Disambiguation is key.