Clustering Psychometric Survey Data

Question

I have created a 15 question survey to understand the financial preferences of people. It asks questions like-

'How comfortable are you making financial decisions?'

with generally 4 or 5 options like 'very comfortable', 'somewhat comfortable', ...., 'makes me anxious'.

I want to cluster this data to identify the groups of people with similar responses/behaviors. I have currently encoded users' responses as numbers with 'very comfortable' being encoded as 5 and so on.

I thought of doing Hierarchical Clustering but I am unsure if doing that on a Likert scale would make sense.

The goal is to cluster people with similar responses. How do I approach this?

Answer 1

I would look to the literature for examples of (1) choice of algorithm, and (if distance-based) (2) choice of metric. For example, Bland 2018 uses a clustering procedure in their psychometric model.