5-point likert scale into low, medium, high categories?

Question

I am designing a study where the objective is to collect the % for low, medium and high to populate probability tables. For example, my objective is to know what is the probability in % to have a {low, medium, high) intention to purchase a new phone given some scenarios where I modify some variables. For the purposes of the data I need % for low, medium and high; however my question is if it would be better to design my data collection questions as:

A) Directly asking for Low, Medium and High, and then I would just do the calculations for the % based on what people responded, very straightforward.

B) Use a 5-point likert scale and then re-scale considering very likely and likely as High, neutral as Medium, and unlikely and very unlikely as Low?

I am worried that for option A it might not be valid due to it not being at least 5 points, or it might be too simplistic that it might deemed not valid; on the other hand I am worried re-scaling as in option B might not be valid either if I do not have enough respondents or it becomes too arbitrary to re-scale as proposed.

I tried to look for other studies as examples but I was unable to find examples of any of both options. Any comment would be much appreciated.

Answer 1

I don't see any reason to use option B. There's no reason why a 5-point scale should be more valid than a 3-point one, and re-scaling is somewhat arbitrary because you could also put point 2-4 in the medium category. What I cannot guarantee is the validity of option A, but this doesn't have to do with whether your scale has 3 or 5 points (and is therefore not helped by option B) but rather with whether intention to buy can be validly measured by just asking people a straight question.

By the way I don't think that in principle the 5-point scale is worse than the 3-point scale (at the end of the day the intention to buy doesn't have a "true" scale resolution, not 3, not 5, not anything else in particular). My preference for option A here is only because you state that at the end of the day you want to analyse data as 3-point low/medium/high, in which case using a 5-point scale for asking is pointless.