These sit next to each other on an org chart and select for genuinely different people.
The work
Trading is decision-making under time pressure with incomplete information. The unit of work is a position and its risk. You are rewarded for good decisions at speed, and for noticing quickly when you are wrong.
Research is finding and validating signal in data. The unit of work is a hypothesis and its evidence. You are rewarded for being right slowly and for not fooling yourself.
Both need quantitative ability. They differ in what they do with it.
The interviews
A trading process leans on mental arithmetic, expected value, market-making games, and how you behave when a position moves against you. Speed matters. Being decisive matters.
A research process leans on statistics, experimental design, programming, and a detailed interrogation of your own past work. Speed matters much less. Being able to say what your analysis does not show matters a great deal.
The role-specific question sets are split along exactly this line.
How to tell which you are
Ask yourself which failure would bother you more: making a fast decision that turned out wrong, or spending three months on an idea that turned out to be noise.
If the first bothers you more, you are probably a researcher. Traders have to be able to be wrong quickly and move on; a temperament that ruminates on a bad decision is a real liability on a desk. Conversely, a researcher who cannot sit with an unresolved question for weeks will struggle.
Can you apply to both?
Yes, and many people do. But prepare for them separately - the overlap is smaller than the org chart suggests. If you are short on time, pick the one whose interview you would enjoy more. That is usually a good proxy for the work.
What about quant developer
A third path, and an under-applied one. It hires on genuine software engineering - systems, performance, correctness - with a lower quantitative bar than either of the above and, at many firms, comparable compensation. If you are a strong engineer who likes markets, it is often the highest-probability route in.