IMO it's not science fiction, but science itself, that you need to learn and research
Absolutely. There is a lot of supposed SF that's not really SF at all, just Magic in Space or Magic in the Future.
Also Logic, Mathematics, Engineering.
Even Fantasy works best if internally logical.
Unless the SF is of the purely fantastical / whimsey end of space opera / humour / satire then at least it shouldn't obviously contradict mathematics and especially logic. They are not going to change. Most of known physics and chemistry isn't going to change.
If you decide to do properly "Hard SF" then you need to research any idea to college level as primary and secondary texts or illustrated books for children may actually have "Lies for Children" or old misconceptions.
Examples:
Filament Light bulbs have a vacuum: Mostly untrue for nearly a 100 years.
Atoms are like little solar systems with Electrons Orbiting the nucleus of Protons and Neutrons: Untrue. Even the "shell" description is very misleading. The Electrons' position is described by a wave function and has a probability of being at a particular place.
We have no idea what gravity is or how it's propagated. We can calculate the attractive force between two masses at a particular distance or how much a mass of a particular density bends passing light (or radio waves). We have no idea why, either in Relativity or Quantum Mechanics, though there are some weird theories.
Evolution says nothing about origin of Life itself, or the Universe. Nor does it imply machines, computer programs, biological organisms get more advanced. It's about how biological entities can change.
Artificial Intelligence doesn't mean what most SF books and popular press suggests. Today A.I. is about better ways to search a database to identify real world objects (images), or answers to a question from the data. No-one is actually researching AI in the Star Trek Data or Asimov "Positronic Brain" sense as we don't even know how to start. The "Singularity" and "self emergent machine intelligence" are not really SF, or Iain M. Banks "AI" but pure fantasy, akin to princes turning into frogs. Calling it A.I. research is marketing. It's just regular computer programming and databases.
A. C. Clarke (or someone like him) suggested it was fine to put one or two totally imaginary & unlikely things in real S.F. If in fact almost all the tech is so unlikely to be really contravening basic logic, mathematics and most existing science, then it's actually not really SF at all, but Fantasy. A lot (but not all) of Space Opera is like this.
Time travel is more Fantasy than SF, as logic suggests it's impossible. But there are lots of fun and/or good Time travel stories.
FTL (Faster Than Light) travel is probably impossible in the normal sense, but we aren't stuck at Light Speed Ships (which would take no time to travel for occupants, but same time as light for external observers) or Generation ships for Space Travel. Space Opera doesn't explain it or worry about it at all. There are though three main Mathematics/Physics ways to get space ships from A to B faster than light without actual FTL travel (which is impossible). Some kind of Warp Drive, Wormholes or some other unspecified way of "folding space", like a walkway across the top of two skyscrapers instead of travelling via the ground. It's best to not explain it and only use it as any existing theory for Warp Drive or Wormholes needs materials that might not exist and infeasible amounts of energy (as much or more than entire visible universe).
Explain nothing, then people can more easily suspend disbelief. All the best SF is about people. How the people react, develop, cope, relate etc. So in a sense if Space Opera isn't Samuari or Cowboys or Detectives & Criminals or Spy Story in space, it's merely boring technobable. People don't want to know how a light sabre works, they want the farm boy that's really a lost prince to succeed in fighting the evil warrior-wizard. Star Wars and Star Trek lost credibility when they started to "explain" stuff in the films / shows.
A contemporary set thriller doesn't stop to explain how a car or aeroplane works (unless maybe some sort of technical sabotage is important to the plot), the writer researches what an Airbus 380 can do, the places, the injuries, medicine etc. Build up a reference library. That's what many good non-SF writers have done. Have organised book marks. Have a folder structure on computer and "Save Website Complete as HTML" so you can have it if it's removed from internet and easy fast search / edit and organise even offline. You can't rely on websites keeping content or even continuing to exist.