I'm simply referring to "Volume" when using the word Levels, the technical word would be "Amplitude" in audio, but no one use this, we talk about recording levels, output levels...
You have several stages of control over the levels, you need to think about the audio signal workflow inside your DAW.
It starts by the level of your source ( output level of a sample or a midi instrument ) that goes through your channel ( individual track ) effects from left to right ( some effects will give you another level knob or/and a wet/dry knob ), then reaching your channel ( track ) level and from there it can either go to group/bus or directly to your Master channel. That's the basic audio signal workflow in every DAW. Best practice is to keep your Master channel level to 0 dBFS and don't touch it and gain stage each track against each other : the goal is to set appropriate level for each track, think about it like balancing your all Mix and make each element sit well together, nothing to weak or to loud. Each track output level are summing on your master channel, if you start to hot with your kick and bass, it's gonna be harder to balance your mix and you will loose clarity since you will need to work with loud levels on each track. Another problem is what you're gonna hear from your monitors, because output levels are summing on those too. Depending of their specs & ability to reproduce sound ( especially at low frequency ) if you start to mix with a loud kick, bass or sub, then your monitors will struggle to reproduce the sound and adding other elements will soon lead to muddiness & lack of clarity.
Try to start with a Kick around -14 dBFS ( using the kick track output level fader ), you will then have more headroom to mix the other elements in. If you feel like you're missing some "juice" for monitoring ( levels are too low ) then rise your sound-card output, not your DAW Master channel level. But iMO, it's good practice to learn to mix at lower levels, it takes some time & habit though.
Another consideration for levels ( when I wrote in my comment "The level at which you hit those devices is also important" is the output levels at which one you reach your effects input. If you're using your DAW devices like compressor, they are 100% digital and more forgiving but if you're using 3rd party plugins that were designed after analog gears, those device are engineered to work best at input levels around -18 / -20 dBFS, if you hit them too hard, they saturate & distort or at least they won't work as expected.
Go back to the source choice first. Kick and bass have to work together almost right from the start. If you're using a long kick like a 909, it's gonna be more hard to use a long bass sound, they be dueling & fighting. There's elements in a kick like Oomph, Thump, Body, Punch, Click ( basically going from lowest frequency range to highest one ) that makes a kick sit better with a bass ( or your full Mix ) than others. It's very common practice to change kick at the end of a track because you realize another kick is working better than the original one you used.
Same goes for bass, there's different frequencies areas, you can use EQ to filtered out the unwanted sub frequency below 45 or 30 Khz with a High Pass Filter ( Low Cut ). But there's other mid-range or higher frequency areas that might need to be accentuated a bit and help the bass to come through. Sometime there's not enough of that in the original sample or synth preset you're gonna use, you can then need to add a "click bass" layer to help the bass to cut through the mix.
Side-chaining your bass to the Kick is of course good to do, you can also use automated filters or plugins like LFO Tools to achieve this, but in the end you won't fix all your issues with compression if the Kick & Bass don't work together from the start. Using side-chaining compression to achieve that "pumping" effect is another story.
If you'd like to dive more into this I suggest you watch this tutorial from Protoculture
Also use the search ( magnifier icon on the top right ) feature on the tutorials main page and type in "Compression", "Compressor", there are series about understanding compression as well as individual videos that can be very useful to watch on the site.
There's several reasons to use a reference track. You can use it to better understand/follow a track structure and arrangement and reference a final mix in terms of final loudness, levels that you should aim for.
You need to use mastered tracks in lossless audio formats like Wav, Aiff, Flac. MP3 are compressed and less reliable, there's a lack of audio information. Pick up a track that you find well mixed in the same genre as yours. You can simply load the track on an audio channel inside your Project and refer to it as needed ( trying to match the kick level, drums elements, bass, synths, vocals, fx...etc ). This can be cumbersome though since you'll need to mute/un-mute the reference track to compare it to your own Mix. Another solution is to use 3rd party plugins that allow you to load the track ( or even several tracks ) and depending of the features of the plugins it will offer a much more user friendly workflow as well as some visualization or options.
One of the best tool for this is Metric AB from Plugin Alliance ( pricey but really powerful ), but there's other options too like "Reference" from Mastering The Mix. Sample Magic AB is now discontinued I think.
Again here's another course from Protoculture about ADTPR Metric AB
It is most of the time but it shouldn't...LOL. As explained before, go back to the source. You won't fix things easily during the mix-down and even less during mastering stage. From sample to preset choice, try to get it right at the source, adjust levels, use filters and EQ to make the sound sit in the mix. It's much more efficient to spend more time choosing your source material than trying to fix things later on in the mix.