I use a metronome to practice strumming on the guitar. I want to test how I'm doing without the metronome. Am I too fast, or too slow, do I hold the beat steady? What tool or technique can I use to test my accuracy?
Easy, but won't work: Record yourself, then play the recording back with a metronome running. This isn't very helpful, though, because if you set the metronome to 60 bpm but play perfectly in-tempo at 59 bpm, then you'll be half a beat off in just a few measures. Unless you're a robot, you're unlikely to be able to hold an exact tempo, not even deviating by 1 bpm. The same problem would be true of this method; it could get out of sync just as easily.
Better: Find a metronome (or metronome-app) with a "tap" feature. This means you hit a button every time you hear a beat, and it tells you the average tempo it's detecting based on those taps. Record yourself, play it back, and hit "tap" as you listen to your recording. If the average tempo varies by only a few bpm, you're amazing. If it varies wildly, you're in trouble.
EDIT: After thinking about this question for a few days, something seems off: For the most part, anybody who can't keep a steady tempo knows it. I realize I might be mistaken in your problem. You mention "beat" and "fast or slow," which made me think the overall concern was tempo—i.e. whether the beats are equally spaced—but you also mention strumming. Which usually means "subdivisions"—a rhythm pattern of smaller notes that fit inside the beat. I wonder whether your concern is more about those small notes being accurately placed within the beat? Or, if you're thinking about beat/tempo, I would suggest that the problem might be that you're focusing on those rhythms. Let me explain:
First: If you're worried about the rhythmic accuracy of the subdivisions: A good practice tool can be to force a metronome to produce those subdivisions. This can be tricky since most metronomes only go up so high. But, say, if your song is at 100 bpm and your metronome goes to 200, then at least set it to 200 and think of it as eighth notes. Then, if you're strumming sixteenth notes, you can at least check your timing halfway through the beat. Some metronomes, like the physical "Dr. Beat" or the app "Tempo," can even output smaller divisions, like can beep sixteenth notes at any tempo, or some rhythm patterns like "an eighth followed by two sixteenths" (or, in the case of Dr. Beat, triplets against sixteenths!). Practicing with these subdivisions can be revealing; sometimes we're not playing what we think we are.
But a lot of these answers have emphasized that it's not always desirable to play perfectly metronomically, and that's true here too. If multiple people are trying to play these rhythms exactly in sync, and want to be mathematically perfect about them, then great. But in general, as long as the big beats are in the right place, who cares whether the subdivisions are accurate? Lots of genres intentionally fudge the subdivisions, stealing from one part of the beat to give to another; that's the whole idea behind "shuffle," or swing, or French baroque notes inegales. And if a whole ensemble can make the same fudgings equally, can get "in the pocket" together, that's when true "groove" is achieved.
Second: even if you're still just talking about keeping the beats steady, it might be that your problem is that you're focusing on those small notes. Like, if your strumming pattern is an eighth note followed by two sixteenths ("one and a two and a three and a four"), then if you think of it this way: "Okay, I play a note that's half a beat long. Then I play a note that's a quarter of a beat long. Then...", then you're likely to wind up wandering in your tempo. Instead, focus on the big beats: "one and a two and a...". Feel the pulse of those, lock onto it, and let the "little notes" fall into place between them.
When I first learned to drive a car, on the first day behind the wheel, I tried to steer by focusing on the road right in front of the car. "Oh crap, I'm inches away from the line; veer left. Oh crap, now I'm too far from it; veer right. Oh crap..." My driving instructor told me to instead look up to the farthest point I could see ahead and aim for that, and like magic, it was much easier to stay on course. Similarly, if we try to make a tempo by adding up dozens of "little notes," we might make a lot of "little mistakes" and add up to being quite off, but if we focus on the "big beats," we have more chances to check in and adjust.
A simple but very efficient method for assessing your own playing is recording yourself. It doesn't need to be a nice clean recording -- just use whatever tool currently at your hand (phone, computer,... anything). Playing back your recordings is (at least for me) a great way to improve, because it makes it possible to weed out problems that you wouldn't be even aware of otherwise.
If you want to test your metronomic precision, just switch the metronome on for a while so you can "catch" the rhythm, and record yourself strumming. After that, you can play the recording back along with the metronome and you will see how much off you are.
(Completely by the way: it's great to work with the metronome, but make sure you don't overdo it. You don't want to sound too "robotic". Keeping some very minor variations in tempo and/or loudness will actually make the music more "living" and less like a machine.)
Recording yourself is the simplest solution. You can listen back and determine if you're "on beat". If recording yourself isn't an option, a practice technique is to try to make the metronome "disappear". Set the metronome to a lowish volume, and when you strum on the beat you should be unable to hear the metronome. If you're practicing complicated strumming patterns, then set the metronome to x2 or x4 as fast and try to make the correct e's, &'s, and a's "disappear"
I don't know whether this would work as well when strumming a guitar, but on a keyboard the best tool I've found for practising precise timing is a delay effect.
If you have an FX pedal, multi-FX unit, or audio plug-in with a delay or echo, set it for exactly one beat at your target speed (e.g. 500ms at 120bpm), and try to play in time with yourself.
Perfect timing gives crisp, exact repeats, while any timing imperfections are immediately obvious. I don't know of any other tool that makes the difference between precise and almost-precise so very clear and immediate; the direct feedback is great for practising to.
It works best for short, staccato notes; when notes last a whole beat or longer, the timing is harder to hear. For more interesting and/or musical results — and for more of a challenge — you can set a delay of 2 beats, or even better 1½ beats; together with a little feedback (so that each note gives a few echoes, fading away), that allows for a range of engaging effects.
I’m going to do against the consensus and say you don’t have to record yourself. I recommend two different things:
First, to test your rhythmic accuracy pick a song with an 8th note pulse. Set the metronome to 8th notes (double the tempo of the song). Strum along with the metronome and listen for the synchronization between the metronome and your strums. If you are using strumming patterns that combine quarter and eighth note strums there will of course be some solo clicks here and there. The clicks should almost disappear into the strums. If you are comfortable playing with the metronome and it is fairly effortless to make the clicks and the strums line up with each other then you are able to strum accurately.
Next pick 3 or 4 songs you like to play and are comfortable with. Use the metronome to assign a tempo you like to each song. For each song listen to the metronome click for 10-15 seconds and try and internalize the tempo, then turn off the metronome and play the song. When you’re done compare your ending tempo to the metronome and see if you’re still in the ballpark of the original tempo. It’s ok to drift a little faster or slower, after all we are not machines. However, if the difference is large then you know you have some work to do.
A metronome is an obvious yardstick, however, you state that you want to do it without.
You'll still have to use some sort of yardstick, otherwise how can your timing be tested?
That could be one of several things.
Play along with a drummer with excellent timing skills - there are those around!
Play along to a well-known recording that you already play, turn down the volume for several bars, then up again, and listen to where you and the recording are - in synch, or not. Beware, though, that a lot of music does actually fluctuate slightly, so that's going to impact on your results.
Use a stopwatch - you can time a section of a song, say a verse, then play it, again on the stopwatch, when the final timings should be the same.
But, there's no way you can check yourself without using some kind of outside agency!
I have personally found playing with a loop station to be beneficial to my own timing. Record a rhythm loop, then play along with that. Any timing issues will be obvious and force you to improve, if you want to enjoy the experience at all.
Start with a short and simple chord progression, and get good at timing the foot taps so your loop is well timed. Once you have that down, play longer progressions with more changes to work on maintaining a consistent tempo.
Next level, record a loop of melody, then play your chord progression along with that. This will really help you learn to maintain a consistently timed internal rhythm.
All of the above is also really fun.
Edit: I realize this will necessitate either a pickup in the guitar, or a mic, and an amp or headphones output on the loop station, so apologies if these are not available to you.
Here’s a story about how Earl Scruggs and his brother would walk laps in opposite directions around their house while playing, with the goal to be in perfect time with each other every time they met. Seemed to work for him, but probably not practical for most of us. Maybe you could do something similar with a recording played in one place while you walk in and out of hearing range.
The author also goes on to promote his own technique using a metronome, but I won’t include that, since the question asks for techniques without it.