Not homework.
Let’s say I have the results to a survey on how many hours respondents sleep on average. I could ask them to self-report, and I believe that would result in a ratio measure (as 0 is theoretically possible and meaningful, even if nobody can truthfully answer that). Some may answer 8 some 3.14159 hours and we know the magnitude differences. Makes sense so far.
I could also ask them to answer categories, a = 0 hours, b = 1 hour, c = 3 hours etc. Would that still be ratio? And would that depend on if the final category is e.g. 10 hours vs. 10 or more hours?
Finally, what if I have ranges, like: a = 0 hours, b = 1-2 hours, etc. That’s not ratio or interval because the sizes are unequal, so ordinal? What if they’re all equal size like 0-1, 2-3, 4-5? What’s throwing me off is that we do know the sizes and magnitudes, just not if someone answered 6 vs. 7.
If they give answers like 8 or 3.14159 then that’s ratio.
If you make it multiple choice a=0 b=1 c=2 that would technically be ordinal but an argument can also be made for interval because you’re implying that if the answer is between 0:30 and 1:29 then they should choose b. You’ve actually created intervals from 0 to 0:29, 0:30 to 1:29, 1:30 to 2:29 etc.
It’s normal for the first and/or last interval to be different sizes than the others. What’s important is whether the distance between the boundaries is constant. In this case, your boundaries are 0:30, 1:30, 2:30, etc. which are all exactly 1 hour apart. If you widen the intervals to 0:00 to 1:29, 1:30 to 3:29, 3:30 to 5:29, etc. then each of your boundaries are exactly 2 hours apart and it’s still interval.
I think your understanding of what “interval” level of measurement is differs from mine. I’m not a statistician but I have taught introductory statistics, and from everything I’ve seen, “interval level of measurement” doesn’t mean that the measurements, or values of the variable, are intervals, but there are precise numerical intervals between values of the variable.
The classic example is temperature (either in degrees Fahrenheit or Celsius). An interval of 10 degrees represents the same temperature difference whether it’s between 25 and 35 or between 90 and 100. But, unlike the ratio level of measurement, there’s no meaningful zero (0 degrees doesn’t mean "no temperature at all) and no meaningful ratios between temperatures (50 degrees isn’t “twice as hot” as 25 degrees).
In all the instances where the OP is talking about dividing responses up into a set of categories, I’d call it ordinal.