I can answer your 1st and 4th question.
The error bars aren't always the same length.
For example, if there are values in the table such at 4.5 +/- 0.2 , 5.6 +/- 0.3 ,
Then definitely, the error bar of the the first value (4.5 +/- 0.2) will be smaller than (5.6 +/- 0.3)
The reading should have the same significant figures as the raw data or ONE MORE significant figure than the raw data.
The absolute uncertainty should also be the same significant figures as the raw data that is given. Most of the times, the uncertainty that is given in the raw data is of one s.f. so the absolute uncertainty should also be of one s.f.
And I'm sorry but I can't answer your 2nd and 3rd question because I'm not sure about those myself
------
And this is how you calculate the uncertainty in l^2:
I'll show you with the first value. And then you can probably do the rest
The first value on the table is 6 +/- o.4
First find the square of l.
l^2
= (6)^2
= 36
And then you find the fractional uncertainty, which is,
(Uncertainty/original value)
so,
(0.4/6) x 2 (It is being multiplied by 2 because they are asking you for a squared value of L )
= 0.13333...
And then you multiply the fractional uncertainty with the square of L that you have found previously.
(0.133333...) x 36
= 4.8
so the final answer will be : 36 +/- 4.8
This process is used whenever they ask you for a square value.
I hope you have understood!
Don't hesitate to ask me if you have any other questions!
Best of luck!