Refractive index experiment/finding error on graphs

[peanut111]

I'm writing a physics practical report on the refractive index of glass. We performed an experiment in which we measured the incident angle and the refractive angle of light entering a glass block. When plotting sin(ϴ_a) against sin(ϴ_g) with the y-intercept as 0

(to satisfy snell's law:
n_asin(ϴa)=n_gsin(ϴ_g)
sin(ϴ_a)=(n_g/n_a)sin(ϴ_g)
where n_g=refractive index of glass and n_a =refractive index of air.
This is the same form as y=mx)

it yields a line with a gradient equal to the refractive index of glass. We also have to calculate the error, using the minimum and maximum gradient method. However, my data would better fit a trend line without a y-intercept of 0, but when I give it this trend line the gradient is 1.1, which obviously can't be the refractive index of glass. My problem is, that when I add my maximum and minimum gradient trend lines to fine error, they are both less than the one I have. Is there any other way to find uncertainty in a graph?

 

[ehild]

Hy peanut111, welcome to PF.
Can you show your measurement data? ehild

 

[peanut111]

I've attached my data :)

 

[ehild]

I've attached my data :)

I can not open xlsx files, can you convert them to xls or txt? or just type in some pairs of angles?
Are you sure you used the angles between the rays and the normal of the plane?

ehild

 

[peanut111]

I hope this works for you :)

Also, I'm pretty sure I measured correctly, and the data correlates, it just works less well when the y-intercept is 0 because the maximum and minimum gradients are both less than the one I have.

 

[ehild]

Thank you for the data. To tell the truth, the measurement was not accurate enough, and even there might have been some systematic error. It is rather difficult to measure angles with satisfactory accuracy. How did you measure them? What was the experimental set-up?

You should include the data for zero incidence. If you see the ray which enters normally to the surface, it should travel in the same direction. Adding the (0,0) point to both graphs it makes the results more acceptable.

ehild