Line of best fit? (1 Viewer)

[mmmm.]

Do we always use line of best fit for graphs or can we join the points. When we use line of best fit how do we know when we keep the line straight and when we use a curved line?

 

[Vitrecan]

• if they say use a line of best fit then do a line of best fit, I don't remember coming across questions where we can connect the dots
• I believe the line is always straight for subjects like physics and maths but I can't comment on the other subjects

 

[turtle67]

Yeah or if one of the questions is asking to predict or describe a trend you'd use a line of best fit. For most questions in chem it's a line of best fit from what I've seen and done

 

[vishnay]

It depends on the context if you can clearly see that the data has a non-linear relationship then look for ways to make the relationship linear. If you cannot do that then you can draw a curve of best fit. Curves of best fit are used in the sciences more than Maths.

 

[jazz519]

Never connect the dots. The whole point of drawing a graph in the sciences is to show a trend between different variables. Therefore if the data:
is linear - use a line of best fit
if the relationship is not linear - use a curve of best fit

 

[CM_Tutor]

Also, watch out for data which involve two lines of bit fit, and extrapolate between them

 

[ultra908]

Never connect the dots. The whole point of drawing a graph in the sciences is to show a trend between different variables. Therefore if the data:
is linear - use a line of best fit
if the relationship is not linear - use a curve of best fit

i got a question in my trial which was like plotting the incidence of a disease over the seasons, where we were told to connect the dots. This was bcos the relationship isnt rlly clear or consistent, like how in climate data or economics data they always have dot to dot graphs

 

[jazz519]

i got a question in my trial which was like plotting the incidence of a disease over the seasons, where we were told to connect the dots. This was bcos the relationship isnt rlly clear or consistent, like how in climate data or economics data they always have dot to dot graphs

If you are told to do that then it's fine but just in a random question 99% of the time it's going to be a normal line of best

 

[mmmm.]

Thankyou everyone

 

[CM_Tutor]

i got a question in my trial which was like plotting the incidence of a disease over the seasons, where we were told to connect the dots. This was bcos the relationship isnt rlly clear or consistent, like how in climate data or economics data they always have dot to dot graphs

As Jazz says, join the dots if that is the instruction, but even for climate or economic data a plot of something like a moving average can be more appropriate.

 

[Eagle Mum]

For the majority of high school science experiments, the line of best fit is expected to be a straight line with the ‘least squares’ such that the square root of the sum of the squares of the distances of each data point to the line is the lowest possible value. Without Excel or some other form of statistical software, at best, you can only draw the line (with a ruler) by visual estimation.

There are also tests of linearity which are extremely tedious to calculate manually, but again Excel and statistical software are easy to use. These software will calculate whether a straight line or a curve (and the order of the polynomial equation for the curve which is the best fit) is a better fit for the data, but I don’t think that’s expected at high school level.