Hello!
How do I draw a line of best fit?
I always have trouble knowing how to draw them and where to stop the line.
Especially for calibration curves in AAS.
Thank you
Whilst in the real world it's possible to determine the equation of the actual line of best fit, as far as HSC science goes your toolbox is quite limited. Given a set of (plotted) data, to draw your line of best fit simply draw in the line that best represents the slope of the curve itself. The following are some tips you may choose to employ
- If almost all of your data can be represented on a smooth line, and by almost all I mean only a few weird ones (maybe 2), you should ignore the weird ones and treat them as outliers or mistakes in the experiment and just draw the line through the other points.
- If your data is very scattered (unlikely to appear in the HSC) and you insist on drawing a line of best fit, focus on the slope of the line of best fit.
- Stop at least the largest value on the x-axis given. We draw lines of best fit for the sake of an 'interpolation' - determining a trend within the extreme values given. Alternatively, stop until you go over the grid that you're given (because in the HSC they usually give grids), because you may need it later for an 'extrapolation'.
- If your data doesn't look linear go for a curve of best fit. Never happens with AAS in particular though.