# Why are some mistakes counted as “experimental error”?

I have always thought (and been told) that experimental errors and uncertainties do not include mistakes which can be fixed by just conducting the experiment more carefully. Yet we still include parallax error and zero error when talking about experimental error. Can these two not simply be removed by conducting the experiment carefully? Or is it that these two might happen despite our care?

Zero error can be usually fixed by re-calibrating measurement device or changing it into a newer not damaged device. It's one-time fix. However story is far more complex about parallax error - you can't be sure that you are looking at device readings at a right angle. Besides usually parallax error will be lost in other more fundamental error source which is half-value of measurement device smallest scale interval. For example. Consider that you are measuring square's perimeter with a ruler which has smallest interval of $$1 mm$$. Your ruler readings are (10 + 10 + 10 + 10) mm. Then because measurement errors have tendency to add-up, your total measurement error of perimeter is $$\frac{1}{2}mm+\frac{1}{2}mm+\frac{1}{2}mm+\frac{1}{2}mm = 2mm$$
Thus your perimeters measurement result would be $$40mm\pm2mm$$