Error bars

In my coursework we have to do error bars on our graphs. What do error bars mean and are they good or bad?
Thanks

When you are carrying out repeats in an experiment, it is very unlikely that they will all be the same because of a number of errors that you can't reasonably remove. For instance, if you're measuring the volume of oxygen being given off by a plant per minute and you do three repeats, you may have results of 1.5mL, 1.4mL, 1.7mL. They're obviously all similar, but they're not exactly the same. Error bars are there to show how much error you've got which is important because the less error there is, the greater the certainty that patterns you've discovered are actually real and hence the more accurate your results are likely to be.

How exactly you draw error bars depends. Traditionally, the standard deviation is often used to show error and the actual point is plotted as the mean (so the error bars cover the range of the mean +/- 1stdv). However, at GCSE you wouldn't be expected to calculate the standard deviation so you'd just use the mean and range. So in the case described above, the point would be plotted at 1.53mL (the mean) and the error bars would cover the range of 1.4mL-1.7mL.