Question: What is the difference between confidence intervals and standard error bars? When is it good to use one over the other?
Answer: First of all, Standard Error tells you the precision of your estimate of the population mean, or how close your sample mean is to the true mean. However, your sample mean is not likely to be exactly equal to the true population mean. Confidence intervals on the other hand are equal to the standard error times the critical value of your test, so instead tell you that if (assuming an alpha of 0.05) you sampled the population 100 times and calculated the mean, 95% of the time your true population mean would fall between the two confidence intervals. This can then be adjusted to whatever confidence level you want (95%, 90%, 99%). So instead of telling you how precise your estimate of the mean is, it gives you a range in which you know the probability that the true population mean is likely to fall.
I think this pdf (starting half way down page 3) describes it pretty well.
http://www-users.york.ac.uk/~mb55/msc/applbio/week5/confint_text.pdf
Now, can anyone explain when we use Standard Deviation instead of Standard Error?