Is the data normally or lognormally distributed? I'm not sure how to give context to this question. We're to use Excel to analyze data and use log base 10 for each column of data that we analyze, which I'm not sure what they want here. Are they wanting us to use the log base 10 for the analyzed data or for the raw data? Then the assignment asks us to determine if the data is normally or lognormally distributed. I do not know how to do this and Google searches do not sound like English.
 A: There's no process by which you determine with any certainty what distribution data is from. 
i) real data is never (or at least essentially never) from any simple well-known distribution, so whatever distribution you're checking for, it isn't exactly that.
ii) even if it were of some reasonably simple form, that simple form has almost no chance (unless the data was deliberately simulated) of being on a short list you might consider trying; there are very large numbers of equally or almost equally simply looking distributions that might fit about as well.
It's no trouble to check diagnostically which of two (or some other small number of) proposed distributions appear more consistent with your data... and this might be what they're after.
So, for example, on could do a Q-Q plot of the data and the log-data. If the first shows a straight line while the second has the curvature that indicates left-skewness, the data are consistent with normality. If the first indicates right skewness while the second is straight, then the data are consistent with lognormality. On the other hand, it's quite possible that both look skew, or even that both look reasonably straight:

If they booth look skew it may be that neither is a tenable model, while if both look straight (as above) either might be tenable.
The point is, none of this (nor any form of formal testing) will tell you what it is. It may in some cases tell you what it probably isn't.
