How can anyone know whether Excel is giving the right answer to anything that is more complex that one can do in one’s own head? I mean if I had to check every formula and function, I don’t need Excel, but couldn’t there be a bug that someone missed that returns an erroneous result in specific situations? That could be potentially catastrophic depending on what the data and results were used for.
In: 0
This is why modern maths curricula are trying to teach kids number sense, rather than rote calculation steps.
It means that when you put your formula into excel, you know how to sense check it.
For example, if I put in:
247 + 4674 * 3566
I know the answer is an odd number that is less than 20,000,000, and more than 14,000,000 Because the 247 is tiny compared to the final result, and I can round up 4674 and 3566 to 5000 and 4000 respectively. Or I can round them down to 4000 and 3500. I know the final answer is odd, because multiplying two even numbers gets an even number, and adding 247 to an even number gets an odd number.
This is step one of the “differences” method used in “new maths” when doing multiplication. The key is that here, I’m letting the computer do all the tricky stuff.
Computers are good at the tricky stuff, but they’re no good at telling if the data entered in is correct. So if they give a wrong answer, it’ll often be very wrong. So a wrong answer is often easy to spot. Humans are much better at knowing what the answer *should* look like, and then investigating if it looks wrong.
Depending on how much I care, I could try for a greater or lesser rigour in my cross-checking. It really depends on the report being run, and what kind of errors you expect to see in your data. A financial statement needs to zero out exactly, while an estimate might be more forgiving, especially if you know the tendency in your data is to overestimate time and costs.
Latest Answers