by StaceyKoprince Fri Nov 04, 2011 2:24 pm
If there are two answers that are extremely close, then you shouldn't expect to be able to use estimation to get to the one right answer (except, on occasion, when the problem explicitly indicates that you can estimate - but, even then, the answers usually have a decent spread).
Estimation *can* be useful to get to the one right answer, but only if there is a decent spread in the answer choices. Otherwise, estimation is a *guessing* technique - it can get you close and improve your odds when you do finally guess.
If you find that you are making choices about how to estimate that then "skew" your results and lead you to the next higher or lower answer in the mix, then there is definitely something you can do about that.
Whenever we have to pick a new number for something (eg, they give me 67% and I choose to use 70% instead), our goal is to try to pick the closest possible number that will make the problem easier, right? First, I need to think very carefully about what will make the problem easier. For instance, in the example I just gave above, 2/3 is even closer to 67% than is 70%, and maybe the problem would be even easier if I switched to fractions instead of percents - so that kind of flexibility in thinking is crucial.
Next, what if I have to make more than one estimate while I'm working my way through the problem? Then I need to think about how I'm going to *offset* my estimates - choose to go a little bigger one time, and then a little smaller the next time, so that the errors I'm introducing will somewhat cancel each other out. If I go bigger the first time and go bigger the second time, now I'm taking the risk that I'll end up too close to one of the other answers.
For example, let's say we have this problem:
67% * 24%
Yuck.
67% --> 2/3 (rounded down)
Because I'm doing multiplication, I want to round *up* on my next one:
24% --> 25% --> 1/4 (rounded up)
Let's contrast that with division:
(67%) / (21%)
67% --> 2/3 (rounded down)
This time, I actually want to round down again. Denominators work "opposite" to the normal way things work. Decreasing a numerator makes the whole thing smaller, but decreasing a denominator makes the whole thing bigger. So if I want to offset my decreased numerator, I have to decrease the denominator as well:
21% --> 20% --> 1/5
Make sense?
Stacey Koprince
Instructor
Director, Content & Curriculum
ManhattanPrep