Describe the relationship between bias and variability through an example. Given that the proportion of all Americans that believe President Bush is doing a good job is 42%, you then conduct a series of 10 polls to confirm or refute this claim of 42%. Describe your results if… a) There is high bias and high variability in your results. b) There is high bias and low variability in your results. c) There is low bias and high variability in your results. d) There is low bias and low variability in your results.
@Hero @phi
@phi ?
Your survey will give you a percent who say good job and a percent who say bad job. you can find the average of the answers. If there is low bias, the average will be close to 42%. if there is high bias, the average will be far higher than 42% (or far lower) variability will be difference between the averages of the 10 polls. If all 10 give roughly the same answer, you have low variability. Notice you could get 10 polls that each say 60%good job. that would be biased (far from 42%) , but low variability (all 10 give about the same answer)
a) There is high bias and high variability in your results.
what would i put for that?
@phi
if there is high bias, the average will be far higher than 42% (or far lower)
but what about with high versus low variability:? @phi
variability will be difference between the averages of the 10 polls. If all 10 give roughly the same answer, you have low variability.
im not sure how i would answer a tho?
Bias is related to skewedness, and variability is related to standard deviation.
if you write down 10 numbers and they are all wildly different from 42, you have high bias and high variability
Okay i get that. what about high bias, low variability
is this right for A?(a) If there is high bias and high variability, the numbers will not be anywhere near the 42 percent. :
@phi
@amistre64
yes, give an example of 10 numbers
Ok what about b, c and d
phi seems to have a better handle on this then I would :)
k. What is high bias and low variabiliity ? @phi
High skewedness, and low standard deviation.
I dont thihnk thats what the question wants.. @phi are u stilll here
What is high bias and low variabiliity ? what does high bias mean ? low variability means all 10 polls give close to the same answer.
So would (b) be: The polls will be around 42.
no, around 42 is low bias
then im confused on high bias with low variability??
first, answer what does high bias mean ?
idk, alot of bias?
if there is high bias, the average will be far higher than 42% (or far lower)
ok but what about that AND low variability
you have 10 polls. each has an average. if they have high bias, the average is far from 42%. variability is how much the answers differ from poll to poll. low variability means they all have close to the same average.
So (b) is: The average is far from 42 percent, with close to the same avergage.. that doesnt make sense
here is an example with 3 polls: 10%, 11%, 9% they have almost the same average. but they are nowhere near 42%. they have high bias, but low variability
can u explain to me how to word b please
lik ehow to word it.
I put If a certain number of polls have just about the same average, but are nowhere near 42 percent, they have high bias and low variability.
how about: b) There is high bias and low variability in your results. the poll results all are close to the same value, example: 20%, but far from 42%
what is low bias and low varibility
ok, c?
what do you think?
Umm. The average will be close to 42%?
all of the polls will be close to 42%,
So c is: All of the polls will be close to 42%?
yes
ok what baout d? im confused on that one
wait, that is for d)
oh ok, yeah i mean what is c?
for c) There is low bias and high variability in your results. if you list the polls they will have a spread around 42%, and if you average the polls you will get 42% low bias, high variablity might be 20% 82% 42% 74% 10%
If there is is low bias and high variability, if you list the polls they will have an average around 42 percent. If you average the polls, you'll get 42 percent low bias. The high variability might be 20, 72% 42% 63% 10%
is that right for c?
yes, but you should use your words, not mine
I will, but is this around right @phi ? (a) If there is high bias and high variability, the numbers will not be anywhere near the 42 percent value. If I wrote down 10 numbers and they were all completely different from 42, I would know that I have high bias and high variability. (b) If a certain number of polls have just about the same average, but are nowhere near 42 percent, they have high bias and low variability. For example, 20 percent, but its far from 42 percent. (c) If there is low bias and high variability, if you list the polls they will have an average around 42 percent. If you average the polls, you'll get 42 percent low bias. The high variability might be 20, 72% 42% 63% 10% (d) If there is low bias and low variability, all of the polls will be close to 42 percent.
Join our real-time social learning platform and learn together with your friends!