2. Meaurement & Statistics
Introduction – PreLab
Name:
Section:
Date:
Statistics
Statistics are frequently the most important part of measurements in science.
Imagine you are trying to discover what causes kidney stones. You survey 1000 people with kidney stones and 960 of them
answer yes to the question “Have you ever consumed alcohol?”.
Your conclusion, 96% of people who have kidney stones
drink alcohol. Does this mean people who drink alcohol are more likely to get kidney stones? Lets say you survey 1000
people without kidney stones and 950 of the 1000 say yes. By surveying the people without kidney stones you have created
a
“control”
group. This is very important in studies but we won't talk about it here.
So 96% of people with kidney stones surveyed who drink alcohol and 95% of those without kidney stones drink alcohol.
(1) Question 1:
Do you think our experiment shows that people who have consumed alcohol are more likely to get kidney
stones? Why or why not? (No right or wrong answer here)
Statistical Error
Error that is inherent to a measurement itself is sometimes called “statistical” error.
How many Skittles are in a package? Is there a right answer? We could count the number of Skittles in a bag and we
wouldn't always get the same number. There is some randomness. Sometimes there would be a couple extra and sometimes
a couple would be missing. What if we counted the number of Skittles in 1000 bags and plotted how many times we
counted each number (a histogram) . It would look something like this (I have generated this assuming we could have
partial Skittles)...
This preview has intentionally blurred sections. Sign up to view the full version.
View Full Document
Teachers should especially recognize the above shape. Its a bell curve!
A histogram of measurements with statistical error should make a bell curve.
So how many Skittles are in a bag on average? The formal scientific value from this data is...
54+3
Notice I did not just give a number, I gave a range, or “error bars” too. If you look at the histogram you can see that most
bags had somewhere around 54. This is the average number of Skittles per bag. If you look at the width of the histogram
(from center to side, halfway to the top), you can see that its about 3. In other words, 54+3 means “On average there are 54,
and most bags lie within 3 of this number.”
Lets go back to our other example. We found that, of the people we surveyed, 96% of people with kidney stones consume
alcohol and 95% of people without kidney stones consume alcohol. Formally, our survey answers should be expressed with
errors. We won't go into detail how these errors are calculated.
This is the end of the preview.
Sign up
to
access the rest of the document.
 Fall '08
 Staff

Click to edit the document details