Download Slide 1

Survey
yes no Was this document useful for you?
   Thank you for your participation!

* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project

Document related concepts

History of statistics wikipedia , lookup

Degrees of freedom (statistics) wikipedia , lookup

Taylor's law wikipedia , lookup

Bootstrapping (statistics) wikipedia , lookup

German tank problem wikipedia , lookup

Misuse of statistics wikipedia , lookup

Student's t-test wikipedia , lookup

Transcript
How many colleges did you apply
to?
Type the number into your clicker and
hit “send”
10-2
Estimating a Population Mean
(σ Unknown)
Confidence Intervals in the
Calculator
•High School students who take the SAT Mathematics
exam a second time generally score higher than on
their first try. The change in the score has a Normal
distribution with standard deviation σ=50. A random
sample of 250 students gain on average x-bar=22
points on their second try.
•Construct a 95% Confidence interval for μ
Confidence Intervals Involving Z
Using the Calculator
What if we don’t know 𝜎?
In common practice, we would never know the
population standard deviation.
• Instead, we would use an estimate of 𝜎: the
sample standard deviation, s.
• We then estimate the standard deviation of 𝑥
𝑠
using σ𝑥 =
𝑛
• This is called the standard error of the sample
mean 𝑥
“Standard error”: You are estimating the
standard deviation…but there will likely be
some error involved because we are
estimating it from sample data.
In other words… the standard error is (most
likely) an inaccurate estimate of a (population)
standard deviation.
The t distributions
When we substitute the standard error of 𝑥
𝑠
σ
( )for its standard deviation ( ) we get the
𝑛
𝑛
distribution of the resulting statistic, t.
We call it the t distribution.
The t-statistic was introduced in 1908 by William
Sealy Gosset, a chemist working for the Guinness
brewery in Dublin, Ireland ("Student" was his pen
name). Gosset devised the t-test as a way to
cheaply monitor the quality of stout.
The t distributions
There is a different t-distribution for each sample
size n.
We specify a t distribution by giving its degrees of
freedom, which is equal to n-1
We will write the t distribution with k degrees of
freedom as t(k) for short.
We also will refer to the standard Normal
distribution as the z-distribution.
Comparing t and z distributions
Compare the shape,
center, and spread of
the t-distribution with
the z-distribution.
As the degrees of freedom k increase, (the
sample size increases), the t-distribution is
increasingly Normal.
Our formula is the
same as it was for zintervals EXCEPT we
replace sigma with s!!!
Finding t with Table C
Suppose you
want to construct
a 95%
confidence
interval for the
mean μ of a
population based
on a SRS of size
n=12. What
critical value t
should you use?
Finding t with Table C
Suppose you want to construct a 95%
confidence interval for the mean μ of a
population based on a SRS of size n=12.
What critical value t should you use?
Finding t with Table C
Suppose you want to construct a 90% confidence
interval for the mean μ of a population based on a
SRS of size n=15. What critical value t should
you use?
Finding t with Table C
Suppose you want to construct a 99% confidence
interval for the mean μ of a population based on a
SRS of size n=34. What critical value t should you
use?
One sample t interval for 𝜇
1)SRS
2) Normality (if you have the raw data you must draw a
boxplot!!!)
- n < 15 : Use t procedures if data are close to
Normal with no outliers
- n ≥ 15 : Use t procedures except in cases of
outliers of strong skew
- n ≥ 30 : Use t-procedures even for clearly skewed
distributions (cannot have extreme
outliers)
3) Independence
One sample t interval for 𝜇
Let’s use our class data to construct a 95%
confidence interval for the true mean number of
colleges that high school seniors applied to in
2013.
One sample t interval for mu
Step 1: STATE
Step 2: PLAN
Step 3: CALCULATIONS
Step 4: INTERPERATION
State: We are estimating ________, the true mean
________________________________
______________________________.
Plan:
Procedure:
Conditions: 1)
2)
3)
Calculations:
Interpretation: We are 95% confident that the
true mean
“Last year, 750,000 applicants submitted 3 million
applications, an average of four per student”
College Decision Day: More Applications, More Problems|TIME.com
http://nation.time.com/2013/05/01/as-college-applications-rise-so-doesindecision/#ixzz2sr0ANbp4
Paired t-procedures
To compare the responses of the two treatments in a
matched pairs design or before and after
measurements on the same subjects, apply the one
sample t procedures to the differences observed
between the pairs.
• µdiff = the mean difference between each pair
Ex) Mrs. Skaff gave a new study tool to her students to
see if it would improve their test scores. She matched
students based on current grade and randomly gave
one student in each pair the study tool.
Paired t-procedures
• µdiff = the mean difference in student grades (given a study tool –
not given a study tool).
Ex) Mrs. Skaff gave a new study tool to her students to see if it would improve their test
scores. She matched students based on current grade and randomly gave one
student in each pair the study tool. She wants to know if the study tool improved test
scores.
Given Study Tool 92
73
81
89
95
90
96
72
85
88
No Study Tool
90
Study tool - none 2
73
84
84
88
91
93
70
80
88
0
-3
5
7
-1
3
2
5
0
State: We are estimating ________, the true mean difference in student grades (given a study
tool – no study tool)
Plan:
Procedure: One Sample (paired) t Confidence Interval for means (σ unknown)
Conditions: 1) Did not state that this was an SRS. Proceed with caution
2) The boxplot appears approx. normal so with a sample size of 10 we
can say that the sampling dist. is approximately normal.
3) Assume that there are at least 10(10) =100 students in the population.
Condition for independence is met.
Paired t-procedures
• µdiff = the mean difference in student grades (given a study tool –
not given a study tool).
Ex) Mrs. Skaff gave a new study tool to her students to see if it would improve their test
scores. She matched students based on current grade and randomly gave one
student in each pair the study tool. She wants to know if the study tool improved test
scores.
Calculations: (80.082, 92.118)
Interperet:
𝒙diff = 86.1
s = 8.4123
n = 10
t* =
We are 95% confident that the true mean difference
in student grades (given a study tool – not given a
study tool) is between 80.082 and 92.118.
Confidence Intervals in the
Calculator
You still need all other steps!!!!
For calculations you must define
ALL variables!!!
Ronald McDonald’s sister Diana Rhea is the
purchasing manager for domestic hamburger
outlets. The company has decided to provide a
free package of Tums to any complaining
customer. In order to estimate monthly demand,
she took a sample of 5 outlets and found the
number of Tums distributed to customers in a
month was
250, 280, 220, 280, 320
(a)Find the sample mean and sample standard
deviation
(b)Construct a 85% confidence interval on the
average monthly demand per outlet.
Homework!