Transcript 09 One Sample t-test

One Sample t-test
Hypothesis Testing:
• Unknown Parameters Requires t-test
• Comparison of One Sample Mean to
a Specific Value
M  0 M  0
t

sM
s/ n
Anthony Greene
1
Outline
Differences between t and z
1.
2.
3.
4.
Unknown parameters
Degrees Freedom
Using the tables
Hypothesis Testing
Anthony Greene
2
What Is a t-test?
• In most research situations, the parameters  and
 are unknown because the test is novel
• Estimates, based on sample statistics must be used
in place of the parameters
• Use of estimation reduces the certainty of the tests
by a quantifiable probability which depends upon
the size of the sample
• For very large samples, the t-test and z-test are
identical
• t-tests are just like z-tests, except that they
compensate for the increasing uncertainty of small
sample sizes
Anthony Greene
3
The larger the df is, the more closely the t
distribution approximates a normal distribution.
Anthony Greene
4
Use of t-test vs. z-test
1. z-test
a)  is known:
b) Mx is computed
c) M is known ( is known)
2. t-test
a)
 is hypothesized or predicted
(not computed and generally not known):
b) Mx is computed
c) M is unknown ( is unknown) :
sM is computed
d) Degrees freedom (d.f.) is computed as the one less than the
sample size (the denominator of the standard deviation):
df = n - 1
Anthony Greene
5
A Note On The Influence of
Sample Size (slide 1 of 2)
•
For the z-test, sample size influences the
shape of the sampling distribution; the
larger the sample size, the more
leptokurtic the sampling distribution
because larger n means smaller M
Platykurtic
Leptokurtic
Anthony Greene
6
A Note On The Influence of
Sample Size (Slide 2 of 2)
For the t-test, sample size has two effects:
1. Makes the sampling distribution more leptokurtic because larger
n means smaller sM
2. t-distribution is more platykurtic than comparable z-distribution
• The smaller the d.f. the more extreme the effect must be to be
detected (rejecting the null hypothesis)
• For df larger than thirty, the
t-distribution is a very close
approximation to
the z-distribution
(see Table B.1)
Platykurtic
Leptokurtic
7
The t-statistic
M  0 M  0
t

sM
s/ n
• This value is used just like a z-statistic: if the
value of t exceeds some threshold or critical
valued, t , then an effect is detected (i.e., the
hypothesis of no difference is rejected)
• Critical values t are found in Table B.2
Anthony Greene
8
Finding Critical Values
A portion of the t distribution table
Anthony Greene
9
Finding Critical Values
The t-distribution for df = 3, 2-tailed α = 0.10
Anthony Greene
10
Finding Critical Values
The t-distribution for df =15, 2-tailed α = 0.05
Anthony Greene
11
Finding Critical Values
The t-distribution for df =15, one-tailed 2-tailed α = 0.05
Anthony Greene
12
The one-sample t-test for a
population mean (Slide 1 of 3)
Step 1 The null hypothesis is H0:  = 0 (the real mean
equals some proposed theoretical constant 0); the alternative
hypothesis is one of the following:
Ha:   0
Ha:  < 0
Ha:  > 0
(Two Tailed) (Left Tailed) (Right Tailed)
Step 2 Decide on the significance level, 
Step 3 The critical values are
±t/2
-t
+t
(Two Tailed) (Left Tailed) (Right Tailed)
with df = n - 1.
Anthony Greene
13
The one-sample t-test for a
population mean (Slide 2 of 3)
Anthony Greene
14
The one-sample t-test for a
population mean (Slide 3 of 3)
Step 4 Compute the value of the test statistic
M  0 M  0
t

sM
s/ n
Step 5 If the value of the test statistic falls in the
rejection region, reject H0, otherwise do not
reject H0.
Anthony Greene
15
Criterion for deciding whether or
not to reject the null hypothesis
Anthony Greene
16
Summary of hypothesis-testing
The null hypothesis H0:  = 0
Type
z-test
t-test
Conditions
Test Statistic
M  0
M  0

/ n
μ0 is known
σ is known
z
μ0 is hypothesized
σ is unknown
M  0 M  0
t

sM
s/ n
Anthony Greene
M
17
Sample Problem
• Given a multiple choice test where each question
has 4 choices, I want to know if a sample of 24
students did better than chance
•
M = 37, s = 14
• What is µ0?
Anthony Greene
18
Sample Problem
• Given a multiple choice test where each question
has 4 choices, I want to know if a sample of 24
students did better than chance
•
M = 37, s = 14
M  0
t
sM
Anthony Greene
19
Sample Problem
• Given a multiple choice test where each question
has 4 choices, I want to know if a sample of 24
students did better than chance
•
M = 37, s = 14
M  0
12
t

 4.2
14
sM
24
Anthony Greene
20
Second Sample Problem
You are conducting an experiment to see if
a given therapy works to reduce test
anxiety. A standard measure of test anxiety
is known to produce a µ = 20. In the
sample you draw of 81 the mean M = 18
with s = 9.
Anthony Greene
21
Second Sample Problem
You are conducting an experiment to see if a given therapy
works to reduce test anxiety. A standard measure of test
anxiety is will known to produce a µ = 20. In the sample
you draw of 81 the mean M = 18 with s = 9.
9
sM 
1
81
Anthony Greene
22
Second Sample Problem
You are conducting an experiment to see if a given therapy
works to reduce test anxiety. A standard measure of test
anxiety is will known to produce a µ = 20. In the sample
you draw of 81 the mean M = 18 with s = 22.
9
sM 
1
81
18  20
t
 2
1
Anthony Greene
23
Third Sample Problem
You are conducting an experiment on ESP.
People who claim to be high in ESP are
asked to guess which of four cards an
experimenter draws from a deck.
Anthony Greene
24
Third Sample Problem
µ = 0.25
tcrit = 1.83
X
X2
0.220
0.048
0.220
0.048
0.220
0.048
0.230
0.053
0.240
0.058
0.240
0.058
SS
n 1
0.270
0.073
0.260
0.068
s
n
0.220
0.048
0.240
0.058
SS  X
s
sM 
t
2
2

X 

n
M 
sM
Σ
2.360
M=
0.236
Σ
SS=
0.003
s=
0.018
sM=
0.006
t=
-2.492
0.560
Anthony Greene
25