Professional Documents
Culture Documents
Single case:
In order to determine probability of obtaining a score, it has to be compared to the appropriate distribution known as the Sampling Distribution. Distribution of sample means
The distribution of sample means is defined as the set of means from all the possible random samples of a specific size (n) selected from a specific population.
m =
The sampling distribution of means has a smaller variance. This is because the means of samples are less likely to be extreme compared to individual scores.
2 = n
M =
IQ Scores - Ranked
71, 76, 76, 77, 79, 80, 81, 82, 83, 83, 84, 84, 84, 85, 85, 86, 87, 88, 88, 88, 89, 90, 90, 91, 91, 91, 92, 92, 92, 93, 93, 93, 93, 93, 93, 94, 94, 94, 94, 95, 95, 95, 96, 96, 97, 97, 97, 97, 97, 97, 97, 98, 99, 99, 99, 99, 100, 100, 100, 100, 101, 101, 101, 102, 102, 102, 102, 103, 103, 103, 103, 103, 103, 103, 104, 104, 104, 105, 105, 106, 106, 107, 107, 107, 107, 107, 108, 108, 108, 108, 109, 109, 110, 111, 111, 112, 112, 112, 113, 113, 113, 113, 114, 114, 115, 115, 115, 117, 118, 120, 121, 121, 121, 123, 123, 125, 125, 126, 131, 136
Sample Size Lowest extreme score M N=1 N=2 N=3 N=4 71 71+76 71+76 + 76 71+76 =76 + 77 71 73.5 74.3 75.0 136 131 + 136 126+ 131 + 136 125 + 126 + 131 + 136 Highest extreme score M 136 133.5 131.0 127
= 15
M =
15 15 = =3 25 5
z=
Area =
.0228 .0456
Z Test
z= M
t Distribution
What happens if we do not have the population standard
deviation? We can use the sample standard deviation as an estimate. Problem Cannot use z and the normal distribution to estimate M probability. z= M
Sample variance tends to underestimate the population
variance
Distribution
t Distribution cont.
Not a single distribution but a family of distributions. One for each degree of freedom (df).
df = n - 1
Sample: Norms:
n = 56 = 100
M = 104.3
s = 12.58
M = 104.3
s = 12.58
M t= sM
s sM = n
z=
M = 104.3
s = 12.58
t=
df = 56 - 1 = 55
Potential Problem
M t= sM
What happens if we do not have a population mean? Two ways of dealing with it. Use a repeated measures design Use two independent samples
Repeated-Measures t test
(In SPSS called the Paired-Samples t-test)
Use one sample, but test at two different times. Occasionally matched samples used. If treatment has no effect what will the outcome be? There will be little difference between the scores on the first and second testing. Allows us to hypothesise what the population mean will be. Mean of the differences will be 0 (zero). Will not be exactly the same due to sampling error So have to test whether the mean difference observed is significantly different from o. Need a sampling distribution
Example: MDI 6 months (X1) 124 94 115 110 . . 126 123 M s n 111.0 13.85 31 MDI 24 months (X2) 114 88 102 127 . . 114 132 106.71 12.95 31 Difference (D) -10 -6 -13 17 . . -12 9 -4.29 16.04 31
D = - 4.29 H0 :
sD = 16.04
n = 31
D = 1 2 = 0
M t = D t= s sM
D
t=
D D 0 D 0 = = sD sD sD n
df = n - 1 df = 31 - 1 = 30
use an independent samples (between subjects) design to get sample data that allows us to evaluate the difference between two populations using the independent samples t test. As with all hypothesis tests, the general purpose of the independent-measures t test is to determine whether the sample mean difference obtained indicates a real mean difference between the two populations (or treatments) or whether the obtained difference is simply the result of sampling error.
Variance of the two Distributions of the Means Standard Error of the Distributions of the Means Mean of the Distribution of Mean Differences Variance of the Distribution of Mean Difference Standard Error of the Distribution of Mean Difference
12
N1
1
N1
and
22
N2
2
N2
and
1 2 = 0
2 X1 X 2 =
12
N1
22
N2
X2
12
N1
22
N2
t=
M sM
= ( M1 M 2 ) ( 1 2 ) s1 n1
2
( M1 M 2 ) ( 1 2 ) t= s M1 M2
s + 2 n2
( M1 M 2 ) s1 n1
2
s + 2 n2
Degrees of Freedom
df = (n1 - 1) + (n2 - 1) = n1 + n2 - 2
Summary
Tests
Z score
z=
z=
M m M sM
t=
t=
t =
D D 0 = sD sD
( M1 M 2 ) ( 1 2 ) s M1 M2 ( M1 M 2 ) 0 s M1 M2
t=
Standard Errors
Z test t test - single sample t test - two matched samples
X =
sM = sD =
N
s n sD n
1 2 1 + sp n2 n1
2 2
s M1 M2 =
(n 1) s1 + (n2 1) s2 = 1 n1 + n2 2
Independent samples
Random sampling The data are measured on at least an interval scale The participants in the two samples are independent The distribution of the populations scores are normal The variances of the two populations are the same (homogeneity of variance)
Robust
Analysis of Variance
What happens when we have more than two groups?
First Method
sj k
s1 + s2 + s3 + s4 + s5 = 5
2 2 2 2
Second Method
( j G ) k 1
MSbetween MSwithin
Systematic variance (treatment effect) + unsystematic variance MSbetween F = = unsystematic variance MSwithin
obtained mean differences (including treatment effects) MSbetween F = = differences expected by chance (without treatment effects) MSwithin