statistics
DESCRIPTION
TRANSCRIPT
Statistics
Z-test (1 var)
T-test (2 vars)
F-test (AVONA)
Correlation Regression Parametric Tests(numeric: mean/stddev)
Kelly Chan | Nov 20 2013
F-test (AVONA)(3 vars)
X-test (Chi Squared)(n vars)
Non-Parametric Tests(category: frequency/proportions)
Statistics
(X)-Tests
Z-test T-test F-test (AVONA) X-test (Chi Squared)
Dependent Independent Goodness-of-Fit Test for Independent
One item
Pre/Post
Longitudinal
Two-Condition
Two items
Observation
Experiential
Kelly Chan | Nov 20 2013
1 variable
Normal distribution
Large sample size
Population StdDev
Non-directional (+/-)
Numeric (mean/stdev)
2 variables
Normal distribution
Small sample size (15~30)
Population StdDev unknown
Non-directional (+/-)
Numeric (mean/stdev)
3 variables
Normal distribution
One-directional (>0)
Numeric (mean/stdev)
Two-Condition
N categories
One-directional (>0)
Category (frequencies/proportions)
X-Statistic
X-CriticalValue
Hypothesis
MeanDifference
Mean
StdDev
Alpha: significance level
df: degree of freedom
2
1
(distances)DifferenceBetweenMeans
(stdDev)StandardError
Formulas Process
Statistics
Calculations
X−Table
Kelly Chan | Nov 20 2013
StdDev
n
Standard
Error
(%) The proportion
of a difference in
means
Decision(stdDev)StandardError
Effect Size
Significance
CI=mean(X )±MarginOfError
CI=mean(X )±CriticalValue(X )∗SE(StandardError)
(distances)DifferenceBetweenMeans
SDpooled
P-Value
CI: Confidence
Intervals
3
4
Statistics
Formulas
Z-test T-test F-test (AVONA) X-test (Chi Squared)
Dependent Independent Goodness-of-Fit Test for Independent
df=n n−1 n(A)+n(B)−2 dfbetween
=k−1
dfwithin
=N −k
(NumberOfRows−1)∗(NumberOfColumns−1)
x x A/B A/B/C A/B/C/ .../N
Sp
2=SS (A)+SS (B)df (A)+df (B)
∑(x −m(x))2+∑( y −m( y))2
∑(x−m (x ))2
n−1∑(x−m (x ))2
n
Kelly Chan | Nov 20 2013
Sp
2=∑(x
i−m(x))2+∑( y
i−m( y))2
df (A)+df (B)
SE=√( Sp
2
n (A)+
Sp
2
n(B))
t=m (A)−m (B)−(u (A)−u (B))
√(S
p
2
n(A)+
Sp
2
n (B))
x2=∑ (
( fo− f
e)2
fe
)
Cramer' sV =√(x2Statistic
n(k−1))
F=∑(n
k∗(m (x
k)−m(G))2)/(k−1)
∑(xi−m(x
k))2/(N −k )
EtaSquare(n2)=SS
between
SStotal
Cohen' sD=m (x
A)−m(x
B)
√(MSwithin
)
Tukey' sHSD=q∗√(MS
within
n)
Cohen' sD=m (x
A)−m(x
B)
√(Sp)
√( variance (x1−x2)
n−1)
mean (x1)−mean (x2)
√(variance(x1−x2)
n−1)
r2=
t2
t2+df
stdDev
√(n)
x−mean (x )StdDev
Statistics
Correlation ���� Regression
t=r∗√(N −2)
√(1−r2)
r=cov (X ,Y)S
x∗S
y
slope=∑((x
i−mean(x))∗( y
i−mean ( y)))
∑ (xi−mean (x )2)
slope=r∗(S
y
S)
Correlation? Hypothesis? (X)-Test? Regression
Kelly Chan | Nov 20 2013
r2<− coefficient of determination
Sx
Y−intercept=mean( y)−slope∗mean(x )