Quantitativeanalysis.correlation

Front 1

How to input data for two variable in taxas Instument
calculate a,b,r ?

Back 1

2nd Data --> data input
2 Stat --> scroll LIN
y = a + bx,
a --> independent variable
b --> dependent variable
correlation r

Front 2

Was erklärt die Korrelation ?

Back 2

quadrat der r (correlation coffezient)--> Determination Coffezient
Mit den könnte die Schwankungen in Werte Y druch die Schwankungen in Werte X werden erkärt.

Front 3

Correlation Hypothesis Test ?
Formula, Degree of Freedon ?

Back 3

t - test
2-seitig when H0 : r = 0;
degree of freedon = N-2
t-statistics = r*(N-2)^0.5/(1-r^2)^0.5

Front 4

Covariance Formula ?

Back 4

r(x,y) = Cov(x,y) / (sigma(x)*sigma(y))
Beta = Covariaz(x,y)/ Varianz(x)

Front 5

What is SEE ?
SEE v/s Standerd Deviation

Back 5

Standerd Error of Estimate
misst eine Schwankung herrum eine geschätze Wert
weil, Standerdardabweichung misst eine Schwankung herrum eine mittelere Wert.

Front 6

Total Variation

Back 6

_
Y(i) - Y

--> lead to
SST

Front 7

Unexplained Variation

Back 7

Y(i) - Y^(i)

--> lead to
SSE

Front 8

Explained Variation

Back 8

_
Y^(i) -Y

--> lead to
SSR

Front 9

Coefficient of Determination R^2

Back 9

SSR / SST

Front 10

Was ist Standardfehler der Schätzung ? Formule ?

Back 10

Da die beobachteten Y-Werte kaum auf der Regressionsgeraden zu liegen kommen ergibt sich eine "Fehler der Schätzung" t-verteilt herum Regressionsgeraden.
Its also the Standerd Deviation of the predicted value for the dependent variable.
SEE = (SSE/(n-2))^0.5 = [(sum(y-y^)^2/(n-2))]^0.5
Ref : Page 12 on Handout + Compendio

Front 11

How to Calculate Standerd Error of Cofficient Beta (from sample)? Hypothesis test ?

Back 11

Beta (Population) = beta(sample) +(-) t( t-value with degree of freedom n-2) * SEE/[sum(x- x-)^2]^0.5
t(value) = (beta - Beta) /Standerd Error Beta

Front 12

Problem :
We would like to prove that b1 (Polulation) is larger then 0.5 , design and prove the appropriate hypothesis test at a 95% conidence level Asume :
SEE = 15.06 , N=10 and
sum(x- x-)^2 = 374.5 and b1^ = 1.188

Back 12

Example is wong,
Null Hypothesis can not be rejected .
t(n=8, significance 5%) = 1.82
t wert = 1.462

Front 13

What is a OLS

Back 13

Ordinary Least Squere .
Graphic from Iphone

Front 14

Relation between SST, SSE and SSR ?

Back 14

SST = SSE + SSR

Front 15

Question 9 from Quantitive workbook (page 57)
Picture from Iphone

Back 15

A. 0.4279
B 0.6542
C 1.178
D 1.544