Mra Exam Notes
Mra Exam Notes
Ho (Null Hypothesis)- It means there is no significant difference between the sample and actual Population.
Ha (Alternate Hypothesis)- It means there is a significant difference Between the sample and the actual population.
1- Frequency Distribution
There are Two Groups for group 1 we do the goodness of fit test (Chi-square Test) and for Group 2 we do the Test of Independence (Chi-Square Test)
2- Proportion
There are two groups for group 1 we will compare the sample and population and in group 2 we will compare two samples only (Z-Test and T-Test)
we will conduct Z – test when the Standard Deviation is not Known and the Sample size is high.
We Will Conduct F- Test when Standard deviation is not Known and the Sample size is Small.
3- Means
There are Two Groups and we will do Z- Test and T-test but if there are multiple samples then an F-test is used.
4- Variance –
There will be two groups for 1st group when we compare samples with the Population we use the Chi-Square Test and for Group 2 when we compare two
samples then we use F- Test.
R ES EA RCH P RO B L EM
-
A R E S E A RC H P R O B L E M I S T H E M A I N I S S U E O R Q U E S T I O N T H AT A S T U DY I S T RY I N G T O S O LV E
There is no Dependent variable in Correlation.
There is No cause-and-effect relation in Correlation.
There exists only a Linear relationship in Correlation
REGRESSION MODEL
Whatever problem exists with the correlation model it gets rectified it with the Regression Model. There is a Cause-and-Effect Relationship because it has
Dependent and Independent Variables.
It is of Two Types
1- Simple Regression (1 dependent variable and 1 independent variable)
2- Multiple Linear Regression( 1 dependent variable and multiple independent variables.
R Square- What impact dependent variable Regression here defines that changes in sales accounts Equation Modelling
show because of the variation in the for 125197.458 because of advertisement and other Y= -0.852.084+19.070X1
independent variable? factors accounts for 13769.208 which is called residual
It is always suggested the Lower the R square is which becomes total.
better
Mean Square= Sum of squares/ df
Adjusted R square- To Compare two different
regression models it is used and the more the F-test = To compare squares
better it is. P Value less than 0.05 (Reject the null hypothesis).
Equation= 103.353-2.645P+0.02I-O.356S
2- FACTORABILITY
Correlation matrix > 0.3
KMO > 0.7
Bartlets test of sphericity-
• Ho- (correlation matrix not equal to identity matrix
• Ha- (correlation matrix = identity matrix .
• Check p value for significance (0.05)
In identity matrix variable 1 has relationship with variable 1 only and not with
any other variable.
In correlation matrix every variable has some relationship with other variable.
EFA STEPS
• Normality test
• If data is skewed remove outliers
• Factorability
• KMO
Scree plot-graphical way to show eigenvalues
• Bartlets test
Only then we will consider this for factor analysis else not • Communalities
• EIGEN Values
• Scree plot ( y- eigenvalues , X – factors)
Scree plot defines an optimal number of
factors retained in the final solution
Unidimensionality (quantitative)
Convergent Validity (quantitative)
In EFA criss-cross is there
In CFA there is no criss-cross among models
AVE>0.5 CR>AVE
MSR= AVE>MSR
ASR= AVE>ASR
Cronbach
alpha>0.7
Then model is
reliable else not
Path analysis with latent variables
Exogeneous
indigenous
Independent variabl
Dependent variable
What are the INDICES OF MODEL FIT?
INDICES are the values which tells whether the proposed model is acceptable or not.
It measures the discrepancies between observed and proposed model implied covariance matrices.
IT IS OF TWO TYPES
1-INCREMENTAL FIT(CFI,TLI,NFI,IFI)
2- ABSOLUTE FIT INDICES( GFI, AGFI, SRMR, RMSEA)