Module 4 - SLP MANOVA/MANCOVA The next stages of the SLP will involve checking assumptions and performing multiple regression for predictors of email. This will involve using the same dataset,GSS.sav,...

1 answer below »

Module 4 - SLP


MANOVA/MANCOVA


The next stages of the SLP will involve checking assumptions and performing multiple regression for predictors of email. This will involve using the same dataset,GSS.sav, which contains 6 variables:


ID, age (years), email (using email measured in weekly hours), children (number of children), sex (0=female, 1=male), and income (0= LT $1000 – 24999, 1= $25000 or more).


*Please refer toSPSS Commandsfor tips on analyzing data.



Modules 3-5



Stage 4. Assumption checking


Check the assumptions of normality, linearity, homoscedasticity, and collinearity. Describe and provide support regarding whether the assumptions were met (Include supporting tables/graphs).



Stage 5.
Multiple regression



  1. Perform multiple regression to identify whether there is an association between the independent variables (age, children, sex, and income) and email, using forced entry and stepwise methods of regression. Describe, compare, and interpret your results (include supporting tables).

  2. Perform regression using only sex as an independent variable. Compare your results to the above regression results. Describe and interpret your findings.



Please submit stage 4-5 of SLP at the end of Module 5.


SLP Assignment Expectations


Length: SLP assignments should be at least 2 pages (500 words) in length excluding tables.


References: Any references used should be from academic sources and cited using APA format.

Answered Same DayJun 04, 2021

Answer To: Module 4 - SLP MANOVA/MANCOVA The next stages of the SLP will involve checking assumptions and...

Sourav answered on Jun 08 2021
148 Votes
Regression Modeling :
Multivariate Regression Analysis: Multiple regression analysis is a statistical method used to predict the value of a dependent variable based on the values of two or more independent variables.
Model Equation:             Y = βo + β1*X1+ β2*X2 + £
Where, Y =
dependent, Xi’s = independent, βi’s = Coefficients and £ = Residual or Error Term
Assumption of Multiple Regressions:
· Linear Relationship
· Normality of Residuals
· Homoscedasticity: Means same variance of residuals.
· Multicollinearity: Means two or more independent variables related to each other.
· Autocorrelation: Residuals are correlated.
Normality and outliers checking for the Dependent variable:
    Tests of Normality
    
    Kolmogorov-Smirnova
    Shapiro-Wilk
    
    Statistic
    df
    Sig.
    Statistic
    df
    Sig.
    Email hours per week
    .274
    87
    .000
    .659
    87
    .000
    a. Lilliefors Significance Correction
    
    
    
    
Conclusion: From Above test we can say that dependent variable is not normally distributed as p-value for test is less than 0.05 and it can also observe from the histogram chart as the data are left side and plot is rightly skewed. Also from Boxplot we can see there are some of outliers.
Collinearity or linear relationship:
    Correlations
    
    
    Email hours per week
    Number of children
    Respondents sex
    Total family income
    Age of respondent
    Pearson Correlation
    Email hours per week
    1.000
    -.247
    -.162
    .172
    -.106
    
    Number of children
    -.247
    1.000
    -.202
    -.246
    .255
    
    Respondents sex
    -.162
    -.202
    1.000
    -.011
    .031
    
    Total family income
    .172
    -.246
    -.011
    1.000
    .142
    
    Age of respondent
    -.106
    .255
    .031
    .142
    1.000
    Sig. (1-tailed)
    Email hours per week
    .
    .011
    .067
    .055
    .164
    
    Number of children
    .011
    .
    .030
    .011
    .009
    
    Respondents sex
    .067
    .030
    .
    .459
    .386
    
    Total family income
    .055
    .011
    .459
    .
    .094
    
    Age of respondent
    .164
    .009
    .386
    .094
    .
    N
    Email hours per week
    87
    87
    87
    87
    87
    
    Number of children
    87
    87
    87
    87
    87
    
    Respondents sex
    87
    87
    87
    87
    87
    
    Total family income
    87
    87
    87
    87
    87
    
    Age of respondent
    87
    87
    87
    87
    87
Conclusion: From above table we can see that there are no such pair of variables which having correlation between them. So its means that all the independent variables are not correlated to each other. Also if we take the threshold to check the correlation for independent and dependent as 0.3, then we can see that there are no such variable having values above 0.3 and below -0.3.
Task 1:
1. Over All Model:
    Coefficientsa
    Model
    Unstandardized Coefficients
    Standardized Coefficients
    t
    Sig.
    95% Confidence...
SOLUTION.PDF

Answer To This Question Is Available To Download

Related Questions & Answers

More Questions »

Submit New Assignment

Copy and Paste Your Assignment Here