Part I

Question 1 :

Think of this as a "virtual" study. Based on the data, specify your imaginary research questions (i.e., what questions are you trying to address?) that will be examined via in multiple regression analysis. As an requirement for the class, you should include at least one independent categorical variable ( e . g ., gender ) , two continuous independent variable s , and the interaction between the categorical variable and each of the continuous independent variables. Of course, you can include more than one categorical independent variable in the model. Write your research question(s) and hypotheses.

Question 2 (20 points) :

Before you do regression analysis on your research questions, perform a preliminary analysis on the data to study and describe what data tells you based on the findings from distribution of variables, their means, and standard deviations.

Question 3 :

Based on your research question(s), fit a multiple regression model to your data and write the model from the analysis (5 points).

Question 4 :

Before you starting interpreting the results, please run residual analyses to:

1. examine assumptions about linearity, normality, and homogeneity of variance (10 points):
2. check whether there are any outliers that may influence your results (10 points):
3. check whether there is multicollinearity among variables in the model (5 ponits):

Question 5 :

1. for the fitted model, to what degree the model fit the data? (5 points)
2. interpret the regression coefficients from the analyses you performed, pay attention to the following (10 points):

--- What are the differences between a standardized and an unstandardized regression solution? Which one you used in your interpretation, and please justify?

Question 6 :

Do results support your hypotheses of the study? (10 points)

Part II

Question 1 :

It has been argued that how much students learn depend on the length of time the students spend on doing schoolwork per week. In the initial experiment, four groups of 10 students each were asked for four different lengths of study time---12, 18, 24, and 36 hours---at the standard school setting. Then, the test score of each of the 30 students was analyzed to determine the effect of study time on the test score, y .

1. Write a quadratic model relating the mean test score, E(y) , to the length of study (10 points).
2. Suppose the research and development department is interested in knowing whether different study times will yield different learning outcome ( y ) Please develop a dummy coding scheme that should be used in the regression model. (10 points)

Question 2 :

The principal of a school that has been in the office five years is scheduling her work load for next year, and she must estimate the number of staff available for work. She asks the school statistician to perform the following analyses based on the data in the table.

1. Fit the model to the data given in the table below (5 points).
2. Interpret the regression coefficients (6 points).
3. Please plot a histogram of absence rate by quarter. What trend do you notice in the data? (5 points)
4. Using quarter as a continuous variable (i.e., do not dummy code quarter variable), do a curve estimation. Is there sufficient evidence to conclude that mean absence rate has a cubic pattern across 5 years? (5 points)

Price: $28.16

Solution: The downloadable solution consists of 15 pages, 1316 words and 7 charts.
Deliverable: Word Document