6
Quick Instructions for Top Assignment Expert:
1. Please answer ALL questions below in American English spelling. (No format, just answer questions)
2. In the Analysis Problem section, please use the .csv link provided to download the data for analysis. Use any method you want (excel, R, etc.) to answer the questions.
3. The solution can be submitted by using this same document since you are just answering questions.
Short-Answer Problems
Some of these questions will appear on the short-answer part of the tests. As part of this homework, answer the following questions, usually just several sentences that include the definition.
Text, Ch 12
1. Identify the steps in the data preparation process.
2. Discuss the processes of validating and coding data.
3. Describe the processes of data entry and data cleaning.
4. Discuss the advantages of pre-coding questionnaires.
5. Compare a one-way tabulation vs. a two-way tabulation.
6. Explain the descriptive statistics that should be used with each type of scale.
Data Analysis
1. What does the business analyst hope to accomplish with a regression analysis? That is, what are the two goals of regression analysis?
1. What is ? What are the two primary situations in which it is applied?
1. How does the graph of X compare with Y vs the graph of X with ?
1. What is the meaning of the slope coefficient in = b0 + b1X1
1. In regression analysis computer output, there are one or more t-tests presented. Describe the purpose of the test(s), including the null and alternative hypotheses. What do you conclude if the associated p-value for a test is less than?
1. In computer output for regression analysis there are one or more confidence intervals presented. Describe the purpose of the confidence interval, and its interpretation.
1. What is the meaning of the residual variable e ?
7. What is the criterion of ordinary least squares regression to obtain the estimated model?
8. Model Fit: The standard deviation of the residuals to interpret model fit
9. Model Fit: R-squared as a relative index of fit
Analysis Problem
Note: This problem is similar to what you do for your project, use regression analysis to analyze the contribution of, here just one, product attribute to satisfaction. All responses obtained with a survey. Next week we do multiple regression where we have many product attributes in the equation, which is exactly the situation for your project, as well as the worked problem on the Final.
Consider a marketing survey of 253 customers of a restaurant called SFG.
Data: http:
web.pdx.edu/~ge
ing/data/SFGsfg.csv
Responses to the individual items are in a 7-pt Likert format, from 1 to 7. Assess the outcome variable of Satisfaction (x22) with the following item:
How satisfied are you with the SFG?
Not Satisfied Very
At All Satisfied
1 2 3 4 5 6 7
What are the customer experiences that lead to customer satisfaction? One potential such variable is the customer’s perceived tastiness of the food. Consider the product attribute of taste, Tastiness (x18).
The food at the SFG tastes excellent.
Strongly Strongly
Disagree Agree
1 2 3 4 5 6 7
Research Question: To what extent does perceived Tastiness contribute to Overall Satisfaction at Restaurant SFG?
Do questions a through q and s, t, u, and x from the template.
Use the following information for Questions c and d.
Scatterplot/Co
elation Matrix
a. Identify the response variable and the predictor variable(s).
. Show the scatterplot matrix (just one scatterplot for a single predictor) and co
elation coefficients of the relationship of each of the variables in the model with each other. From only this visual information, develop some intuition for the subsequent analysis.
i. Relevance: Do the predictor variables relate to the target (response) variable? Explain.
ii. Uniqueness: [If multiple predictor variables] Could collinearity be a problem? Explain.
iii. Model Selection: [If multiple predictor variables] Given the co
elations, what is the most likely candidate for the final model? Explain.
Estimated Model
c. Write the estimated regression model.
d. Specify and interpret the sample slope coefficient.
e. Manually calculate the fitted/predicted value for the given values of predictor variables X.
f. Manually calculate the associated residual. Interpret for the given values of predictor variables X and response variable y.
Hypothesis Test: Applied to the one specified predictor variable
g. Specify the null hypothesis and its alternative for the hypothesis test of the slope coefficient.
[answer with respect to the specifics of this analysis, e.g., not Predictor 1 but the actual name of each predictor in this specific analysis]
h. Show and label the calculation of how many (estimated) standard e
ors the estimated slope coefficient, b, is from the hypothesized population value.
[define the concept with the relevant numbers of this specific analysis, with or without a formula]
i. Include and apply the definition of the p-value with the relevant numbers for this specific analysis.
[include the relevant numbers in this specific analysis as an application of the general definition]
j. Specify the basis for the statistical decision for the hypothesis test and the resulting statistical conclusion for alpha=0.05.
[be specific with the numbers from this analysis as to the evaluation of the null hypothesis]
k. Hypothesis Test: Interpretation, as an executive summary you would report to management.
[applied to the relevant numbers of this specific analysis to generalize the results to the population, with no jargon like p-value or t-value or null hypothesis]
Confidence Interval: Applied to the one specified one predictor variable
l. Specify the value that the confidence interval estimates.
[do not provide the confidence interval, which is the estimate, not the value that it estimates]
m. Apply the definition of the 95% margin of e
or for its computation using the relevant numbers of this analysis with 2 approximating the t-cutoff.
[show the definition in words of the concept by applying the relevant numbers of this specific analysis, with or without a formula]
n. Show the computations of the 95% confidence interval illustrated with the specific numbers from this analysis.
[show the definition of the concept but apply the relevant numbers of this specific analysis, formula optional]
o. Confidence Interval: Interpretation, as an executive summary you would report to management.
[no jargon, which includes the phrase “slope coefficient”, nothing about hypothesis tests]
p. Demonstrate the consistency of the confidence interval and hypothesis test using the specific numbers for this analysis for both results.
[comparison includes the specifics of the numbers for this specific analysis for both inferential results]
Model Fit
q. Evaluate fit with the standard deviation of residuals.
. Evaluate fit with R2 and PRESS R2, including their comparison. Does this value indicate reasonable fit?
s. Show any potential outliers and explain why they are outliers.
Model Selection [if multiple predictor variables]
t. Consider all the predictor variables simultaneously. Based on the p-values of the slope coefficients, are any predictor variables much less useful for predicting the response variable (target)? Why or why not?
u. Any collinearity problems? Why or why not?
v. Based on this information and the best subset analysis, which model do you recommend? Why?
Prediction Intervals
w. For the 95% prediction interval of [response variable y] for [the values of predictor variables X], show the interval including its calculation (can approximate with the t-cutoff of 2).
x. Interpret the prediction interval.
Conclusion
y. What decision do you recommend to management based on these results?
c. value of the predictor variable: 4
d. value of the response variable: 5