MOTEL CASE

Survey Background

On a certain day of the week Motel Researchers and Consultants, Inc. (MRCI) conducted a survey that was given to travelers passing through the security checkpoints at a large multi-terminal airport. Ten researchers were assigned to each checkpoint at each domestic terminal. They were interested in only domestic travelers that were ending their travels.

Every 50 ^th traveler was asked by one of the researchers to complete the motel questionnaire. As soon as the selected traveler was chosen, the next researcher started counting for the next 50 ^th traveler. If the next 50 ^th traveler refused to answer the questionnaire or if the traveler was not ending their travels, the researcher simply waited for their next turn in the rotation.

The survey instrument questioned a number of aspects about the traveler’s motel stay as detailed in the attached Excel work book. Respondents answered a series of questions pertaining to their experience and aspects of the quality of their motel, overall quality, intent to return, and demographic information. Among other things, the first Excel work sheet describes three types of variables as follows:

• Reception variables (recept 1 – 4)
• Accommodation variables (accom1 – 7)
• Departure variables (depart1 – 4)
  [Note: the data file has 2 different intent variables – intent1 measured on a 0 –9 semantic differential scale and intent2 measured on recorded classification of intent with 3 categories.]
  The second Excel work sheet shows all of the data that has been collected thus far, which are 198 surveys.
  A motel association has hired Motel Researchers and Consultants, Inc. (MRCI) to conduct the survey, and evaluate the data regarding the most important determinants of overall motel quality.
  Q1
  What are the most important determin ants of overall motel quality ? Use regres sion analysis to determine what are the most important variables influencing a traveler’s rating of overall quality of the motel (oqual) based on the independent variables:
• Reception variables (recpt 1 – 4 )
• Accommodation variables (accom 1 – 7 )
• Departure variables (depart 1 – 4 ) .
  Instructions:
  (Q1a) Assuming that the data entry process is still continuing, "clean -up" the data as required such as eliminating data with missing data points or errors? Note that a hold-out or cross-validation subset is not to be created; starting out with 198 usable surveys in the data file, describe any procedures and "cures" prior to data analysis regarding the variables of interest.
  (Q1c) T est the assumptions of the respective technique and report the findings; also, explain (1) how the assumptions were tested for the technique, (2) any actions taken or not taken (including tests for outliers and any actions taken), and (3) the resulting impact of the findings;
  (Q1d) Do a regression analysis including a summary section, ANOVA table, and regression coefficients table; Comment on the basic regression findings with respect to the full model including Rsquared, adjusted Rsquared, and F-Stat from the ANOVA; report should include a description of procedures used and cures (if any) used prior to data analysis of variables of interest
  (Q1e) Do regression on a reduced model/final model including the Run Summary (Rsquared, adjusted Rsquared, etc.), Regression Coefficients and ANOVA table; the goal is to be parsimonious while maximizing Rsquared; the default setting under multiple regression is (none) or forward
  (Q1f) Discuss in detail the reduced regression model /final model including what criteria was set for removing variables from full model; report should include a description of procedures used and cures (if any) that were used prior to data analysis of variables of interest (include any plots that help to explain);
  (Q1g) In plain English explain findings to the motel association so that motel managers could benefit (reference any plots, etc. that may help to explain);
  (Q1h) Create a " new " hotel data file to include (1) original data, (2) any new variables created or used, and (3) any stored output such as residuals from the final regression runs or items used for outlier detection, etc.; change variables names from "C##" to something descriptive or provide an appropriate legend; delete any columns of "saved variables" which are not relevant such as failed transformations, interim regression runs, etc.
  Q2
  What are the most important determin ants of a traveler ’s intention to return to the motel ? Use regres sion analysis to estimate a traveler’s intention to return to the motel based on the independent variables as follows :
• Reception variables (recpt 1 – 4 )
• Accommodation variables (accom 1 – 7 )
• Departure variables (depart 1 – 4 ) .

[If an influential observation (outlier) is thrown out in Q1, it does not mean it is thrown out for Q2 (It’s based on the residuals of the dependent variable). Also, as outliers are removed more observations are flagged as potential outliers. So it is important to justify eliminating outliers.]

Instructions (These instructions are the same as Q1):

(Q2a) Assuming that the data entry process is still continuing, "clean-up" the data as required such as eliminating data with missing data points or errors? Note that a hold-out or cross-validation subset is not to be created, and we are starting out with 198 usable surveys in the data file; describe any procedures and "cures" prior to data analysis regarding the variables of interest;

(Q2b) Perform descriptive statistics and get to know the data;

(Q2c) Test the assumptions of the respective technique and report the findings; also, explain (1) how the assumptions were tested for the technique, (2) any actions taken or not taken (including tests for outliers and any actions taken), and (3) the resulting impact of the findings;

(Q2d) Do a regression analysis including a summary section, ANOVA table, and regression coefficients table; Comment on the basic regression findings with respect to the full model including Rsquared, adjusted Rsquared, and F-Stat from the ANOVA; report should include a description of procedures used and cures (if any) used prior to data analysis of variables of interest;

(Q2e) Do regression on a reduced model/final model including the Run Summary (Rsquared, adjusted Rsquared, etc.), Regression Coefficients and ANOVA table; the goal is to be parsimonious while maximizing Rsquared; the default setting under multiple regression is (none) or forward

(Q2f) Discuss in detail the reduced regression model/final model including what criteria was set for removing variables from full model; report should include a description of procedures used and cures (if any) that were used prior to data analysis of variables of interest (include any plots that help to explain);

(Q2g) In plain English explain findings to the motel association so that motel managers could benefit (reference any plots, etc. that may help to explain);

(Q2h) Create a " new " hotel data file to include (1) original data, (2) any new variables created or used, and (3) any stored output such as residuals from the final regression runs or items used for outlier detection, etc.; change variables names from "C##" to something descriptive or provide an appropriate legend; delete any columns of "saved variables" which are not relevant such as failed transformations, interim regression runs, etc.

Q3

Is there any difference with respect to overall quality (oqual ) between the different types of motel (type) , gender, and purpose of the trip (as well as any interaction) ? Is it possible to test whether the mean ratings are equal among the 4 classifications? Make a decision on how best to analyze the data.

Instructions:

(Q3a) Test the assumptions of the respective technique and report the findings; also, explain (1) why the technique was chosen, (2) how the assumptions were tested for the technique, (3) any actions taken or not taken (including tests for outliers and any actions taken), and (4) the resulting impact of the findings;

(Q3b) In plain English explain and summarize findings to the motel association so that motel managers could benefit (reference any plots, etc. that may help to explain);

Price: $49.99

Solution: The downloadable solution consists of 37 pages, 2837 words and 41 charts.
Deliverable: Word Document