您的位置：首页 >留学课业>Report课业辅导>Report 案例展示

2016-08-29 03:42:25 已有 635 人浏览 作者：admin 类别：Report 案例展示

**TABLE OF CONTENTS**

**Executive Summary……………………………………………………………………………………………****2**

1. **introduction****………………………………………………………………………………………………****2**

2. **Approach and research questions****…………………………………………………………………****..2**

3. **Research design and procedures****……………………………………………………………………****3**

4. **Sample description****………………………………………………………………………………………****3**

5. **Analysis of results****…………………………………………………………………………………………****7**

6. **Recommendations****…………………………………………………………………………………………****21**

**Reference list****………………………………………………………………………………………………………****22**** **

**Executive Summary**

Michael Jenkins’ dream is to one day own his own restaurant, which he wishes to call “Food For Fork”. However, due to the limited experience and financial resources, Mr. Jenkins is advised to run a thorough analysis, which can help him know more about the demand for an upscale restaurant, price level, characteristics, location and the way to promoting. This report will solve these problems with the data analysis using SPSS.

At last, we have some recommendations for him with the analysis of the result.

**Introduction**

Michael dreams of a fine, upscale restaurant featuring the finest entrées, drinks, and desserts in an elegant atmosphere.

He knows that, although he has learned quite a bit about restaurant operation in his area and also quite a bit about upscale restaurants, he is not sure if there is an interest in his city for such a restaurant. Even though the metro area is nearly 500,000 in population, he has no assurances that there are enough persons with the income and tastes necessary to make his business successful.

He needs some additional information whether a market exists for his service. There are also other decisions for which he feels he needs additional information. He is not certain how to promote the restaurant in his town. Where will he promote the restaurant when he first opens? Also, there are many choices to make about the design of the restaurant, the price the market is willing to pay for an upscale entrée, the best location, and so on.

**Approach and research questions**

We can know the need and interest of people by the overall research of the market. This objective can be broken into the following questions:

1. How many total dollars do you spend per month in restaurants (for your meals only)?

2. How likely would it be for you to patronize this restaurant (new upscale restaurant)?

3. What would you expect an average evening meal to be priced?

4. To which type of radio programming do you most often listen?

5. Which newscast do you watch most frequently?

6. Which section of the local newspaper would you say you read most frequently?

7. Prefer Waterfront View?

8. Prefer Drive Less than 30 Minutes?

9. Prefer Formal Waitstaff Wearing Tuxedos?

10. Prefer Unusual Desserts?

11. Prefer Large Variety of Entrees?

12. Prefer Unusual Entrees?

13. Prefer Simple Décor or Prefer Elegant Decor?

14. Prefer String Quartet or Prefer Jazz Combo?

15. Year Born?

16. What is your highest level of education?

17. What is your marital status?

18. Including children under 18 living with you, what is your family size?

19. Please check the letter that includes the Post Code in which you live (coded by letter).

20. What is your gender?

21. What is your annual salary?

**Research design and procedures**

Conclusive research seeks to both quantify the data and generate results that can be generalized to the population of interest. Traditional telephone interviews involve an interviewer using a paper questionnaire to record the respondent’s answers. A single cross-sectional design was chosen for the research wherein one sample of the respondents was drawn from the target population with information obtained only once.

**Sample description**

1. Figure 1 shows the relationship between marital status and the price of an evening meal.

2. Figure 2 shows the relationship between marital status and total money spend per month in restaurant.

3. Figure 3 shows different peoples’ choice about newscast and Décor.

Figure 4 shows peoples’ choice about postcode.

**FIGURE 1**

**FIGURE 2**

**FIGURE**

**FIGURE 4**

**Analysis of results**

H_{0}: µ $19;

H_{1}: µ $19

We use the one sample t-test: it tests the significance of the difference between the mean of sample and the designated value for the population mean.

Assumptions for one sample t-test:

- The data values are independently collected;

- The sample is randomly selected;

- The size of the sample is less than 10% of the population which is 50,000 in this case;

- The population is normally distributed;

- The significant level is 5%.

The results from SPSS are included in Table 1, we can get:

- The sample number is 400 and has Mean is $19.2300 with the standard deviation of $0.37797.

- t-stat=0.609; df = 399;

- P-value = 0.543

The sample got has the mean a little higher than the forecast value, while the dispersion of the data is comparatively narrow. However, the null hypothesis cannot be rejected at the significance level of 5%. The average amount that people are willing to pay for a meal may be equal or more than $19.

| ||||

N | Mean | Std. Deviation | Std. Error Mean | |

What would you expect an average evening meal to be priced? | 400 | $19.2300 | $7.55943 | $0.37797 |

| ||||||

Test Value = 19 | ||||||

t | df | Sig. (2-tailed) | Mean Difference | 95% Confidence Interval of the Difference | ||

Lower | Upper | |||||

What would you expect an average evening meal to be priced? | .609 | 399 | .543 | $0.23000 | -$0.5131 | $0.9731 |

H_{0}: µ= $70,000;

H_{1}: µ $70,000

Also using the one sample t – test, we can analyze whether the average income of people surveyed is significantly different from $70,000.

The assumptions are same as analysis 1.

The results from SPSS are included in Table 2:

- The sample number is 400 and has Mean is $77087.5000 with the standard deviation of $28896.92760.

- t-stat=4.905; df = 399;

- P-value = 0.000

The sample got has the mean higher than the forecast value, while the dispersion of the data is comparatively wide. With this result, it is concluded that the null hypothesis can be rejected. Therefore we can say that the average income of the population is significantly different from $70,000.

| ||||

N | Mean | Std. Deviation | Std. Error Mean | |

What is your annual salary? | 400 | 77087.5000 | 28896.92760 | 1444.84638 |

| ||||||

Test Value = 70000 | ||||||

t | df | Sig. (2-tailed) | Mean Difference | 95% Confidence Interval of the Difference | ||

Lower | Upper | |||||

What is your annual salary? | 4.905 | 399 | .000 | 7087.50000 | 4247.0371 | 9927.9629 |

H_{0}: µ _{simple }= µ _{elegant }

H_{1}: µ _{simple}≠µ _{elegant}

To test whether the difference for two dependent variables is not zero, we can use paired sample t-test to solve it.

The assumptions for paired sample t-test are (Paula, 1995):

- The differences between these two variables are all independent;

- The differences is a normal distribution;

- The significant level is 5%.

The results from SPSS are included in Table 3:

- The sample number N is 400;

- The mean of simple décor is 2.26 and elegant décor has the mean of 3.64;

- t-stat=-40.398; and

- P-value = 0.000 (0.05)

Therefore, the differences between the preferences of these two styles of décor are significant from zero, and we can find that people surveyed prefer elegant décor to simple décor.

| |||||

Mean | N | Std. Deviation | Std. Error Mean | ||

Pair 1 | Prefer Simple Decor | 2.26 | 400 | 1.212 | .061 |

Prefer Elegant Decor | 3.64 | 400 | 1.183 | .059 |

| ||||

N | Correlation | Sig. | ||

Pair 1 | Prefer Simple Decor & Prefer Elegant Decor | 400 | .838 | .000 |

| |||||||||

Paired Differences | t | df | Sig. (2-tailed) | ||||||

Mean | Std. Deviation | Std. Error Mean | 95% Confidence Interval of the Difference | ||||||

Lower | Upper | ||||||||

Pair 1 | Prefer Simple Decor - Prefer Elegant Decor | -1.380 | .683 | .034 | -1.447 | -1.313 | -40.398 | 399 | .000 |

H_{0: }there is no difference in the mean amount spent in restaurant each month across people with different marital status (µ_{single }_{=}µ_{married }_{=}µ_{rock }_{=}µ_{other(divorced,widow,etc>)});

H_{1}: at least one marital group has significantly different mean from the others.

One-way ANOVA test (analysis of variance) is a technique that the mean differences among three or more samples (Heiberger & Neuwirth, 2009).

The assumptions of one-way ANOVA are (Heiberger & Neuwirth, 2009):

- The population of differences is a normal distribution

- The data values are independently collected;

- All variables have the same variances;

- The significant level is 5%.

The results from SPSS are included in Table 4. F-stat is 99.939 (P-value is 0.000). We can conclude that the null hypothesis is rejected and there is significant differences existing between at least two of these three variables. According to the third table of Appendix 4, the Post hoc multiple comparisons test shows that all of the differences between any two variables are significant as the p-values are all 0.000 for all comparisons, while the descriptive table shows that married people spent the most on restaurant (the highest mean), compared with the other two groups.

| ||||||||

How many total dollars do you spend per month in restaurants (for your meals only)? | ||||||||

N | Mean | Std. Deviation | Std. Error | 95% Confidence Interval for Mean | Minimum | Maximum | ||

Lower Bound | Upper Bound | |||||||

Single | 146 | $180.8904 | $30.42280 | $2.51781 | $175.9141 | $185.8668 | $101.00 | $240.00 |

Married | 175 | $232.6114 | $29.87614 | $2.25842 | $228.1540 | $237.0689 | $174.00 | $301.00 |

Other (Divorced, Widow, etc.) | 79 | $203.2785 | $42.00792 | $4.72626 | $193.8692 | $212.6877 | $117.00 | $307.00 |

Total | 400 | $207.9400 | $40.11945 | $2.00597 | $203.9964 | $211.8836 | $101.00 | $307.00 |

| |||||

How many total dollars do you spend per month in restaurants (for your meals only)? | |||||

Sum of Squares | df | Mean Square | F | Sig. | |

Between Groups | 215060.863 | 2 | 107530.431 | 99.939 | .000 |

Within Groups | 427157.697 | 397 | 1075.964 | ||

Total | 642218.560 | 399 |

| ||||||

Dependent Variable: How many total dollars do you spend per month in restaurants (for your meals only)? Tukey HSD | ||||||

(I) What is your marital status? | (J) What is your marital status? | Mean Difference (I-J) | Std. Error | Sig. | 95% Confidence Interval | |

Lower Bound | Upper Bound | |||||

Single | Married | -$51.72102 | $3.67668 | .000 | -$60.3705 | -$43.0715 |

Other (Divorced, Widow, etc.) | -$22.38807 | $4.58142 | .000 | -$33.1660 | -$11.6101 | |

Married | Single | $51.72102 | $3.67668 | .000 | $43.0715 | $60.3705 |

Other (Divorced, Widow, etc.) | $29.33295 | $4.44614 | .000 | $18.8732 | $39.7927 | |

Other (Divorced, Widow, etc.) | Single | $22.38807 | $4.58142 | .000 | $11.6101 | $33.1660 |

Married | -$29.33295 | $4.44614 | .000 | -$39.7927 | -$18.8732 | |

*. The mean difference is significant at the 0.05 level. |

| ||||

Tukey HSD | ||||

What is your marital status? | N | Subset for alpha = 0.05 | ||

1 | 2 | 3 | ||

Single | 146 | $180.8904 | ||

Other (Divorced, Widow, etc.) | 79 | $203.2785 | ||

Married | 175 | $232.6114 | ||

Sig. | 1.000 | 1.000 | 1.000 | |

Means for groups in homogeneous subsets are displayed. | ||||

a. Uses Harmonic Mean Sample Size = 118.945. | ||||

b. The group sizes are unequal. The harmonic mean of the group sizes is used. Type I error levels are not guaranteed. |

H_{0}: µ _{string quartet} = µ _{Jazz combo}

H_{1}: µ _{string quartet} ≠ µ _{Jazz combo}

Using the same paired sample t - test as analysis 3, we can analyze whether there is a difference between string quartet and Jazz combo.

The assumptions are same as analysis 3.

The results from SPSS are included in Table 5:

- The sample number N is 400;

- The mean of string quartet is 3.39 and Jazz combo has the mean of 2.40;

- t-stat=33.469; and

- P-value = 0.000 (0.05)

We get the result that there are differences between the preferences of string quartet and Jazz combo, and we can find that people surveyed prefer string quartet to Jazz combo.

| |||||

Mean | N | Std. Deviation | Std. Error Mean | ||

Pair 1 | Prefer String Quartet | 3.39 | 400 | 1.231 | .062 |

Prefer Jazz Combo | 2.40 | 400 | 1.232 | .062 |

| ||||

N | Correlation | Sig. | ||

Pair 1 | Prefer String Quartet & Prefer Jazz Combo | 400 | .885 | .000 |

| |||||||||

Paired Differences | t | df | Sig. (2-tailed) | ||||||

Mean | Std. Deviation | Std. Error Mean | 95% Confidence Interval of the Difference | ||||||

Lower | Upper | ||||||||

Pair 1 | Prefer String Quartet - Prefer Jazz Combo | .987 | .590 | .030 | .929 | 1.046 | 33.469 | 399 | .000 |

H_{0}: There is no relationship between likelihood of attending and the section of the newspaper that is read;

H_{1}: There is a relationship between them.

To test the association between two and more variables, we can use the chi-square test by the ways of comparison of the expected and actual frequency of the goodness of fit (Satorra & Bentler, 2001).

The assumptions of chi-square test are (Satorra & Bentler, 2001):

- Data values are independently collected;

- All samples are randomly chosen;

- The counts are all more than 1;

- The significant level is 5%.

The results from SPSS are included in Table 6:

- chi-square value is 11.612; Phi value is 0.170 and Cramer’s V is 0.098 ;

- P-value = 0.477 (0.05);

With these results, we get that the null hypothesis cannot be rejected, which means that there is no sufficient evidence to prove that advertising on one section of newspaper can significantly improve the likelihood of visiting the restaurant than other sections.

| ||||||

Cases | ||||||

Valid | Missing | Total | ||||

N | Percent | N | Percent | N | Percent | |

How likely would it be for you to patronize this restaurant (new upscale restaurant)? * Which newscast do you watch most frequently? | 400 | 100.0% | 0 | 0.0% | 400 | 100.0% |

| |||||||

Which newscast do you watch most frequently? | Total | ||||||

7:00 am News | Noon News | 6:00 pm News | 10:00 pm News | ||||

How likely would it be for you to patronize this restaurant (new upscale restaurant)? | Very Unlikely | Count | 24 | 18 | 20 | 16 | 78 |

% within How likely would it be for you to patronize this restaurant (new upscale restaurant)? | 30.8% | 23.1% | 25.6% | 20.5% | 100.0% | ||

Somewhat Unlikely | Count | 10 | 28 | 19 | 19 | 76 | |

% within How likely would it be for you to patronize this restaurant (new upscale restaurant)? | 13.2% | 36.8% | 25.0% | 25.0% | 100.0% | ||

Neither Likely Nor Unlikely | Count | 19 | 18 | 21 | 22 | 80 | |

% within How likely would it be for you to patronize this restaurant (new upscale restaurant)? | 23.8% | 22.5% | 26.2% | 27.5% | 100.0% | ||

Somewhat Likely | Count | 19 | 22 | 18 | 21 | 80 | |

% within How likely would it be for you to patronize this restaurant (new upscale restaurant)? | 23.8% | 27.5% | 22.5% | 26.2% | 100.0% | ||

Very Likely | Count | 24 | 22 | 23 | 17 | 86 | |

% within How likely would it be for you to patronize this restaurant (new upscale restaurant)? | 27.9% | 25.6% | 26.7% | 19.8% | 100.0% | ||

Total | Count | 96 | 108 | 101 | 95 | 400 | |

% within How likely would it be for you to patronize this restaurant (new upscale restaurant)? | 24.0% | 27.0% | 25.2% | 23.8% | 100.0% |

| |||

Value | df | Asymp. Sig. (2-sided) | |

Pearson Chi-Square | 11.612 | 12 | .477 |

Likelihood Ratio | 12.006 | 12 | .445 |

Linear-by-Linear Association | .034 | 1 | .854 |

N of Valid Cases | 400 | ||

a. 0 cells (0.0%) have expected count less than 5. The minimum expected count is 18.05. |

| |||

Value | Approx. Sig. | ||

Nominal by Nominal | Phi | .170 | .477 |

Cramer's V | .098 | .477 | |

N of Valid Cases | 400 | ||

a. Not assuming the null hypothesis. | |||

b. Using the asymptotic standard error assuming the null hypothesis. |

**Analysis 7: Postcode**

H_{0}: there is no difference in the average price people are willing to pay for a meal across postcode (µ_{A}=µ_{B}=µ_{C}=µ_{D});

H_{1}: at least one postcode has significantly different mean from the others.

The same one-way ANOVA test applied in analysis 4 can be used in this analysis.

The same assumptions as analysis 4 apply to this analysis as well.

The results from SPSS are included in Table 7. F-stat is 3.099 (P-value is 0.027). We get that the null hypothesis is rejected and there is significant differences existing between at least two of these four variables. According to the third table of Table 7, the Post hoc multiple comparisons test suggests that there are only differences between B and C, and the descriptive table shows that people living in B (postcode 3, 4 & 5) are willing to spend the most on the meal, compared with the other three areas.

| ||||||||

What would you expect an average evening meal to be priced? | ||||||||

N | Mean | Std. Deviation | Std. Error | 95% Confidence Interval for Mean | Minimum | Maximum | ||

Lower Bound | Upper Bound | |||||||

A (1 & 2) | 93 | $19.4409 | $8.09114 | $0.83901 | $17.7745 | $21.1072 | $1.00 | $39.00 |

B (3, 4, & 5) | 109 | $20.7523 | $7.16881 | $0.68665 | $19.3912 | $22.1133 | $1.00 | $38.00 |

C (6, 7, 8, & 9) | 94 | $17.5745 | $8.39417 | $0.86579 | $15.8552 | $19.2938 | $1.00 | $35.00 |

D (10, 11, & 12) | 104 | $18.9423 | $6.34880 | $0.62255 | $17.7076 | $20.1770 | $1.00 | $36.00 |

Total | 400 | $19.2300 | $7.55943 | $0.37797 | $18.4869 | $19.9731 | $1.00 | $39.00 |

| |||||

What would you expect an average evening meal to be priced? | |||||

Sum of Squares | df | Mean Square | F | Sig. | |

Between Groups | 522.971 | 3 | 174.324 | 3.099 | .027 |

Within Groups | 22277.869 | 396 | 56.257 | ||

Total | 22800.840 | 399 |

| ||||||

Dependent Variable: What would you expect an average evening meal to be priced? Tukey HSD | ||||||

(I) Please check the letter that includes the Post Code in which you live (coded by letter). | (J) Please check the letter that includes the Post Code in which you live (coded by letter). | Mean Difference (I-J) | Std. Error | Sig. | 95% Confidence Interval | |

Lower Bound | Upper Bound | |||||

A (1 & 2) | B (3, 4, & 5) | -$1.31143 | $1.05879 | .603 | -$4.0431 | $1.4202 |

C (6, 7, 8, & 9) | $1.86639 | $1.09699 | .324 | -$0.9638 | $4.6966 | |

D (10, 11, & 12) | $0.49855 | $1.07044 | .966 | -$2.2632 | $3.2603 | |

B (3, 4, & 5) | A (1 & 2) | $1.31143 | $1.05879 | .603 | -$1.4202 | $4.0431 |

C (6, 7, 8, & 9) | $3.17783 | $1.05575 | .015 | $0.4540 | $5.9016 | |

D (10, 11, & 12) | $1.80999 | $1.02813 | .294 | -$0.8426 | $4.4625 | |

C (6, 7, 8, & 9) | A (1 & 2) | -$1.86639 | $1.09699 | .324 | -$4.6966 | $0.9638 |

B (3, 4, & 5) | -$3.17783 | $1.05575 | .015 | -$5.9016 | -$0.4540 | |

D (10, 11, & 12) | -$1.36784 | $1.06743 | .575 | -$4.1218 | $1.3861 | |

D (10, 11, & 12) | A (1 & 2) | -$0.49855 | $1.07044 | .966 | -$3.2603 | $2.2632 |

B (3, 4, & 5) | -$1.80999 | $1.02813 | .294 | -$4.4625 | $0.8426 | |

C (6, 7, 8, & 9) | $1.36784 | $1.06743 | .575 | -$1.3861 | $4.1218 | |

*. The mean difference is significant at the 0.05 level. |

| |||

Tukey HSD | |||

Please check the letter that includes the Post Code in which you live (coded by letter). | N | Subset for alpha = 0.05 | |

1 | 2 | ||

C (6, 7, 8, & 9) | 94 | $17.5745 | |

D (10, 11, & 12) | 104 | $18.9423 | $18.9423 |

A (1 & 2) | 93 | $19.4409 | $19.4409 |

B (3, 4, & 5) | 109 | $20.7523 | |

Sig. | .297 | .324 | |

Means for groups in homogeneous subsets are displayed. | |||

a. Uses Harmonic Mean Sample Size = 99.550. | |||

b. The group sizes are unequal. The harmonic mean of the group sizes is used. Type I error levels are not guaranteed. |

**Analysis 8: Gender**

H_{0}: µ _{male} = µ _{female}

H_{1}: µ _{male} ≠ µ _{female}

Also using one-way ANOVA test which is applied in analysis 4&7 can be used in this analysis.

The assumptions are same as analysis 4&7.

The results from SPSS are included in Table 8. F-stat is 8.820 (P-value is 0.003). We get that the null hypothesis is rejected and there are differences between male and female spending. According to descriptive table, female spent more than male in relations to their average expenditure on restaurant on monthly basis.

**Table**** 8**

| ||||||||

How many total dollars do you spend per month in restaurants (for your meals only)? | ||||||||

N | Mean | Std. Deviation | Std. Error | 95% Confidence Interval for Mean | Minimum | Maximum | ||

Lower Bound | Upper Bound | |||||||

Male | 204 | $202.1569 | $42.77268 | $2.99469 | $196.2522 | $208.0615 | $101.00 | $298.00 |

Female | 196 | $213.9592 | $36.29690 | $2.59264 | $208.8460 | $219.0724 | $111.00 | $307.00 |

Total | 400 | $207.9400 | $40.11945 | $2.00597 | $203.9964 | $211.8836 | $101.00 | $307.00 |

| |||||

How many total dollars do you spend per month in restaurants (for your meals only)? | |||||

Sum of Squares | df | Mean Square | F | Sig. | |

Between Groups | 13923.906 | 1 | 13923.906 | 8.820 | .003 |

Within Groups | 628294.654 | 398 | 1578.630 | ||

Total | 642218.560 | 399 |

H_{0}: There is no relationship between income and total expenditure;

H_{1}: There is a relationship between them.

Using the same chi-square test applied in analysis 6 in this analysis.

The assumptions are same as analysis 6.

The results from SPSS are included in Table 9:

- chi-square value is 276.584;

- Phi and Cramer’s V are both 0.613;

- P-value = 0.613 (0.05).

With these results, we can get that the null hypothesis cannot be rejected. In other words, it lacks the sufficient evidence to show that there is correlation between the income and total expenditure on the restaurant.

| |||

Value | df | Asymp. Sig. (2-sided) | |

Pearson Chi-Square | 276.584 | 284 | .613 |

Likelihood Ratio | 292.583 | 284 | .350 |

Linear-by-Linear Association | .125 | 1 | .723 |

N of Valid Cases | 400 | ||

a. 360 cells (100.0%) have expected count less than 5. The minimum expected count is .19. |

| |||

Value | Approx. Sig. | ||

Nominal by Nominal | Phi | .832 | .613 |

Cramer's V | .416 | .613 | |

N of Valid Cases | 400 | ||

a. Not assuming the null hypothesis. | |||

b. Using the asymptotic standard error assuming the null hypothesis. |

**Analysis 10: Multiple ****R****egression Analysis**

In order to analyze the relationship between one dependent variable and multiple independent variables, the most common statistical methodology used is to conduct a multiple regression analysis. In this case, the regression can be expressed as:

the average amount people spend on food each month = b_{1} + b_{2} x “the average price people are willing to pay for meals” + b_{3} x “age” + b_{4} x “marital status” + b_{5} x “gender” + b_{6} x “income”

The assumptions of the multiple regression analysis are:

- The data values are independently collected;

- The population of errors is a normal distribution;

- The significant level is 5%.

The results from SPSS are included in Table 10:

- Adjusted R square is 0.145 which means that there is only 14.5% of the dependent variable’s movement can be explained by these five variables;

- Since P-values of b_{2 }for “the average price people are willing to pay for meals” is 0.882 (>0.05), this variable should be deleted from this regression;

- Since the coefficient for “income” is 0.000, this variable should also be removed.

Therefore, the regression is:

The average amount people spend on food each month = 125.557 + 0.795 x “age” + 14.553 x “ marital status” + 9.059 x “gender”

In other words, it can be interpreted that when keeping other variables unchanged, when the age is increased by 1 year, the average amount people spend on food each month will increase by $0.795. Considering marital status and gender, since each category has number to represent it, the change of the category will change the average expenditure as well. For example, married people will spend $14.553 more on food than single people, while this category has lower expenditure than “other (divorce, widow, etc.)” by $14.553. Furthermore, female is more likely to spend $9.059 more on the restaurant than male in this area.

However, this multiple regression analysis suggest that regarding the marital status, “other (divorce, widow, etc)” spent the most on the food, which is contradict to the finding of analysis 4. This is the red flag indicating that analysts should exert extra caution on the quality of the data collected. However, since the adjusted R square is comparatively low for this multiple regression, it can be suggested that the variable “marital” may not be used for this multiple regression.

Hence, the regression should be rewritten as: the average amount people spend on food each month = 125.557 + 0.795 x “age” + 9.059 x “gender”

| ||||

Model | R | R Square | Adjusted R Square | Std. Error of the Estimate |

1 | .394 | .155 | .145 | $37.10752 |

a. Predictors: (Constant), What is your annual salary?, Age, What would you expect an average evening meal to be priced?, What is your gender?, What is your marital status? |

| |||||||||||||||

Model | Sum of Squares | df | Mean Square | F | Sig. | ||||||||||

1 | Regression | 99693.034 | 5 | 19938.607 | 14.480 | .000 | |||||||||

Residual | 542525.526 | 394 | 1376.968 | ||||||||||||

Total | 642218.560 | 399 | |||||||||||||

a. Dependent Variable: How many total dollars do you spend per month in restaurants (for your meals only)? | |||||||||||||||

b. Predictors: (Constant), What is your annual salary?, Age, What would you expect an average evening meal to be priced?, What is your gender?, What is your marital status? | |||||||||||||||

| |||||||||||||||

Model | Unstandardized Coefficients | Standardized Coefficients | t | Sig. | 95.0% Confidence Interval for B | ||||||||||

B | Std. Error | Beta | Lower Bound | Upper Bound | |||||||||||

1 | (Constant) | 125.557 | 11.847 | 10.598 | .000 | 102.265 | 148.848 | ||||||||

What would you expect an average evening meal to be priced? | .037 | .246 | .007 | .149 | .882 | -.447 | .520 | ||||||||

Age | .795 | .214 | .176 | 3.718 | .000 | .374 | 1.215 | ||||||||

What is your marital status? | 14.553 | 2.593 | .266 | 5.613 | .000 | 9.455 | 19.650 | ||||||||

What is your gender? | 9.059 | 3.753 | .113 | 2.414 | .016 | 1.680 | 16.437 | ||||||||

What is your annual salary? | .000 | .000 | .116 | 2.487 | .013 | .000 | .000 | ||||||||

a. Dependent Variable: How many total dollars do you spend per month in restaurants (for your meals only)? |

In conclusion, in addition to the last multiple regression analysis, we can summary all the findings of the first nine analyses as follows:

1. Mr. Michael Jenkins can expect that the average amount that people are willing to pay for a meal can be higher than $19 in that area;

2. The average annual income in that area is significantly above $70,000;

3. People are more likely to prefer elegant décor to simple décor in the restaurant;

4. The married people spent the most on restaurant, compared with the other two groups: single and other (divorced, widow, etc.);

5. People are more likely to prefer string quartet to Jazz combo;

6. There is no sufficient evidence to prove that if Mr. Jenkins advertises on newspaper, it does not matter that which section he should choose as there is no significant difference between any two of various sections on the newspaper;

7. The restaurant should open in Area B (postcode 3,4 &5);

8. Female spent more than male in relations to their average expenditure on restaurant on monthly basis;

9. There is correlation between people’s income and total expenditure on the restaurant. Therefore, the target customer should not be categorised by income.

**Recommendations**

Based on the summary above, the recommendations for Michael and the decisions regarding his restaurant are:

1. There is demand for an Upscale Restaurant; the price level can be higher than $19.

2. The characteristics of the restaurant should be elegant with string quartet. He should also pay more attention to the interest and like of the married people, the female.

3. Michael should open his restaurant in Area B (postcode 3,4 &5).

4. There is no need for Michael spending money on advertises on newspaper.

Paula Diehr. (1995). Breaking the matches in a paired t-test for community interventions when the number of pairs is small, *Stat Med*, 14(13): 1491-1504.

Heiberger, R.M. & Neuwirth, E. (2009). *R Through Excel*, Springer Science+Business Media, LLC, ISBN: 978-1-4419-0051-7.

Satorra, A. & Bentler, P.M. (2001). A scaled difference chi-square test statistic for moment structure analysis, *Psychometrika*, Vol. 66, No. 4, pp. 507-51

- 上一篇：Table of Contents
- 下一篇：没有了