Example Data A Survey of 50 Companies
In January ‗08, fifty customers of a lumber manufacturer were surveyed regarding their satisfaction with products and service. These customers buy from the supplier and sell to retail chains like Home Depot and Lowes. Shortly after, the manufacturing company was sold. In June ‗08, the customers were telephoned and interviewed interviewed and were asked to rate overall satisfaction again.
Variable id delivery
Position Label 1 ID 2
Delivery Reliability
3
Product Satisfaction
4
Technical Support
Prodsat
Techsat
Salesat 5 Salesforce Size Usage
6 Firm Size 7 Usage Level
Satjan 8 Satjun 8 Structure OwnType PurType
10 11 12
Overall Satisfaction in January Overall Satisfaction in June Structure of Procurement Type of Ownership Type of Purchasing
Variables in the working file
Measurement Level Scale Participant ID number On a scale of 1 to 10, how would you Scale rate the reliability of delivery of your orders? On a scale of 1 to 10, how would you rate your satisfaction with the quality Scale of your most recently purchased products? On a scale of 1 to 10, how would you Scale rate your satisfaction with the technical support? On a scale of 1 to 10, how would you Scale rate your satisfaction with the sales support? 0 = small (less than 100 emp.) 1 = Ordinal large (100 or more) What percent of your purchases are Scale from our company? On a scale of 1 to 7, rate your overall Scale satisfaction with your most recent purchasing experience. On a scale of 1 to 7, rate your overall Scale satisfaction with your most recent purchasing experience. How your purchasing is structured? Nominal 0 = Decentralized; 1 = Centralized 0 = Publicly Traded; 1 Privately Nominal owned 1 = Private Label; 2 = Company C ompany Nominal Brand; 3 = Both
, describe in your Microsoft Word document the application F or each r esearch questi on of the seven steps of the hypothesis testing model. Step 1: State the hypothesis (null and alternate). Step 2: State your alpha (unless requested otherwise, this is always set to alpha = .05). Step 3: Collect the data (use one of the data sets). Step 4: Calculate your statistic and p-value. (This is where you run spss and examine your output files.) Step 5: Accept or reject the null hypothesis. (This is where you report the results of your analyses t (df) = t-value, p = sig. level.) Step 6: Assess the Risk of Type I and Type II Error. (Did the data meet the assumptions of the statistic, effect size, and sample size?) Step 7: State your results in APA style and format.
Example 7 Question 1: Is there a relationship among the variables measuring different aspects of customer satisfaction? 1. Run a Pearson correlation matrix using delivery reliability, product satisfaction, technical support, sales satisfaction, overall satisfaction in January and overall satisfaction in June. Correlations
Delivery
Product
Reliability Satisfaction Delivery Reliability
Pearson Correlation
1
Satisfaction
Support
force
January
in June
**
.206
.628
.180
.002
.439
.152
.000
50
50
50
50
50
50
Pearson Correlation
.193
1
.317
.726
Sig. (2-tailed)
.180
.025
50
*
50 **
*
.000
.174
.000
50
50
50
50
1
.133
.340
.555
.356
.016
.000
50
50
50
50
1
.173
.326
.229
.021
50
50
50
1
.317
Sig. (2-tailed)
.002
.025
50
50
Pearson Correlation
.112
.726
.133
Sig. (2-tailed)
.439
.000
.356
50
50
50
N Overall Satisfaction in
Pearson Correlation
.206
.195
.340
*
.173
January
Sig. (2-tailed)
.152
.174
.016
.229
50
50
50
50
N
**
**
**
*
.479
50 *
50 **
.628
.484
.555
.326
.479
Sig. (2-tailed)
.000
.000
.000
.021
.000
50
50
50
50
50
**. Correlation is significant at the 0.01 level (2-tailed). *. Correlation is significant at the 0 .05 level (2-tailed).
**
*
**
.000
Pearson Correlation
N
**
.484
.436
**
**
.195
Pearson Correlation
N
Overall Satisfaction in June
**
.112
N
Salesforce
Satisfaction in
Technical Sales
.436
N
Technical Suppor
Overall
.193
Sig. (2-tailed)
Product Satisfaction
Overall
1
50
2. Create a scatter plot for the following pairs: (1) delivery reliability — overall satisfaction in June; (2) product satisfaction — overall satisfaction in June; and delivery reliability — product satisfaction.
The scatter diagram suggest that there is a weak positive correlation between the two variables.
The scatter diagram suggest that there is a weak positive correlation between the two variables.
The scatter diagram suggest that there is a weak positive correlation between the two variables.
3. Report the descriptive statistics, assumptions tests, as well as tests of statistical significance identify of positive and negative relationships. Descriptive Statistics Mean
Std. Deviation
N
Delivery Reliability
4.34
1.673
50
Product Satisfaction
5.34
1.154
50
Technical Support
2.88
.895
50
Sales force
2.66
.848
50
Overall Satisfaction in January
3.58
1.372
50
Overall Satisfaction in June
4.70
.953
50
Student‘s t test is adopted to check whether there is any significant positive correlation between the variables. H0: Correlation coefficient =0 H1: Correlation coefficient >0 (One sided hypothesis) Test Statistic used is t test
r t
n2
1
2
r
Significance level =0.05 Decision rule : Reject the null hypothesis if the p value is less than the significance level. The Correlation coefficient with p value of the one tailed test is given below. Correlations
Delivery
Product
Reliability Satisfaction Delivery Reliability
Pearson Correlation
1
Sig. (1-tailed) Product Satisfaction Pearson Correlation
Technical Support
Sales force
.193
Technical
Sales
Satisfaction in
Satisfaction
Support
force
January
in June
**
.436
.112
.206
0.090
.001
0.219
0.075
*
1
**
.317 .726
0.090
Pearson Correlation
.436
.317
Sig. (1-tailed)
.001
0.0125
Pearson Correlation
.112
.726
Sig. (1-tailed)
*
**
1
.133
.340
0.178
.016
.000
1
.173
.326
0.1145
0.015
.133
.206
.195
.340
.173
0.087
.016
0.114 5
in January
Sig. (1-tailed)
0.075
Overall Satisfaction
Pearson Correlation
.628
.484
in June
Sig. (1-tailed)
.000
.000
**
**
.484
0.087
0.178
Pearson Correlation
.000
.000
.000
Overall Satisfaction
.195
**
.628
0.0125
0.219
**
Overall
.193
Sig. (1-tailed)
**
Overall
*
**
*
.555 .326 .000
.0.015
*
.000 **
.555
*
**
1
.479
.000 **
.479
.000
**. Correlation is significant at the 0.01 level (2-tailed). *. Correlation is significant at the 0 .05 level (2-tailed).
Conclusion The t test for the significant correlation indicates that the correlation between Product satisfaction- Delivery reliability, Sales force- Delivery reliability, Overall satisfactionDelivery reliability, Over all satisfaction – Product satisfaction ,Sales force – Technical support, Overall satisfaction in January – Technical support are insignificant.
1
Question 2: Does delivery reliability impact overall satisfaction in June? 1. Run a simple regression using delivery reliability as the independent v ariable and overall satisfaction in June as the dependent variable. a
Coefficients Unstandardized
Standardized
Coefficients
Coefficients
Std. Model 1
B
Error
(Constant)
3.147
.297
Delivery Reliability
.358
.064
Beta
.628
t
Sig.
10.595
.000
5.596
.000
a. Dependent Variable: Overall Satisfaction in June
The estimated regression model is Overall Satisfaction in June = 3.147 +0.358 * Delivery Reliability
b
Model Summary
Model 1
R a
.628
Adjusted R
Std. Error of the
R Square
Square
Estimate
.395
.382
.749
a. Predictors: (Constant), Delivery Reliability b. Dependent Variable: Overall Satisfaction in June
2
The model adequacy measure R suggests that 39.5% variability in Overall Satisfaction in June can be explained by the regression model.
2. Report the descriptive statistics, assumptions tests (scatter plots), as well as tests of statistical significance. Descriptive Statistics Mean
Std. Deviation
N
Overall Satisfaction in June
4.70
.953
50
Delivery Reliability
4.34
1.673
50
Correlations
Pearson Correlation Sig. (1-tailed)
Overall Satisfaction in June
Delivery Reliability
Overall Satisfaction in June
1.000
.628
Delivery Reliability
.628
1.000
.
.000
.000
.
Overall Satisfaction in June
50
50
Delivery Reliability
50
50
Overall Satisfaction in June Delivery Reliability
N
The Correlation coefficient between Overall Satisfaction in June and delivery reliability is positive with 0.628. The regression coefficient of Delivery Reliability on Overall Satisfaction in June can be interpreted as
For a unit increase in Delivery Reliability, the Overall Satisfaction in June increase by 0.358 units’
The significance of this regression coefficient is tested using the t test H0: Regression coefficient =0 H1: Regression coefficient > 0 Significance level =0.05 Decision rule: Reject the null hypothesis if the p value is less than the significance level. Details T statistic =5.596 P value =0.000 Conclusion: Reject the null hypothesis. The sample provides enough evidence to support the claim that Delivery Reliability has a significant effect on Overall Satisfaction in June. The assumption for the validity of regression analysis is checked using the residual analysis. The histogram and normal probability plots suggest that the residuals are normally distributed. The homogeneity of variance assumption is valid as the plots of residuals against the predicted values are random.
Question 3: Does delivery reliability and product satisfaction impact overall satisfaction in June? 1. Run a multiple regression using delivery reliability as the independent variable and overall satisfaction in June as the dependent variable. a
Coefficients
Standardized Unstandardized Coefficients Model 1
B
Std. Error
(Constant)
1.662
.479
Delivery Reliability
.316
.058
Product Satisfaction
.312
.084
Coefficients Beta
t
Sig.
3.467
.001
.556
5.464
.000
.377
3.712
.001
a. Dependent Variable: Overall Satisfaction in June
The estimated regression model is Overall Satisfaction in June =1.662+ 0.316 * Delivery Reliability+0.312* Product Satisfaction b
Model Summary
Model 1
R a
.729
Adjusted R
Std. Error of the
R Square
Square
Estimate
.532
.512
.666
a. Predictors: (Constant), Product Satisfaction, Delivery Reliability b. Dependent Variable: Overall Satisfaction in June 2
The model adequacy measure R suggests that 53.2 % variability in Overall Satisfaction in June can be explained by the regression model.
2. Report the descriptive statistics, assumptions tests (scatter plots), as well as tests of statistical significance. Descriptive Statistics Mean
Std. Deviation
N
Overall Satisfaction in June
4.70
.953
50
Delivery Reliability
4.34
1.673
50
Product Satisfaction
5.34
1.154
50
Correlations
Pearson Correlation
Sig. (1-tailed)
N
Overall Satisfaction in June
Delivery Reliability
Product Satisfaction
Overall Satisfaction in June
1.000
.628
.484
Delivery Reliability
.628
1.000
.193
Product Satisfaction
.484
.193
1.000
.
.000
.000
Delivery Reliability
.000
.
.090
Product Satisfaction
.000
.090
.
Overall Satisfaction in June
50
50
50
Delivery Reliability
50
50
50
Product Satisfaction
50
50
50
Overall Satisfaction in June
The Correlation coefficient between Overall Satisfaction in June and delivery reliability is positive with 0.628 and Overall Satisfaction in June and Product satisfaction is 0.484 . The regression coefficient of Delivery Reliability on Overall Satisfaction in June can be interpreted as
For a unit increase in Delivery Reliability, the Overall Satisfaction in June increase by 0.358 units,.
For a unit increase in product satisfaction, the Overall Satisfaction in June increase by 0.312 units,.
The significance of this regression coefficient is tested using the t test H0: Regression coefficient =0 H1: Regression coefficient > 0 Significance level =0.05 Decision rule: Reject the null hypothesis if the p value is less than the significance level. Details Delivery Reliability Product Satisfaction T statistic 5.464 3.712 P value 0.000 0.001
Conclusion: Reject the null hypothesis. The sample provides enough evidence to support the claim that Delivery Reliability and product satisfaction has a significant effect on Overall Satisfaction in June.
The assumption for the validity of regression analysis is checked using the residual analysis. The histogram and normal probability plots suggest that the residuals are normally distributed. The homogeneity of variance assumption is valid as the plots of residuals against the predicted values are random.
Write a brief conclusion statement summarizing your results. What can you tell this manufacturing company about the relationship among satisfaction variables? Are there any areas they need to improve? Does adding a second variable to the regression equation increase prediction of customer satisfaction?
The regression analysis indicates that both Delivery Reliability and Product Satisfaction Satisfaction variables have a significant effect on overall satisfaction in June. The multiple regression models is able to explain 52.3% variability in overall satisfaction in June. We may add more explanatory variables to improve the model adequacy to a higher level. It can be noted that the model adequacy increase from 39.5% to 52.3% due to the addition of Product Satisfaction as the second explanatory variable .
Example 8 Question 1: Before the change of ownership, the company was encouraging its customers to reduce private labeling as a way to reduce cost of goods sold. Explore the distribution of customers by purchase type. Does the distribution of customers (private label, brand label, or both) differ from what one would expect by chance? Does if differ they expect more brand labeling?
Type of Purchasing
Observed N
Expected N
Residual
Private label
18
16.7
1.3
Company Brand
18
16.7
1.3
Both
14
16.7
-2.7
Total
50
H0: There is no significant difference in the number of customers in the three categories. H1: There is significant difference in the number of customers in the three categories. Test Statistic used is Chi square test for goodness of fit. Significance level =0.05 Decision rule: Reject the null hypothesis if the p value is less than the significance level. Details
Test Statistics
Type of Purchasing Chi-Square df Asymp. Sig.
a
.640 2
.726
a. 0 cells (.0%) have expected frequencies less than 5. The minimum expected cell frequency is 16.7.
Conclusion: Fails to reject the null hypothesis. The sample does not provides enough evidence to support the claim that there is significant difference in the number of customers in the three categories.
Question 2: Run a chi square goodness of fit using purchase type as the variable with “all categories equal” for the expected value. 1. Run a chi square goodness of fit using purchase type as the variable with ―all categories unequal‖ with 12, 26, and 12 as the expected values.
Type of Purchasing Observed N
Expected N
Residual
Private label
18
12.0
6.0
Company Brand
18
26.0
-8.0
Both
14
12.0
2.0
Total
50
2. Report the observed and expected values and the tests of statistical significance.
H0: The number of customers in the three categories are (12,26,12). H0: The number of customers in the three categories are different from (12,26,12).
Significance level =0.05 Decision rule: Reject the null hypothesis if the p value is less than the significance level. Details Test Statistics Type of Purchasing Chi-Square df
a
5.795 2
Asymp. Sig.
.055
a. 0 cells (.0%) have expected frequencies less than 5. The minimum expected cell frequency is 12.0.
Conclusion: Fails to reject the null hypothesis. The sample does not provides enough evidence to support the claim that there is significant difference in the number of customers in the three categories are different from (12,26,12).
Question 3: Is there a relationship between the company size and type of procurement? Run chi-square independence test (crosstabs) using company size and type of procurement. Use the chi square and the phi coefficient to evaluate the relationship and statistical significance. Report the observed and expected values and the tests of statistical significance.
H0: There is no association between company size and type of procurement. H1: There is association between company size and type of procurement. Test Statistic used is Chi square test for independence Significance level =0.05 Decision rule: Reject the null hypothesis if the p value is less than the significance level. Details
Firm Size * Structure of Procurement Crosstabulation Count Structure of Procurement
Firm Size
Decentralized
Centralized
Total
Small
14
13
27
Large
12
11
23
26
24
50
Total
Firm Size * Structure of Procurement Crosstabulation Expected Count Structure of Procurement
Firm Size Total
Decentralized
Centralized
Total
Small
14.0
13.0
27.0
Large
12.0
11.0
23.0
26.0
24.0
50.0
Value
df
Asymp. Sig. (2sided)
Pearson Chi-Square
.001
a
1
.982
Continuity Correction
.000
1
1.000
Likelihood Ratio
.001
1
.982
.001
1
.982
Fisher's Exact Test Linear-by-Linear Association N of Valid Cases
50
Conclusion: Fails to reject the null hypothesis. The sample does not provide enough evidence to support the claim that there is association between company size and type of procurement. The phi and chi square coefficients indicate jointly the strength and the significance of a relationship. The value of Phi is very small indicating that there is no relationship between company size and type of procurement. Symmetric Measures
Nominal by Nominal
N of Valid Cases
Value
Approx. Sig.
Phi
-.003
.982
Cramer's V
.003
.982
50