Sample Management Report Paper on A Data Analysis of a Customer Satisfaction Survey

A Data Analysis of a Customer Satisfaction Survey

Executive summary

In the recent past, Computers R Us Company that specializes in manufacturing and retailing computer devices has launched a service and repair division known as CompleteCare. Although the division aims at providing quality services to the customers, it experiences various challenges. To establish the background of the challenges, a research was conducted to evaluate the effectiveness of its current strategies in the division. A survey method was utilized to collect the data, and t-test, correlation test, regression test and one-way ANOVA tests were utilized to analyze the data. It was established that the overall customer satisfaction was slightly greater than six. However, only one determinant of the customer satisfaction namely the level of advices that the members of staff offer to customers was greater than six. The other three determinants of customer satisfaction do not meet the minimum threshold. In relation to these findings, it has been recommended that the company should initiate a new loyalty reward program, improve the way its employees communicate with employees and decrease the response time in the new division.  

A Data Analysis of a Customer Satisfaction Survey

Introduction

The Computers R Us Company specializes in manufacturing and retailing computer devices. In order to improve the services it offers to its customers, the company recently launched a CompleteCare division. The division is focused on offering quality services, but it experiences distribution challenges and parts availability challenges. In addition, the division lacks enough trained personnel in it telephone center to handle queries from customers. As a result of this, the company has received numerous complaints from customers. After evaluating the above issues, the company has established that customer satisfaction contributes significantly to the challenges facing CompleteCare division. In order to address the challenges, the company has developed several initiatives aimed at driving customer satisfaction to a minimum of 6 out of 10. The several initiatives include improving the methods of communication at the division; improving the level of advices at the division; decreasing the response time; and initiating a new loyalty reward program. Before altering the current practices, the company wishes to evaluate the effectiveness of its current strategies and the present research is focused on evaluating these issues.   

Research Design

The survey method that involved identifying research participants from a pool of customers and issuing the customers with questionnaires to fill was utilized to collect the data. A sample of 500 research participants was identified. However, only 420 of these research participants were able to fill the questionnaires and return them. The rest did not respond. In terms of sampling, a simple random sampling method was utilized. This method involved picking customers randomly from a pool of customers. The method was utilized to offer customers equal chances of participating in the research.

Hypothesis Development

This part of the research identifies the research questions and their hypotheses that were utilized to test the current strategies.  

Question 1

            Does the present level of customer satisfaction deviate from the management’s goal of 6 out of 10?

Null hypothesis (H0): the present level of customer satisfaction is equal management’s goal of 6 out of 10.

Alternative hypothesis (HA): the present level of customer satisfaction is not equal to management’s goal of 6 out of 10.

Question 2

Does the mean satisfaction level of the male customers with the loyalty reward program differ from the satisfaction level of the female customers with the loyalty reward program?

Null hypothesis (H0): the mean satisfaction level of the male customers with the loyalty reward program is equal to the mean satisfaction level of the female customers with the loyalty reward program.

Alternative hypothesis (HA): the mean satisfaction level of the male customers with the loyalty reward program is not equal to the mean satisfaction level of the female customers with the loyalty reward program.

Question 3

For the five age groups evaluated in the study, is there any difference among their overall customer satisfaction? 

Null hypothesis (H0): the means for the age groups: 1, 2, 3, 4 and 5 are all equal.

Alternative hypothesis (HA): at least one of the means for the age groups: 1, 2, 3, 4 and 5 is different from the means for the other age groups.

Question 4

Is there any disparity in customer satisfaction between the response for the initiative of ‘increasing the level of advice at the division’ and the response for the initiative of ‘the loyalty rewards program at the company?

Null hypothesis (H0): the mean for the response to the initiative of ‘increasing the level of advice at the division’ is equal to the mean for the response to the initiative of ‘the loyalty rewards program at the company.  

Alternative hypothesis (HA): the mean for the response to the initiative of ‘increasing the level of advice at the division’ is not equal to the mean for the response to the initiative of ‘the loyalty rewards program at the company.

Question 5

Do any of the initiatives of ‘decreasing response times in the CompleteCare division’, ‘the level of advice CompleteCare staff provide on Computers R Us products and services’, ‘the level of communication with staff and management’ and ‘new loyalty rewards program’ influence the overall satisfaction of Computers R Us customers?

Null hypothesis (H0): the slope of the regression line is equal to zero. That is, H0: βracl=0.

Alternative hypothesis (HA): the slope of the regression line is not equal to zero. That is, at least of βi is not equal to zero.

Statistical Technique and Justification

To test the first research question, a univariate t-test will be utilized as this is the appropriate test for testing hypothesis involving observed mean against specified values (Zikmund, Babin, Carr, & Griffin, 2012). In this case, the specified values will be six whereas observed values will be calculated from the sample. If calculated mean will be equal to six, then the null hypothesis will not be rejected. On the contrary, if calculated mean will not be equal to six, then the null hypothesis will be rejected.  

To test the second research question, independent sample t-test will be utilized as this is the appropriate test for comparing the difference between two means from two independent samples (Zikmund, Babin, Carr, & Griffin, 2012). The t-test will be utilized because the variances for the samples are unknown.

To test the third research question, one-way ANOVA test will be utilized as this is the appropriate test for comparing the mean difference of more than two groups (Zikmund, Babin, Carr, & Griffin, 2012). The groups in this case will be the five age groups.

To test the fourth research question, a two-sample t-test will be utilized as this is the appropriate test for comparing the mean difference between two groups (Albright, & Winston, 2016). The first group in this case will concern itself with the level of advice that the members of staff provide to the customers whereas the second group will concern itself with the reward program.

To test the fifth research question, both correlation test and regression test will be utilized as these are the appropriate tests for testing the relationship between a dependent variable and multiple independent variables (Zikmund, Babin, Carr, & Griffin, 2012). The dependent variable will be the overall satisfaction whereas the other four variables will be the independent variables.  

Results and Statistical and Non-Statistical Interpretation

The results for the first research question are as shown below in Table 1.

Table 1: t-Test results
 Observed meanAssumed man
Mean6.2428571436
Variance2.7141493350
Observations420420
Hypothesized Mean Difference0
df419
t Stat3.021055174
P(T<=t) one-tail0.001336636
t Critical one-tail1.648498411
P(T<=t) two-tail0.002673271
t Critical two-tail1.965641764 

Statistical Interpretation

From Table 1 above, it is evident that t-calculated (3.021055174) is greater than t-critical (1.965641764) and p-value (0.002673271) is less than the significance level (5%); thus, we can reject the null hypothesis that the level of customer satisfaction is equal to 6 out of 10 (Johnson, 2002).

Non-Statistical Interpretation

The average level of customer satisfaction (6.242857143) is greater than 6 out of 10. Therefore, we should conclude that the level of customer satisfaction is not equal to 6 out of 10.

The results for the second research question are as shown below in Table 2.

Table 2: t-Test results.
 MaleFemale
Mean3.9453551915.708860759
Variance4.6563381973.436065222
Observations183237
Hypothesized Mean Difference0
df359
t Stat-8.823857201
P(T<=t) one-tail2.45541E-17
t Critical one-tail1.649109151
P(T<=t) two-tail4.91083E-17
t Critical two-tail1.966593866 

Statistical Interpretation

From Table 2 above, it is evident that t-calculated (8.823857201) is greater than t-critical (1.965655386) and p-value (4.91083E-17) is less than the significance level (5%); thus, we can reject the null hypothesis that there is no significant difference between the satisfaction of male and female customers with the new loyalty rewards program at Computer R Us (Quirk, 2015).

Non-Statistical Interpretation

The average satisfaction level (3.945355191) for the males is less than the average satisfaction level (5.708860759) for the females with the new loyalty reward program. Therefore, we should conclude that there is significant difference between the satisfaction level of the males and females with the reward system. The female customers appear to be more satisfied with the reward system than the male customers.

The results for the third research question are as shown below in Table 3.

Table 3: Anova: Single Factor
SUMMARY
GroupsCountSumAverageVariance
Age group 11006016.013.98979798
Age group 21086756.252.376168224
Age group 31127216.43752.050112613
Age group 4643946.156252.451388889
Age group 5362316.4166666672.707142857
ANOVA
Source of VariationSSdfMSFP-valueF crit
Between Groups11.2385714342.8096428571.0355347610.3884996572.393438223
Within Groups1125.994152.713228916
Total1137.228571419    

Statistical Interpretation

From Table 3 above, it is evident that F-calculated (1.035534761) is less than F-critical (2.393438223) and p-value (0.388499657) is greater than the significance level (5%); thus, we should fail to reject the null hypothesis that there are no significant differences in the overall customer satisfaction across the age groups: 1, 2, 3, 4 and 5 (Lee, Lee, & Lee, 2000).

Non-Statistical Interpretation

Based on the fact that the average satisfaction level for the five groups are 6.01; 6.25; 6.4375; 6.15625; and 6.416666667 respectively, then we should conclude that there is no significant difference in the overall customer satisfaction across the five age groups (Bajpai, 2009). 

The results for the fourth research question are as shown below in Table 4.

Table 4: t-Test results
 Question 6Question 8
Mean6.790476194.94047619
Variance4.4428685084.724372088
Observations420420
Assumed Mean Difference0
df837
t Stat12.52209686
P(T<=t) one-tail2.1287E-33
t Critical one-tail1.646676169
P(T<=t) two-tail4.25739E-33
t Critical two-tail1.96280223 

Statistical Interpretation

From Table 4 above, it is evident that t-calculated (12.52209686) is greater than t-critical (1.96280223) and p-value (4.25739E-33) is less than the significance level (5%); thus, we can reject the null hypothesis that there is no significant difference between the response of increasing the level of advice and changing the loyalty reward system (Weiers, 2010).

Non-Statistical Interpretation

The average mean for the level of advice provided to the customers at the division (6.79047619) is greater than the average mean for the loyalty reward system (4.94047619). Therefore, we should conclude that there is significant difference between the satisfaction levels for the two responses with the level of advices having higher mean (Siegel, 2011).

The results for the fifth research question are as shown below in Table 5.

Correlation results

 Question 4Question 5Question 6Question 7Question 8
Question 41
Question 50.3588661
Question 60.4071280.717511
Question 70.4122450.6308010.7354431
Question 8-0.43117-0.22801-0.27049-0.350811
Table 6: Regression results
Regression Statistics
Multiple R0.537620274
R Square0.289035559
Adjusted R Square0.282182889
Standard Error1.39580186
Observations420
ANOVA
 dfSSMSFSignificance F
Regression4328.699495582.1748738742.17853591.10557E-29
Residual415808.5290761.948262834
Total4191137.228571   
 CoefficientsStandard Errort StatP-valueLower 95%
Intercept5.4408392550.35994420115.11578531.94358E-414.733298147
Question 50.0834498840.0566217411.4738134690.141289929-0.027851282
Question 60.1313276870.0545112082.4091868680.0164227620.024175188
Question 70.1132774320.0614454231.8435454770.065962185-0.00750563
Question 8-0.2453427390.03350773-7.3219743451.27951E-12-0.311208771

Statistical Interpretation

From table 5 above, the response time has a correlation coefficient of 0.358866 with the overall satisfaction. The level of advice has a correlation coefficient of 0.407128 with the overall satisfaction. The level of communication has a correlation coefficient of 0.412245with the overall satisfaction whereas the loyalty reward program has a correlation coefficient of -0.43117with the overall satisfaction. This indicates that the response time, level of advice and level of communication have positive weak relationships with the overall satisfaction. It also indicates that the loyalty reward program has a weak negative relationship with the overall satisfaction.

From Table 6 above, it is evident that F-calculated (42.1785359) is greater than F-critical (1.10557E-29) and p-value (1.94358E-41) is less than the significance level (5%); thus, we can reject the null hypothesis that the slope of the regression line is equal to zero (Anderson, Sweeney & Williams, 2012).

Non-Statistical Interpretation

The average satisfaction levels for the four initiatives are 5.883333; 6.790476; 5.57381; and 4.940476 respectively. Based on these averages, it is evident that majority of the customers are satisfied with the advices they receive from the members of the staff. Conversely, it is evident that majority of the customers are dissatisfied with the loyalty reward program (Beri, 2010). In relation to these findings, it is evident that the level of advices that the members of staff provide to customers contribute significantly to the overall customer satisfaction whereas satisfaction with the loyalty reward system contribute insignificantly to the overall satisfaction. This being the case, we would conclude that the level of advices that the members of staff provide to customers influence the overall satisfaction in the positive direction whereas satisfaction with the loyalty reward system influences the overall satisfaction in the negative direction.

Analysis and Summary of the Statistical Results

The current degree of customer satisfaction stands at 6.242857143. At age group level, the degree of satisfaction stands at 6.01; 6.25; 6.4375; 6.15625; and 6.416666667 for the five age groups that were interviewed. This indicates that there is no significant difference among the five age groups with the degree of satisfaction. However, with regard to determinants of customer satisfaction that were evaluated, the response time has an average of 5.883333 satisfaction; the level of advice has an average of 6.790476 satisfaction; the level of communication has an average of 5.57381 satisfaction; whereas the loyalty reward program has an average of 4.940476 satisfaction. This indicates that majority of the customers are satisfied with the advices they receive from the members of the staff. In terms of dissatisfaction, the results indicate that majority of the customers are dissatisfied with the loyalty reward system, the level of communication and the response time. Based on these results, Computers R Us should focus its attention on improving its reward system. In addition, the company should also focus its attention on improving the level of communication of its employees and their response time to customers’ queries.

Recommendations

Based on the above results, Computers R Us should do the following so that the degrees of satisfaction for the other three determinants of customer satisfaction can meet the management’s goal of 6 out of 10.

  • First, the company should decrease the response time in CompleteCare division by either increasing the number of employees in this department or considering other ways that can decrease the response time.
  • Second, the company should improve the level of communication of its employees in the division. In this case, the company can train its employees how to communicate with customers effectively.
  • Third, the company should change its current loyalty reward system.

As the company implements the above recommendations, it should look at the factors that contribute to the differences between the satisfaction level of the male and female customers. If the company would do this, it would improve the overall customer satisfaction. At the same time, the mean satisfaction level for each determinant of satisfaction would be equal to six or higher than six. 

References

Albright, S., & Winston, W. (2016). Business analytics: data analysis & decision making. Stamford: Cengage learning.

Anderson, D., Sweeney, D., & Williams, T. (2012). Essentials of modern business statistics. Australia: South-Western / Cengage Learning.

Bajpai, N. (2009). Business statistics. Delhi: Pearson.

Beri, G. (2010). Business statistics. New Delhi: Tata McGraw-Hill.

Johnson, G. (2002). Research methods for public administrators. Westport, CT [etc.: Quorum.

Lee, C., Lee, A., & Lee, J. (2000). Statistics for business and financial economics. Singapore: World Scientific Publ. Co.

Quirk, T. (2015). Excel 2013 for business statistics: A guide to solving practical problems. Cham: Springer.

Siegel, A. (2011). Practical Business Statistics. Burlington: Elsevier Science.

Weiers, R. (2010). Introduction to business statistics. Stamford: Cengage learning.

Zikmund, W., Babin, B., Carr, J., & Griffin, M. (2012). Business research methods. South-Western: Cengage Learning.