Internship Report on Customer Service of HSBC

Executive Summary:

Practical knowledge has no alternatives especially for BBA students. Through the report students are given a chance to acquire practical knowledge through a little period of time. It helps the students to provide an on-the-job exposure and an opportunity for transaction of the theoretical canephor into real life situation. The project is a part of the BBA program for BBA students of the International Islanic University Chittagong. So like every year, Department of Business Administration of the university has provided us to prepare a report in order to acquire practical knowledge in a particular area. The time for preparing the report in HSBC Limited was very insufficient, yet I tried my level best to find out relationship between knowledge of the HSBC customers and HSBC service within this short period of time.

I was assigned to study the HSBC Customer service “How may I help you” Here I learnt a little about the Customer service of HSBC.

1. How many times do you use your ATM card?              Quarterly        Once a year         Twice a year          Once in every month         Every week

1. To what extent do you have problems with the location of the ATM machines?

Very much           1          2          3          4          5          Not at all

Please rate each of the following traits amongst the HSBC Customer Service Employees, 1 being the strongest and 5 being the weakest:

1. To what extent do you have problems with the number of the ATM machines?

Very much           1          2          3          4          5          Not at all

1. How often are you annoyed with the ATM machine services?

Very Often     1          2          3          4          5          Not at all

Do you use credit card?       Yes                            No

1. If yes, how many times do you use your Credit card?  Quarterly         Once a year         Twice a year          Once in every month            Every week

1. How often are you annoyed with the customer service employees at HSBC?

Very Often     1          2          3          4          5          Not at all

Customer Satisfaction in different dimensions (1 being strongly agree, and 5 being strongly disagree). Customer service employees

Personal Information

Gender :             Male                                      Female

Age        19-28                   29-38                39-48              49-58                   59 and above

Marital Status                 Married              Single

Your Profession               Student         Businessman            Employee

Housewife                               Others

Educational Level         SSC Level              HSC Level       Under Graduate 

 Post Graduate :

Where do you stay?      Banani          Dhanmondi          Uttara          Gulshan  Others (Please Specify)

Thank you!! You’ve been a great help J

Figure 1.   Gender Distribution of the HSBC Customers Who Were Surveyed

Table 1.  Gender Distribution of the HSBC Customers Who Were Surveyed

I surveyed 30 customers of which 46.7% were females and 53.3% were males. This can be explained through various spheres. Number one, being that there has been an increasing number of women literacy. More women are now found in the work field with their male co-workers. Hence, they need banking facilities.    

Figure 2.   Age Distribution of the Customers

Table 2.  Age Distribution of the Customers

Majority of the customers I surveyed fell in the age group of 29 to 38 years. Next were from 19 to 28 years old, the rest being insignificant. This can portray many things of which is younger customers tend to bank more than the elder ones.

Figure 3.   Frequency Distribution of the Marital Status

Table 3.  Frequency Distribution of the Marital Status

There isn’t much difference between the numbers of married customers and unmarried ones. This can be explained as the trend that people tend to get married at a much later age in the recent times.

Figure 4. Distribution of Profession of the Customers

Table 4. Distribution of Profession of the Customers

As we can see that majority of the customers fall under the category of employees and businessmen compared to the unemployed category (students and housewives). This makes sense as it is evident that banking is more important for those who are engaged in the daily transactions more.

Figure 5. Education Level of the Customers

Table 5.  Education Level of the Customers

The difference is not worth while though, but it should be mentioned that in Bangladesh HSBC, the trend is that the customers are seen to have undergraduate degrees compared to postgraduate degrees. This also confirms that none of the customers I surveyed are illiterate or educated of below standards.

Table 6.  Distribution of the Residence of the Customers

Table 6.  Distribution of the Residence of the Customers

Majority of the customers stay in Gulshan and Banani. This can be explained as such that the survey was conducted in the Gulshan Branch of HSBC. Since, most of the customers of that branch come from that area; it is evident that this will be the result.

Figure 7. Number of Years with HSBC

Table 7.  Distribution of the Number of Years the Customers are with HSBC

I thought the number of years a customer banks in HSBC might reveal or relate with a lot of the other variables in the questionnaire. It is seen that majority of the customers have been banking with HSBC for 1-3 years and the next in the table is 4 to 6 years. It is same amount of customers less than one year and from 7-9 years. There can be several explanations for this.

Figure 8.  If the customer maintains accounts with other banks

Table 8.  Frequency distribution of the customer if they maintain accounts with other banks

This is another important component for the base of customer knowledge. It can be assumed that if customers have accounts with other banks, they are likely to be well informed about the current services available in the market at their due prices with benefits.

Figure 9. Customers Maintain Accounts With Which Other Banks

Table 9.  Customer Maintain Accounts With Which Other Banks

Customers who maintain accounts with other banks mostly bank with SCB. The rest insignificant ones bank with BRAC and DBBL. This shows a rising trend of the Bangladeshi investors and borrowers to invest and save in multinational banks. This can result due to the vulnerable political commotion in the country, social prestige, etc. This can also pose a fierce competition amongst the two big giants.

Table 10.   What The Customers Think About Reading Instructions

Majority of the customers interviewed agree that it is moderately important to read instructions provided by the bank. Then comes the ‘Not Sure’. But a healthy portion of the customers also claims that they feel it’s moderately less important to read the instructions and they can almost do without reading them! This attitude can be explained by the poor reading habits of the Bengalis. Much of the education system is to be blamed for this. The typical Bengali Medium education system discourages reading and unfortunately encourages memorizing. This highly dampens the spirit of creativity and results as such.

Table 11.  How Well Customers Read Instructions Provided by HSBC

I put this question in order to verify if the customers comply with their answers in the question 10. And it worked just as I predicted. Majority of the customers don’t read instructions that often. During my internship term, I found plenty of customers who ran to the bank without properly reading the instructions and later on complained if we told them that for the service they wanted, they had to go elsewhere. This dissatisfaction would not have been as much if the instructions were read.

Table 12.  Frequencies of Employee Accounts Amongst the Customers

Many companies pay their salary through banks. HSBC also has the option and name the accounts as employee accounts. When the employee changes the job, she/he can just change the address and maintain the same account with HSBC. Around 40% of the customers I surveyed were employee accounts.

Table 13.  Frequency of Visits to HSBC

I asked the customers how frequently they visited the bank. Most of them regularly visit the bank in a monthly basis. This indicates that they come into contact with the services frequently.

Table 14. Frequency Distribution of Customer Ratings on HSBC’s Services

This question directly asks how the customers would rate HSBC’s services. Maximum customers agreed that they think the service of HSBC is good. Although there are a significant number of customers saying that they are not sure and think the service is neither good nor bad.

Table 15.  Type of Accounts Customers Maintained in HSBC

The type of account most popular in HSBC is Local currency savings account. Next, nearly one third of the Local currency savings account is the current account (both local and foreign currency ones). Company account also accounts in this table but it grabs only a small part of the pie.

Table 16.  Amount Maintained in the Accounts

This question is very crucial as well in determining which customers are more satisfied. I kept this question in order to find relation with other variables and draw conclusions. It can been seen that around 90% of the customers surveyed keeps from BDT 5001.00 to over 200,000.00 on an average. Only 10 percent maintains from BDT 1000.00 to 5000.00.

Table 17.  ATM Card Used by the Customers

Almost 97% of the customers maintain ATM Card with HSBC. This can tell a little bit about the recent trend towards Bangladeshi Fast Life. The movement towards modern living leads to use ATM Cards and move away from traditional banking. The faster the country will adopt the fruits of modernization the faster the standard of living will increase leading to a rapid economic growth.

Table 18. How Frequent the Card is Used

Only querying about the ATM card if used or not really does not give much of a meaning. It is up to the frequency of the card used that will help open the window and give meaning to it. From this question I can be seen that almost all the customers use the card o a weekly basis. This can reflect that a huge portion of the services consumed by the customers fall under ATM. Faults in the machines can be a part of the dissatisfaction amongst the consumers.

Table 19.  Extent to Which There is a Problem With ATM Location

Majority answers that they do not face problems regarding the location of ATMs. This can also be explained numerously. There can be some errors, since I collected data from only Gulshan Branch and disregarded customers from other six branches/locations. Customers who come to Gulshan branch may be content with the location since they travel in Gulshan and Banani area for their work, and there are ATMs in both Gulshan and Banani. On the other hand, those who work in Mirpur can face severe problems due to lack of a single ATM in Mirpur area. Right after the 43.3% saying not that much problem is faced, 26.7% says they have a healthy amount of problem.

Table 20.  Extent to Which There is Problem With Number of ATMs

This question is different from the previous one in the dimension that the problem faced in this one is regarding the number of ATMs. Compared to the response of the previous question, customers are more satisfied with the number of ATMs. This can also be addressed with the similar error faced before (customer base is limited in the sense that only Gulshan Branch customers were surveyed).

Table 21.  Customers Rating HSBC Employees in Terms of Knowledge of the Subject Matter

This is a vital criterion in determining customer satisfaction level. This points out what customers think about the HSBC employees’ knowledge level about the subject matter they deal with. Customers rated this very generously. Almost all agreed strongly that the employees are well qualified for the job they do.

Table 22.  Patience of HSBC Customer Service (CS) Employees

Patience is another important dynamics of customer services. With customers, it is extremely pivotal to deal with patience. Customers have also been very generous while rating patience of the employees serving them. No one rated the customer service employees otherwise. During my internship, I encountered numerous cases where I was pushed to the limit! But tried to calm myself and worked with patience. And later on, some customers did come and apologize later. This in turn enhances the image of the organization.

Table 23. Oral Communication Skills of the CS Employees

Oral communication skill determines a lot about the customer satisfaction level since verbal medium is used to present. An organization can be perfect but if it employs people who can not properly verse with the customers, may God help the organization! In HSBC, customers rated oral communication skills of the customer service employees very well. But it has still scope to perfect this trait.

Table 24.  Affability of CS Employees

Affability is an essential part of dealing with customers from a wide group of people. People coming from different background reacts differently. But universally it is found that it is hard to behave bad with a friendly person or a cheerful one. This key point is very positive as well amongst the customers when asked to rate.

Table 25. Situation Adaptability of CS Employees

Compared to previous dimensions, situation adaptability is scrutinized heavily. Although majority of the customers filled up that the trait is very strong amongst the customer service employees, but there is a small percentage saying there is a scope for poor situation adaptability. During my internship period, there was customer whose ATM card was mailed from Singapore. I tore his Envelope in order to get a copy of his received acknowledgement, but I did not do that in front of him. Later when I gave him the acknowledgement to sign, he was furious and was shouting at the top of his lungs. I was so embarrassed that I did not know what to do. I was poor in adapting the situation in that case.

Table 26.  Punctuality of CS Employees

Although majority of the customers voted for better performance of the HSBC customer service employees, but few cases show that there is a room for improvement for the employees in terms of punctuality. With fast paced lifestyle, time is a priceless commodity whose value is appreciating in a daily basis. Hence, care should be taken when it comes to killing customer’s time.

Table 27.  Interest Shown by CS Employees to Customer Queries

A space for improvement is also in this trait – interest shown by the customer service employees for customer queries. It is an essential role determining customer satisfaction. Customers will not feel safe in their transactions and neither would they want to bank somewhere where they do not feel wanted. Hence, measures should be taken in order to make the customers feel that they are important. The employees are interested in helping their queries.

Table 28. The Extent to Which CS Employees perform the Right Service at the First Time

In every service it is extremely important to perform the right service at the first time. This can be explained through many dimensions. To begin with, performing the service right the first time eliminates the chance of making errors the first time. Customers are not irritated. Secondly, it is less time consuming. Less time is needed to spend after each customer.

Table 29.  CS Employees Keeps the Customers Informed to What Extent

One of the major tasks of the customer service employees is to sell their products. HSBC is none the less different in this case. It is seen that majority customers are confused about this trait. This leaves a room for improvement. The service employees can sell their products once they inform the customers. This can benefit all the three parties engaged in the business (customers, employees, and the organization). Customers may get a better deal out of the new products, employees performance can be appraised better for increments &/or promotions, lastly but not the least HSBC can widen its customer base and sell more products by keeping the customers informed.

Figure 30.  The Extent to Which CS Employees are Willing to Help

Table 30.  The Extent to Which CS Employees are Willing to Help

After quite some time the customers have rated the customer service employees very modestly in terms of willingness to help the customers. This trait talks a lot about the employees. To start with, they must be in a healthy environment. Next, it can be told that the organization is customer sensitive for which it maintains an employee base with willingness to attend customers.

Figure 31. CS Employees if too Busy to Respond to Queries

Table 31. CS Employees if too Busy to Respond to Queries

Right after the question of willingness to help, I placed this question to find out if there are any discrepancies. Customers rated very well in the previous question regarding willingness, but here, although majority agrees that the employees are not too busy to respond to their queries, there are significant cases where customers moderately disagrees to the statement. This can be explained as such that the branch is over crowded and may be more employees are required to attend the customers.

Figure 32.  CS Employees If Consistently Courteous With the Customers

Table 32.  CS Employees If Consistently Courteous With the Customers

Customers agree gaily that customer service employees are consistently courteous with the customers. This is a key factor in determining the satisfaction level of the customers. They would not like to deal with CS employees who are obstinate.

Table 33.  CS Employees If Neat

We judge people through their appearances even though we don’t intend to at times. Neatness is an extremely important psychological factor in determining customer satisfaction. I saw that HSBC has uniforms for both the male and female employees. There is no written rule as to how neat & clean the employees should be, but in my term, I found the employees to be neat on average and so did most of the customers I surveyed.

Table 34. How Annoyed Customers are With ATM Service

Since ATM covers a big portion of the services offered by HSBC, dissatisfaction in the machines poses a big threat to the image of the bank. Customers when asked how annoyed they are with ATM services answers ‘Not that much’ as majority. But significant amount also answers that they are annoyed with the machine. I also faced almost regular cases regarding ATM failures. One of the things the customer gets annoyed is the few seconds the machine waits for the PIN after which it retains the card. The customers have to wait at least for a day to get back their cards.

Table 35.  Credit Card Used

Only 23% of the customer I surveyed uses credit cards. HSBC has affiliation with Prime Bank when it comes to credit card. Figure 36 depicts that most of the people who use credit cards uses it very little, like once a year. This can be explained with the limited usage of credit cards in the country. But much to modernization, a change in the pattern is seen to have followed until very recently.

Table 37.  How often annoyed with HSBC CS

I placed this question in the middle to verify the question I asked in the beginning about the Customer Service of HSBC. Majority of the customers went for ‘not that much’. Next in the list is ‘not at all’. Both are good news for the organization, but concern should be there to improve. A room for improvement is strongly present regarding customer services.

Cross Tables

Cross tabulation 1.  Read instructions * Annoyed with ATM machine service

Cross tabulation 2. Read instructions * Profession

Cross tabulation 3. Read instructions * Education Level

Cross tabulation 4.  Read instructions *Customer’s Age

Cross tabulation 5.  How often annoyed with HSBC CS * Average amount of BDT maintained in HSBC

Cross tabulation 6. Number of years with HSBC * Rate Customer Service of HSBC

      Number of years with HSBC * Rate Customer Service of HSBC Crosstabulation

                                 Chi-Square Tests

a  14 cells (93.3%) have expected count less than 5. The minimum expected count is .40.

Cross tabulation Analyses:

According to the figures in cross tabulation 1, it can be seen that those who read instructions are more satisfied than those who rarely do. This can help to correlate because the chi-square test helped me be sure that there exists some relationship between the variables. I am sure because the assume value is less than 0.05. In the second cross table, it is found that those who are employees and have accounts in HSBC tend to read instructions way more than those of other profession like businessman, students, housewives and others. This is also made sure by the chi-square test. I undertook many variables with which I tested through cross tables, but I considered only those which the passed the chi-square test. There is a relationship between reading instructions and education level of the customers according to the chi-square test, but the percentage in the table does not show much difference between those who are undergraduate and those who are postgraduates. The next cross table is regarding reading the instructions with their respective ages. The result is surprising, since the younger customers tend to read much more than the elder ones. The age ranges from mainly 29-38 years. That whose age ranges from 39-48 mostly reads instructions for the sake of reading. In cross table 5, it can be clearly derived that those who maintain limited amounts with HSBC on an average basis tend to be more annoyed with the Customer Service (CS) compared to those who maintain greater amounts. The table depicts that majority customers with amount BDT1,000.00 to 5,000.00 has a healthy amount of problems with the CS whereas most customers with more than 50,000.00 BDT seem to have less problem or none at all. The sixth cross table portrays the relationship between the numbers of years a customer has been banking with HSBC with rating the CS of HSBC. The table represents that customers who banked for 4-6 years have rated HSBC as either ‘Excellent’ or ‘Good’. Although it can be said that the trend is that higher the number of years that a customer banks, better is the ranking for CS. But a technical problem persists; customers who’ve been banking for more than 7 years tend to rate the CS equal amounts of ‘excellent’, ‘Good’ and ‘Bad’. So it is not proper to state that pattern. These are the few insights taken from the tables.

Regression analysis:

Regression analysis is a modeling technique that identifies, quantifies, and describes the relationship between at least two variables, which depend on one another because of an underlying theoretical reason. In a multiple regressions, we have several independent variables, which are used to predict the dependent variable.

Statement of the hypothesis:

It is hypothesized that the level of customer satisfaction about the services provided by HSBC is positively related with the level of being informed about its service package of the customers. In particular, the customers who are more informed about the services tend to be more satisfied than those who are less informed. In addition, it is also hypothesized that the employed customers (who are working in any organization) are more satisfied than the unemployed customers (students, housewife etc.).

Description and specification of the model:

The dependant variable of this model is CSI (Customer Satisfaction Index) and the independent variables are CII (Customer Informed Index) and customer’s employment condition, a dummy variable. CSI index (Customer Satisfaction index) includes 8 dimensions of customer’s satisfactions. The independent variable CII (Customer Informed Index) is measured by using 3 critical factors that strongly indicates customers’ degree of being informed of the services provided by HSBC. This model also includes a dummy variable as another independent variable in order to test the relation whether customer’s employment status has any relation. Usually, a dummy variable is a numerical variable used in regression analysis to represent subgroups of the sample in your study. In research design, a dummy variable is often used to distinguish different treatment groups. Dummy variables are useful because they enable to use a single regression equation to represent multiple groups. This means that we don’t need to write out separate equation models for each subgroup. The dummy variables act like ‘switches’ that turn various parameters on and off in an equation. Here, a 0, 1 dummy variable is used where the variable takes the value 1 for employed customer and it takes the value 0 for unemployed customers. However, It is expected CII is positively related with CSI and dummy variable is also expected to be positively related with CSI. More importantly, it is assumed that this two independent variable is expected to be linearly related with the dependent variable, CRI.

The population multiple regression equation is as follows:

Y= b₀ +b₁X₁ +b₂X₂ + e

Here,

Y = CSI Index

b₀  = intercept of the regression line.

b₁ = coefficient of independent variable X₁ (CII)

b₂ = coefficient of independent variable X₂ (Dummy Variable)

X₁ = CII.

X₂  = Dummy variable; takes the value 1 for employed customer and it takes the value 0 for unemployed customers.

e         = Random error with the mean 0.

Assumptions of the model:

            For this model we are assuming that other variables that has influences on CSI Index as constant. We have assumed some standard clauses to use ordinary least square method, which are known as classical group of assumptions: These are:

• Xi’s are fixed or if Xi’s are independent of e_i‘s.             [ i=1,2,…..,n]
• e_i‘s are random variables with mean = 0.                        [ i=1,2,…..,n]
• V (e_i)=E [e_i -E (e_i)]² = E (e_i ²) = s²_E is constant.             [ i=1,2,…..,n]
• Ei’s are independent of each other.                                 [ i=1,2,…..,n]
• There is no set of nonzero constant Co, C1, …..,Ck such that, C0+C1X1+………..+CkXk = 0. So, X₁,……………X_K are linearly independent.

      [ i=1,2,…..,n]

Data collection:

The data for the regression analysis is collected from the research questionnaire. Each questionnaire contains 42 questions and information obtained from some of the questions has been used to measure different variable of the model. The sample size is 30. The dependent variable Y (CSI) is measured using 8 dimensions of customer’s satisfactions such as customers rating on employees knowledge, patience, oral communication skills, affability, situation adaptability, responds toward customers, sincere interest and employee neatness, these 8 dimensions of customer satisfaction have been assigned with  weights according to their relative importance and thereby the CSI is calculated. The independent variable X₁ (CII) is measured using 3 critical factors that strongly indicates customers degree of being informed of the services provided by HSBC such as the time line of customers banking experience with HSBC, the extent to which costumer read the instructions provided by HSBC and frequency of their visit to HSBC. Finally this factor has been assigned with a weight according to their relative importance about customer’s degree of being informed and thereby the CII is calculated. Both the index, CSI and CII is calculated in a way that the upper limit of them is 500 and lower limit is 0. Another independent variable X₂ is a Dummy variable where the variable takes the value 1 for employed customer and it takes the value 0 for unemployed customers. (The details of the data collection is given in the Appendix)

Model Estimation AND INTERPRETATION:

The regression equation is

Y = 331 + 0.234 X₁ – 11.5 X₂

β₀ = 331

β₀ is the intercept of the regression line. Here, b₀ =331 means if all the independent variable takes the value zero, on an average, the CSI (Customer Satisfaction Index) will be 331.

β₁’s estimate b₁ =0.234

Other things remaining the same, if the CII (Customer Informed Index) increases by 1, on an average, the CSI (Customer Satisfaction Index) will go up by an additional 0.234.

Β₂’s estimate b₅ = -11.5

Other things remaining the same, for employed Customer CSI (Customer Satisfaction Index) is 11.5 lower than unemployed customers. It basically indicates that on an average the less satisfied with the services than the unemployed customers, surprising it is the opposite of the prior expectation.

Co-efficient of determination R² = 22.5%

R-sq is 22.5% means almost 22.5% of the observed variability in the dependent variables is explained by the independent variables of our regression model. So, this regression equation has a very low explanatory power. However, in cross sectional data R-sq is usually very low. Importantly, A problem of using R² as an overall measure of the quality of a fitted equation is that when we add more and more variables in a model, even if these variables are irrelevant to the model, it raises the value of R². So, R² is not a good indication of the model’s overall quality. Rather the adjusted coefficient of determination should be used for measuring the overall performances of the model.

Adjusted R² = 16.8%

R-sq (adj) is the adjusted or corrected co-efficient of determination. R² (adj) is 16.8% means almost 16.8% of the observed variability in the dependent variables is explained by the independent variables of our regression model.

Hypothesis Testing:

In this section of regression analysis, the methodology of hypothesis testing is used to test a null hypothesis to find out whether the combination of all the variables is useful predictors of the dependant variable. Another tests of hypothesis for partial regression coefficients (That is, H₀: b_j = 0) are used to determine if a specific independent variable is conditionally important in the multiple regression model. By using the Student’s t statistics and p-value it is concluded whether or not a particular predictor variable is conditionally significant, given the other variables in the regression model. (The following two tables provides the required value for the hypothesis test & interval estimation)

Test of Significance for the whole Regression:

First, hypothesis test is to test a null hypothesis to find out whether the combination of two independent variables is useful predictors of the dependant variable. Accepting this hypothesis will conclude that none of the predictor variables regression model is statistically significant and thus it provide no useful information.

Null Hypothesis Ho: b₁= b₂= 0

Alternative Hypothesis H_1: At leastone b_j  ¹0    (j = 1,2,3,4,5 )

Decision Rule: We reject the null hypothesis if

F_cal >F tab

Or, (SSR/k) / [SSE/(n-k-1)] >F tab

F_cal = 3.92

F _tab= F_{K, n-K-1, α}  = F _{2, 30-3, 0.05} = F _{2, 27, 0.05} = 3.35

Here, F _cal >F tab. Therefore, I reject the null hypothesis that all coefficients are zero against the alternative hypothesis at 5% level of significance. The p-value or the smallest significance level at which the null hypothesis can be rejected is 0.031. So, the entire variable as a whole can predict the dependable variable (CSI) and the combined effect of these two variables do improve the model that predicts CSI.

Now, the hypothesis testing for each of the coefficients is done to analyze if individually they have any influence on the dependent variable keeping the other variable constant.

For the intercept the Hypothesis Testing: (Two sided test)

Null Hypothesis Ho: α = 0

Alternative Hypothesis H₁: α ≠ 0

Decision Rule: We reject the null hypothesis if T cal> T tab.

t _cal= 12.09

t _tab= t _{n –k-1, α / 2} = t _{30 -3, .0.25} = t _{27, .0.25} =2.052

Here t _cal is greater than t _tab. therefore I fail to reject the null hypothesis against the alternative hypothesis at 5% level of significance with the two-sided test. The p-value or the lowest level of significance at which the null hypothesis can be rejected is 0.00.

Hypothesis Testing For b₁:

Null Hypothesis Ho: b₁ ≤ 0

Alternative Hypothesis H₁: b₁ > 0

Decision Rule: We reject the null hypothesis if, t _cal> t _tab

t _cal = (b₁-b₁) / S_b1 = 2.85

t _tab= t _{n –k-1, α} = t _{30 -3, .0.5} = t _{27, .0.5} = 1.701

Here t _cal is greater than t _tab. Therefore I reject the null hypothesis against the alternative hypothesis at 5% level of significance. Therefore, the CII is statistically significant predictors of the dependant variable (CSI); and it has a positive relationship with the CSI (Customer Satisfaction Index). However, The p-value or the smallest significance level at which the null hypothesis can be rejected is 0.001

Hypothesis Testing For b₂:

Null Hypothesis Ho: b₂ ≤ 0

Alternative Hypothesis H₁: b₂ > 0

Decision Rule: We reject the null hypothesis if, t _cal> t _tab

t _cal = (b₂-b₂) / S_b2 = -.65

t _tab= t _{n –k-1, α} = t _{30 -3, .0.5} = t _{27, .0.5} = 1.701

Here t _cal is less than t _tab. Therefore I fail to reject the null hypothesis against the alternative hypothesis at 5% level of significance. However, the p-value or the smallest significance level at which the null hypothesis can be rejected is 0.52. This indicates that the customer’s employment status is not a statistically significant variable for predicting customer’s satisfaction about the service.

confidence interval:

Another way of determining whether a specific independent variable is important in the multiple regression models is to find the confidence interval. The 95% confidence intervals for b₀, b₁, b₂, is estimated in the following.

95% confidence interval for b₀:

The 95% confidence interval for Y-intercept (b₀) ranges from 274.60 to 386.82. Here the 95% confidence interval for b₀ does not include 0, and thus I reject the two-tail hypothesis that the intercept coefficient is 0. Based on this confidence interval, it can be conclude that the intercept a statistically significant predictor in the multiple regression model.

95% confidence interval for b₁:

The coefficient for the CII has a 95% confidence interval 0.06< b₁ <0.41. Here the 95% confidence interval for b₁ does not include 0, and thus I reject the two-tail hypothesis that this coefficient is 0. Based on this confidence interval, it can be conclude that X₁ (CII) is a statistically significant predictor variable in the multiple regression model.

95% confidence interval for b₂:

The coefficient dummy variable has a 95% confidence interval -48.03< b₂ <25.02. As the confidence interval includes 0, and thus I cannot reject the two-tail hypothesis that the intercept is 0. Based on this confidence interval, it can be conclude that the dummy variable (customers employment status) is not a statistically significant predictor variable in the multiple regression model.

analysis of residual:

The normality plot indicates an approximate linear relationship, thus it is impossible to reject the assumption of normally distributed residuals.

The next residual analysis examines the relationship between the residual and the dependent variable (both the observed and predicted value). But, this plot does not provide any clear relationship or any useful information. However, There is no relationship between the residuals and the predicted value, thus the model errors is stable over the range.

Problems with the model:

Before moving to the findings of this regression analysis it is highly recommended to the reader to the aware about some the problems/limitation of this analysis.

Specification Bias:

Other than the two independent variables included in this model, their may some other variables that are important predictors of CSI (costumer satisfaction Index) which has been omitted. Omitting them from this model causes Speciation bias may be the reason for the poor explaining power of this model.

Measurement Error:

The calculation of CSI and CII from a qualitative measure to a quantitative measure is somewhat controversial, as the dimensions of each variable are chosen arbitrarily and the weights to each variable is also given arbitrarily. Moreover, there may be some other omitted dimensions that are useful to measure both CSI and CII. Another important limitation is that both the CSI and CII has been measure in a way that they can take the maximum value 500 and the minimum 0. This also limits the acceptance of this analysis to some extent.

FINDINGS:

Hypothesis test of the coefficients of this model indicates that the independent variable CII is significant; where as another independent variable the employment status of employees is not statistically significant. The confidence interval analysis also supports the above findings. Therefore, the findings indicate that the level of customer satisfaction about the services provided by HSBC is positively related with the level of knowledge about its service package by its customers. The hypothesis of this regression analysis that “The customers who are more informed about the services tend to be more satisfied than those who are less informed.” proved by the above findings. However, another hypothesis that the employed customers (who are working in any organization) are more satisfied than the unemployed customers (students, housewife etc.) cannot be proved by our findings.

Conclusion:

Apart from all the outputs’ interpretation to put it shortly, I found numerous relationships between the variables I chose in my questionnaire. I do not want to repeat the findings again and hence conclude my paper as such that it has been a pleasure to have done this report under the supervision of Mr. M. Sayeed Alam. Even after all the obstacles and limitations, I am successful in carrying out the regression analysis. There is a lot I have learned from this report which otherwise would have been left unlearned. I truly thank the Faculty Advisor of my Internship, Mr. M. Sayeed Alam, Branch Manager,HSBC, Motijheel Branch, Mr. Md. Saidul Amin, Customer Services Manager Cristopher Raju Gomes and Shaikh Alam, Rifat Akhter, Richard Rudolf Gomes, Saiful Bari  Customer Service Department, HSBC, Motijhee Branchl, to have helped me out of their way. The completion of this report would not have been possible if not for them.