QUESTION
ANSWER
Executive Summary
Newcastle Bank has decided to conduct a research around the level of satisfaction and the instances of online frauds that its customers face. This is a major issue that the banks have to resolve for its customers and in addition to monetary loss these frauds lead to mental and social loss to the customers as well. Thus various statistical tests like the ttest, anova, least square regression etc. were applied by the research department of the bank to understand the difference of online frauds in demographics, the level of general satisfaction of its customers and the factors that affect that level of satisfaction.
It was found that the older female customers are most susceptible to these online frauds and thus the bank should dedicate extra resources in education this demographic against the possibility of these frauds.
The bank also realized that the level of satisfaction in the customers is generally low and thus is had to improve that. The easies, quickest and the most effective method for that will be increase the speed of bank’s response towards the complaints of the customers and the resolution of the frauds themselves. The report urges the banks to increase the staff strength to decrease the response time for the complaints filed for online frauds. The bank also needs to device methods to increase the speed of the actual resolution of the frauds to increase the level of satisfaction of the customers.
Introduction
Online fraud us a real concern for banks around the world. The issue of online frauds is even greater in the country of Australia and the banks there have been pondering of the possibilities of increasing their security measures to protect them. They also have to educate the general public and their clients about the possibility of frauds and how to evade them.
Despite all the measures from both the banks and general public, there is a remote possibility of a fraud and at times people have to face such issues. Thus, the banks also have to be active in their response to online frauds. There are two aspects to this, first the banks have to be vigilant and detect such frauds themselves. In most cases the detection should be quick enough to reverse the fraudulent transactions even before the thieves can make any material damage to the victim.
On the other hand the banks should be quick in response to the complaints by the customers. They have to not only help them recover the damages to the maximum aspect possible but they also have to provide social and mental comfort to the victims. In this regard the bank needs to understand the pattern in frauds and what can they do to mitigate the damage and at first prevent that itself. The banks also need to understand the level of customer satisfaction at various stages in the process. The overall satisfaction is just one aspect but the satisfaction at stages like help post fraud and information dissemination before fraud. The study below is aimed at the same.
Research Design
For the purpose of the research the bank has to collect the data form the clients of the bank around their level of satisfaction from the process at various stages if they experience a fraud and how they think the bank performs with respect to making them aware of the possibility of the frauds. An anonymous questionnaire is one of the simplest and the most suitable method to understand the implications. The questionnaire has to be short and comprehensive to keep the respondent from swaying away in the process and at the same time collect enough data that the bank is able to draw meaningful conclusions from the survey.
The questionnaire will have an initial set of basic questions which will be able to determine the demographics of the respondents. The questions will be around the gender and the age group of the respondents. The actual age is not material but the age bracket. One obvious question will be whether the respondent has actually experienced an online fraud or not.
The next section of the questionnaire is about the perception of the perception of the clients/customers of the bank about how the bank is serving them in terms of help against digital and online frauds. The first question will ask about the overall satisfaction of the customers towards Newcastle bank on its online security and the services they provide to help the customers. The next question will be the estimation of the satisfaction of the customers from the services that the bank provides for prevention of the frauds. The next three questions are directed towards the assistance that customers receive in case of an online fraud and the speed of complaint resolution. The last question is towards the satisfaction of the customers towards the social and mental support that the bank provides to its customers in case of a fraud.
The sampling of the customers in such a survey has to be from a complete dataset of the bank [1]. It has to be made ass random as possible and should entail the complete demographics of the customers of the bank. The number of invited participants should be more than double of the required respondents to account for people not filling the survey [2].
It must also be noted that no such information should be collected from the respondents that will even make it remotely possible to identify the person who filled the form [3]. This is the part where the anonymity of the survey is essential. Online fraud might be a very critical aspect for the people who have had to face the problem and thus a person who can access individual responses might take unfair advantage of the responses. In such a case the complete research will be highly unethical and invalid. The research and thus the data collection methods should solely be aimed at improving the security and response methods of Newcastle Bank and no other purpose.
Hypothesis Development
The research aims to answer multiple questions and for each question the following section will aim at developing a hypothesis. First the question will be presented that the research has to answer and following that the hypothesis will be developed. Following are the questions and hypothesis along with the same.

Is the number of online fraud experience same across the gender?
Null hypothesis (H_{0}): There is no difference between the number of online frauds among male and female customers.
Alternate hypothesis (H_{A}): There is difference between the number of online frauds among male and female customers.

Is the number of online fraud experience same across the age group?
Null hypothesis (H_{0}): There is no difference between the percentages of people experiencing online frauds among customers of different age groups.
Alternate hypothesis (H_{A}): There is difference between the percentages of people experiencing online frauds among customers of different age groups.

The bank believes that an overall customer satisfaction score below 6 in a 10point scale is an indication that customers are not satisfied with the online security system and services provided by the bank. Does the current level of overall customer satisfaction differ from management’s goal of 6 out of 10?
Null hypothesis (H_{0}): The mean overall satisfaction level of customers of bank is six.
Alternate hypothesis (H_{A}): The mean overall satisfaction level of customers of bank is not six.

Is there any difference between the level of customer satisfaction with regard to prefraud and postfraud (mental and social) support provided the bank?
Null hypothesis (H_{0}): There is no difference between the level of customer satisfaction with prefraud and postfraud support. by
Alternate hypothesis (H_{A}): There is difference between the level of customer satisfaction with prefraud and postfraud support.

Can customer satisfaction in relation to any of the following explain the overall customer satisfaction/dissatisfaction with regard to online security system:
Null hypothesis (H_{0}): level of advice is not a significant factor in explaining the overall satisfaction of the customers
Alternate hypothesis (H_{A}): level of advice is a significant factor in explaining the overall satisfaction of the customers

the level of assistance provided by online fraud resolution team in the case of an online fraud,
Null hypothesis (H_{0}): the level of assistance provided by online fraud resolution team is not a significant factor in explaining the overall satisfaction of the customers
Alternate hypothesis (H_{A}): the level of assistance provided by online fraud resolution team is a significant factor in explaining the overall satisfaction of the customers

quick response towards an online fraud complaint,
Null hypothesis (H_{0}): quick response towards an online fraud complaint is not a significant factor in explaining the overall satisfaction of the customers
Alternate hypothesis (H_{A}): quick response towards an online fraud complaint is a significant factor in explaining the overall satisfaction of the customers

quick resolution of the online fraud,
Null hypothesis (H_{0}): quick resolution of the online fraud is not a significant factor in explaining the overall satisfaction of the customers
Alternate hypothesis (H_{A}): quick resolution of the online fraud is a significant factor in explaining the overall satisfaction of the customers

the level of mental and social support provided to victims of fraud.
Null hypothesis (H_{0}): the level of mental and social support provided is not a significant factor in explaining the overall satisfaction of the customers
Alternate hypothesis (H_{A}): the level of mental and social support provided is a significant factor in explaining the overall satisfaction of the customers
Statistical Technique and Justification
A total of five hypotheses have to be tested in the research project. As the aim of the tested hypothesis in each case is different thus the associated test will also be different and subsequently the rationale. In the following section we present the statistical techniques that will be used and the reason for using those techniques.
To test hypothesis one, we will use the independent sample ttest [4]. This is an appropriate choice of test as we have to compare the means of the number of online frauds between male and female respondents. The numbers of online frauds are represented by a ratio scale variable whereas the control variable will be the variable identifying gender which is nominal.
To test hypothesis two, we will be using Anova as the choice of test [5]. The test is appropriate as the number of online frauds experienced has to be divided into multiple groups based on the age group of the respondents. We will have to test the difference of the means in the different groups. The control variable identifying the age group of the respondent is an interval scale variable whereas the actual number of online frauds is represented by a ratio scale variable.
Hypothesis three will be tested using the ttest [6]. We have to test the mean of the overall satisfaction of the customers against the benchmark of 6 that the Newcastle Bank has set for its respondents which it desires to achieve. As the mean is to be tested against a benchmark, ttest will be appropriate. The overall satisfaction is captures by an interval scale variable. As we have not tested the distribution of the variable itself thus, the more lenient ttest has to be used. In case of a normal distribution of responses the test could have been replaced with the more suitable ztest which gives better and more concrete results.
To test hypothesis four, we will have to perform a paired sample ttest [7] to identify whether there is a different in the level of satisfaction between prefraud and postfraud response of the bank. We have to test the difference in men of the same sample in two different scenarios and thus a paired sample ttest will have to be used. Both the variables are interval scale variables.
To test hypothesis five, we will have to use least square regression [8] or OLS. Hypothesis 5 has been further split into five subhypothesis which will have to be tested. In our case the overall satisfaction of the respondent is the dependent variable for which we have to identify the variables that have an impact on the dependent variable. All the variables included for the test of the hypothesis including the overall satisfaction of the respondents are interval scale variables. A single regression analysis will be able to identify all the variables which are influencing factors for the satisfaction level. We will select the variables which have a significant impact on the dependent variable and reject the others.
Results, and Statistical and nonstatistical interpretation
Hypothesis 1
tTest: TwoSample Assuming Unequal Variances 


Female 
Male 
Mean 
4.35468 
3.218274 
Variance 
4.131005 
3.212317 
Observations 
203 
197 
Hypothesized Mean Difference 
0 

df 
394 

t Stat 
5.935552 

P(T<=t) onetail 
3.21E09 

t Critical onetail 
1.64873 

P(T<=t) twotail 
6.43E09 

t Critical twotail 
1.966003 

The results of the independent sample ttest are displayed in the table above. The calculated tstatistics is higher than the two tailed critical value. The p value is also much less than the significance level of 0.05. Thus we have enough evidence to reject the null hypothesis. Thus we accept the alternate hypothesis that there is a significant difference between the number of frauds than male and female respondents experience.
From the test we can say that female respondents experience more online frauds than the male respondents. Thus, females are more susceptible to online frauds than males.
Hypothesis 2
Anova: Single Factor 

SUMMARY 

Groups 
Count 
Sum 
Average 
Variance 

Age group 1 
101 
361 
3.574257 
3.306931 

Age group 2 
76 
288 
3.789474 
3.741754 

Age group 3 
81 
271 
3.345679 
3.579012 

Age group 4 
66 
281 
4.257576 
5.024942 

Age group 5 
76 
317 
4.171053 
4.303684 

ANOVA 

Source of Variation 
SS 
df 
MS 
F 
Pvalue 
F crit 
Between Groups 
46.14684 
4 
11.53671 
2.945619 
0.020244 
2.394533 
Within Groups 
1547.043 
395 
3.916565 

Total 
1593.19 
399 




The table above shows the results for single factor anova. The pvalue is lower than the significance level of 0.05 thus we have sufficient evidence to reject the null hypothesis and accept the alternate hypothesis that says that there is a significant difference between the number of online frauds experienced among people of different ages.
Based on the test we can say that older people experience more online frauds when compared to younger people.
Hypothesis 3
Mean 
3.68 
Hypothesized Mean 
6 
Standard Deviation 
1.905697 
Standard Error 
0.095285 
tvalue 
24.34805 
t Critical onetail 
1.64873 
t Critical twotail 
1.966003 
The table above clearly shows that the tvalue calculated is much higher than the critical tvalue. Thus we have enough evidence to reject the null hypothesis and we accept the alternate hypothesis that there is a difference between the actual mean of the overall satisfaction and the bank’s benchmark of 6.
Based on the test we can see that the actual satisfaction level of the customers is much below 6. The mean satisfaction of the sample is just 3.68 and thus the bank needs to work hard on improving the satisfaction of its customers.
Hypothesis 4
tTest: Paired Two Sample for Means 


Question 5 
Question 9 
Mean 
4.55 
4.405 
Variance 
3.892231 
2.988446 
Observations 
400 
400 
Pearson Correlation 
0.135141 

Hypothesized Mean Difference 
0 

df 
399 

t Stat 
1.187999 

P(T<=t) onetail 
0.11777 

t Critical onetail 
1.648682 

P(T<=t) twotail 
0.235541 

t Critical twotail 
1.965927 

We can see that based on the table above there is no significant evidence to reject the null hypothesis. Thus there is no difference between the level of satisfaction of customers between prefraud and postfraud response of bank.
The customers are not highly impressed by the response of the bank compared to what they do for education them regarding online frauds. If the bank has the policy of equal concentration on both the groups then it is doing fine otherwise it might want ot improve concentration on one of the aspects.
Hypothesis 5
SUMMARY OUTPUT 

Regression Statistics 

Multiple R 
0.789344038 

R Square 
0.62306401 

Adjusted R Square 
0.618280559 

Standard Error 
1.177405009 

Observations 
400 

ANOVA 


df 
SS 
MS 
F 
Significance F 
Regression 
5 
902.8447 
180.5689 
130.2541 
3.47E81 
Residual 
394 
546.1953 
1.386283 

Total 
399 
1449.04 




Coefficients 
Standard Error 
t Stat 
Pvalue 

Intercept 
1.370692819 
0.263823 
5.19551 
3.28E07 

Question 5 
0.197644735 
0.03701 
5.340324 
1.57E07 

Question 6 
0.005660186 
0.058193 
0.09727 
0.922565 

Question 7 
0.377086426 
0.036732 
10.2659 
4.55E22 

Question 8 
0.548858677 
0.04105 
13.37062 
6.99E34 

Question 9 
0.120825923 
0.038765 
3.116877 
0.001962 
Based on the table above we can reject the null hypothesis for four of the five variables but we have to accept the null hypothesis for satisfaction with the level of assistance provided by online fraud resolution team in the case of an online fraud. We also notice an intercept which is significant.
Based on the analysis we can say that there is a general level of dissatisfaction in the customers. The other relevant factors lead to the satisfaction rising. The speed of resolution seems to impact the overall satisfaction the most.
Analysis and Summary of the statistical results
The analysis conducted on the response of the respondents has resulted in concrete and actionable results. The first result that the analysis showed was that there is a clear difference in the amount of frauds that different demographics experience. When looked according to the gender, female customers are more susceptible to online frauds than male respondents. Similarly, when the candidates were segregated by age groups we saw that the older candidates are more in danger of online frauds than younger candidates. When we club both the results then is can be said that older female customers of the banks might be the prime target of such frauds.
The analysis was also conducted on the overall level of satisfaction of the customers of Newcastle Bank and it was found that the level of satisfaction of the customers is ver low when compared with the target satisfaction level of 6 on a ten point scale. The mean satisfaction level of the respondents was a mere 3.68 which shows that there is a large scope of improvement. It was also found that the level of satisfaction of the clients before the frauds and after the frauds did not change much. On one hand it is good that the satisfaction level has not decreased but at the same time the level itself is low and thus needs to be worked upon.
The research further tried to analyze the factors which are responsible for leading to the overall satisfaction level. In other words what aspects of the bank’s process, prefraud and postfraud, are a significant factor in impacting the overall level of satisfaction. It was found that in the absence of all the factors the actual level of satisfaction is negative.
It was found that the level of assistance by the online fraud resolution team was not a significant factor whereas the other like prefraud response and the postfraud moral support were. It was also found that the quick resolution of an online fraud was the most impacting factor, followed closely by the quick response to the complaint of such frauds. Overall we can say that the speed of response to the complaint and frauds is the most important to the clients. Thus, these should be the factors that the bank should concentrate on to have a better satisfaction level among the customers. This will also show the best results with minimum investments.
Recommendations
The prime motive of the bank should be prevention of such frauds the level of satisfaction is immaterial if the bank can prevent the frauds altogether. The analysis has showed that female and older customers of the banks are more susceptible to the online frauds. Thus, the bank should have special seminars and educations sessions for the customers in this demographic. There can be additional checks in place for old female customers in case of online transactions which can help reduce online frauds.
As the level of satisfaction for the customers is very low thus, the bank should also focus on improving that. The speed of response to both the fraud itself and the response to the complaint of the fraud will be the main factors that can help the bank improve the level of satisfaction. The bank should begin by increasing the speed in both the cases. The staff size can be increased to increase the speed of response to such complaints. The actual resolution of the frauds is more dependent on the technical capability of the bank which it will have to improve to detect and reverse fraudulent transactions at a greater pace.
REFERENCES

Yates, Daniel S.; David S. Moore; Daren S. Starnes (2008). The Practice of Statistics, 3rd Ed. Freeman. ISBN 9780716773092

Meng, Xiangrui (2013). “Scalable Simple Random Sampling and Stratified Sampling”. Proceedings of the 30th International Conference on Machine Learning (ICML13): 531–539

Peter, Bacchetti, Leslie, E. Wolf, Mark, R. Segal, Charles, E. McCulloch (2005). “Ethics and Sample Size”. American Journal of Epidemiology, Volume 161, Issue 2, Pages 105–110

Welch, B. L. (1947). “The generalization of “Student’s” problem when several different population variances are involved”. Biometrika. 34 (1–2): 28–35

Anscombe, F. J. (1948). “The Validity of Comparative Experiments”. Journal of the Royal Statistical Society. Series A (General). 111 (3): 181–211

Mankiewicz, Richard (2004). The Story of Mathematics (Paperback ed.). Princeton, NJ: Princeton University Press. p. 158. ISBN 9780691120461

Derrick, B; Toher, D; White, P (2017). “How to compare the means of two samples that include paired observations and independent observations: A companion to Derrick, Russ, Toher and White (2017)”. The Quantitative Methods for Pschology. 13 (2): 120–126.

Kenney, J. F. and Keeping, E. S. (1962) “Linear Regression and Correlation.” Ch. 15 in Mathematics of Statistics, Pt. 1, 3rd ed. Princeton, NJ: Van Nostrand, pp. 252–285
Looking for Research Assignment Help. Whatsapp us at +16469488918 or chat with our chat representative showing on lower right corner or order from here. You can also take help from our Live Assignment helper for any exam or live assignment related assistance.