Hypothsis
H1: Perceive usefulness will have a significant positive influence on customers’ acceptance
H2: Perceive ease of use and service quality will have a significant positive influence on customers’ ability to accept e-banking.
H3: Perceive trust will positively influence customers’ acceptance.
H4: Perceived security and Privacy has a positive influence on customers’ ability to accept e-banking.
H5: Perceive reliability will positively influence customers’ acceptance.
H6: Perceive accessibility influences customers’ acceptance positively
H7: Convenience will positively influence customers’ acceptance
H8: customers’ intention to adopt e-banking will be positively influenced by customers’ satisfaction.
Chapter 4: Data Analysis and Discussion of Findings
4.1 Introduction
The previous chapters outlined the methodology used for this study. The questionnaire was developed and based on the conceptual framework; this chapter addresses in details the statistical procedure and presents the results. This chapter will focus on the data analysis and also detailed discussion of the findings.
According to Coorley (1978), the main goal of “the statistical techniques are to assist in establishing the plausibility of the theoretical model and to estimate the extent to which the various explanatory factors seem to be influencing the dependent variable” (p.13). The primary purpose of this research study was to identify and investigate the factors that enables user accept online banking information systems. In order to achieve these objectives, this thesis used two different statistical software tools. Statistical Package for Social Sciences (SPSS) was used for analysing the preliminary data, explained in the following sub-section. The Analysis Moment of Structures Software (AMOS) for Structural Equation Modeling (SEM) was used for measurement model analysis and structural model to test the proposed hypothesised model.
Statistical Package for Social Science will be used to analyse the quantitative data obtained from this survey questionnaire. This a wildly used software package and a highly acceptable one used by many researchers in different disciplines including social sciences, business studies and information system research (Zikmund, 2003). This software will be used in this research for date coding, missing data treatment, testing and identification of data normality. In addition, SPSS will be used to perform descriptive statistics such as frequencies, percentages, mean values, and standard deviations. Therefore, this tool has been used to screen the data of this research study in terms of data coding, treatment of missing data (i.e., using ANOVA), identification of outliers (i.e., Mahalanobis Distance (D2)) test and find out the data normality (i.e. using kurtosis and skewness statistics).
Each one of these techniques are explained and discussed in the following sections. In addition, SPSS was also applied to perform descriptive statistics such as frequencies, percentages, mean values, and standard deviations. These analyses were performed for each variable separately and to summarise the demographic profile of the respondents in order to get preliminary information and the feel of the data (Sekaran, 2000). Furthermore, before applying SEM, SPSS was used to conduct exploratory factor analysis (EFA) for the first stage of data analysis to summarise information from many variables in the proposed research model into a smaller number of factors, which is known as factor / dimension reduction (Hair et al., 2006). Data collection in this quantitative survey mainly used nominal and ordinal scales, which would return data in a form appropriate for this technique (Kline, 2005).
4.2 Data preparation and collection process
The collection process took longer than expected because it was not easy to find enough respondents to participates in thesame place at a said time. Researcher had to visit banks several times to find enough participants. This resulted in obtaining only 302 completed questionnaire forms out of the 350 distributed. Each collected form was reviewed for completeness necessary to the analysis. After data cleaning and screening a total of 282 of the completed forms were found useable for analysis, resulting in 80.6% response rate. The following section discusses pre-analysis data processing.
Pre-analysis Data Processing
After completion of data collection, it was very important to have them examined through conversion into a form suitable for data analysis to ensure their integrity, significance, accuracy and representability.
Data Coding
Coding refers to “the process of assigning numerals or other symbols to answers so that responses can be put into a limited number of categories or classes” (Kothari, 2004, p.123). This means that each category of answers in the questionnaire will be allocated a specific code that will help the researcher transfer it into a form identifiable by computer and make subsequent analysis easier (Saunders et al. 2012). In this study, the continuous response scale (questions 13-24) used a pre-coded technique by allocating numbers for each question, with No. 1 meaning ‘strongly disagree’ and No. 5 ‘strongly agree’, which facilitated respondents task. The questions 1-12 and 25-30 were entered into the coding scheme prior to being entered into the computer software. The collected data were entered into SPSS and the codes were labelled for each variable as to illustrate the meaning of codes.
Data cleaning and filtering
Data cleaning and filtering prior to any further statistical analysis in this study will be to ensure that the input data is free of coding errors or any loss of data or any inappropriate responses . This process is very important to ensure that the input data includes only the exact value is used to check bulk theory is essential. Descriptive statistics and frequency tables using SPSS to identify the range of values and inconsistent responses ( Sanders et al. , 2012). The missing data must be taken into account to decide how to deal with it. According to Dong Peng and (2013) Lost data can be on two levels: unit level and project level. Unit level means that all who do not take or reject the respondents, and the project level are those who return the survey is not a complete answer. Item Level occurred for two main reasons. First, the defendant may not be the lack of information in the case, reluctant to answer a number of “sensitive” issues or lack of answer some questions to answer part of the questionnaire (S). Second, the respondent may not have time to complete the answer Questionnaire (Sanders et al., 2012). . In addition, three modes Sanders et al. (2012) defined missingness: The Missing completely random (MCAR), missing at random (MAR), and no missing random (NMAR). MCAR missing values occur when the variable is not associated with the variable itself or any other variables of interest. As MAR, it occurs in variable missing values are not the variable itself, but related to other variables. In NMAR, the value of the variable value of the missing variable itself as well as with other variables. It is therefore necessary for this study to address the problem of data loss in order to avoid down the wrong result, damage to internal validity, resulting in loss of statistical power and external invalidity when research results are to be generalized. There are different approaches to address the missing data such as list-wise deletion, pair-wise deletion, mean substitution, estimation of conditional means, imputation using the expectation maximization algorithm (EM), multiple imputation and regression-based imputation (Dong and Peng 2013). In this study, the percentage of missing data was identified before conducting further statistical inferences. Out of the 302 responses, 20 had missing data ranging between 0.05% and 18% of the survey. In average, this accounts for approximately 11% of all responses. Excluding such forms was considered inappropriate for this research because it reduces the samples size, which in turn affects the generalizability of data findings. Although, there was no agreement in related literature about the acceptable percentage of missing data, many studies agree that 10% is considered acceptable (Bennett, 2001; Schlomer et al., 2010). Therefore, 6 forms were excluded for exceeding the 10% of missing data while 14 were retained due to not exceeding that percentage.
Missing data
Missing data is a very common problem in all type of survey research because it usually involves a large number of samples (Bryman and Cramer, 2005). Hair et al. (2006) note that missing data causes two main problems: (a) it minimises the ability of statistical test to imply a relationship in the data set, and (b) it creates biased parameter estimates. The potential effect of missing data depends on the frequency of occurrence, the pattern of missing observations, and the reasons for the missing value (Tabachnick and Fidell, 2001). Hair et al. (2006) point out that if the pattern of missing data is systematic (i.e. non-ignorable or is not missing at random), any technique used to treat this missing data could possibly generate biased results whereas, if the missing data is scattered in a random fashion with no distinct pattern (i.e. missing completely at random = MCAR), any remedy to treat this problem is assumed to yield acceptable results.
Although there are no clear set guidelines regarding what constitutes a large amount of missing data: Kline (1998, p. 75) suggested that missing values should probably constitute less than 10% of the total data. According to Cohen and Cohen (1983), 5% or even 10% of missing data on a particular variable is not large. Olinsky et al. (2013) point out that if the percentage of cases with missing observations is less than approximately 5%, and the pattern is ignorable, most simple analyses should yield reliable results.
This study followed steps suggested by Byrne (2001) for dealing with incomplete (missing) data, which were: (1) Investigation of the total amount of missing data, (2) Investigation of the pattern of missing data, (3) and finding out appropriate techniques to deal with missing data.
Assessing Non-response Bias
As discussed in Chapter five, the non-response bias is important to be addressed especially that the response rate in this study was 80.6%. This bias occurs when respondents in the sample refuse to participate in the survey due to certain characteristics they may have. The existence of non-response bias is prone to result in a major problem in the study because it would generate bias in the sample, which undermines its validity and quality (Linder et al., 2001; Ygge and Arnetz, 2004).
Non-response bias was evaluated by comparing the responses of early and late respondents. Lindner et al. (2001) suggested that the early and late comparison respondents’ is the most widely useful method in quantitative research to identify non- response bias. They argue that if there are no significant differences between early and late respondents, the study results can be generalized to the population. This study considered the first 40 responses as early responses because they responded fast without any further efforts by the researcher, while the last 40 responses are considered late responses due to efforts exerted to obtain them. Independent t-test was used to compare early and late respondents. This result showed that (p>0.05) in all variables, which indicates that there were no significant differences between early and late respondents.
4.3 Outliers
Kline (2005) and Hair et al. (2006) described outliers as cases with scores that are distinctively different from rest of the observations in a dataset. Researchers have warned that problematic outliers can have dramatic effects on the statistical analysis such as model fit estimates and parameter estimates and they can create a negative variance (Finney and DiStefano, 2006). There are two main types of outliers i.e. univariate and multivariate outliers. A univariate outlier is the case that has an extreme value on one variable whereas a multivariate outlier is a case with an unusual combination of values on two or more variables (Tabachnick and Fidell, 2001; Kline 2005). Although, there is no absolute judgment of an extreme value, a commonly accepted rule of thumb is that scores more than three standard deviation away from the mean may be considered as outliers (Kline, 2005). The univariate outlier can be detected easily by diagnosing frequency distributions of Z-scores (Kline, 2005).
Tabachnick and Fidell (2013) stated that univariate and multivariate outliers could be present among dichotomous and continuous variables. In this study, univariate outliers were not identified because the study utilized a Likert scale with 5 categories ranging from 1 – strongly disagree to 5- strongly agree. However, if respondents answered strongly disagree or strongly agree; these response options might become outliers, as they are the extreme points of the scale.
Presence of multivariate outliers in data can be checked by Mahalanobis distance (D2) test, which is a measure of distance in standard deviation units between each observation compared with the mean of all observations (Byrne, 2001; Kline, 2005; Hair et al., 2006). A large D2 identifies the case as an extreme value on one or more variables. A very conservative statistical significance test such as p < 0.001 is recommended to be used with D2 measure (Kline, 2005; Hair et al., 2006). In this research study, researcher measured Mahalanobis distance using SPSS version 16.0 and then compared the critical χ2 value with the degrees of freedom (df) equal to number of independent variables and the probability of p < 0.001.
4.4 Normality Test
Normality is defined as the “shape of the data distribution or an individual metric variable and its correspondence to the normal distribution, which is the benchmark for statistical methods” (Hair et al., 2006; p. 79). Violation of normality might affect the estimation process or the interpretation of results especially in SEM analysis. For instance, it may increase the chi-square value and may possibly cause underestimation of fit indices and standard errors of parameter estimates (Hair et al., 2006). One approach to diagnose normality is through visual check or by graphical analyses such as the histogram and normal probability plot that compare the observed data values with a distribution approximating the normal distribution. If the observed data distribution largely follows the diagonal lines then the distribution is considered as normal (Hair et al., 2006).
Beside the shape of distribution, normality can also be inspected by two multivariate indexes i.e. skewness and kurtosis. According to Tabachnick and Fidell (2013) skewness refers to the symmetry of distribution while kurtosis refers to the peakedness of distribution. The skewness portrays the symmetry of distribution whereas the kurtosis refers to the measure of the heaviness of the tails in a distribution (also known as peakedness or flatness of the distribution) compared with the normal distribution. In normal distribution, the scores of skewness and kurtosis are zero. Hair et al (2006) point out that skewness scores outside the -1 to +1 range demonstrate substantially skewed distribution. However, Kline (2005) suggest that values of the skew index greater than three (3.0) are indicated as extremely skewed and score of the kurtosis index from about 8.0 to over 20.0 describe extreme kurtosis. In this study, the researcher set the maximum acceptable limit of observation values up to ±1 for the skewness and up to ±3 for the kurtosis. Thereafter, the researcher used factor analyses and structural equation modeling for inferential statistical analyses.
4.5 Factor Analysis
Factor analysis (FA) techniques are used to address the problem of analysing the structure of the correlations among a large number of measurement items (also known as variables) by defining a large set of common underlying dimensions, known as factors. FA (Factor Analysis) takes a large set of variables and summarises or reduces them using a smaller set of variables or components (factors) (Hair et al., 2006). The main purposes of the FA therefore include: (a) understanding the structure of a set of variables, (b) constructing a questionnaire to measure any underlying variables, and (c) reducing a data set to a more manageable level (Field, 2006, p.619). Therefore, at first, the researcher identifies latent dimensions of the structure of the data and then determines the degree to which a test item (variable) is explained by each factor. This is then followed by the primary uses of FA: summarisation and data reduction (Hair et al., 1995). This purpose can be achieved by either exploratory factor analysis or confirmatory factor analysis techniques. However, the exploratory factor analysis technique is used for “take what the data give you”; whereas the confirmatory factor analysis technique involves combining variables together on a factor or the precise set of factors for testing hypotheses (Hair et al., 2006, p.105).
In this research study, the researcher first conducted exploratory factor analysis (EFA) to examine the dimensions of each construct (herein called as a factor) and then confirmatory factor analysis (CFA) was performed for testing and confirming relationships between the observed variables under each hypothesised construct (Zikmund, 2003; Hair et al., 2006). The Next section explains exploratory factor analysis performed by using SPSS version 16.0.
Reliability and Validity Analysis
Reliability and validity are important concept in research and should be measured to ensure that the instruments in the survey are valid and reliable which leads to a better quality data. The following sections show in details the measurement of these two concepts.
Initial Reliability Assessment
Reliability refers to the stability of measurement instrument through time. In the current study, the constructs in the survey were measured by multiple item scale. Therefore, internal consistency was used to measure the reliability of this study through measuring correlations between items within a scale of a given construct. Cronbach’s alpha was used to calculate the internal reliability or homogeneity formed of a multiple items scale (Creswell, 2012). Cronbach’s alpha value ranges between 0 and 1, where coefficient alpha is closer to 1, being the greater degree of items’ reliability. However, there has been no agreement among researchers on an acceptable cut-off value for reliability. Many considered that value 0.7 or above highly acceptable (Pallant, 2007; Field, 2009) while some have confirmed the value of 0.6 as fair (Moss et al., 1998;Yong et al., 2007) and others argued that a value above 0.5 is poor but acceptable (Nunnally, 1978; Bowling,1997). George and Mallery (2003, P.231) presented a rule of thumb for Cronbach’s alpha categorizing reliability values, as shown in the Table
Cronbach’s Alpha Internal Consistency
0.9 ≥ α Excellent
0.8 ≤ α< 0.9 Good
0.7 ≤ α< 0.8 Acceptable
0.6 ≤ α< 0.7 Questionable
0.5 ≤ α< 0.6 Poor
α
Table: Rule of thumb for Cronbach’s alpha
4.5.1 Findings from bank managers’ interviews
Respond Rate
Interview was carried out with 2 managers of each of the ten selected banks. A total of 20 managers were supposed to be interviewed; however, only 14 managers were available for interview (70%). Interview lasted 10-15 minutes on average. Responses collected are shown below.
Banks Responses
Ecobank 2
Afriland First Bank 1
SGBC 2
Standard Chartered 1
UBA 2
UBC 1
Atlantic Bank 1
BICEC 2
CBC 1
CitiBank 1
Total 14
Table 4.1: Number of managers interviewed in each bank
Demographic Characteristics
Demographic Characteristics Intervals Frequency Percentage (%)
Gender Male
Female 8
6 57
43
Age 60 0
2
6
4
1
1 0
14
43
29
7
7
Education High school
Diploma
Bachelors
Postgraduate 0
0
4
5
5 0
0
28
36
36
Occupation Director
Manager 5
9 36
64
Table 4.2: Demographic characteristics of respondents
From table 4.2, it is clear that 57% of the interviewees were male while 43% were female. Most of them were between ages 31 to 40, managers (64%) and had either a bachelor’s degree or were postgraduates hence, university-level degrees. This shows that males are most dominant in the management field.
This shows that middle age people are highly educated and have idea on the modern developments of the society. One of the questions related to continence and wiliness towards implementing e-banking strategy in their banks proved that most financial activities take place in the main cities. It showed that banks preferred to open up more branches and do businesses in Yaoundé and Douala, which are two of the busiest, developed, and financial cities of the country. Another aspect that was noted from these interviews was the fact that most of their institutions already make use of e-banking services and are still implementing other e-banking services that meets up with customer demand. Reasons stated for implementing these services included cost reduction, quality improvement, efficiency, competition with other banks providing same services, and last but not least, to attract customers. It was also noted that not much have been done since the implementation of these services but anxious to know factors that will assist in making decisions towards implementing better services.
Some factors were noted that happened to be negative and hinder the success of implementing better electronic banking services in Cameroon. However, all these factors did not hold same level of severity while others were more fundamental than others. A breakdown of the findings from the interviews will be shown below.
4.5.2 Managerial and organizational factors
From the interviews with bank managers, some important facts were noted and will be shown below.
Lack of strategic plan
This is one of the most important factors that serve as a hindrance to better electronic banking services in Cameroon. Interviewee stated this as a hindrance and explained that Cameroon has potentials to benefit from modern electronic banking services but there is a lack of clear national strategy that slows down the technology adoption rates of banks and the country as a whole. Banks managers and directors stated that they do not have enough facilities for better e-banking technology adoption or national plan for their banks because senior officials of the government have not yet realised the value of e baking technology therefore, this is not made a government priority. However, a few of the interviewees stated that the government is just now realising the importance of national strategy to the country’s finances and is making efforts towards providing better support and protection to both financial institutions as well as the people of Cameroon. The government is also helping by supporting hardware and software importation into the country and making affordable to banks that cannot afford these facilities. They predict that with help from the government, a lot can be done to better e-banking services that will be able to attract more people to accept and make use of these services.
Lack of E-laws and Legislation
There is a lack of strong e- laws to protect both banks and Cameroonians. Some interviewees stated their fear of using information technology and the Internet as a whole even though they are knowledgeable enough to use these systems. There are not enough laws to protect people using the Internet in Cameroon therefore; it is left for individuals to take care of their own safety. And because many are not sure hoe to protect themselves from these services, they become reluctant to accept an make use of e-banking services.
Resistance to change
This was another important factor worth noting. Bank managers and directors who were interviewed indicated that not only are people reluctant to accept change but also are in denial of accepting new change. Some even stated that some staffs were not happy with the adoption by their banks to provide people with electronic banking services as many people are already use to the traditional method of banking or cash handling. A conclusion of what was drawn from these interviews was a reluctance of acceptance of electronic banking as it will affect the normal performance and activities of the banks and most of them were scared of loosing customers due to these changes.
Lack of IT knowledge and awareness
It was reported there was a lack of IT knowledge. Most people have limited knowledge of computer uses, manipulation as well as low level of awareness. The work force found it had to manipulate the new technology and many difficulties were faced and delays in transactions. One banks manager mentioned training a few of their staffs in information technology so that there was a better service provision. This is one of the factors that affect banks from carrying out a better electronic banking service provision. It was noted that IT knowledge is one of he factors being considered for staff recruitment. One manager stated, “We have made IT knowledge, internet banking awareness out number one priority when looking for new recruits”.
Expensive and Shortage of IT training courses
This was stated as one of the factors that affect banks. Most interviewees mentioned how short and inefficient the IT courses they took were and how expensive it was for them. Therefore most people still lack in IT knowledge even after taking IT courses and others shying away from these courses due to the cost. One manager explained how disgusted it was with the cost. “ It was crazy how much money I paid to take up a course that lasted me three months. What I learned at the end was something I could learn at home in the space of a week. Sometimes I even had to share computers with a classmate because there was not enough for every individual”. A manager spoke out with great emotion about the cost of taking up online banking. “I some times have pity for these people who make us of our services”. When the researcher asked why he said that, he added “some people come in because they have been informed on how secured banking online would be for them but then to start up an online banking account, they need to pay an upfront fee of 50.000FCFA. This is a lot of money and you will be surprised to learn some people do not make up to that amount a month and to others that is their monthly salary”.
4.6 Findings from Customers’ Questionnaires
Respond Rate
This research was aimed at collecting 30 responses from each of the ten selected banks. However, only 282 out of 300 questionnaires that were supposed to be distributed and this represented 94% of the total. The responses are shown in the table below.
Banks Responses
Ecobank 45
Afriland First Bank 31
SGBC 29
Standard Chartered 21
UBA 37
UBC 13
Atlantic Bank 30
BICEC 42
CBC 15
CitiBank 19
Total 282
Table 4.3: Responses from the banks
Demographic characteristics Elements Intervals Percentage (%) Users Percentage (%) Non-users Percentage (%)
Gender Male
Female 179
103 64
36 95
66 53
64 84
37 47
63
Age 60 14
108
87
45
22
06 5
38
31
16
8
2 6
57
51
19
9
1 42
53
59
42
41
17 8
51
36
26
13
5 58
47
41
58
59
83
Education High school
Diploma
Bachelors
Post graduate 5
9
74
98
96 2
3
26
35
34
2
5
48
61
72 40
56
65
62
75 3
4
26
37
24 60
44
35
38
25
Occupation Student
Gov’t employee
Private employee
Business 68
87
49
78 24
31
17
28 23
50
21
49 35
57
43
63 45
37
28
29 65
43
57
37
Income (CFA Frs) <100,000 100,000-200,000 200,001-30,000 300,001-40,000 400,001-50,000 >50,000 5
18
39
52
64
104 2
6
14
18
23
37 1
6
13
23
37
57 14
33
33
44
58
57 6
12
26
29
27
45
86
67
67
56
42
43
Descriptive analysis of demographic characteristics
Table 4.4: details of participants with respect to gender, income, occupation, education and age.
Gender
The figure below shows respondents by gender. It shows that 63% of respondents were male while 37% were female. The number of Cameroonian male that makes use of banking services in Cameroon is almost double the number of female users.
Figure 4.1: Percentage by gender of respondents
The following figure shows the parentage of gender of non-users and users of online banking in Cameroon.
Figure 4.2: Gender of users and non-users
Based on the figure above, the number of female users is more than the number of electronic banking users. We noticed previously that the number of male using banking services in general is more than the number of female users however, even though there are fewer females using banking services, more of them do make use of electronic banking.
Demographic
Characteristics Users Non-users
Gender
Male
Female
59
41
69
31
Age
60
4
40
36
13
6
1
6
37
26
19
9
3
Education
High school
Diploma
Bachelors Degree
Post Graduate
1
3
26
32
38
3
4
28
39
26
Occupation
Student
Government worker
Private worker
Business
16
35
15
34
32
27
20
21
Income (CFAFrs)
500.001
1
4
9
17
27
42
4
8
18
20
19
31
Table 4.3.1: Percentages of respondents according to users and non-users by demographic characteristics
Age
Figure 4.3: Ages of respondents.
The Pie chart above shows that 38% of respondents were between the ages of 20-30 years old, while there was only 2% respondents from customers of age above 60 years. This therefore draws us to the conclusion that the majority of bank customers are young. This is clear as the younger generation are beginning to accept and embrace these new methods of banking.
Figure 4.4: Respondents of users and non-users of electronic banking according to age.
Figure 4.4 above shows the percentage of non-users of electronic banking is greatest amongst people who are 60 years and older. There is a slight difference between non-users of respondents between the ages 41 and 50 and 51 and 60, the greatest number of electronic banking falls between the ages 31 and 40. This therefore proves that older people avoid adopting electronic banking services and will rather stick to the traditional methods of banking of which they use at their convenience.
Level of Education
Figure 4.5: Shows respondents according to level of education.
The figure above shows that 35% f respondents have a Bachelors degree while 34% of the respondents have a Diploma. The least number of respondents were a holder of certificates lower than high school level. Therefore, most of the users of electronic banking facilities are holders of university degrees or its equivalents.
Figure 4.5: Shows the responses of users and non-users of electronic banking with respect to level of education.
From figure 4.5 above, it is clear that electronic banking users are higher amongst the respondents who are postgraduate (38%) degree holders followed by respondents with a bachelor degree (32%) and then those with diplomas (26%). This shows that there is a positive relationship between the level of education and the user of electronic banking. Therefore, the higher the level of education attained, the greater the probability of the customers adopting Internet banking (Mohammed et al., 2009).
Occupation of Respondents
Figure4.6: The percentages of respondents by their occupations.
From the above figure, it is clear that most respondents are government employees (31%). The difference between government employees and people with businesses (28%) was not so great. Only 17% of respondents were privately employed while 24% were made up of students. This ties with Karjaluoto (2002) who observed that online services are mostly used by well educated and people who have better occupations than non-users.
Distribution of users and non-users according to occupation
Figure4.7: Occupation of users and non-users
It can be noticed from the above figure that the highest number of non-users of electronic banking services fall amongst students while respondents with businesses make use of electronic banking than any other occupation.
Income
Figure 4.8: show the income of respondents.
Figure 4.8 shows that 37% of respondents are earners of 500,001CFA FRS and above followed by earners of 400,001-500,000 CFA FRS (23%). The least number of respondents were had monthly incomes lower than 100,000CFA FRS (2%). This ties with the manager’s comment earlier that stated how expensive it is to start up an online baking account. Because wanting to start up this service for people who earn 100.000CFAF means giving away half of their monthly earnings. This explains why acceptance for this group of people is less.
Distribution of Users and Non-users of e banking by income
Figure 4.9: Income disparity between users and non-users
The above figure illustrates the income disparity between users and non-users and indicates that income levels seems to be the major factor affecting people from taking up electronic banking. It clearly shows that the people who earn more, in the case above, 42% of users earn 500,001CFA FRS an above followed by earners of 400,001-500,000 CFA FRS. Hence, respondents who earned less fell amongst the non-users of electronic banking. Therefore, the higher the income level, the more affluent the people and the more likely they are to possess a personal computer, thus to use electronic banking (Mohammed et al., 2009). It was noticed that only 9% of respondents used online banking while 18% do not use online banking but both groups earn 200.001-300.000 CFA FRS. Generally, less people make use of online banking even thought they earn more money except those who earn 500.001CFA FRS and above. This is clearly understandable for this group of people for that is a lot of money and in this case, online banking is a suitable way to save their earnings.
Internet Usage
Participants were also asked if they use the Internet or not. The following table shows the number of respondents who said yes and those who said no to using the Internet.
Respondents Yes
161 No
121
Percentage (%) 57 43
Table 4.5: Percentages of users use of the Internet
From table 4.5, we see that 57% of respondents make use of the Internet whether at home or in cyber cafés while 42% do not use the Internet. This shows that the number of Cameroonians that are Internet users is low compared to other countries. This can be shown on a pie chart as per below:
Figure 4.10; Respondents use of the Internet
Frequency of Internet Usage
Table 4.6: Respondents who agreed that they use the Internet were then asked how often they made use of it and the results are as follows.
Frequency Everyday Once a week Once a month Periodically
Respondents 38 67 34 22
Percentage (%) 24 42 21 13
Figure 4.11: Frequency of respondents use of the Internet by percentage
From figure 4.11, we notice that respondents who accepted to using the Internet at least once a week had the highest percentage (41%). 22% of respondents stated that they use the Internet at least once a month. This is less than the 24% who admitted to using the Internet at least once a day. It was great to notice that only 13% of the respondents make us of the Internet periodically.
Where respondents use the Internet
Variables Frequency Percentage (%)
Home 31 19
Work 17 11
Library 15 9
Café 37 23
School/University 21 13
On mobile 40 25
Table 4.7; show percentage of respondents with respect to where they make use of the Internet.
Figure 4.12: respondents percentages with respect to where they make use of the Internet
From the figure above, 25% of respondents who use the internet accept to using the internet on their mobile phones while, 23% said they make use of the internet at paid cyber cafes. While 19% of respondent make use of the Internet from home, 9% will visit libraries to use the Internet.
Table 4.8; Respondents’ perception of online banking
Perception Frequency Percentage (%)
I like online banking 108 67
I do not like online banking 53 33
Figure 4.13: percentage of respondents’ perception of online banking
Usage of different e banking services
E-banking Services Percentage (%)
Branch banking 20
Telephone banking 6
Phone banking 22
SMS 10
Mail banking 3
ATMs 39
TOTAL 100%
Table 4.9: Shows the percentage of users by the different banking services
Most respondents still go into branch to do their day to day banking (39%), followed by phone banking (22%), ATMs (20%), SMS (10%), telephone (6%) and finally, mail banking (3%). Most services offered includes ATMs, phone banking, SMS and in branch services (Nasri, 2011). These findings are consistent with Foon and Fah’s observations from a research conducted in Malaysia in 2010.
Figure 4.14: Percentage of usage of e banking services.
Satisfaction of consumers who use electronic banking
Rate 1 2 3 4 5
Percentage (%) 2 10 25 29 34
Table 4.10: Consumers’ satisfaction of e banking by percentage
From the table 4.9, we notice that most electronic banking consumers rates their satisfaction to this services as 5 of 5 (34%). This shows they are very satisfied with the services. 29% rated their satisfaction on 4 of 5, 25% gave a 3/4 rating, 10% were somewhat dissatisfied while only 2% said they are dissatisfied with the electronic banking service.
Those who considered using e-banking in the future
Response Percentage (%)
YES 56
NO 44
Table 4.11: percentage of users who plan to use e baking in the future
From the table above, 56% of non-users who were asked if they will consider taking up e-banking in the future said yes while 44% of non-users said no to considering future use of e banking. These numbers are promising and tells us that in the future, more people will take us electronic banking. Respondents who said no to taking up electronic banking mentioned cost of internet network problems, cost of difficulties to use as the factor that discourage them from using electronic banking services.
Figure 4.15: percentage of those who consider using e banking in the future
Factors encouraging the use of e-banking
As shown on the table below, up to 25% of respondents indicated that they take up e-banking due to convenience, 21% indicated that they take up e-banking due to ease of use, while 19% said trust is an important factor for them while using e-banking. Privacy (16%), 11% stated accessibility is an important factor that encourage them take up e banking while just 8% indicated reliability as an encouraging factor to using electronic banking.
Factors Percentage (%)
Trust 19
Accessibility 11
Ease of use 21
Privacy 16
Reliability 8
Convenience 25
TOTAL 100
Figure 4.12: Factors that encourage e-banking usage
Figure 4.16; shows factors encouraging the use of e-banking (%)
Does e-banking make banking more convenient?
This question was asked to determine if e banking is perceived to be convenient by the respondent and whether or mot this influence its usage. The table below shows that a total of 72% of respondents agreed that Internet banking makes banking more convenient while 28% disagrees with this.
Does e banking make banking more convenient?
Frequency Percentage (%)
YES 116 72
NO 45 28
Table 4.13; shows percentage of respondents who said yes or no to convenience being a main factor for e-banking adoption
Figure 4.17; shows how convenience influence e-banking adoption
Are you looking for a similar paper or any other quality academic essay? Then look no further. Our research paper writing service is what you require. Our team of experienced writers is on standby to deliver to you an original paper as per your specified instructions with zero plagiarism guaranteed. This is the perfect way you can prepare your own unique academic paper and score the grades you deserve.
Use the order calculator below and get started! Contact our live support team for any assistance or inquiry.
[order_calculator]