Perception and Awareness of Young Internet Users towards Cybercrime: Evidence from Malaysia

: Cybercrime is a criminal (unethical and unlawful) activities using internet facilities such as virus infections, identity theft and hacking. There is high risk of becoming a victim especially for young internet user. The purpose of this study is to protect them by providing empirical evidence to the policy makers in combating cybercrime. The study examines the relationship between perception and gender, age and knowledge as well as the relationship between awareness and gender, age and knowledge towards cybercrime. A field survey is conducted among 342 students in the faculty of accountancy of Universiti Teknologi MARA (UiTM) with a structured questionnaire that covers demographic information and seven most known cybercrimes. Percentile analysis, correlation matrix, multivariate regressions are done to test the hypotheses. In addition, Post Hoc test is conducted to locate where the significant differences lies. The study finds: (1) Female students are more aware and have affirmative insights than male, (2) students in the age group of 18-23 years have lower perception and awareness than those aged 24 years and above and (3) those with higher academic qualifications are more aware at cybercrime and perceived the issue of risk differently. The study provides empirical evidence to the top management of the higher level institutions on the needs to improve their policies and procedures to protect young generation reducing the high risk of becoming a victim.


Introduction
The rapid changes in computer connectivity and innovation in digital technology provide numerous benefits to human life but it is not out of side affect such as cybercrime. Cybercrime is a new wave of crimes using internet facilities, which needs to be addressed urgently and earnestly by policy planners to protect the young generation as there is a high risk of becoming a victim of this crime (Asokhia, 2010;Mensch and Wilkie, 2011). Center for Strategic and International Studies (CSIS) reported that every year there is a financial loss of 445 billion dollar in world economy due to cybercrime (The Daily Amar Desh, 2014). Cyber Security Malaysia reported that the total number of cybercrimes was 1038 in 2007 and it increased to 2123 in 2008 (The Star,200). Cybercrime statistics along with an increasing number of research studies indicate that young people do not always behave ethically in online activities and hence, there is a chance of every internet user becoming a victim (McQuade, 2009). Chen et al. (2008) state that human factors are involved in security awareness process. Human beings are usually the first line of defence to secure information assets, no matter how advanced and rigid the security technology solutions may be. All the security breaches such as virus infections, identity theft and hacking are the direct cause of carelessness and lack of knowledge and action on the part of users (Chen et al., 2008). A high level awareness about information security and cybercrime issues amongst users at home, in government and educational institutions, especially young people, would decrease the occurrence of cybercrime (Sembok, 2003). The effectiveness of combating cybercrime among users' especially young users will work if they are familiar and adroit while using online. Therefore, human factors such as gender, age, knowledge and skills (experience) may assist in augmenting the levels of awareness among young people. Students' perceptions of risk and awareness on security of the internet and information should be profoundly addressed (Wang et al., 2008). Past studies have examined students' attitude towards computer skills and ethics among educators (Shariff and Deni, 2005;Aris et al., 2004;Sembok, 2003). The majority of earlier studies are focused on approaches of law and enforcement agency tools in combating cybercrime. However, only few studies are conducted to examine students' awareness and their perception of risk against cybercrime. Thus, this study is carried out with an aim of examining the actual level of awareness and the perception of students at the faculty of accounting, UiTM, Shah Alam, Malaysia about cybercrime rise. The purpose of this study is to investigate the awareness of the respondents about cybercrime and how they perceived the risk of such crime. The specific objectives are to examine (1) the level of awareness towards 'Cybercrime' in terms of gender, age and knowledge; (2) the level of perception towards 'Cybercrime' in terms of gender, age and knowledge and (3) the relationship between level of perception and level of awareness on cybercrime.
The following sections present review of literature and hypotheses development, conceptual model, method, test of hypotheses and findings and conclusion. Colfer (2007;Li, 2006) state that there are dissimilar perceptions and awareness between men and women. According to Titi (2003) women are more aware of cyber regulations and have superior ethical values compared to men. Women are less likely to become victims as compared to men. Lifestyle Theory states that sex is an often-mentioned demographic characteristic that is associated with difference in lifestyle (Reyns, 2010;Ngo and Paternoster, 2011;Wolak et al., 2006;Choi, 2008). Neiss et al. (2009) state that perception and awareness of young people are dissimilar between age groups. It is because young people and older people have different perspectives. Young people give more negative emotional perception than older adults. Lifestyle Theory suggests that individual of different ages participate in different kinds of lifestyle. These lifestyle differences, therefore, expose individuals of different ages to varying levels of risk of victimisation (Reyns, 2010;Ngo et al., 2011;Wolak et al., 2006;Choi, 2008). It has been persistently reported that younger people are more likely to be victimised than older people. The Australian Youth Affairs Coalition (AYAC, 2010) submitted a report on cyber safety to the office of Privacy Commissioner in which they state that the students in the age group of 18-24 are in high risk environments when expose to online activities.

Review of Literature and Hypotheses Development
Knowledge is very important for young people to prevent cybercrime (Curtis and Colwell, 2000;Wang et al., 2008). Chawki (2005) states that educating young people would help decrease the risk of students in cyberspace. Asokhia (2010) finds that the level of education contributes significant differences to the students' perceptions of cybercrime. Knowledge helps people to be more aware on cybercrime (Levin et al., 2008). The number of cybercrime victims could be reduced by introducing proper awareness activities such as training programs, sufficient resource for compliance, develop policies & regulations and sufficient protection of personal information (Choi, 2008;Levin et al., 2008;Chawki, 2005;Bougaardt and Kyobe, 2011). Choi (2008) emphasises on the effectiveness of university programs in promoting knowledge and values about cybercrime as these programs could improve future behaviour of students' towards cybercrime in terms of safety and security. This would establish norms and adjust prospects for illegal or delinquent behaviours. Based on the review of above literatures it is anticipated that gender, age and knowledge have significant influences on cybercrime. The following six hypotheses are developed based on the above assumption and in line with the objectives of this study. Hasan et al. (2015) present a table providing the definition and measurement of the variables and the range of their values. Likewise, the definition and measurement of the variables and the range of their values for this study are presented in Table 7 (Appendix-1).
H 1 = There is an association between perception of cybercrime and gender H 2 = There is an association between awareness of cybercrime and gender H 3 = There is an association between perception of cybercrime and age group H 4 = There is an association between awareness of cybercrime and age group H 5 = There is an association between perception of cybercrime and knowledge level H 6 = There is an association between awareness of cybercrime and knowledge level

Conceptual Framework
A diagram is designed to form a graphical view of the diagnostic pattern of the problem under study. The diagram shows that six hypotheses are tested such as hypothesis 1 test the relationship between gender and perception on cybercrime, hypothesis 2 tests the influence of gender on awareness, hypothesis 3 tests the impact of age on perception, hypothesis 4 tests the link between age and awareness on cybercrime, hypothesis 5 tests the association between knowledge and perception and hypothesis 6 tests the bonding between knowledge and awareness on cybercrime. Finally, we examine the correlations between the level of awareness and the level of perception. The conceptual framework of this study can be viewed in Fig. 1.

Research Approach
Primary data is used to examine the influence of age, gender and knowledge on both awareness and perception in this study. A field survey is conducted in the faculty of accountancy of Universiti Teknologi MARA (UiTM), Shah Alam, Malaysia. For this, a structured questionnaire is developed and distributed randomly among 342 students who are pursuing diploma, under-graduate and post-graduate programmes. They completed the questionnaires and returned it back to us in time. In the questionnaire we covered demographic information and seven most known crimes such as (1) unauthorised access to computer material, (2) unauthorised access to internet to commit further offences, (3) unauthorised modification of the contents of any computers, (4) wrong communication, (5) people involved in abetments and attempts would be punished, (6) cyber terrorism and (7) cyber nuisance.
Frequency and percentile analysis are done to examine the level (very low, low, moderates, high and very high) of perceptions about above-mentioned most known crimes. Multicollinearity analysis is done to examine the multicolllinearity among age, gender, knowledge of threats and knowledge of offences. Multivariate analyses are done to examine the level of perception and level of awareness of age, gender and knowledge. A further analysis is done digging of age level, education level and knowledge level in order to have a concrete outcome of the study. We use Post Hoc Games-Howell (multiple comparison test) technique to examine the relationship between perception and age, level of education and knowledge of offences. In case of awareness, we use Post Hoc Least Significant Difference (LSD) test technique to realize the relationship between awareness and age, level of education and knowledge of offences. Finally, we check the correlation between level of perception and level of awareness of cybercrime.

Results and Discussion
If we look at the percentile analysis of demographic data we find that 63.2% of the respondents are female and 36.8% are male. 25.4% (majority) of the respondent are in the age group of 22-23; 42.73% (majority) of the respondents are undergraduates; 31.6% (majority) of the respondents have 3-5 years' experiences; 31.3% (majority) of the respondents' use internet for 2-4 h; 46.8% (majority) of the respondents have internet access through broadband; 41.8% (majority) have knowledge of the most dreadful known virus threats; 32.2% (majority) of the respondents have knowledge of worst known threats hackers and only 26% (majority) have knowledge of fraud.
If we look at the percentile analysis on respondents' knowledge about seven common crimes, We use Pearson's pair wise product moment correlation coefficient (r) to examine the multicollinearity between variables such as age, gender, knowledge of threats and knowledge of offences. A correlation matrix table is produced for all the variables using SPSS and the results are presented in Table 2.
If we look at Table 2, we see that the highest level of correlation existed between knowledge of offences and knowledge of threats (0.167). Some authors (Judge et al., 1985;Bryman and Cramer, 1997) suggest that simple correlation between variables should not be considered harmful until and unless it exceed 0.80 or 0.90. Thus we find no multicollinearity between variables. As such, there is no problem with these variables in the interpretation of the results of multivariate analysis.
Next, we use multivariate analysis technique (Ztest and F-test) to have overall outcome of accepting or rejecting the hypotheses. We use z-test for gender and f-test for age and knowledge. We analyse the results from both aspects such as level of perception and level of awareness for gender, age and knowledge. The results of multivariate analysis for gender are shown below. If we look at Table 3, we find p-values of both level of perception and level of awareness show that there are significant differences between male and female regarding the level of perception as well as the level of awareness of cybercrime at 5% level of significance. It indicates that the male and female of accounting students in UiTM have different level of perception and awareness of cybercrime. These findings support hypothesis 1 and hypothesis 2. Therefore, H 1 and H 2 are accepted. The results of multivariate analysis for age are shown below. Table 4 reveals that the p-values of both level i.e., level of perception and level of awareness are significant differences among students in different age groups relating to their perception and awareness of cybercrime at 1% level of significance. It indicates that there are significant difference between young people's perception and their age as well as between awareness of cybercrime and their age. Therefore, H 3 and H 4 are accepted. The results of multivariate analysis for knowledge are shown below.
If we look at Table 5, the p-values show that the level of education and knowledge of offences contributed significantly to the differences of the students' perception of cybercrime. Similarly, the level of education and knowledge of offences contributed significantly to the differences of the students' awareness of cybercrime. Having good knowledge can help people be more aware of cybercrime (Levin et al., 2008). People who have knowledge will have different perception and awareness on cybercrime and would affect their reaction Perl, 2009;Bougaardt and Kyobe, 2011;Choi, 2008;Chawki, 2005;Levin et al., 2008). There is significant dissimilarity between young people's perception and awareness of cybercrime and their knowledge. Therefore, H 5 and H 6 are accepted.
At this stage of analysis, we only know that all of our hypotheses are accepted. Therefore, we could say that there are significant relationships between perception and age, gender and knowledge. Similarly, there are significant relationship between awareness and age, gender and knowledge. But this is not enough for us to identify the influence of different age groups, knowledge of offences and level of education on perception as well as awareness. So, we need to carry out further analysis to portray a conclusive scenario of this study. Games-Howell post hoc multiple comparison test is carried out to examine the relationship between perception and different age groups, knowledge of offences and level of education. Least Significant Difference (LSD) post hoc multiple comparison test is carried out to examine the relationship between awareness and different age groups, knowledge of offences and level of education. Figure 2 it portrays the summarised result of Games-Howell post hoc multiple comparison test between age groups and perception of cybercrime. There are significant differences between age groups towards perception of cybercrime. Figure 2 reveals that the perception of different age groups are statistically varied between students in the age group of 18-19 years (M = 3.73) and those of 22-23 years (M = 4.05) and 24-25 years (M = 4.07). Students in the age group of 22-23 years and 24-25 years reported significantly higher perception on cybercrime compared to students of 18-19 years. There are no significant differences between other groups.  Figure 3 portrays the summarised result of Least Significant Difference (LSD) post hoc multiple comparison test between age and awareness of cybercrime. There are significant differences between different age groups regarding awareness of cybercrime. Figure 3 reveals that the awareness of different age groups are statistically varied between students of 18-19 years (M = 3.26) and those of 24-25 years (M = 3.54) and more than 25 years (M = 3.53). Other age groups of 20-21 years (M = 3.19) and 22-23 years (M = 3.25) also reveals significant differences with those of 24-25 years and more than 25 years. The result reports that students in the age group 24-25 years and more than 25 years have significant higher awareness of cybercrime compared to those of 18-23 years. There are no significant differences between other groups.
The summarised result of Games-Howell Post Hoc multiple comparison test between knowledge of offences and perception of cybercrime are shown in the following Fig. 4. The results demonstrate significant difference between perception of cybercrime and knowledge of offences. The perception of different knowledge groups are statistically varied between students' with very low knowledge of offences (M = 4.11) and those who have moderate knowledge (M = 3.81) and very high knowledge (M = 4.74). Students' with moderate and very high knowledge of offences have significant differences in perception of cybercrime. The results indicate that students who have high knowledge about offences would have high/positive perceptions of cybercrime. The summarised results of Games-Howell post hoc multiple comparison test between levels of education and perception of cybercrime are shown in the following Fig. 6. The perception of different levels of education were statistically dissimilar between diploma students (M = 3.77) and those in undergraduate (M = 4.01) and postgraduate levels (M = 3.99). The results state that undergraduate and postgraduate students have higher significant perception compared to other level of education Thus students with higher level of education have higher perception of cybercrime than those with lower level of education. However, there are no significant differences in other groups.
The summarised results of LSD post hoc multiple comparison test between levels of education and awareness of cybercrime are shown in the following Fig.  7. It is observed that diploma students (M = 3.77) have significant difference with postgraduate students (M = 3.99) and undergraduate students (M = 4.01). Thus postgraduate students have high significant difference in awareness of cybercrime compared to others group. The results suggest that students with higher level of education have better awareness of cybercrime than those at lower levels. However, there are no significant difference in other groups.
At this stage of analyses we know the specific relationship between perception and age, gender, knowledge and level of education. We also know the relationship between awareness and age, gender, knowledge and level of education. But at this moment we do not know the relationship between awareness and perception of cybercrime. We believe that there is a positive relationship between awareness and perception of cybercrime. Therefore, we examine the Pearson correlation between them. If we look at the following Table 6, it shows that there is a positive correlation between respondents' perception level and awareness level of cybercrime at 1% level of significance. This suggests, the higher the level of awareness, the higher the level of perception.

Conclusion
This study concludes that gender, age and knowledge have significant influences on the level of perception and awareness of cybercrime. We test three hypotheses on the relationships between perception and age, gender and knowledge. We also test other three hypotheses on the relationship between awareness and age, gender and knowledge. All six hypotheses are supported by this study. We find female students are more aware than male students about cybercrime. For age group, we find students in the age group of 23-25 years have significantly higher perception of cybercrime than those of 18-19 years and the age group of 24-25 and more than 25 years have significantly higher awareness of cybercrime compared to those of 18-23 years. This suggests students with greater awareness and perceptions of cybercrime are more cautious about the same. For knowledge of offences and level of education, we find students with knowledge of cyber offences and high level of education have high perceptions and awareness of cybercrime. More revealing fact is that perception and awareness are positively correlated at 1% level of significance. The findings indicate that the level of awareness and the level of perception of the respondents regarding cybercrime are equal. The study suggests students having higher knowledge of cybercrime also have better awareness of the cybercrime. Curtis and Colwell (2000;Atkinson et al., 2009) mentioned that unambiguous knowledge of cybercrime would help young people in preventing and avoiding this delinquency. By education, students can develop positive attitudes towards cyberspace (Wang et al., 2008). Thus the result of this study is to some extent similar to the findings of previous studies.
This study recommends necessary policy measures to be taken by the higher learning institutions including UiTM to prevent alarming cybercrime as well as cyber war. Educating young people with some knowledge on cybercrime would help increase their level of awareness and also perceptions on cybercrime. Further investigation and experimentation into cybercrime is strongly recommended. Conceivably the growth of cybercrime in Malaysia, as all over the world is on the rise and to curb its scope and complexity is the pertinent need today. Hence factors that strongly influence level of perceptions and awareness on cybercrime among students in Malaysia higher institutions could be more relevant and feasible to be explored for the benefit of all.