Academia.eduAcademia.edu

Figure from:Assessing the Effects of COVID-19 on Online Routine Activities and Cybercrime: A Snapshot of the Effect of Sheltering in Place

See full PDFdownloadDownload figure
Naive Bayes Directed Acyclic Graphical Model for Cybercrime Victimisation (Pre/Post-COVID- Given the nature of the natural experiment being done with different groups, the ability to compare the outcomes can be questioned. However, to demonstrate the validity of the results, propensity score matching (PSM) was also performed. PSM is a quasi-experimental method where statistical techniques are used to construct an artificial control group by matching each treated unit with a non-treated unit of similar characteristics to artificially remove selection bias (Ali et al., 2019;

Figure 2 Naive Bayes Directed Acyclic Graphical Model for Cybercrime Victimisation (Pre/Post-COVID- Given the nature of the natural experiment being done with different groups, the ability to compare the outcomes can be questioned. However, to demonstrate the validity of the results, propensity score matching (PSM) was also performed. PSM is a quasi-experimental method where statistical techniques are used to construct an artificial control group by matching each treated unit with a non-treated unit of similar characteristics to artificially remove selection bias (Ali et al., 2019;

Related Figures (5)

Note. For all t-test the alternative hypothesis is that Group | is not equal to Group 2 Caribbean Journal of Multidisciplinary Studies Volume 1, No. 1, pp. 36-60, May 2022
Note. For all t-test the alternative hypothesis is that Group | is not equal to Group 2 Caribbean Journal of Multidisciplinary Studies Volume 1, No. 1, pp. 36-60, May 2022
[he results shown in Table 4 suggest that all forms of cybercrime have decreased in prevalence after the COVID-19 pandemic. However, the x* test suggests that while the overall decrease in sybercrime victimisation is related to the COVID-19 pandemic, not all cybercrimes were affected n the same way. While the decreases in the prevalence of victimisation for unauthorised access, -yberbullying and unsolicited content are related to the pandemic, the decrease in victimisation ising malware is not related to the pandemic. results of x7- tests Comparing Pre- and Post-COVID-19 Samples
[he results shown in Table 4 suggest that all forms of cybercrime have decreased in prevalence after the COVID-19 pandemic. However, the x* test suggests that while the overall decrease in sybercrime victimisation is related to the COVID-19 pandemic, not all cybercrimes were affected n the same way. While the decreases in the prevalence of victimisation for unauthorised access, -yberbullying and unsolicited content are related to the pandemic, the decrease in victimisation ising malware is not related to the pandemic. results of x7- tests Comparing Pre- and Post-COVID-19 Samples
Figure 2 activities. To further define the role or dependence of cybercrime victimisation on COVID-19 pandemic measures, Bayesian dependency modeling was done using the summated victimisation data. The results of this are presented in Figure 2. This will identify the possible causal relationship that may exist between the COVID-19 pandemic and cybercrime victimisation and our online routine activities.
Figure 2 activities. To further define the role or dependence of cybercrime victimisation on COVID-19 pandemic measures, Bayesian dependency modeling was done using the summated victimisation data. The results of this are presented in Figure 2. This will identify the possible causal relationship that may exist between the COVID-19 pandemic and cybercrime victimisation and our online routine activities.
Comparison of Predictors of Cybercrime Victimisation Identified Pre- and Post-COVID-19
Comparison of Predictors of Cybercrime Victimisation Identified Pre- and Post-COVID-19