CrimeFace Video Analytic System and Reoffending Rates in Retail Stores

Published: Last Edited:

This essay has been submitted by a student. This is not an example of the work written by our professional essay writers.

CrimeFace Video Analytic System and Reoffending

Rates in Retail Stores

This particular report will attempt to recognise the relationship between the increased use of the CrimeFace system and its effectiveness in reducing recidivism rates. Thus, by taking into consideration as a directional hypothesis method, it could be argued that an observational statement could be produced which dictated the expected direction in relation to the two chosen variables (Singh, 2007). Furthermore, this report will also explore the specific research design and any potential limitations of the research.

Research Hypothesis

Research Question: "Does the implementation of the CrimeFace system have a significant positive effect on the prevention and disruption of retail theft in the stores where it is installed?"

Hypothesis (H1): The implementation of CrimeFace video analytic system within selected retail stores will lead to a significant reduction in theft recidivism rates.

During this research, it was imperative that the hypotheses developed followed certain specific guidelines. Hence, to ensure the success of the research; the hypothesis produced needed to be concerned with something that the financers of the research would be interested with, thus generalizable and to a certain extend broad enough to show a causal relationship between the two selected variables. Therefore, as previously stated ‘’ a good hypothesis states as clearly and concisely as possible the expected relationship (or difference) between two variables and defines those variables in operational, measurable terms" (Gay, 1996:62). Consequently, it was extremely vital that the hypothesis produced indicated a relationship between usage of CrimeFace and decrease in recidivism rates.

Therefore, by analysing the project requirements it became abundantly clear that leaders of this research were in a sense keen to observe a relative decrease in thefts regarding the stores that had this system installed. The issue of shoplifting is one that definitely high on the agenda of many; including the manufacturing company and the organisation providing substantial financial support towards the research being conducted. Consequently, whilst current prevention methods may be considered as to a certain extent advanced; they are still without a doubt lacking when it comes to decreasing shoplifting rates. Thus, offenders have boasted that even though they ‘’ generally recognise that security is commonplace, it does not stop them stealing and most claim that they are caught infrequently’’ (Gill, 2007:7). Therefore, it could be argued that it is due to this supposed arrogance from shoplifters combined with the almost archaic preventative methods of stores that has led to shoplifting still being considered as a major issue. Hence, it was noted in particular research that ‘’ we’re only catching 6 million of the more than 275 million who commit retail theft (globally). That’s about 1 in 46’’ (Pillar, 2013:1). Thus, it is evident that major advancements need to be made to ensure the reduction of theft in stores; advancements that may be amplified by the implementation of the CrimeFace system.

For that reason, it was vital to consider the precise conditions when formulating the hypothesis. The argument in question needed to be relatively future orientated, whilst also taking into consideration the relationship between the two variables (Singh, 2007). Thus, it was decided that it would be most beneficial if the hypothesis took into consideration recidivism rates and how the implementation of this system affected them. Thus, by being permitted access to the store’s own CCTV images of individuals who have previously been caught stealing in their premises, this would allow the hypothesis to be easily testable (Prasad et al, 2001). The system being tested should significantly be able to detect previous offenders, leading to my hypothesis possibly being able to also answer the research question; due to the research predominately seeking to indicate a positive effect on the prevention and disruption of this criminal act. Therefore, it was clear that although the hypothesis selected slightly digressed from the research question, by using the data made available to us; this hypothesis could potentially still establish whether or not there will be a significant positive effect on the prevention and disruption of retail theft in the stores where the CrimeFace system is installed.


To further increase the validity of both the hypothesis presented and the research question provided previously, it was imperative that the evidence required to test the hypothesis was appropriately analysed. Thus, it is clear that the overall function of a specific research design is to ensure that the evidence obtained throughout the research is able to unquestionably answer the initial research question (Bryman, 2008). Therefore, to ensure that the hypothesis selected is both testable and future orientated, variables such as environmental conditions and geographical crime rates needed to be taken into consideration (Singh, 2007).

Hence, to ensure a successful hypothesis, certain specific evidence needed to be analysed to ensure that the hypothesis could be accepted. Alongside the data regarding recidivism rates; information regarding store location and crime rates of the locality were also of upmost importance. Thus, in a sense; the evidence needed to test this particular hypothesis would be data regarding recidivism rates; and the success of the software in identifying previous repeat offenders before they can reoffend. Therefore, the main purpose of hypothesis testing is to in a sense use specific sample information in a way that allows the researcher to decide whether it is most productive to reject the null hypotheses concerning a population parameter. Thus, by developing both a null and statistical hypothesis, this would allow the study to be further reliable and valid, this is due to both features analysing population values and not observed statistics (May, 2010).

Alterative Statistical Hypothesis

This particular notion has been claimed to be the statement that the researcher is expecting to prove is true. Thus, in a sense it is the complete opposite of the null hypothesis (Bhattacherjee, 2014).

Alternative Hypothesis (HA): There will be an increase in the detection of reoffenders in stores that install the CrimeFace system.

Null Hypothesis

The null hypothesis has been noted as being a claim of no difference. Hence, the null hypothesis is statement concerning factors such as mean value and population parameters. (Bhattacherjee, 2014).

Null Hypothesis (H0): There will not be an increase in the detection of reoffenders in stores that install the CrimeFace System.

Research Design

Design components

To successfully ensure that the research is completed in an effective manner, a productive research design needed to be implemented. Thus, to guarantee that the CrimeFace system was successful in terms of both detecting reoffenders and in regards to this specific research; very sophisticated close-circuit television cameras that operated with the CrimeFace software needed to be installed throughout the stores in specific ‘hotspots’. These, hotspots would without a doubt include entrances and areas of the store that included high value products. Hence, once individuals would enter the store or act in a sense ‘suspiciously’ in areas where high value products were located; the CrimeFace system could be instantly used to detect and individuals face and indicate whether they were in fact previous offenders. Thus, if the database was the able to reveal that those suspicious customers were reoffenders then special attention could be paid to them; either in the form of security guards being informed or through close analysis through the CCTV system to prevent any thefts and in as sense disrupt the individual.

Data collection method

The data collection method will be focussed around the information provided by the CrimeFace database. Thus, the information provided by the Aldesco company will without a doubt prove to beneficial as it will grant the research with primary data which could be used to further add validity and reliability to the study. Furthermore, the data received from the company could also be analysed in a method that allows us to indicate whether or not there has in fact been a substantial decrease in thefts by reoffenders.

Justifications: This particular method of data collection whilst being our only genuine method of collecting data; could also prove to be the most effective. This is due to primary data collection allowing us to be able to engage in a comparison between our two variables. Thus, indicating whether there is a causal relationship between our independent and dependent variable.

Data analysis method

Once it was determined that I there would be two main variables within the research being completed; a one-way ANOVA method seemed to be the most suitable. Thus, due to the research being based on the comparison of our results to the expect results; an analysis of variance would allow us to significantly test one variable against another. Hence, a one-way ANOVA ‘’is the analysis of the variance of values of a dependent variable by comparing them against another set of values, the independent variable’’ (Griffith, 2010:234).

Justifications: Therefore, by implementing an analysis of variance, this would allow me as the researcher to effectively be able to test the selected hypothesis and that ‘’the mean of the tested variable is equal to that of the factor’’ (Griffith, 2010:234).


Although this hypothetical research study indicates a clear and straightforward process, there still without a doubt several shortcomings that could potentially impact the report in a negative manner. One vital aspect that needs to be taken into consideration; is the effect of geographic crime rates on our intended results. This is because, even though the system would be placed within only selected stores initially; the crime rates will without a doubt differ from location to location. Thus, this would impact on not only the research but also the hypothesis to be tested as the crime rates of a specific location could potentially impact the reduction of recidivism rates regarding thefts within stores using the CrimeFace system (May, 2010). Therefore, this variable could potentially distort the results as regardless of the impact of the system; certain areas are still likely to have higher recidivism rates than others.

Furthermore, the sophistication of modern shoplifters could also impact the success of the hypothesis generated. Therefore, in a contrasting manner to previous shoplifting tactics, the entire concept has undergone a slight transformation, unlike previously where store theft was considered to be an individualistic act; recently shoplifters have been noted as conducting their actions in groups (Mullich, 2013). Thus, these teams will employ “blocking units, people pretending to be shoppers who “accidentally” get in the way of store security personnel when they try to respond’’ (Mullich,2013:1) , this could essentially also effect the CrimeFace System. Therefore, whilst the system may be focussed on other more ‘’suspicious’’ individuals, certain reoffenders could potentially be allowed to engage in the theft of store products, thus impacting the results and effecting the hypothesis. Furthermore, offender’s facial aspects may also alter throughout the years through factors such as age, plastic surgery and even possibly gender modifications. Therefore, this could affect both the CrimeFace System and the hypothesis; as the system may not be able to detect previous offenders, further distorting our results.


Bryman, A. (2008). Social Research Methods, 3rd eds. Oxford: Oxford University Press.

Bhattacherjee, A. (2014). Social Science Research: Principles, methods, and practices. Available at: Accessed on: (28/12/14)

Gay, L.R. (1996). Educational Research: Competencies for analysis and application. California. Merrill Publishing.

Gill, M. (2007). Shoplifters on shop theft: implications for retailers. Perpetuity Research and Consultancy International (PRCI) Ltd. Available at: (Accessed on: 28/12/14).

Griffith, A. (2010). SPSS for Dummies. New Jersey: John Wiley & Sons, Inc.

May, T. (2010). Social research: issues, methods and process. 4th eds. Berkshire: McGraw-Hill Education.

Mullich, J. (2013). Reducing the Risk of Retail Theft. Smarter Insights on Security. Wall Street Journal [Online], (Last updated on: 18th November 2013). Available at: Accessed on: (02/01/15).

Pillar, M. (2013). Retail Theft Study: Only 1 In 46 Gets Caught. Reintegrated Solutions for Retailers [Online], (Last updated on: 19th March 2013). Available at: (Accessed on: 28/12/14).

Prasad, S. Rao, A and Rehani, E. (2001). Developing hypothesis and Research questions. Arizona State University. Available at: Accessed on: (02/01/15).

Singh, Y.K. (2007). Research Methodology: Techniques and Trends. New Dehli. APH Publishing.

1 |