Introduction
The critique will start with an analysis of the research problem and purpose of the research. Thereafter, an analysis of the suitability of the method will be done as well as the reliability and validity of the article. Ethical considerations and conclusions made will also be analysed so as to determine if the study is valuable.
Critical review
Research problem
The research problem is too broad. Consequently, the research question contains too many variables that cannot be effectively analysed. The author should have narrowed down his research paper. He states that the purpose of the research is to study shoplifting. It would be more effective to focus on a causability problem or the relationship between two variables (Bassett & Bassett, 2003). For instance, he could have focused on the effect of CCTV equipment on shoplifting as his research problem. Instead, the researcher decided to use the latter relationship as a subset to his main problem. Choosing a wide research problem stretches the researcher’s resources and causes him to analyse the problem superficially (Tanner, 2003).
Given the overall direction of the research, it is evident that the research purpose is not clear. In certain instances, the author appears to focus on the problem of shoplifting in only one store. However, in other instances, he generalises the results and talks about shoplifting in general. If he was conducting a case study research, then he should have indicated this in the research title and research objectives (Dawson, 2009). He merely stated that he wanted to study shoplifting, but did not specify geographical limitations of the study or the central area of focus.
Methods
The researcher used a combination of quantitative and qualitative methods. The qualitative methods took the form of unstructured interviews. Quantitative methods were employed through the use of official inventory records in the concerned company as well as through questionnaires. Quantitative methods are appropriate for descriptive questions or questions that start with ‘why’ and ‘what’. In the study under analysis, the researcher wanted to find out why customers shoplifted, the extent of shoplifting and the effect of CCTV equipment on shoplifting. The first question was a qualitative one (why customers shoplift), so this should have been the only one which utilised unstructured interviews (Miles, 1994).
To some extent, the researcher observed this rule because he interviewed a shoplifter concerning reasons behind shoplifting. However, he was wrong when he used semi structured interviews to answer quantitative questions like the extent of shoplifting and the effect of CCTV equipment on shoplifting. For instance, he interviewed the line manager about the rate of increase of shoplifting, yet he could have used official records as a more reliable source of information (Punch, 2005).
Conversely, one should use quantitative research methods when answering research questions that start with ‘who’ and ‘how’. When analysing the extent to which something affects another variable or how two variables are related (causal relationships), then it is appropriate to use quantitative methods. Additionally, one can use quantitative research methods when analysing the magnitude of a problem. For instance, the writer wanted to measure the extent of shoplifting. The author used inventory analysis, which is quantitative research and suitable to the question type.
He also compared unaccountable inventory before and after installation of CCTV equipment to determine the effect of these cameras on shoplifting. The latter was also a quantitative method and was appropriate. However, the author used questionnaires to collect data about qualitative questions. Questionnaires are better suited to non descriptive research objectives, or questions that do not analyse ‘why’ something happens (Davies, 2007). By limiting respondents’ answers to what the author could think about in the multiple choices, then the research’s findings became inadequate.
When analysing the research method, one must look at the effectiveness of the sample size. The author explains that only ten people responded to the questionnaires, yet he had prepared 30 of them. Good sampling sizes should effectively reflect the demographics of the population under analysis (Robson, 2002). Furthermore, the larger the sampling size, the more representative the research findings will be.
The researcher did not specify whether the store under analysis had a high number of customers or not. Regardless of the lack of this information, one can still assume that 10 is a relatively small number for store attendants. The researcher should have striven to get as many people as possible in the research. 30 would have been a representative sample; so he should have sent out 100 questionnaires. Nonetheless, the investigator made a point of noting the ages and gender of the respondents. He should have ensured that these samples represented the general population in the store (Langton & Hollinger, 2005).
There were problems with the research instruments as well. The author did not design his questionnaire appropriately. Since his research was partly a qualitative one, then he should have allowed the participants to give their own unadulterated reasons for shoplifting (Savin-Williams, 2002). Therefore, he should have asked them open-ended questions about why they shoplifted. Furthermore, the multiple choices he provided were not exhaustive. For instance when asking people why they shoplifted he only provided three academically useful options: “for the thrill”, “money” and “want goods”.
Additionally when he asked respondents what would stop them from shoplifting, he only gave then two useful options: CCTV and more security guards. If he really wanted to use structured questionnaires or a questionnaire with closed-ended questions, then he should have created a comprehensive list of why people shoplift. For instance he would have added factors such as peer pressure, psychological temptation, I won’t get caught, as a relief for boredom, to punish storeowners and capitalists, it harms no one, has minimal fines and as a status symbol (Burrows, 1988).
The author’s observation time was insufficient to make reliable inferences. When studying the relationship between CCTVs and shoplifting, he only considered three months yet he needed more time to make tangible conclusions. Furthermore, he only asked five questions in the questionnaire. He should have added more questions that would further demystify the process of shoplifting (Silverman, 2010).
Data type
The nature of results employed also makes the research conclusions doubtful. The author did not perform statistical analyses of the findings on inventory. He only read the amount of unaccounted inventory and deduced that it was reflective of an increase in shoplifting. However he did not focus on the initial inventory as a fact that could affect the amount of unaccounted inventory. In the first table, which contained inventory details for the months of May, June, July and August, the author did not calculate the percentage of ‘inventory in’ that unaccounted inventory represented. It is possible the higher the amount of inventory, the higher the extent of shoplifting among staff members, who probably know how much inventory there is in the store.
A percentage calculation of the percentage of inventory that unaccounted inventory represented shows that shoplifting was not always on the rise during these four months. In May, unaccounted inventory represented 5.54%, in June it increased to 6.041% then decreased to 4.748% and increased again to 7.709%. It is always better to work with ratios rather than absolute figures as the author did (Buttle, 1996), (Huck & Cormier, 1974).
Another major challenge with the analysis of the data was the lack of statistical inference tests. The author was analysing the relationship between CCTV installation and shoplifting. For reliable results, he should have calculated the statistical significance of the relationship between these two variables, yet he did not. He should also have represented the relationship graphically so as to show whether shoplifting increased with time or not. From the percentages quoted in the previous paragraph, this increase is not clear. It would be more apparent in graphical representations (Creswell, 2008.).
The same pattern is prevalent in the second table which looks at the effect of CCTVs on shoplifting. The author only read the absolute unaccountable inventory without calculating ratios and statistical correlational values. While absolute unaccountable inventory may have increased between December and February, the percentage value of unaccountable inventory as a part of “inventory in” remained constant at 5% for both months. Therefore, one can argue that an increase in absolute unaccountable inventory corresponded to an increase in ‘inventory in’. Customers had a higher number of goods to shoplift in December than they did in February. This means that the CCTV cameras may not have led to a decrease in shoplifting. The reduction may simply be due to reduction in the number of goods (inventory) that could be shoplifted.
Validity and reliability
The reliability of this research is quite doubtful. Investigators need to ensure that they eradicate as many rival causal factors as possible when carrying out research. Sometimes these issues may emanate from the internal environment of the study or the external one. Selection bias is one of the ways in which a study can be rendered invalid. Investigators who choose non equivalent groups or non representative groups introduce this error into their findings (Carnell, 1997). One can assert that a selection bias occurred in the analysis because the author used inventory data as a measure of the extent of shoplifting that customers engaged in. He did not consider the fact that sometimes staff members could shoplift.
The author concluded that there was an inverse relationship between CCTV installation and shoplifting. To prove this, he measured the amount of unaccounted inventory prior to installation of the cameras and after installation of the cameras. However, something could have happened during that month to cause an increase in inventory. Consequently, it is not accurate to simply deduce that the reduction in unaccounted inventory was caused by the presence of shoplifting. This deduction would have been more reliable if the author had studied the unaccounted inventory for a prolonged period of time before the installation of the CCTV cameras (like 6 months) and after installation of the equipment. In fact, it is a known fact that the festive season is busier than other shopping seasons, so higher unaccountable inventories should be expected. The author simply compared the amount of inventory during the festive season (December) and the off-season in February.
The generalisability of this research is also uncertain because although the author purported to conduct a study on shoplifting, he only focused on one store. As if this was not enough, the author used a series of methods that were not appropriate to the nature of the research. For instance, interviewing Stella about the effect of CCTV was misguided because Stella probably used her own biases to establish this relationship (Noaks, 2004). Quantitative results would have been more effective to answer that research question.
The validity of the results is debatable because no statistical test was done. This means that results may have occurred by accident rather than as a result of real differences that exist within the population. All quantitative studies need to have a p value of greater than 0.05 (Heppner & Claiborn, 1989). Readers have no way of knowing this value because the author did not calculate it.
The author may also have altered the respondents’ results by pre-empting them. For instance, when talking to Stella, he asked her whether she thought the CCTVs help the store by reducing shoplifting and Stella affirmed that they did not. However, he shared his research findings with her thus causing her to change her answers. The investigator also did the same thing with his line manager when he revealed his research findings before asking about his opinion. The researcher biased the manager’s response by quoting official results. Respondents may try to act smart when an authoritative figure (like the investigator) takes a stand on a research matter (Bruce et al., 2008). They are unlikely to oppose his assertions as the case was with two of the three interviewees.
Ethical considerations
This researcher tried to follow most of the rules of ethical research, but still broke some. For instance, he ensured the autonomy of the participants in the interviews and questions by voluntarily asking them to participate (Cooper & Dewe, 2001). In one of the interviews- with the shoplifter – the author lied to her that he would make it easier for her is she answered his questions. This was unethical conduct because the author could not guarantee her immunity from conviction. She was unwilling to participate and it was only through his coercion that she did.
On the flipside, the author did not cause physical or psychological harm to any of the participants. He did not give the participants preferential treatments based on their position in the store. Nonetheless, the investigator appears to give staff members more respect over customers because he assumed that in-store personnel do not shoplift. Perhaps the location of the interviews was also questionable because they were conducted at a back room. Members of the opposite gender may object to the morality of such a choice especially if the store is in a conservative community.
Whether statements made by the researcher follow from methods.
The researcher affirmed that he wanted to look at the extent of shoplifting in the research. This means that he was not analysing the pattern of shoplifting; he simply wanted to know the degree of shoplifting that occurred (or the number of shoplifters).
However, when making his conclusions, he talks about a pattern of rising shoplifting. Not only did the investigator fail to specify the time period of his analysis, but he also assumed that an increase in shoplifting reflected a high number of shoplifting cases. This is mathematically untrue; something can increase at a high rate but may still have small absolute value. For instance, a half-filled glass may fill up at a higher rate under a leaking tap than a half-filled bucket, but the amount of water stored in the glass after 10 minutes may still be much less in the glass than in the bucket. Likewise, the instances of shoplifting in the store may be rising at a high rate, but their absolute values may be low (Jupp, 1989).
Even the very notion of rising cases of shoplifting in the study is highly doubtful. The researcher only measured the number of inventory unaccounted for by the retailer. Unaccounted inventory is not the same as shoplifted inventory.
A lot of other causes may have contributed to the unaccountability. For instance, paperwork errors could have occurred in favour of unsuspecting customers. Damage may have occurred during the transit of goods within the store, and this may not be documented. Some vendors can defraud retail stores, as well (The Cooperative Group, 2008). Here, a retailer may think that he or she received a certain amount of inventory, yet the vendor may have overstated his order. Furthermore, goods may be lost at the retailer’s warehouse. Opening counters without making a sale can also lead to unaccountable inventory. If a product return occurs or faulty items are sold, then the retail store may loose inventory too. As a result, the statements the author made at the end of the report could not have emanated from his methods.
In the discussion section, it was pointed out that shoplifting is quite prevalent or there were high instances of shoplifting. The basis upon which the author makes this assertion is not clear. He did not quote national or district shoplifting numbers so that he could deduce that the shoplifted products found in the store were just as bad or worse than the standard.
The author also concluded that females are likely to play down their offenses because one respondent appeared to do so. It is not acceptable to use assertions of one respondent as a representation of the whole population. The attitude displayed by the shoplifter could have been a result of her personality. This investigator should not stereotype all females based on one observation. He ought to have interviewed a number of shoplifters-both male and female in order to make these conclusions.
In the conclusion, the researcher claims that shoplifting is “a real problem and is only getting worse”. While this finding may be true for the store under consideration, it may not hold for other stores in the country (Belrin, 1996). The researcher should have included the limits of his research in this concluding statement.
Recommendations
First, the author should use another measure to assess the amount of shoplifting cases. Unaccountable inventory is not directly related to shoplifting, so it should not be considered. Instead, the author should focus on the number of shoplifters caught within the store. Alternatively, a CCTV monitoring system should be placed on all parts of the store and the number of shoplifters observed per day should be noted. In fact, most retailers depend on surveillance staff monitoring to detect fraudulent activities.
Secondly, the researcher should narrow down the research problem. He should not try to focus on assessment of three issues: extent of shoplifting, effect of CCTV on shoplifting and why people shoplift. This wide focus has yielded ineffective results because the researcher is stretched too thinly. He should pick one aspect alone (Rodriguez, 1998).
Problems were manifested in the definition of the research purpose as well. Consequently, the investigators should decide on whether to do a case study of shoplifting or not. If he chooses the case study approach, then he should always add that his observations were restricted to the store. Even his research objectives and questions should be phrased in a way that incorporates the retail store as the only focus of the paper (Tennen et al., 1995). However, if he chooses to do a more general research, then he needs to consider widening his sample size. He could collect data from a series of retail stores within his district or chosen geographical location. Alternatively, he could use police reports as these will have a wider selection of shoplifters and would provide him faster access to them.
During administration of the research, the investigator should avoid asking quantitative information (number of shoplifters) in qualitative data collection forums (unstructured interviews) (Clegg, 1990). He should only rely on police reports, surveillance tallying and retail reports for quantitative findings. Furthermore, he should expand the questions in his questionnaire, by asking open-ended questions on why people shoplift, as well as provide comprehensive multiple choices.
The ethical issues raised should be avoided. When conducting the paper, shoplifters should willingly participate in the research. To eradicate the need for coercion, the researcher could interview persons arrested for shoplifting at police stations. Furthermore, the author should not exhibit selection biases. After collection of data, material should be analysed using statistical tests and represented on graphs.
Conclusion
Owing to the definitional and methodological flaws, this research does not pass the test of a valid, generalisable and reliable study. Ethical issues as well as the lack of a connection between the methods and the author’s conclusions make it inconsequential to shoplifting literature.
References
Bassett, J. & Bassett, C. (2003). Reading and critiquing research. British Journal Perioper Nursing, 13(4), 163-164.
Belrin, P. (1996). National report on shoplifting. Jericho, NY: Shoplifters Alternative.
Bruce, N., Pope, D. & Stanistreet, D. (2008). Quantitative methods for health research: A practical interactive guide to epidemiology and statistics. NY: Wiley Blackwell.
Burrows, J. (1988). Crime prevention through crime analysis. Web.
Buttle, F. (1996). SERVQUAL: Review, critique, research agenda. European Journal of Marketing, 30(10), 8-32.
Carnell, R. (1997). Critiquing research. Nursing Practitioners, 8(12), 16-21.
Clegg, F. (1990). Simple statistics: A course book for the social sciences. Cambridge: Cambridge University Press.
Cooper, C. & Dewe, P. (2001). Organisational stress: A review and critique of theory, research, and applications. Oxford: OUP.
Creswell, J. (2008.). Research design; Qualitative, quantitative, and mixed methods approached. London: Sage Publications.
Davies, M. (2007). Doing a successful research project: Using qualitative or quantitative methods. London: Palgrave Macmillan.
Dawson, C. (2009). Introduction to research methods: A practical guide for anyone undertaking a research project. NY: How To Books Ltd.
Heppner, P. & Claiborn, C. (1989). Social influence in counselling: A review and critique. Journal of Counselling psychology, 36(3), 365-387.
Huck, S. & Cormier, W. (1974). Reading statistics and research. NY: Routledge.
Jupp, V. (1989). Methods of criminological research. London: Routledge.
Langton, L. & Hollinger, R. (2005). Correlates of crime losses in the retail industry. Security Journal, 18, 27-44.
Miles, H. (1994). Qualitative data analysis. Thousand Oaks, CA: Sage.
Noaks, L. (2004). Criminological research: Understanding qualitative methods. London: Sage.
The Cooperative Group (2008). Retail Crime Survey. Web.
Punch, K. (2005). Introduction to social research: Quantitative and qualitative approaches. London: Sage.
Robson, C. (2002). Real world research. Oxford: Blackwell Publishing.
Rodriguez, A. (1998). The dangerous discourse of invisibility: A critique of the National Research Council’s national science education standards. Journal of Research in Science Teaching, 34(1), 19-37.
Savin-Williams, R. (2002). A critique of research on sexual minority youths. Journal of Adolescence, 24(1), 5-13.
Silverman, D. (2010). Doing qualitative research: A practical handbook. London: Sage.
Tanner, J. (2003). Reading and critiquing research. British Journal Perioepr Nursing, 13(4): 170-174.
Tennen, H., Hall, J. & Affleck, G. (1995). Psychology: A review and critique. Journal of Personality and Social Psychology, 68(5), 870-884.