GUARDING DOCUMENTS BY INCREASING RISKS AND REDUCING CRIMINAL OPPORTUNITIES: APPLYNG ENVIRONMENTAL CRIME THEORIES TO FORGERY CRIMES By Roy Fenoff A DISSERTATION Submitted to Michigan State University in partial fulfillment of the requireme nts for the degree of Criminal Justice Doctor of Philosophy 2015 ABSTRACT GUARDING DOCUMENTS BY INCREASING RISKS AND REDUCING CRIMINAL OPPORTUNITIES: APPLYNG ENVIRONMENTAL CRIME THEORIES TO FORGERY CRIMES By Roy Fenoff Signature forgery is a prominent crime problem that is rarely discussed by criminologist s. Additionally , according to the National Research Council of the National Academies of Sciences (2009), it is an area of forensic science that is in need of more rigorous research. T he pres ent study has two pa rts. In the first part , the number and types of changes that typically result when a writer disguises his or her sig nature were investigated . In the second part , the effect that the level of guardianship has on the interaction between t he offender and the guardian in forgery and document fraud crimes was examined . Environmental crime theories suggest that the presence of a guardian reduces crime because it decreases the opportunity to offend (Felson & Clarke, 1998). However, in certain k inds of crimes, the presence of a guardian may not be enough to deter an offender from engaging in criminal behavior; the guardian must also be perceived by the offender as capable and skilled. In the case of forgery crimes and document fraud, it is unclea r how offenders respond to and determine capable guardianship. By understanding how offenders interpret and adapt to different levels of guardianship, effective prevention strategies may be developed. A secondary data source was used for this study. The d ata w ere collected from a sample of college students a t Michigan State University. Three primary statistical techniques were used for the data analysis portion of the project: (1) Chi - squ are test of independence, (2 ) Tukey - Kramer Pairwise Comparisons, and (3 ) Exploratory Factor Analysis (EFA ). The results for t he first part of the study sho wed that letter form, legibility, size, spelling, and spacing were the most frequent ly altered handwriting features , and the second part show ed that the participants were influenced by the scenario (i.e., level of guardianship) . Copyright by ROY FENOFF 2015 v ACKNOWLEDGEMENTS I would like to recognize my dissertation committee members, Drs. Steve Dow, David Foran, Edmund McGarrell , and Mahesh Nalla for their guidance throughout the process . vi TABLE OF CONTENTS LIST OF TABLES... ........................................................................... .............................viii LIST OF FIGURES............................................................................................................ .x LIST OF ACRONYMS.. .................................................................... ............................... x i i CHAPTER 1: INTRODUCTION ................................ ................................ ....................... 1 P roblem ................................ ................................ ................................ ............................ 3 CHAPTER 2: CURRENT PROBLEMS THAT THE STUDY WILL ADDRESS ........ 1 0 The Value of Combining Environmental Crime Theories and Forensic Science ......... 1 0 Challenges In Solving Forgery and Document Frau d Crimes ................................ ...... 12 Forensic Document Examiners and Disguised Handwriting ................................ ........ 15 Protecting the Victims by Keeping the Investigators Involved ................................ ..... 18 CHAPTER 3: THEORETICAL FRAMEW ORK ................................ ............................. 2 0 Environmental Crime Theories ................................ ................................ ..................... 2 0 Routine Activities Theory ................................ ................................ ......................... 2 1 Rational Choice ................................ ................................ ................................ ......... 22 Situational Cr ime Prevention ................................ ................................ ..................... 23 Preventing Forgery and Document Fraud ................................ ................................ ..... 24 Offender - Guardian Relationship ................................ ................................ ................... 26 Capable Guardianship and Increasing the Offender's Risk of Being Caugh t ............... 29 CHAPTER 4: METHODOLOGY ................................ ................................ .................... 37 Research Questions ................................ ................................ ................................ ....... 37 Operationalization for Level of Guardianship ................................ .............................. 39 Survey Questionnaire ................................ ................................ ................................ .... 39 Part I ................................ ................................ ................................ .......................... 39 Part II ................................ ................................ ................................ ......................... 4 0 Part III ................................ ................................ ................................ ........................ 4 1 Research Design ................................ ................................ ................................ ............ 4 1 Sample Strategy ................................ ................................ ................................ ............. 42 Procedure ................................ ................................ ................................ ....................... 47 Coding Scheme ................................ ................................ ................................ ............. 50 Analysis Techniques ................................ ................................ ................................ ..... 53 Chi - Square Tes t of Independence ................................ ................................ .............. 53 Tukey - Kramer Pairwise Comparisons ................................ ................................ ....... 53 Exploratory Factor Analysis ................................ ................................ ...................... 54 CHAPTER 5: RESULTS.................................................... . .............. .................. .......... . ... 57 Research Que stion 1 ................................ ................................ ................................ ...... 57 Research Que stion 2 ................................ ................................ ................................ ...... 6 0 vii Research Que stion 3 ................................ ................................ ................................ ...... 6 1 Research Que stion 4 ................................ ................................ ................................ ...... 7 0 Research Que stions 5 and 6 ................................ ................................ .......................... 7 1 Research Que stion 7 ................................ ................................ ................................ ...... 7 2 Research Que stion 8 ................................ ................................ ................................ ...... 7 3 CHAPTER 6: DISCUSSION AND CONCLUSION ................................ ....................... 7 6 Part I: Disguised Handwriting (Research Questions 1 4) ................................ ............ 76 Primary Disguise Strategies ................................ ................................ ....................... 7 7 Unintended Handwriting Changes ................................ ................................ ............. 7 8 Number of Handwriting Changes Made by the Study Participants ........................... 81 Participants' Success in Executing Their Intended Disguise Strategies .................... 84 Part II: Guardianship (Research Questions 5 8) ................................ ........................... 85 The Guardian's Effect on the Participants' Disguise Strategy ................................ ... 85 Participants' Thought - process Behind a Specific Disguise Strategy ......................... 86 Limitations of Study ................................ ................................ ................................ ...... 89 Directions for Future Research ................................ ................................ ..................... 91 Signature Forgery ................................ ................................ ................................ ...... 91 Environment and Guardianship ................................ ................................ ................. 94 Policy Implications ................................ ................................ ................................ ........ 94 Conclusion ................................ ................................ ................................ ..................... 97 AP PENDICES ................................ ................................ ................................ .................. 99 APPENDIX A: Definitions and Measurements of Guardian(ship) ............................ 10 0 APPENDIX B : Survey Questionnaire ................................ ................................ ........ 1 07 APPENDIX C : Consent Form ................................ ................................ .................... 11 1 APPENDIX D : Documents used for the Handwriting Exerc ise ................................ . 11 2 APPENDIX E: Case Scenarios ................................ ................................ ................... 11 3 APPENDIX F: Handwriting Coding Scheme ................................ ............................. 11 4 APPENDIX G: List of Handwriting Variables ................................ ........................... 1 19 APPENDIX H: Images Showing Intentiona l Handwriting Changes .......................... 12 1 REFERENCES ................................ ................................ ................................ ............... 12 9 viii LIST OF TABLES Table 1 . Descriptive statistics for sample population .........................................................45 Table 2 . Scenario count of the hand writing features that t he study participants said they changed ...............................................................................57 Table 3 . Top five handwriting features intentionally changed by the study participants ................... ..........................................................................59 Table 4 . Number of handwriting features changed by scenario ..........................................60 Table 5 . Results of Chi - square Test for handwriting features changed ..... ..........................61 Table 6 . Counts of Handwriting features not changed, intentionally changed, and unintentionally changed across each scenario .................................62 Table 7 . Summary of handwriting features changed by scenario .. ......................................64 Table 8 . Orthogonally rotated component loadings for the baseline condition ...................65 Table 9 . Total variance explained for the baseline ..............................................................66 Table 10 . Orthogonally rotated component loadings for the guardianship condition ...........67 Table 11 . Total variance explained for the guardianship condition .......................................68 Table 12 . Comparison of baseline and guardianship fac tor solutions ...................................70 Table 13 . Table of success rates and 95% confidence intervals for each scenario ................71 Table 14 . Table of pairwise comparisons with 95% confidence intervals for each scenario ............... ...........................................................................................71 Table 15 . Results of the chi - square test for assessing the relationship between guardianship and disguise strategies used ..................................... .........................72 Table 16 . Results of chi - square test for assessing the relationship between guardianship and the top five disguise strategies used ..........................................72 Table 17 . Ways in which the scenarios influenced .........73 Table 18 . C hi - square test for how participants were influenced by the scenarios .................73 Table 19 . Influence that the s cenarios had .................74 ix Table 20 . C hi - sq uare test for why participants were influenced by the scenario ..................75 Table 21. Definitions and Measurement s of Guardian(ship) ............... ................................ 100 x LIST OF FIG URES Figure 1. The Crime Triangle .................................................................... ............................27 Figure 2 a. Naturally written capital letter "S" ......................................................................... 52 Figure 2 b. Failed attempt to alter the capital letter "S" ...........................................................52 Figure 3 a . Naturally written capital letter "D" ........................................................................52 Figure 3 b . Successfully altered capital letter "D" ...................................................................52 Figure 4. Overview of the handwriting features intentionally changed by participants......................................................... ...............................................59 Figure 5a. Naturally written signature ....................................................................................82 Figure 5b. Altered signature ................................................... .................................................82 Figure 6a. Natural signature created for scenario 3 ................................................................83 Figure 6b. Altered signature created for scenario 3 .............................. ..................................83 Figure 7a. Natural signature created for scenario 3 ................................................................89 Figure 7b. Altered signature created for scenario 3 .............................................. ..................89 Figure 8a. Scenario 1 - Restaurant Setting............................................................................112 Figure 8b. Scenario 2 - Restaurant Setting..................................................................... .......112 Figure 8c. Scenario 3 - Restaurant Setting............................................................................112 Figure 9. Examples of altered signatures in scenario 1.......................................................121 Figure 10. Examples of altered signatures in scenario 2.......................................................122 Figure 11. Examples of altered signatures in scenario 3.......................................................123 Figure 12. Example #1........... ...............................................................................................124 Figure 13. Example #2..........................................................................................................125 Figure 14. Example #3. .........................................................................................................126 x i Figure 15. Example #4..........................................................................................................127 Figure 16. E xample #5..........................................................................................................128 xii LIST OF ACRONYMS ACFE Association of Certified Fraud Examiners ASCLD American Society of Crime Laboratory Directors CCTV Closed Circuit Television EFA Exploratory Factor Analysis FBI Federal Bureau of Investigation FTC Federal Trade Commission GAO Government Accountability Office ISMG Information Security Media Group MSU Michigan State University MSU FCU Michigan State University Federal Credit Union NCVS National Crime Victimization Survey NIBRS National Incide nt - Based Reporting System NRC/NAS National Research Council of the National Academies of Sciences UCR Uniform Crime Reports 1 CHAPTER 1 : INTRODUCTION Despite being a common and prominent crime problem with serious financial costs to society, forgery 1 is rarely discussed by criminologists. In fact, of the limited studies that have been conducted on thi s crime, the most useful ones are only found in the forensic science literature , and they lack research on the human element (i.e. , the forgers and the victims). Since these studies have focused primarily on the technical aspects of forgery, they are of li ttle use to criminologi sts. As a result, the aim in this study was to bridge the gap between criminologists and forensic scientists by advancing the knowledge on the effect that the human element has on the technical aspects of a forgery , and how this info rmation can be useful in preventing this crime or identi fying the offenders. T he offender - guardian interaction in forgery crimes was examined by looking at the forgery choices that offenders ma d e when they we re forced to interact with a guardian 2 in order to carry out their crime. Furthermore, there was a focus on a type of forgery referred to as an auto - forgery 3 (disguised signature ), which is commonly used by 1 A forgery can be defined as the fake making or altering of handwritten or electronically produced documents, artwork, or cultural artifacts with the intent to deceive or defraud (Fenoff, 2013). In this study, the term forgery refer red to a non - genuine signature that f ell into one of following categories: (1) simple, (2) simulated, (3) tracing, (4) cut - and - paste, or (5) auto - forgery. 2 In general, a guardian can be identified as a person who watches over a crime t arget and may intervene or prevent a crime from happening (Felson & Boba, 2010). In this study, guardianship wa s about scrutiny. In other words, how closely a manager - guardian examine d an individual's signature at the transaction point. More specifically, at a restaurant, retailer, and bank setting. 3 The term auto - forgery is a controversial term used to describe a signature that has been deliberately altered by the author to later deny its authenticity ( ASTM, 2007; Harrison, Burkes, & Seiger, 2009; Levinso n , 2001; Scientific Working Group for Forensic Document Examination, 2013). In their book on handwriting identification, Huber and Headrick (1999) argued that auto - forgery is a contradiction of terms because while it defines a forgery of one's signature cr eated by oneself, by definition a forgery is created by another person. Nonetheless, auto - forgery is the conventional term used to describe an intentionally disguised signature , and it w as used throughout this paper. It wa s beyond the scope of this study t o clarify the auto - forgery concept. 2 people who want to obtain goods and/ or services without paying for them. The rationale for this f ocus was advancing the understanding of a rarely studied kind of forgery, by identifying the handwriting features that we re intentionally and unintentionally altered by the forger. The four specific goals of this study we re as follows. The first goal was to understand the thought process of individuals as they devis ed a strategy to disguise their signature ( creat ed an auto - forgery) under different conditions and circu mstances. This knowledge would be valuable to forensic document examiners and investigator s who encounte red cases involving disguised handwriting, as it will provide them with additional information about the factors that forgers take into consideration when disguising a signature. The second goal was to identify the primary (intended handwriti ng changes) and secondary (unintended handwriting changes) 4 disguise techniques used by individuals who disguis ed their signature in order to later deny signing a document. This information will enable investigators to recognize the primary and secondary c hanges that typically result from employing a particular disguise strategy. Furthermore, one of the weaknesses identified by the National Research Council of the National Academies of Sciences (NRC/NAS) in their 2009 report titled Strengthening Forensic Sc ience in the United States: A Path Forward , was addressed. The report highlighted the need to increase scientific rigorous research (NRC/NAS, 2009). T he third goal was to enhance criminal justice of manager - guardianship 5 , particularly in crimes of forgery and document fraud. This knowledge to detect 4 See Appendix G for a list of the handwriting variables used in this study. 5 A manager - guardian looks after a place, securing it from crime (Felson, 2006). In this study the manager - guardians wer e waiters, waitresses , cashiers, and bank tellers. 3 their disguise strateg y , and how they adjus ted their behavior accordingly. The fourth goal was to examine the value of marrying social science (e.g., environmental criminology theories) with forensic science (e.g. , forensic document examination) to study a specific crime proble m , and develop a crime prevention strategy that incorporates both a traditional crime prevention component and a behavior alteration element targeted specifically towards the undeterred offender. Considering the persistent and growing threat of check and c ard fraud on society, the implications for understanding the offender - guardian relationship in these types of crimes wer e profound. Knowing how offenders perceive and are affecte d by the guardian will enable criminal justice and criminology scholars to dev ise potentially simple and cost effective measur es that prevent these crimes, or alter the behavior of those who remain undeterred in such a way that forensic document examiners and investigators can successfully evaluate the signature(s) in question and d etermine authenticity. The Problem The available information on the financial costs of fraud is limited and not very helpful in understanding the scope and scale of specific types of fraud. Unfortunately, official sources of crime data such as the Nati onal Incident - Based Reporting System (NIBRS), the Justice Department's Bureau of Justice Statistics National Crime Victimization Survey (NCVS), and the FBI's Uniform Crime Reports (UCR) are not helpful in identifying the number and the different types of f orgery crimes that occur each year. For instance, even though the UCR contains arrest data on forgery crimes (Barnett, 2000 ; Federal Bureau of Investigation, 2014 ) , it does not identify the prevalence and type of forgery committed. Also, while the NIBRS ca ptures more offense types than the UCR and has the ability to collect more specific incident characteristics (i.e. , location type, property description, offender information, and type of victim), its 4 information is limited to known criminal incidents and t he specific types of forgery crimes are not well defined and are largely unknown. As for the NCVS, it collects information on property crimes like the UCR and the NIBRS, but it does not collect or report on the number and types of forgery crimes. In summar y, official sources of crime data cannot be used to understand the scope and scale of forgery crimes. The costs associated with financial fraud , and their impact on businesses and the greater economy have been investigated by several organizations over the past few years. In 2003, Klynve ld Peat Marwick Goerdeler , a company that provides worldwide audit, tax, and advising services, released the results of its telephone - based fraud survey in which 459 executives from private, public, and government organizati ons took part. Klynveld Peat Marwick Goerdeler (2004) found that 75 percent of the organizations that participated had experienced an instance of fraud in the past 12 months, representing a 13 percent increase from their 1998 fraud survey . Furthermore, whe n compared to earlier studies, the different types of fraud occurrences reported almost doubled. For example, check fraud increased from 26 to 40 percent, credit card fraud from 13 to 20 percent, expense account abuse from 13 to 36 percent, and payroll fra ud from 3 to 12 percent. For those companies surveyed, the total estimated losses resulting from fraud were more than $296 million annually. Identifying fraud risks and implementing effective preventive and deterrent measures are paramount steps to mitigat e financial fraud. The Information Security Media Group (ISMG), a media company that focuses on information security, risk management, fraud, and compliance, released the results of its 2014 fraud survey in which more than 300 banks, credit unions , and non - banking financial institutions participated. ISMG found that 78 percent of the respondents reported that financial losses from fraud had either increased or remained steady over the 5 previous year , and that the most common types of fraud experienced by the surveyed organizations were credit/debit card fraud (65 percent), phishing (46 percent), check fraud (43 percent), A utomated C learing H ouse /wire (26 percent), and third - party P oint O f S ale skimming (19 percent). The surveyed organizations indicated that t heir two biggest challenges in preventing fraud was lack of customer awareness (50 percent) and insufficient resources (42 percent). Although 73 percent of respondents rat ed themselves at average or below average in developing and providing customer awaren ess programs, only 20 percent of them pla nned to increase customer awareness programs, and 29 percent pla nned to invest more fraud prevention resources into fraud detection and monitoring systems in 2015 (ISMG, 2014). Financial fraud crimes have become mor e prevalent in recent years and they keep growing as technological advancements continue to flourish. For instance, with bills being conveniently paid online and goods and services being easily purchased with the swipe of a debit or credit card, the amount of debit/credit card fraud has rapidly increased. In this regard, in 2011, the Federal Reserve Bank of Boston's Payment Strategies Group conducted a study on payments - related fraud experienced by financial institutions in the Federal Reserve First Distric t. A total of 70 financial institutions in New England responded to the survey. Of all the different types of payment methods, the signature debit card was reported by 93 percent of the financial institutions as the type that had the most exposure to fraud . PIN debit card and check fraud attempts were the next highest categories at 49 percent. While the percentage of check fraud attempts reported by banks was 67, thrifts and credit unions only reported 33 percent (Federal Reserve Bank of Boston, 2012). Othe r authors have reported similar findings. For example, the Federal Reserve Bank of Minneapolis identified signature debit cards as the biggest problem , with 79 percent of the financial institutions reporting this type of fraud , followed by checks at 43 6 per cent and credit cards at 18 percent (Federal Reserve Bank of Minneapolis, 2012). Likewise, the Federal Reserve Bank of Dallas reported signature debit cards as the main type of fraud experienced by financial institutions (72 percent), followed by checks (5 1 percent), and credit cards (19 percent) (Federal Reserve Bank of Dallas, 2012). Also pervasive in society is occupational fraud. As defined by the Association of or personal enrichment through the deliberate misuse or application of the employing the Nations on Occupational Fraud and Abuse, which considered 1,483 cases of fraud reported by certified fraud examiners. The survey participants estimated that a typical organization loses five percent of its annual revenue to fraud. When this wa s applied to the US estimated gross domestic product, which was $16.8 trillion in 2 013 (World Bank, 2014), the losses translated into about $840 b illion annually. The ACFE also found asset misappropriation to be the most common form of fraud, encompassing 85 percent of cases; the government, banking and financial services, public adminis tration, and the manufacturing sector to be the most common victims of occupational fraud; and small businesses to be disproportionately victimized by this kind of fraud (ACFE, 2014). Along with businesses and organizations, individuals are also victims of the current fraud epidemic. In 2009 , the Deputy Assistant Attorney General for the Department of Justice's Criminal Division branded identity theft as one of the fastest growing crimes in the US (D epartment of Justice , 2010). This was supported by the F ederal Trade Commission's (FTC) claim that more than 9 million people have their identity stolen each year , with financial identity theft specifically being the m ost common (White, 2012). T oday, identity theft is the most 7 common form of fraud reported to t he FTC (2014a). It has been estimated that over 900 million since 2005 , and the monetary losses from these breaches can reach hundreds of millions of dollars (Pr ivacy Rights Clearinghouse, 2014). However, the National White Collar Crime Center warn ed that it is difficult to estimate or predict the number of compromised records that will be, or may already have been, used fraudulently ( National White Collar Crime C enter , 2008). T he sensitive nature of the information contained in these stolen records, especially social security and credit card numbers, makes consumers vulnerabl e to identity theft (Guy, 2014), which has been the top consumer complaint for thirteen ye ars in a row (FTC, 2014a) . Furthermore, credit card fraud was the second most common form of identity theft reported to the FTC in 2012 (FTC, 2014b) . The trend toward an increased use of debit and credit cards wa s further illuminated by a retail point - of - sale study conducted by Javelin Strategy & Research in 2012. Data were collected through an online survey of 3,210 randomly selected US consumers. The authors found that although purchases with cash were numerous, these generally had a lower transaction v alue. While cash was only used 27 percent of the time for higher value point - of - sale transact ions, debit and credit cards were used 60 percent of the time. The authors argued that if the use of debit and credit cards becomes more prevalent by 2017, the ove rall number of cash sales, no matter the transaction value, will decrease (Javelin Strategy & Research, 2012). Although some forms of financial fraud, such as phishing schemes, telemarketing fraud, and computer related crimes increase in number over time (IC3, 2013; Liesik, 2010; N ational C onsumers L eague , 2010), no other kind of financial fraud has been as persistent and successful as the crime of forgery, which can be traced back to Roman law under the Code of Justinian in 8 539 AD (Baker, 1955) . According to Baker (1955), forgery was a lucrative business during the Middle Ages, especially for people holding high office positions in Rome and throughout Eu rope. In fact, the author claimed that forgery was practiced in every country where handwriting was a me diu m of communication . Currently , most financial frauds and cases of identity theft include a forged or altered document of some sort (FTC, 2014b). With the rapid advancements in computer and printer technology, the creation of counterfeit and forged docum ents has increased and become an easier and more profitable criminal occupation than in the past (Agar, 2011; Sparacello, 2012). Although forged 6 and altered documents such as tax returns, wills, deeds, and trusts can have a considerable and immediate im pact on the families and individuals involved, forged checks have received the most public attention as far as questioned documents are concerned (Foster, 2012; Tuohy, 2 013). In this regard, the United States Secret Service estimated that in 2007 check fra ud in the US amounted to more than $5 billion annually (Shreiner, 2013). In fact, the chief of the Secret Service Financial Crimes Division claim ed that check fraud wa s the number one way criminals attac ked ational C heck F ra ud C enter , 2013). Furthermore, the National White Collar Crime Center (2009) claim ed that 500 million checks are forged annually in the U . S ., equating to about $10 billion in losses . The F ederal B ureau of I nvestigation lure Report for the fiscal years 2006 and 2007 indicated that the number of suspicious activity reports (SARs) for check related fraud accounted for 27 percent of the 3,186,213 SARs filed by US financial institutions, resulting in 6 In addition to the different types of forged documents discussed in this section, there are many other kinds of documents that are commonly forged including: social security applications, immigration applications, drive license applications, tax returns, and property conveyance documents (deed s ), among others. Wherever documents are relied upon for identification, the potential for forgery to happen is enormous. Unfortunately, the lack of available data makes it impos sible to quantify the true cost of this crime in all of these document areas. 9 more than $21.4 billion i n losses ( F ederal B ureau of I nvestigation , 2008). Furthermore, the U nited S tates Department of the Treasury Financial Crime s Enforcement Network's SAR Activity Review, By the Numbers sh owed that 27 percent of the SAR s filed by depository institutions in 20 09 could be attributed to fraud - related activities, and that check fraud was one of only two categories that ha d experienced an increase in SAR reports between 1996 and 2009 ( Financ ial Crimes Enforcement Network , 2010). A survey completed by the Associatio n of Financial Professionals indicated that 60 percent of corporate member organizations experienced either actual or attempted payment fraud in 2013 ( Association of Financial Professionals , 2014). From those, 82 percent were victims of check fraud , follow ed by credit/debit card fraud at 43 percent ( Association of Financial Professionals , 2014). Moreover, the Ame rican Bankers Association found that 37 percent of banks reported losses due to check fraud to be an estimated $648 million , and 54 percent of bank s reported industry losses from credit/debit card fraud (e.g., point - of - sale signature and pin number) to be an estimated $943 million in 2012 ( American Bankers Association , 2013). The Nat ional Check Fraud Center claim ed that checks we re the payment metho d most frequently targeted by criminals to commit payment fraud , and that 80 percent of business - to - business transactions we re made by check ( Nat ional Check Fraud Center , 2013). Although the Ame rican Bankers Association predict ed that the number of checks used in business transactions will decrease slightly in the coming years, it warn ed that the number of fraudulent checks may actually increase. 10 CHAPTER 2 : CUR RENT PROBLEMS THAT THE STUDY ADDRESS ED The Value of Combining Environmental Crime Theories 7 and Forensic Science While most of the information available on forgery and document fraud c ame from surveys conducted on businesses or information reported to government agencies such as the FBI, the United S tates Secret Service, and the FTC, these stud ies and reports d id not provide any details regarding the conditions under which check and card fraud take place. As a result, most of the knowledge about these crimes wa s deduced from the limited information that wa s available. For example, the studies co mpleted by the Federal Reserve B anks clearly show that check and biggest problems. This suggest ed that most check and card fraud involves altered and forged signat ures. Although this deduction wa s helpful, it d id not explain how these signatures were produced and under which circumstances the documents were signed. Although disguised handwriting is regularly encountered in the field of forensic document examination , distinguishing disguise from different authorship remains a problem area for many investigators (Bird, Found, & Rogers, 2010 ; Bird, Found, & Rogers, 2012 ), which in part may be due to a serious lack of empirical work on this subject area. The findings of the current study help ed fill in the gap in the criminological literature, more specifically in the environmental crime literature, in regards to crimes of forgery and document fraud by exploring the offender - signature fraud, and the kind of disguise strategy employed by the offender based on this 7 Environmental criminology (also referred to as crime science), is the study of crime places and events (Brantingham & Brantingham, 1991). The term "environmental" is often co nfused with crimes related to wildlife and environmental harm such as poaching, illegal logging, or pollution. However, the terms that criminologists often use to describe this area of study is conservation (Gibbs, Gore, McGarrell, & Rivers III , 2010 ) or g reen (Lynch & Stretesky, 2003) criminology. 11 capability assessment. Ultimately, this deeper understanding will assist researchers in designing additional research studies an d developing practical crime prevention measures. Although t he research described here could have been conducted using environmental crime theories or forensic science independently, it wa s by combining these two that the crimes of forgery and document f raud c ould be truly understood in a holistic manner. Traditionally, criminology has focused on offenders and the circumstances that prompt individuals to break the law. Theories of criminal disposition, which have dominated the field of criminology, are ba sed on the principle that people commit crimes for reasons that are generally outside of their control (e.g., disorganized social environment, family dysfunction, poverty, inequality, peer pressure, lack of self - control, genetics, etc.). In contrast, envir onmental criminologists argue that crime and specific crime events should be examined, not the criminal tendencies of individuals (Clarke, 1980). In this regard, the field of environmental criminology has emerged as an alternative approach to understand cr ime by focusing on criminal opportunity and the places of crime events. Clarke (1995), along with other environmental criminologists, argued that because crime will persist rather than be eliminated from society, prevention is the best approach. However, s ince some crimes will persist regardless of the kind of prevention program that is employed, albeit at a lower level, additional information should be collected to understand the undeterred offender (i.e. , his behavior and attitudes). As a result , the impa ct that incorporating behavior alteration strategies into crime prevention programs would have on increasing the likelihood of an offender being caught after committing a crime was explored . According to the FBI's 2012 UCR , a vast majority of offenders get away with their crimes, thus the need to improve the current crime prevention strategies is enormous. T he clearance rates for violent and property crimes were 46.8 percent and 19 percent , respectively (FBI, 2013). These statistics sugges ted 12 that non - viole , compared to 53 percent of violent offenders. Unfortunately , these low clearance rates are not new. In fact, they have been relatively consistent for the last 18 years (FBI, 2014). Moreov er, according to the U.S. Justice Department's Bureau of Justice Statistics , only 36 percent of non - violent crimes were reported to the police in 2013 (Truman & Langton, 2014). There is no question that crime prevention programs work (N ational I nstitute o f J ustice , 2014; Sherman et al., 1998). However, the numerous published studies that demonstrat ed their success in deterring some crimes fail to explain why some offenders remain undeterred and often get away with the crime (FBI, 2012). This is where the u nion of social science (i.e., environmental crime theories) with the technical aspects of forensic science (e.g ., forensic document analysis , fing erprint analysis, firearm and tool mark analysis, footmark analysis , DNA analysis , etc.) could be used to unde rstand crime events in a more meaningful way. Social scientists do not know much about the kind of evidence a forensic scientist or investigator needs to solve a crime, nor should they be expected to know. However, social scientists can collaborate with fo rensic scientists to better understand what kind of evidence is necessary to investigate and solve specific kinds of crime and develop crime prevention strategies that (1) deter potential offenders and (2) manipulate the physical environment so that the un deterred offenders leave behind some useful form of physical evidence (forensic evidence) that would increase their chances of being identified after committing a crime. Challenges in solving forgery and document fraud crimes Due to the prevalence of che ck and card fraud, and the overall pervasiveness of forgery and document fraud throughout society, it is expected that local law enforcement is receiving complaints about these crimes. However, to effectively help the victims, law enforcement 13 personnel mus t be knowledgeable about forgery detection and document fraud. Since the training received by police officers is predominantly geared towards the use of force and criminal law enforcement (Walker & Katz, 2005), one can assume that their understanding of fo rgery detection is limited ( U.S. GAO, 2003 ; U.S. GAO, 2008). Therefore, when they receive a report from a victim of fraud, whether it is a private citizen or a business, police agencies may have to turn to their crime laboratories for assistance in investi gating and solving the crime. Although more than 18,760 police agencies are estimated to exist in the U nited S tates and multiple state crime laboratories do not even have a q uestion ed document section, such as Montana, Tennessee, Vermont, and Wyoming (ASCLD, 2014; Brady, 2007). Considering the limited number of questioned document examination laboratories and the prevalence of forgery and document fraud, local law enforcement agencies are under extreme pressure to either move cases through the system or direct victims to seek redress in court through civil action (NBC News, enforcement agencies decide to move forward with an investigation without the assistance of a crime laboratory or a forensic document examiner, it is possible that they may be wasting valuable resources by chasing non - existent offenders or using biased information, due to their lack of training and eagerne ss to close the case (Miller, 1984; Miller, 1987; Rossmo, 2013). When police officers assume that the person claiming to be the victim of a forgery or document fraud crime is indeed a victim, their investigation focuses on finding the offender. However, in many cases of forgery and document fraud, the crime wa s actually committed by the complainant (Huber & Headrick, 1999). These - disguise their signature so that, at a later date, they can deny signing the document. By failing to consider t he victim as the suspect, investigators will not identify the 14 actual signer of the signature in question , and if they do suspect the victim, they will most likely have difficulty providing evidence that supports their suspicion. Without any meaningful trai ning in forgery detection, conducting a simple side - by - side comparison of the signatures is usually not very helpful because the differences between the known and questioned signatures will necessarily result from the intentional disguise strategy used by the off ender (Ellen, 2006). Consequently , the investigators may attribute the observed differences to different authorship, when in fact the differences may be the product of disguise. If the investigators recognize that they a re unable to identify the aut hor of the signature in question , they may inform the complainant that they cannot solve the crime due to lack of evidence, or they may assist the complainant in clearing his/her name of any wrongdoing by writing a police report that indicates that she or he was the victim of identity theft or check fraud. However, if the complainant is actually the offender, this report becomes his/her get - out - of - jail free card since it essentially prevents the holder from being pursued for that crime. In the absence of pu nishment, offenders are empowered to continue targeting businesses and individuals, while the real victims are left absorbing the loss or spending more money trying to pursue the matter through civil action. When a business is victimized , rather than seeki ng monetary compe nsation through court action, the business owner will generally opt for raising prices and passing the loss to the consumer, who then becomes a victim of the crime as well (Finklea, 2012 ; Whitt, 2008 ). As briefly mentioned above, the gene ralist nature of police work puts police officers at a disadvantage when it comes to investigating and solving crimes that require specialized skills and training. While experts in questioned documents and forgery detection complete a 2 - year training progr am and many ha ve a college degree (ASTM, 2012 ), most law enforcement personnel do not pursue a formal 2 - year training program and many are not required to have a 4 - 15 year college degree 8 (Luzer, 2010; U.S. Department of Labor, 2012). Therefore, when law enfo rcement agencies are faced with a crime they have no expertise in, and the state, county , or federal crime laboratories are unable to assist them, they must opt for one of the following options: 1) do their best at investigating and solving the case on the ir own, 2) contact a forensic document examiner in private practice to assist th em, or 3 ) advise the victim to pursue the case through civil action. Forensic Document Examiners and Disguised Handwriting While the prevalence and economic impact of disguis ed signatures is unknown, its potential impact on society can be gauged by looking at a closely related crime problem: charg eback fraud or "friendly fraud." This kind of crime occurs when consumers use their credit card to pay for a product or service , and then request a chargeback from their bank claiming that they did not make the purchase, when in fact they did (Brown, 2014). In 2012, Mindwave Research released the findings of their Online Fraud Report, which surveyed U.S. and Canadian online merchants t o assess online payment fraud trends. The merchants participating in the study reported a total fraud loss of $83 billion in 2011 , with chargeback fraud accounting for 41 percent of the ir losses (Cybersource, 2012). The 2013 LexisNexis True Cost of Fraud S tudy reported that merchants pay $2.79 in costs for each dollar of fraud losses (LexisNexis, 2013). Although chargeback fraud sometimes does not include a signature, the thought process and scheme (getting something for nothing) is similar to the one used by individuals who disguise 8 A 4 - year degree from an accredited college or university and a 2 - year apprenticeship style training program is required to become certified in forensic document exa mination . Many law enforcement officers do not have a 4 - year college degree, and many law enforcement agencies do not have the resources to support an extensive 2 - year training program. Therefore, it is difficult for law enforcement officers to become certified in forensic document examination . 16 their signature in a document so they can later deny signing it. However, unlike chargeback fraud, in cases of disguised handwriting the offender always leaves physical evidence behind. Unfortunately, the quantity and quality of this evidence often varies , and in many cases may not be useful for identification purposes. Dwindling resources and lack of training in forgery detection prevent most law enforcement agencies from adequately investigating forgery and document fraud cr imes, leaving victims defenseless (COPS, 2011; Hamari, 2010; Howard, 2005; New York Post, 2010). Moreover, in cases of disguised handwriting where the complainant can be the actual offender, it seems likely the police are generally unprepared to respond. A s a result, law enforcement professionals often seek the assistance of an expert in handwriting identification to assist in forgery cases. When evaluating documents in which the writing is suspected to have been disguised, the forensic document examiner is trained to look for certain handwriting features that are commonly found when an individual attempts to disguise his/her writing in order to avoid being identified. In the literature of forensic document examination , there are two kinds of sources that gu ide the examiner when evaluating a questioned writing for disguise: books and journal articles. As for books, in many of the ones written on this topic (Conway, 1959; Ellen, 2006; Harrison, 1958; Hilton, 1982; Huber & Headrick, 1999; Osborn, 1929; Ro bertso n, 1991), the authors relied on their personal experience or the experience of others to identify the handwriting features that we re commonly found when a writer employed a disguis e strategy. In addition, there we re empirical studies that investigate d disg uised writing, many of which focus ed on anonymous letters and notes ( Alford, 1970; Harris, 1953; Keckler, 1997; Konstantinidis, 1987; Leung, Chung, Tsui, & Cheung, 1988; Regent, 1977; Seaman - Kelly, 1999; Tsui, 1997; Willard, 1988 ) , rather than disguised si gnatures. The few studies that investigate d disguised 17 signatures had small sample sizes and did not consider how intentional changes may have impact ed other aspects of the writing (Herkt, 1986; Michel, 1978; Mohammed, 1993 ; O'Block, 1991 ). Although the di sguise strategy used to write an anonymous letter and sign a document may seem similar, they can be quite different. In the case of disguised writing for anonymous letters (e.g., bomb threat) , individuals are trying to avoid being identified and are not co ncerned about getting the writing past a guardian. For example, the author of an anonym ous letter is not concerned about getting his disguised handwriting past a bank teller or retail cashier; he just wants to conceal his identity. As a result, he will alt er his handwriting as much as possible and choose a technique that allows him to maintain his concentration long enough to maintain the disguise without reverting back to his normal handwriting habits (Ellen, 2006 ; Huber & Headrick, 1999 ). As for disguised signatures , the individual may or may not be concerned about getting his signature past a guardian. If he is not, he can change his signature in such a way that it looks nothing like his regular one. For instance, the individual may scribble illegible loo ps and lines that do not resemble his original signature in any way. However, if the individual is worried about getting his signature past a guardian, he can alter his signature in a way that it maintains some pictorial similarity 9 to his normal signature , but is dissimilar enough that at a later point in time he can claim that the signature is not his. If the latter happens, it can become a serious challenge for the forensic document examiner to interpret the similarities and differences found between the questioned and normal signatures. For example, a forensic document examiner may interpret a strategically planned disguise, such as a change in letter - form or a change in the legibility of the signature, as a notable difference and wrongly conclude that t he normal and 9 Pic torial similarity refers to the handwriting features that give a body of writing a certain "look ," such as letter form, slant, spacing, and size. 18 disguised signatures are not of common authorship. Furthermore, the forensic document examiner may wrongly interpret that a difference is the result of disguise when it could actually be the result of different authorship. In fact, various re search conducted by scientists at the Forensic Expertise Profi ling Laboratory at La Trobe Univer sity in Victoria, Australia has indicated that many forensic document examiners have difficulty identifying disguised handwriting (Bird, Found, Ballantyne, & Ro gers, 2010; Bird et al. , 2012; Dyer, Found, & Rogers, 2008; Found & Rogers, 2008). Therefore , it should not come as a surprise that law enforcement agencies, which are not skilled in forgery detection unlike forensic document examiners, are ill equipped to identify disguised handwriting. Protecting the Victims by Keeping the Investigators Involved In many cases of forgery and document fraud , it is unlikely that the investigator will be able to identify the forger (Huber & Headrick, 1999; Osborn, 1929). Th is is especially true in cases where the differences between a known and a questioned handwriting sample are grotesque (i.e. , the signature looks nothing like the original one). In these cases, the forensic document examiner, or investigator, cannot comple tely identify or eliminate the complainant as the author of the questioned signature. Therefore, if the complainant is truly a victim, he or she will receive no compensation , and if the complainant is the offender (i.e. , disguised signature), he or she wil l receive no punishment. However , in cases where the differences between a known and a questioned signature are minor, some sort of case closure is likely. In other words, a person disguising his signature by simply changing one or two handwriting features will more likely be identified as the author of the signature in question and subsequently punished by being made to pay for the goods and services received , or worse, being prosecuted for the fraud. Considering this, it would be advantageous to create an environment in which the forger believes it is 19 necessary to avoid deviating too far from the normal signature because he knows the recipient (guardian) will compare the signature with an identifying signature on a driver's license or another document. By making the forger believe that the recipient will be taking a closer look at the signature, the forger may change his behavior. For example, instead of creating a signature that looks nothing like his normal one (e.g., an X mark or a straight line), the fo rger may only employ a few deviations from his normal signature , such as changes to the capital letters. As a result, when the forensic document examiner or investigator is asked to evaluate the questioned signature, he will have little difficulty in deter mining its authenticity. 20 CHAPTER 3 : THEORETICAL FRAMEWORK The theoretical focus of the current study wa s on the guardianship component of routine activity theory and the role that guardians play in preventing forgery crimes. This theore tical framework was developed from environmenta l crime theories . Environmental Crime Theories Grounded in opportunity theories, environmental criminology is a term used to describe a group of theories about crime events (Cullen & Agnew, 2006). In contras t to traditional crime theories, which focus on the offenders and the circumstances that prompt them to break the law (i.e. , theories of criminal disposition), environmental criminologists argue that crimes and specific crime events should also be examined . Since the offender is only one element of a crime, by studying all the aspects of a specific crime event, researchers and practitioners can be in a better position to create crime prevention strategies that reduce opportunities for crime. A considerable amount of theoretical developments have taken place over the last 30 years in the field of criminology. In environmental criminology, these developments have been used to explain crime events and guide practitioners in the creation and implementation of c rime prevention programs (Clarke, 1992; Knutsson, 2006; Sherman, Farrington, Welsh, & MacKenzie, 2002). Three theoretical perspectives that have been developed to understand crime events are: routine activit y theory, rational choice, and situational crime prevention. When combined, these perspectives capture the thought process of offenders: how they make the decision of when and where to commit a crime, what makes a suitable target, and how guardians can protect targets (Eck & Weisburd, 1995; Felson & Boba , 2010). Below is a brief description of each perspective. 21 Routine Activity Theory While exploring the relationship between humans and their environment, Cohen and Felson (1979) develop ed the routine activity approach to explain the changes in crime rate s over time. They claimed that the increase of crime rates could be attributed to the societal and technological changes that took place after World War II. Cohen and Felson (1979) argued that uld be understood. Accordin g to the routine activity theory , three factors must be present for a crime to occur: a motivated offender, a suitable target, and the absence of a capable guardian. The probability of these three elements coming together in spac e and time is influenced by routine activi ty patterns such as work , travel, and recreation (Cohen & Felson, 1979). Therefore, structural changes in these patterns can impact crime rates by increasing crime opportunities (Felson, 1987). Since its origin i n 1 979, the routine activity approach experienced several major developments 10 . The most remarkable of these was the expansion of the elements needed for a crime to occur: place, controllers, and super controllers (Sampson, Eck, & Dunham, 2010). In this reg ard, place managers have been added to regulate behavior in the places they control (Eck & Weisburd, 1995) , and while the guardian exercises control over the target or victim, an intimate handler has been added to exert some control over a motivated offend er (Felson, 1986; Tiller & Eck, 2011). This expansion of the routine activity theory has not only increased the understanding of crime events, but it has provided additional potential points of intervention for other crime problems (Bossler & Holt, 2009; H ollis, Fejes, Fenoff, & Wilson, 2014; Hollis & Wilson, 2014; Madensen & Eck, 2008; Pires & Clarke, 2012; Reyns , 2013; Tillyer & Eck, 2011; Warchol, 2011). 10 For example, Danielle Reynald elaborated on the concept of guardianship and provided a new means of measurin g it by directly observing human behavior (Reynald, 2009; Reynald, 2010; Reynald, 2011a; b; c). 22 Rational Choice While routine activity theory identifies the necessary elements for a crime to occu r, the rational choice perspective describes how offenders make decisions (Tillyer & Eck, 2011). Based primarily on economic theories of crime, the rational choice perspective asserts that crimes are the result of rational choices made using a benefit - cost analysis; an offender will commit a crime if the expected utility exceeds the expected costs (Cornish & Clarke, 1987; Jacobs, 2010). Consequently, deciding to participate in crime is a two - part process. First, individuals must be willing to commit a crime (the involvement decision), and then they must decide what crime they want to commit (the event decision), which is strongly influenced by the immediate situation (Cornish & Clarke, 1986; Kroneberg, Heintz, & Mehlkop, 2010). For instance , after making the decision to perpetrate a robbery, the offender determines when, where, and how the robbery will - making (Garcia - Retamero & Dhami, 2009; Jacobs & Wright, 2010), the rational choice perspective acknowledges that the information an individual uses to make a decision may be inaccurate or impaired by situational changes, such as drug and alcohol use or moving to an unfamiliar location (e.g., new neighborhood, different facility, etc.). - making ability evolves over time since it is shaped by experience. Through experience, offenders learn which discriminative cues are associated with good targets , and they use these positive cues to make decisions about future cr ime events (Jacobs, 2010). Al though the positive reinforcement obtained from previous crime events can lead to an increased frequency of offending (Tillyer & Eck, 2011), Cornish and Clarke (1986) pointed out that an offender may also choose to st op committing crimes if more attractive alternatives to crime are available. Furthe rmore, the authors argue d that the decision to discontinue criminal activity may be influenced by a negative 23 experience during a criminal event or by a change in the offende Supporters of rational choice theory argue that crime prevention strategies should focus on eliminating crime opportunities and/or shifting the perceived benefits and costs, rather than focusing on the criminal tendencies of ind ividuals. Situational Crime Prevention Situational crime prevention aims to reduce crime opportunities through the manipulation of the immediate environment. It is grounded in the rational choice perspective , but it draws from the routine activity theory to deal with specific forms of crime (Tillyer & Eck, 2011). The concept of situational crime prevention was first introduced in 1980 by Ronald Clarke as an alternative to traditional criminological approaches to crime prevention. Clarke (1980) argued that According to Clarke, changing criminal dispositions is practically impossible because rehabilitation is difficult, and trying to eliminate the root causes o f crim e is unfeasible . Therefore, manipulating situational factors is the most practical strategy to prevent crime (Clarke, 1995). Clarke (1983) argued that crime involvement decisions are rational because individuals weigh the costs and benefits of committing a crime. Thus, prevention strategies could change the offender's decision - making process by manipulating the immediate environment and altering the perceived benefits and costs of crime (Clarke, 1983; Tillyer & Kennedy, 2008). Unlike the generic approaches to crime prevention, which apply various techniques to a broad range of problems, situational crime prevention focuses on a precise problem, in a specific place, under particular circumstances, and it uses this information to devise ways to make the targe t unattractive to criminals. In this regard, Brantingham and Brantingham ( 2005 ) and Cornish and Clarke (2003) identified some factors which include removing excuses, increasing risks and 24 efforts, and reducing rewards and provocations , that could be leverag ed to persuade potential offenders that committing a particular crime would be a bad decision. Preventing Forgery and Document Fraud While crimes such as burglary and robbery have received a vast amount of attention by environmental criminologists (Berna sco & Block, 2009; Bernasco & Block, 2011; Bernasco & Luykx, 2003; Braga, Hureau, & Papachristos, 2011; Brown & Bentley, 1993; Coupe & Blake, 2006; Coupe & Blake, 2011; Garcia - Ratamero & Dhami, 2009; Jacobs, 2010; Palmer, Holmes, & Rollins, 2002; Tilley, T seloni, & Farrell, 2011; Tseloni, Wittebrood, Farrell, & Pease, 2004; Wright, Logie, & Decker, 199 5), other crimes have received little attention. Considering the tremendous impact that forgery and document fraud have on society, it is surprising that thes e kinds of crimes are rarely discussed in the environmental crime literature. In fact, only a few articles related to forgery and document fraud (Knutsson & Kuhlhorn, 1992; Levi, 1998; Levi, Bissell, & Richardson, 1991; Lacoste & Tremblay, 2003; Masuda, 19 96; Mativat & Tremblay, 1997; Newman, 2003; Smith, 1997; Tremblay, 1986; Webb, 1996) were found after conducting a review of this literature, and a majority of these studies focused on credit card fraud. Furthermore, most of these articles were written at least 15 years ago , and many aspects of society have changed since then. Unlike robbery and burglary crimes, which employ prevention strategies that were effective 15 years ago and still are today (Coupe & Blake, 2011; Felson & Boba, 2010; Wright & Decker, 1996; Wright & Decker, 1997), crimes that are greatly impacted by changes in technology must be investigated regularly so that researchers and practitioners can keep pace with the criminals and devise crime prevention strategies that are relevant to the p has changed the way in which some crimes are committed, and social media sites such as 25 Facebook and Twitter have created new kinds of crimes for law enforcement and security professionals (Giles, 2013; Griggs, 2012; IC3, 2011; Peachey, 2012). The same is true for forgery and document fraud. Al though forgery is an old type of crime and the handwriting identification process has remained unchanged, the technology used to create documents and complete financial transactions has changed (Iannacci & Morris, 2000). For instance, high quality fake identity documents can be created without difficulty using inexpensive computer software and printers that are readily available to the general public (Fenoff & Spink, 2013). Furthermore, the rapid advances in technology have resulted in paperless documents requiring electronic signatures. To ensure the validity and legal effect of contracts entered electronically (PL 106 - 229, 2000) , Congress enacted the Electronic Signatures in Global and National Commerce Act ("ESIGN") on June 30, 2000, which facilitates the use of electronic records and signatures in interstate and foreign commerce. In his book, Crime and Nature , Marcus F elson (20 06) stated that in order to understand crimes and document fraud require, in most situations, the direct interaction between the offender and the guardian. Therefore, it would seem as though an active guardian would play a vital role in detecting and deterring forgery and document fraud crimes. However, criminal guardian, and h ow the latter may prevent forgery and/ or document fraud, is extremely limited due to a lack of research in this area. Consequently, further research on how environmental criminology and guardianship can be used to explain forgery and document fraud would c ontribute greatly to the literature on environmental criminology. 26 Offender - Guardian Relationship Some scholars argue that routine activity theory has attracted the attention of researchers who have focused on target suitability and offender motivation, but have overlooked capable guardianship (Knuts son & Clarke, 2006; Sampson et al. , 2010). This is surprising considering that guardianship is the most important component of the crime event model , as suggested by Felson and Boba (2010). Indeed, by disrupti ng the interaction between a motivated offender and a suitable target, a guardian can directly or indirectly prevent a crime from occurring (Hollis - Peel, Reynald, van Bavel, Elffers, & Welsh, 2011). From its original conception, guardianship has evolved i nto three types: handlers, managers, and target - guardians (Sampson et al., 2010), which collectively are referred to as controllers. Handler - guardians are those individuals who have an emotional connection to the offender, and can be a parent, sibling, fri end, school teacher, or religious leader (Eck, 1994; Felson, 1995; Felson & Boba, 2010; Sampson et al., 2010; Tillyer & Eck, 2011). Manager - guardians are individuals who are responsible for controlling and monitori ng behavior at specific places , such as st ore clerks, landlords, home owners, bank tellers, waitresses, flight attendants , and janitors (Eck, 1994; Sampson et al., 2010). Target - guardians are those individuals who protect targets. They are the most widely studied and discussed form of guardianship , and they include ordinary citizens and groups of people who may or may not be known to the target of the crime, for example, a neighbor (Sampson et al., 2010). Taking into consideration this new concept of guardianship, Eck (1994; 2003) proposed that t he "crime triangle" sh ould consist of two layers ( Figure 1): the inner layer representing the three elements necessary for a crime to occur and the outer layer representing the three types of guardians (controllers) that may prevent a crime. According to E ck (2003), the intervention of 27 any one controller can be enough to prevent a crime from taking place. Felson and Boba ( 2010) assert ed that the three forms of guardianship are interconnected in the sense that a crime is accomplished when a motivated offende r moves away from handlers, toward a place without a manager, and a target w ithout a guardian. Figure 1. The Crime Triangle Adapted from Clarke and Eck, 2009 While the studies that have been conducted on guardianship have improved criminal justice and cr have done so in a narrow way. Most of these studies have focused on target - guardians and their role in preventing specific types of property crime (e.g., burglary, robbery, auto - theft). Although research on property crime has consistently shown that target - guardianship is an important factor in reducing the likelihood of crime ( Barclay, Buckley, Brantingham, Brantingham, & Whinn - Yates, 1996; Barclay & Donnermeyer, 2011; Benne tt, Holloway, & Farrington, 2006; Bernasco & Luykx, 2003; Brown & Bentley, 1993; Coupe & Blake, 2006; Felson, 1995; Felson, 2006; Garofalo & Clark, 1992; Lynch & Cantor, 1992; Madensen & Eck, 2008; Pennell, Curtis, Henderson, & Tayman, 1989; Reynald , 2009; Reynald, 2010; Reynald, 2011a; Wilcox, 28 Madensen, & Tillyer, 2007; Wortley & McFarlane, 2011; Wright & Decker, 1996), some scholars have realized that effective target - guardianship requires more than the simple insertion of a warm body for surveillance pur poses (Kenney, 1986; Reynald, 2010; Reynald, 2011a; Tilley, & Webb, 1994; Welsh, Mudge, & Farrington, 2010). According to Lynch and Cantor (1992), active monitoring is a crucial component of capable target - guardianship. In a study of residential burglary c onducted by Brown and Bentley (1993), the authors found that in some cases the risk of being seen was not sufficient to deter burglars; they also needed to know that the s observation that active guardianship is necessary to prevent crime has been further supported by other researchers. For instance, Reynald argued that monitoring alone is not good enough to prevent and deter crime over the long run; the guardian must be w illing to intervene when necessary (Reynald, 2010; Reynald, 2011b). A good analogy of the necessity of active guardianship is the scarecrow. The scarecrow cr op s . However, if the scarecrow remains in the same location and position for too long, the crows eventually learn that it is not a real threat , so they ignore it and attack the crops. Likewise, in the event of a crime, a guardian (e.g., a neighbor) may act as an initial deterrent to some burglars, but if the guardian is not willing to intervene, the deterrent effect will eventually wear off and a burglary will be more likely to occur. Similar observations have been reported in studies conducted on shoplifti ng and the effectiveness of closed circuit television ( CCTV ) surveillance cameras (Carroll & Weaver, 1986; Farrington, Bowen, Buckle, Burns - Howell, Burrows, & Speed, 1993; Phillips, 1999; Welsh & Farrington, 2009). Although environmental criminology has ad vanced the understanding of crime and crime prevention in many remarkable 29 ways, the current knowledge on guardianship comes from research focused primarily on crimes such as burglary, robbery, theft, shoplifting , and auto - theft, which have received a lot o f attention from environmental criminologists. However, there is no evidence indicating that the offender - guardian relationship in these crimes is similar to the one in other types of crimes. Furthermore, the literature on guardianship fails to address how guardians can prevent crime in situations when they must interact with the offenders. One of the central tenets of guardianship is that crime is more likely to happen when a guardian is absent than when she or he is present (Felson, 1995). This is because watching, thus she or he avoids carrying out the offense (Felson & Boba, 2010). However, in forgery and document fraud crimes , the offender often interacts directly with the guardi an, thus the traditional views of guardianship are insufficient in understanding and explaining those crime problems that occur in places where the guardians directly interact with the offenders. Indeed, an offender who is planning to forge a document expe cts to interact with a "watchful" guardian. y be to prevent the forgery and/ or document fraud crime being caught after committing the crime increases considerably. Consequently, in order to understand how guardians influence the forgery and document fraud crimes that happen at specific places, the current definitions of guardianship must be expanded upon to include indivi duals who often have to directly interact with the undeterred offender. Traditionally, the definition of guardianship has acknowledged the role of target - guardians in preventing crim es. In their groundbreaking article, which introduced routine activity theory to the field of criminology, Cohen and Felson (1979) operationalized (or 30 measured) guardianship by a person's participation in the labor force and living in a single - adult house hold. This way of operationalizing guardianship allowed the authors to test their hypothesis that, "the dispersion of activities away from households and families increase opportunities for crime, which results in higher crime rates" (Cohen & Felson, 1979 , p. 588 ). In other words, individuals who are employed outside of the home and who are living in single - adult households are less likely to spend time at home, which results in a higher rate of victimization. Shortly after their 1979 article, Felson and Cohen (1980 , p. 392 ) defined guardianship as "any spatiotemporally specific supervision of people or property by other people which may prevent crim inal violations from occurring." Since then , guardianship has been defined in numerous ways and multiple var iables have been identifie d to operationalize it. The various ways in which guardianship has been defined and measured in the environmental crime theory literature are shown in Appendix A . G uardianship was defined and measured in a relatively consistent ma nner in the first 25 years after Cohen and Felson (1979) first introduced this concept. During this time, many researchers emphasized the effectiveness of individuals or objects in preventing crime when defining guardianship (Cohen & Cantor, 1980; Cohen, K luegel, & Land, 1981; Eck & Weisburd, 1995; Felson, 1995; Lynch, 1987; Tseloni et al. , 2004). While individuals could be housewives, friends, neighbors, private security, police officers , or pedestrians, objects could be burglar alarms , locks, barred windo ws, cameras , or other devices that can prevent or inhibit victimization . In those studies where guardianship was not clearly defined, the manner in which the researchers operationalized this concept in order to understand how they viewed it was analyzed . F or example, Stahura and Sloan III (1988) measured guardianship by female labor force non - participation, police exp enditures , and police employment, while Bennett (1991) measured it by female labor force participation. Meithe, 31 Stafford , and Sloane (1990) us ed the number of people in the household and daytime and nighttime activity, while Garofalo and Clark (1992) measured guardianship by household occupancy and the presence of lighting, dogs, alarm system s and neighborhood watch es . Rice and Smith (2002) look ed at the number of owner - occupied places, the number of multifamily buildings, the number of single - parent homes , and the number of commercial places. Coupe and Blake (2006) used occupancy of housing at the time of burglary, detection of the burglary (i.e . , reporting the crime) , and security devices. Although these authors did not all directly define guardianship, they operationalized the concept in a way that is consistent with how it has been defined. Later on, Felson (2006) re - defined guardianship and i dentified guardians as ordinary citizens (i.e., anyone who passes by or is specifically asked to act as a guardian), who keep an eye on potential targets of crime. More recently, a guardian has been defined as "someone whose presence serves as a gentle re minder that someone is looking or those who engage in natural surveillance, including ordinary citizens going about their daily lives but providing by their presence some degree of security" (Felson & Boba, 2010 , p. 28 ). Guardians have also been described as "crime control agents who are most likely to assume primary responsibility for their residential space, including the people and property contained therein " (Reynald, 2010 , p. 363 ) . Although these definitions are more refined than previous ones, they ar e still consistent with the research that has been completed on guardianship over the last 35 years and have two common key components: (1) the supervision of people or property by other people, and (2) the prevention of crime. As a result, most of the res earch that has been conducted on guardianship is related to target - guardians and their influence on specific property crimes (e. g., bur glary, robbery, car theft, shoplifting ). 32 Even though the original conceptualization of guardianship has been expanded u pon in the last 35 years, the way it has been studied and defined during this time has remained relatively the same 11 . In this regard, the focus has largely been on target guardians and traditional crime problems (burglary, robbery, auto - theft, theft, and s hoplifting) , and other forms of guardianship (handler - gu ardians and manager - guardians) have been rarely considered. Hollis - Peel, Reynald, van Bavel, Elffers, and Welsh (2011 , p. 57 ) attempted to refine the guardianship construct by proposing the followin g formal definition of guardianship: "A guardian is any person and every person on the scene of a potential crime that may notice and intervene (whether they intend to or not )." formal guardians, handlers, managers, and target guardians) is consistent with previous definitions of guardianship, it is problematic in two ways: (1) it is too general to explain guardianship in a meaningful way , and (2) it does not fully acknowledge the importance and role of hand ler - guardians and manager - guardians who, according to Felson and Boba (2010), may have the most important guardianship role of the three types (handler - guardians, manager - guardians, and target - guardians). Later, Hollis, Felson, and Welsh (2013 , p. 76 ) prop osed, what they considered to be, a clear definition of guardianship for use in routine activity theory. They defined guardianship as, "the presence of a human element which acts, whether intentionally or not, to deter the would - be offender from committing a cr ime against an available target . " While this definition of guardianship appears to be more concise than previous versions, its focus is on deterring offenders . As a result , other key elements of guardianship are overlooked. For example, some crimes (e .g., forgery and document fraud) require the offender to interact with the guardian in order to carry out the crime. Therefore, a more focused definition of guardianship will be 11 Some exceptions include Reynald 's Guardianship In Action and the introduction of controllers and super - con trollers (Sampson et al. , 2010). 33 helpful when studying specific kinds of crimes , and the role that specific typ es of guardians play in preventing crime. According to environmental crime theorists , the tenets of routine activity theory and situational crime prevention are all about focusing on specific types of crime. For example, someone studying the five differen t types of signature forgery (simple, simulation, tracing, cut - and - paste, and auto - forgery) will find that each one has different patterns. The same is true when studying guardianship. There are different kinds of guardians , and their effect on the offende r may vary depending on the type of crime, where it takes place , and the capability (real or perceived) of the guardian. Considering this, having so many different definitions of guardianship (Appendix A) creates a problem when trying to understand a guard preventing crime. Since a focused approach to crime prevention requires a comprehensive understanding of the specific types of crime being studied, it is crucial to understand the impact that handler - guardians and manager - guardians have on c rime. In his classic article on situational crime prevention, Clarke (1980) argued that it is possible to prevent crime by increasing the risks of being caugh t. T his can be accomplished by increasing the chances of an offender being seen by someone who is likely to intervene. Although this idea still holds true today, the purported deterrent effect of being seen by a guardian is not as strong in crimes like forgery and document fraud. Since most times the offender will be unable to complete the fraud (e.g., cash a forged check, pass a fake document, or disguise a signature on a receipt) without a guardian approving the transaction, it is the extent and nature of the interaction between the offender and the guardian that is important. The offender knows that to rea p the benefits of the fraud, she or he must first get the fake document or disguised signature past the guardian. The offender and 34 , and the offender tries to predict the forgery detection abilities of the guardian. If the guardian is well trained in forgery detection, then the than if the guardian is inept or just does not care at all . Furthermore, the offender must make a which requires taking calculated risks (Buckner & Carroll, 2006; Cornish & Clarke, 1986; Jacobs & Wright, 2010). This battle between offenders and guardians has been noted elsewhere. In a study of shoplifters, Carroll and Weaver (1986) found that a change in behavior by sales rs and retailers is won or lost depending on the activity and commitment of store employees. In forgery and document fraud crimes , it is rare for an offender to be caught in the act unless a well - trained guardian is present. However, even if the guardian i s not proficient at detecting these kinds of crimes, if the offender perceives the guardian to be well trained, she or he may be less likely to commit forgery or document fraud. Indeed, guardian capability may be less important than guardianship credibilit y (Tilley, 2009). Furthermore, if the offender remains undete rred and commits the crime, his/her chances of being identified increase considerably. For example, an individual who enters a bank with the intention of disguising his signature to withdraw mone y from his account to later deny making the withdrawal must first make a judgment decision as to the forgery detection ability of the bank teller. If he perceives the bank teller as a capable guardian, in the sense that she will carefully evaluate the sign ature, then his disguise strategy may only be a slight deviation from his normal signature in order to get it past the bank teller. However, once this signature becomes part of a forgery investigation, the investigator will have enough evidence (i.e. , comp arable handwriting features) to complete a 35 thorough forensic analysis of the signature and make a determination regarding its authenticity. guardian increases t he risks of the offender being caught. In this regard, the evidence left behind by the offender will assist investigators in making the appropriate charging decisions. To expect businesses, law enforcement, and security professionals to prevent all forge ry and document fraud crimes is unrealistic. No matter what prevention strategies are employed, there will always be undeterred offenders who will successfully pass forged and fake documents some prevention strategies can - making process in a way that makes the handwriting a useful piece of evidence in identifying the of fender will tip the scale in favor of the victim and the investigatory team (police, forensic document examiner, security personnel). Thus, the notion of capable guardianship must be expanded upon to include not only deterring crime, but also altering the behavior of undeterred offenders in such a way that they leave physical evidence behind that increases the chances of identifying them. In this regard, a capable guardian can be defined as an individual(s) whose presence either prevents a crime from occurr ing or alters the crime scene , enabling investigators to identify the offender at a later date. In other words, the crime from taking place , and the secondary task is to alter the behavior of the undeterred offender. Combining environmental crime theories with technical knowledge of forgery and document fraud (i.e. , ars and practitioners better understand these crimes. This deeper understanding will promote the development of more 36 effective crime prevention strategies and improve the investigatory and forensic analysis techniques being used. C rime prevention strategie s should integrate behavior alteration techniques so that the few undeterred offenders have an increased risk of being identified after th e crime has been committed. The current study attempted to combine crime prevention strategies with forensic technical knowledge to better understand the crime of forgery and document fraud. 37 CHAPTER 4: METHODOLOGY The data used in the current study, which came f rom a secondary data source , were collected using the methods described below. The methodolo gy section wa s divided into four parts. First, the research questions and the operationalization for level of guardianship were presented. Second, the survey questionnaire used in this study was discussed. Third, the research design and sampling strategy w ere reviewed. Fourth, the procedures and coding scheme were examined. Research questions Handwriting is the result of a complex perceptual - motor skill that becomes habit - formed over time (Caligiuri & Mohammed, 2012 ; Huber & Headrick, 1999 ). Therefore, it is extremely difficult for individuals to intentionally change several features of their handwriting while also trying to maintain some degree of pictorial similarity. This is because they must consciously suppress habit while also concentrate on those sp ecific aspects of their handwriting that they want to change. Consequently, the best disguise strategy would be to make only minor, but notable Unfortunately, the amount and kind of handwritin g features ( Appendix G ) that writers alter when disguising their signature cannot be thoroughly identified given the enormous gap in the forensic document examination literature regarding signature disguise. Furthermore, the threats to validit y and reliability, which plague the few disguised handwriting studies that have been completed, strengthening forensic science in the U nited S tates acknowledged this prob lem , and stressed the need to increase and improve the quantity and quality of the research studies on handwriting identification (National Research Council, 2009). 38 The goal of this study was to determine how many and what types of changes typically resul t when a writer employs a disguise strategy. This information will provide investigators (i.e. , forensic document examiners, police officers, security professionals, etc.) with a comprehensive list of the primary (intended handwriting changes) and secondar y (unintended h andwriting changes) features that result from disguise. A dditionally , this knowledge will assist them in determining whether a questioned signature is the result of disguise or different authorship. To further the current knowledge on disgui sed handwriting, the role of guardianship, and the strategies that can be used to prevent forgery crimes and document fraud, eight research questions were addressed . The first four questions , which investigate d the forensic asp ects of disguised handwriting , we re as follows: Q 1: What are the primary disguise strategies used by individuals who purposely alter their signature with the intention of later denying ever signing a document? Q 2: How many handwriting features do writers attempt to change when disg uising their handwriting? Q 3: What are the unintentional (i.e. , secondary) handwriting changes that occur as a direct result Q 4: How successful are writers in executing their intended disguise strategy? The remai ni ng four research questions explore d the effect that the level of guardianship (i.e. , how thoroughly the guard ian evaluates the signature) had on the interaction between the offender and the guardian in forgery and document fraud crimes. In this regard, e nvironmental crime theories suggest that the presence of a guardian reduces crime because it decreases the opportunities to offend (Cohen & Felson, 1979; Felson & Clarke, 1998). However, in certain kinds of crimes, the presence of a guardian may not be eno ugh to deter an offender from 39 engaging in criminal behavior; the guardian must also be perceived by the offender as capable and skilled. In the case of forgery crimes and document fraud, it is unclear how offenders determine capable guardianship. To gain a better understanding of how offenders interpret and adapt to different levels of guardianship, and to develop effective prevention strategies, the following questions regarding the offender - guardian relationship were investigated : isguise strategy change depending on the level of guardianship? Q 6: If yes, how does it change? - process when choosing a specific disguise strategy? Q 8: Why do writers choose a specific disguise strategy? Operationalizat ion for Level of Guardianship T he variable , level of guardianship , was operationalized by how thoroughly the guardian evaluate d tudy , there were three scenarios, each one with a different level of guardiansh ip; low (scenario 1), medium (scenario 2), and high (scenario 3) . A low level of guardianship was re presented by a guardian who did not evaluate the signature thoroughly. A me dium level of guardianship was represented by a guardian who may or may not have thoroughly evaluate d the signature, and a high level of guardianship was rep resented b y a guardian that thoroughly evaluate d . Survey Questionnaire A survey questionnaire (Appendix B) was developed to identify and better understand t he thought process and techniques used by individuals who, under different levels of guardianship, disguised their signature . T he survey questionnaire consisted of three parts. Part I Part I of the survey questionnaire was designed to elicit specific infor mation about the 40 , cashiers, bank teller, waitresses) at well - known local businesses, which were chosen based on their popularity with the study 's population (col lege students). Meijer, Wal - Mart, Kroger, K - Mart, and Target were selected to represent local retailers. Olive Garden and Old Chicago were selected to represent local restaurants, and the MSU Federal Credit Union (FCU) was the bank example. This part of th e questionnaire served two purposes. The f irst purpose was to understand tellers (gua rdians) at these businesses, and the second purpose was to give participa nts the opportunity to reflect on their experiences regarding their signatures being examined by cashiers/waitresses/bank tellers when they visit these business establishments. Knowing what participants think a guardian looks for when evaluating a signatur e and how they determine the tect a forgery will help criminal justice scholars and practitioners understand the hypothesized self - projecting behavior that writers use when devising a disguise strategy. For instance, if a participan t believes that the bank tellers at the MSU FCU are better disguise strategy under the bank scenario would be different than the strategy used in the reta il store scenario. Furthermore, giving participants the opportunity to share what they think about guardians look for when examining a signature, will further schola r s ' and practitioner s ' situation under which they sign a document. Part II Part II of the surv ey que stionnaire consisted of six short - answer open - ended questions; 41 tw o for ea ch of the three study scenarios . These questions were designed to allow participants to identify and describe the handwriting changes that they made when asked to disguise their signature, and explain why they chose a specific disguis e strategy. Th is information was then used to determine whether or not study participants were successful in executing their intended disguise strategy, how many different intentional changes were made, and the effect that the hip had on their disguise strategy. Part III This part of the surv ey questionnaire was designed to gather demographic information about the participants. These demographic variables include d age, gender , race, country of citizenship, handedness, and educa tion. T hree questions were also included that address ed the participants who might have received some training or who had experience in forgery detection, participan ts were specifically asked if they had any training in forensic handwriting analysis or forgery detect ion , and whether or not they had ever been victims of forgery or iden tity theft. If a participant had been a victim of forgery or identity theft, she or h e may have learned some of the features that investigators look for when evaluating signa tures. This inside knowledge could have potentially influence d their disguise strategy. Research Design T he one - group pretest - posttest quasi - experimental design was us ed. This research design is structurally represented as follows: O1 X O2 Where O1 is the pretest, X is the treatment, and O2 is the posttest. 42 Sample Strategy A non - probability purposive sample of college students was chosen. Despite the long - standing cont roversy among scholars regarding the use of students as subjects, there is a dearth of empirical evidence either supporting or refuting the notion that college students are different from non - college students in their behavior and attitudes (Peterson, 2001 ). However, much of what is known about human behavior comes from the field of social psychology, which has relied primarily on young American college students tested in university laboratories (Bargh, Chen, & Burrows, 1996; Chartrand & Bargh, 1999; Darley & Batson, 1973; Darley & Latane, 1968; Deutsch & Krauss, 1962; Festinger & Carlsmith, 1959; Gilbert, Tafarodi, & Malone, 1993; Sears, 1986; Stasser & Tutus, 1985; Tajfel, Billig, Bundy, & Flament, 1971; Wegner, Schneider, Carter, & White , 1987). Some clas sic studies on human behavior that have used students as subjects include: Milgram's (1963) behavioral study of obedience; Zimbardo's Stanford Prison Study (Musen & Zimbardo 1991); Asch's (1951) group pressure study; Ross, Greene , and House's (1977) resear ch on the false consensus effect; Snyder, Tanke , and Berscheid's (1977) experiment on the self - fulfilling nature of social stereotypes ; Nisbett and Cohen's (1996) culture of honor work which investigated the psyc hology of violence in the south; and Quattro ne and Tversky's (1984) study of self - deception. In fact, according to Payne and Chappell (2008), the use of student subjects in the fields of psychology and social psychology is so common, that many colleges and universities have specific policies as to h ow students in these courses and degree programs can participate as subjects in research studies. Other disciplines that have also incorporated the use of students as study subjects include information - systems, sociology, political science, and criminal ju stice (Compeau, Marcolin, Kelley, & Higgins, 2012; Jiang, Lambert, & Wang 2007; Koehler & Thompson, 2006; Payne & Chappell, 2008; Wiecko, 2010; 43 Williams, Fletcher, & Ronan, 2007). The purpose of showing the precedent established by prior studies and emphas izing Payne and Chappell's (2008) point regarding the widespread use of stude nt subjects in social research wa s to illustrate that much can be learned from using students as research subjects. The decision to use college stude nts as research subjects in t he current study was made after taking into consideration various fact ors. First, students have been legitimately used in social research. Second, because financial resources were not available to conduct this study, cost and convenience had to be consider ed. I n order to obtain a relatively large sample size without having to offer participants any kind of financial incentive or compensation, using college students appeared to be a feasible option. Finally , to alleviate some validity concerns, a literature review on how college students differ from non - college students on factors related to crime, criminology, behavior, and attitudes was performed . Considering that only a few empirical studies evaluating the use of student subjects in research have been comp leted, the literature review returned some interesting and encouraging results. In a study conducted by Flere and Lavric (2008) on the validity of cross - cultural social studies using student samples, the authors found that with regard to various sociologic al and psychological measures, student samples can be used in cross - cultural studies because they are relatively reliable predictors of general cross - cultural differences. Furthermore, in a critical review of criminological studies using student subjects, Payne and Chappell (2008) concluded that even though students cannot and should not be used in all criminological research projects, there are certain types of studies that are particularly responsive to college student samples. Some notable examples we re tests of criminological theory and studies that assess ed student attitudes and behaviors. In a more recent empirical study of student sample v alidity, Wiecko (2010) found that questions surrounding the 44 validity of student samples may not be as warranted as once thought. In fact, when compared to non - college students, the author reported almost no statistically significant difference in behaviors , and reported only minor differences in the frequency of behaviors and attitudes. Because college students are a convenient source of information, and they are also closely representative of the rest of the population when it comes to criminal and delinquent acts, Wiecko (2010) encouraged their use. I n concluding the article, the author asserted that while college st udents may be culturally unique, this uniqueness does not seem to affect validity. In light of these finding s, the use of student subjects wa s a ppropriate for the current study . When selecting the sampling method, various factors and challenges surroundi ng the study were considered. First, although the random selection of a sample i s considered to be the gold standard in research design, it is not always the most efficient and cost - effective method (Gravetter & Forzano, 2012). For this study, obtaining a random sample of Michigan State Un iversity (MSU) students was possible, but getting a large number of them to volunteer an hour or more of their time to complete a handwriting exercise and travel to and from the study location without any in centive or comp ensation was difficult. Therefore, a non - random purposive sample , which consisted mainly of students from the Introduction to Criminal Justice class , (and a few students from the Criminal Justice and Forensic Science Graduate Programs), was considered. Thi s particular class was selected for three primary reasons. First, the Introduction to Criminal Justice course typically enrolls 600 students who are divided in three sections of approximately 200 students . With a class this size, o btaining a large sample w as convenient. Second, the fact that this class wa s composed of criminal justice and non - criminal justice students provide d diversity to the study sample. Third, students were offered extra credit points to incentivize their participation. 45 Since college s tudents typically fall within the peak age group for criminal activity (Hirschi & Gottfreds on, 1983) , a non - probability purposive sample of college students was selected to participate in this study. More specifically, three hundred and eight college stude nts from MSU in East Lansing, Michigan were selected to participate. All of them were interviewed during independent sessions that lasted approximately 30 minutes. Table 1 shows the demographic characteristics of study participants. The participants includ ed 156 males (51%) and 152 females (49%). The self - reported race/ethnicity of the participants included 74% White, 8.1% Asian, 8.1% Hispanic, 7.1% Black, .3% American Indian or Alaska Native, .3% Native Hawaiian or Other Pacific Islander, and 1.9% identifi A majority of the study participants were between the ages of 17 and 24 (93.2%), followed by 25 30 (4.5%), 31 35 (1.3%), and 36 or older (.9%). The sample was composed of graduate (8%) and undergraduate students (92%) that represe nt ed seventeen different countries from five different geographic regions with 91% from North America, 5% from Asia, 2% from Europe, 1% from Mexico and the Caribbean , and 1% from Africa. The diversity of maj ors represented in the sample was quite impress iv e considering that it included students majoring in the social sciences (62.7%), business (12%), art s and humanities (6.5%) , biological sciences (6.5%), forensic sciences (2.6%), physical sciences (1.9%), education (1%), and other (6.8%). Table 1 . Descrip tive statistics for sample population Variable Total (N=308) Percentage Gender Male Female 156 152 51 49 Race / Ethnicity White - Non Hispanic White - Hispanic 228 25 74 8.1 46 Table 1 A sian 25 8.1 African Amer ican or Black 22 7.1 American Indian or Alaska Native 1 0.3 Native Hawaiian or Other Pacific Islander 1 0.3 Other 6 1.9 Age Range 17 - 24 25 - 30 287 14 93.2 4.5 31 - 35 4 1.3 36 or older 3 0.9 Geographic Reg ion USA or Canada 282 91.6 Asia 15 4.9 Europe 6 1.9 Mexico and Caribbean 3 1.0 Africa 2 0.6 Highest Degree Earned High School Diploma 275 89.3 Bachelor Degree 14 4.5 Master Degree 11 3.6 Associate Degree 8 2.6 Major Social Science 193 62.7 Business 37 12 Art & Humanities 20 6.5 Biological Science 20 6.5 47 Table 1 Forensic Science 8 2.6 Physical Science 6 1.9 Education 3 1 Other 21 6.8 Degr ee Pursuing Bachelor 283 91.9 Master 12 3.9 PhD 13 4.2 Procedure Students who volunteered to participate in this s tudy were asked to choose a date and time that was con venient for them to complete the exercise at Baker Hall, located o n the MSU campus. Upon arrival, participants were asked to read and sig n a consent form (Appendix C). After signing the consent form, they were informed that the survey questionnaire consisted of three sections , and they were encouraged to ask questions if they were un sure about what they were being asked to do in each section. As the participants received the first part of the survey questionnaire (Appendix B), their attention was directed to questions 4 through 12. Participants were then told that althoug h the business examples in those questions were chosen because of their popularity among college students, if they were unfamiliar with them, they could cross them out and replace them with similar businesses in the Lansing area that they were familiar wit h. For example, a retailer ( Meijer, Walmart, Target, K - Mart) could be replaced with another re tailer, and a restaurant ( Olive Garden, Old Chicago) could be replaced with another restaurant. Participant s were also instructed to ask for validation of their r eplacement example if they were unsure about its appropriateness. Although this situation was not recurrent, it did happen with 48 some international students. For example, a participant from China who was unfamiliar with Old Chicago replaced this example wit h another restaurant he was familiar with. After completing this part of the survey, participants were asked to place it face down on their desk and were informed that they were now moving onto a handwriting exercise. Participants were then given the follo wing instructions: I am going to give you three different scenarios. After hearing each scenario, I want you to take a minute to think about the strategy that you will use to disguise (alter) your signature in such a way that, (1) you will be able to get i t past the guardian (i.e., the person who will be taking your signature), and (2) also be able to later deny ever signing the document. Because each scenario is different, I would like you to think about them independently and base your signature disguise strategy on the scenario at hand. The scenarios 12 were designed to simulate three commonly encountered situations in which the chances of signatures being compared increase with each situation. In the first scenario, represented by a restaurant setting, p articipants were led to think that their known and disguised signatures would not be compared. In the second scenario, represented by a retail store setting, participants were led to believe that it was uncertain if their signatures would be compared. In t he third scenario, represented by a bank setting, participants were led to think that their signatures would definitely be compared. Although the guardianship characteristics between a retail establishment and a restaurant setting may appear to be similar, the differences in the level o f guardianship in each are remarkable. For example, in a restaurant, after a customer finishes his/her meal, the guardian (waiter or waitress) often leaves the check with the customer 12 The measures of g uardianship and the scenarios that were created were drawn from the field of social psychology. This wa s the first study to use this kind of technique to study how increasing levels of guardianship affect forgery crimes. Furthermore, most of the guardiansh ip research does not pay much attention to manager - guardians. Therefore, there was not much prior research to draw on in terms of constructing scenarios directly from the guardianship literature. 49 and walks away before the signature is pl aced onto the document and the transaction is complete. However, in the retail setting, since the guardian never walks away before the signature is added to the sales receipt and the transaction has been completed, the offender has to interact with the gua rdian when attempting to commit a forgery. This necessary interaction, which is under an uncertain and inconsistent level of guardianship, forces the offender to make an immediate, on - the - on the sales receipt. It is unknown whether or not the study participants had ever tried to disguise their signature when signing a document with the intention of later denying ever signing it. As a result, a priming technique was used to get particip ants in the "right - state - of - mind" (Bargh & Chartrand, 2000). They were asked to think in a way out of the ordinary , and to act as though their intent was to truly disguise their signature. Following this instruction, participants were given six documents f or each scenario. Each document simulated the kind of receipt/bill/check that they would encounter in each setting (Appendix D). For the restaurant scenario, participants ey received six identical receipts from Wal - Mart. Lastly, they received six identical checks from the MSU FCU for the bank scenario. For the pretest, participants were asked to sign in their natural handwriting five of the six documents given to them for e ach scenario and place them face down on the ir desk. These signatures serve d as the baseline for how each participant would normally sign his/her name. After giving participants a short narrative (Appendix E) of the specific scenario under which they would disguise their handwriting, they were asked to sign the sixth unsigned document using their chosen disguise strategy. For this posttest portion of the study, the subjects received an additional blank sheet of paper where they could practice their disguise strategy before making 50 the actual attempt. This process was repeated for each scenario (restaurant, retail store, and bank). Following the handwriting exercise, participants were asked to complete Part II of the surv ey questionnaire , which asked them spec ific questions about the disguising strategy that they used for each scenario and the reasons they chose that strategy. Finally, participants were asked to complete the third part of the s urvey questionnaire , which sought to elicit demographic information about the participants, as well as their experience with crimes of forgery and ID theft . Coding Scheme In each of the thre e scenarios , participants were asked to produce five natural (also referred to as known ) signatures and one disguised. Therefore, each participant produced a total of fifteen known signatures and three disguised. It has been determined by experts in the field of the imprecision with which the habits of the writer are executed on repeated occasions, at least 15 to 20 signatures are necessary for analysis (Conway, 1959; Hilton, 1982; Huber & Headrick, 1999). By asking participants to naturally sign their name five times for each sce nario, 15 known signatures were obtained and then used for comparison with the three disguised signatures. This enable d the coder intentional changes. To code the collected data, a coding scheme (Appendix F) was developed based on the ASTM International standard guide for the examination of handwritten items (2007) and Huber - one discriminating elem ents. The coding was completed in two stages. The signatures in t he three scenario exercises were coded first, followed by the This order was used 51 to alleviate any bias that could result from coding the participants' self - described intentiona l handwriting changes before scientifically examining and coding the handwriting changes actually performed 13 . A ll the signatures were examined using the attached coding scheme ( Appendix F), and evaluated using the follow ing steps. First, th e fifteen known signatures of each participant were examined and compare d to his/h er three disguised signatures . The comp arison result ed in a list, per participant, of changes associated with each of the scenarios. This list identified all of the handwriting features chan ged by the p articipant. Next, ategy claims for each scenario were reviewed , and a list of the changes that the subject reported making was created. The list of changes that were identified was then compared to the list of the self - identified alterations to answer research questions 1 through 4. After all the sig natures of all participants had been examined, t hree comprehensive lists were created (one for each scenario). These c omprehensive lists identified all of the prim ary (intentional) and secondary (unintentional) handwriting features changed by the participants. In order to code the success (change) or failure (no change) of the disguise strategies, first, each writer's range of natural variation was determined. This information was used as the baseline measure for the participant's self - reported intentional changes. If a study participant employed a disguise strategy that fell within his or her normal range of writing variation, then the altered handwriting featu re was coded as a failure (Figure 2 b ). Conversely, if the disguise strategy fell outside the writer's range of natural variation, then the changed handwriting feat ure was coded as a success (Figure 3 b ). 13 This study investigated how people attempt to disguise th eir handwriting under different levels of guardianship. In no way did 52 Figure 2a. Naturally written Figure 2 b. Failed a ttempt to alter the capital letter "S" capital letter "S" Figure 3a. Naturally written Figure 3 b. Successfully altered capital letter "D" capital letter "D" The data collected in the first and third part of the survey questionnaire were coded using a numeric coding system , and along with the informat ion obtained in Part II, were used to answer research questions 5 through 8 ( which focused on guardianship ) . In this regard, if the level of guardi anship affected the participant s ' choic e of disguise stra tegy, then observable differences among the three scenarios were expected . For example, in scenario one, participants were led to think that the guardian would not closely scrutinize their signature. As a result, the subjects may have dec ided to make their signature grotesquely different from their natural one because the guardian was not going to evaluate it , and it would be easier for them to deny signing it at a later date. However, under the third scenario, since subjects were led to t hink that the guardian would closely scrutinize their signature, then t heir disguise strategy was likely minor. In other words, it di d not deviate too much from their natural signature , so that they could avoid being caught by the guardian. 53 Analysis Tec hniques For the data analy sis portion of this project, three primary statistical techniques were used: (1) Chi - Square Test , (2) Tukey - Kramer Pairwise Comparisons, and ( 3 ) Exploratory Factor Analysis. Chi - Square Test of Independence Chi - 2 ) is a common test statistic used to examine the independence of categorical data (Vogt & Johnson, 2011), which can be converted into one of several measures of association. The simplest use of chi - square is to test for statistically significant differences betw een the observed and hypothesized frequencies of variables p resented in a contingency tabl e . The larger the difference between the observed and hypothesized frequencies, the larger the chi - square statistic is. The larger the statistic, the less likely the observed differences are due to chance ( Vogt & Johnson, 2011 ). In the current study, the chi - sq uare test of independence was used to address research questions 2, 5, 7 , and 8. More specifically, to investigate the relationship between the levels of guardi anship and the number of handwriting features changed, the disguise strategies used, and es were influenced by the scenarios they were presented with. An alpha level of .05 was used for all statistical ana lyses. Tukey - Kramer Pairwise Comparisons Tukey - Kramer Pairwise Comparison is an analysis used to examine all the pairs of means in order to determine any difference between two means (Abdi & Williams, 2010; Tukey, 1949). One advantage of the Tukey - Kramer t est is that it keeps the level of Type I errors (i.e., false computation of confidence intervals for the difference between the means (Abdi & Williams, 54 20 10). In the current study, Tukey - Kramer Pairwise Comparisons were used to answer research question 4, which aimed to determine whether or not the participants were successful in executing their intended disguise strategies. In this regard, the participants ' self - reported changes , and what was actually noted as being changed during the coding process , were compared and used to assess the statistical significance of the differences i n success rates across the scenarios. Exploratory Factor Analysis Factor anal ysis is a method of data reduction that is used to simplify complex sets of data so that relationships between variables can be more easily interpreted and understood (Yong & Pearce, 2013). More specifically, the purpose of factor analysis is to assist res earchers in identifying and understanding the nature of the latent constructs underlying the variables of interest so that they can be explained in terms of a few underlying factors (Bandalos & Finney, 2010). Ex ploratory Factor Analysis (EFA) is typically used in two ways: (1) to explore a specific area of interest in order to discover the main latent constructs or dimensions that reflect the most variation in the data, and (2) to discover if the original variables are organized in a particular way, reflect ing another latent variable (Kline, 1994). According to Bandalos and Finney (2010), many of the decisions made in EFA are subjective , and are made on the basis of congruence with theory and/or previous research. However, some decisions are based on stati stical guidelines, such as determining the number of factors to retain (Bandalos & Finney , 2010; Preacher, Zhang, Kim, & Mels, 2013). Most experts agree that EFA should be used for situations in which little research has been conducted regarding the struct ure of the construct or measure of interest (Bandalos & Finney, 2010; Fabrigar, Wegener, MacCallum, & Strahan, 1999; Tabachnick & Fidell, 2001). Variables used for EFA can take a variety of forms , such as items from a scale or direct measures of subject 55 c haracteristics (Bandalos & Finney, 2010). The variables used in this study are non - continuous, but according to Finney and DiStefano (2006) , this was not a problem since the items had at least five scale points and we re reasonably inter - correlate d. Extrac tion is the process by which the parameters of the factor solution, which include the factor pattern coefficients, are estimated (Bandalos & Finney, 2010). Although there are many different methods that can be used to conduct a factor analysis, P rincipal A xis (PA) and M aximum L ikelihood (ML) methods are commonly used (Bandalos & Finney, 2010; Kline, 1994). For the current study , PA, which uses a least - squares type solution and attempt s to minimize the residuals , was the extraction method used , since it is c onsidered a fairly straightforward method . Several solutions were compared that extract ed various plausible number s of factors. T he final number of factors was determined in part by theory, and by running the analysis extracting different numbers of factor s in order to see which number of factors yield ed the most interpretable results. Extracted factors are often difficult to interpret on the basis of their factor loadings (Field, 2000). Therefore, the factors are rotated to alter the pattern of the factor loadings so that their interpretation can be improved (Field, 2000). While many rotation methods exist, the primary difference is between those th at produce orthogonal (uncorrelated) and oblique (correlated ) factors (Band alos & Finney, 2010). Since it was unknown whether the factors were correlated , both orthogonal (varimax) and oblique (promax) factors were obtained for comparison. In the promax solution, it was generally noted that the factors were not correlated when the minimum criteria for loading was set at r = .4, so it was decided that the model using orthogonal rotation provided the most interpretable results. While it is not uncommon for variables to be removed from the model due to low structure, pattern , or communality (Bandalos & Finney, 2010), all 56 remaining variables were included so that results could be compare d between the differences in factor loadings from the baseline condition to the combined guardianship conditions. To complete the factor analysis, a Principal Axis Factor (P AF) with a Varimax rotation of 28 of 29 items 14 from the handwriting analysis was conducted on the data collected . Prior to t he factor analysis, the data were screened for any outliers and missing data. An examination of the Kaiser - Meyer - Olkin measure of sampling adeq uacy for the baseline condition was performed and found to be factorable (KMO = .591) . Bartlett's Test of Sphericity was highly significant 2 = 1596, p < .001. Initially, 10 factors with eigenvalues greater than one were extruded. 14 The handwriting variable speed was not rotated because it is difficult to determine absolute handwriting speed f rom a static signature. Therefore, this variable was not coded. 57 CHAPTER 5: RESULTS This chapter present s the findings of the current study. All analysis was performed using SPSS v22.0. Research Question 1: What are the primary disguise strategies used by individuals who purposely alter their signature with the intention of la ter denying ever signing a document? A total of 21 handwriting features were purposely cha nged by the study participants. Table 2 provides a scenario count of these features where each column identifies the total number of study participants that changed the associated handwriting feature. Each participant reported changing one to three features per scenario. Table 2 . Scenario count of the handwriting features that t he study participants said they changed Variable Scenario #1 Scenario #2 Scenario #3 N (Column %) N (Column %) N (Column %) Letter Form 165 (29.46) 157 (29.07) 176 (30.45) Legibility 119 (21.25) 118 (21.85) 124 (21.45) Size 69 (12.32) 58 (10.74) 60 (10.38) Spelling 49 (8.75) 37 (6.85) 51 (8.82) Spacing 19 (3.39) 28 (5.19) 26 (4.50) Sla nt 17 (3.04) 25 (4.63) 21 (3.63) Abbreviations 18 (3.21) 14 (2.59) 24 (4.15) Terminal Stroke 11 (1.96) 20 (3.70) 15 (2.60) Speed 17 (3.04) 15 (2.78) 13 (2.25) 58 Table 2 Style of Allograph 22 (3.93) 6 (1.11) 16 (2.77) Punctuation 9 (1.61) 13 (2 .41) 11 (1.90) Embellishment 11 (1.96) 8 (1.48) 10 (1.73) Pressure 10 (1.79) 10 (1.85) 3 (0.52) Hand 2 (0.36) 6 (1.11) 9 (1.56) Lateral Expansion 6 (1.07) 5 (0.93) 5 (0.87) Line Quality 5 (0.89) 5 (0.93) 4 (0.69) Connecting Strokes 3 (0.54) 5 (0.93) 4 (0.69) Alignment 3 (0.54) 5 (0.93) 2 (0.35) Initial Strokes 4 (0.71) 1 (0.19) 1 (0.17) Order of Name 1 (0.18) 1 (0.19) 1 (0.17) Pen lifts 0 (0.00) 2 (0.37) 1 (0.17) No Change 0 (0.00) 1 (0.19) 1 (0.17) Total Responses 560 540 578 Figure 4 il lustrates the percentage of participants that altered handwriting features under the three scenarios. It is clear from this graph that although multiple features were intentionally altered, only five were consistently changed by participants in all three s cenarios: letter form, legibility ( neater and messier), absolute size, spelling, and spacing (see Appendix H for examples of handwriting changes) . 59 Figure 4 . Overview of the handwriting features intentionally changed by participants Table 3 shows the n umber and percentage of participants who changed these five handwriting features , and the scenario under which they did so. For scenario 1, 39.2% of the study participants changed letter form, 28.3% changed legibility, 16.4% changed size, 11.6% changed spe lling, and 4.5% changed spacing. A similar pattern was observed for scenario 2 and scenario 3 where 39.4% and 40.3% changed letter form, 29.6% and 28.4% changed legibility, 14.6% and 13.7% changed size, 9.3% and 11.7% changed spelling, and 7% and 5.9% chan ged spacing , respectively . Table 3 . Top five handwriting features intentionally changed by the study participants Handwriting Feature Scenario #1 Scenario #2 Scenario #3 N (%) N (%) N (%) Letter Form 165 (39.2) 157 (39.4) 176 (40.3) Legibility (nea ter and messier) 119 (28.3) 118 (29.6) 124 (28.4) Size 69 (16.4) 58 (14.6) 60 (13.7) 60 Table 3 Spelling 49 (11.6) 37 (9.3) 51 (11.7) Spacing 19 (4.5) 28 (7.0) 26 (5.9) Total Responses * 421 398 437 * The counts and percentages are based on the top 5 handwriting features that were intentionally changed. Research Question 2: How many handwriting features do writers attempt to change when disguising their handwriting ? Participants reported that their intention was to change one to three handwriti ng features. For this reason, three categories for number of changes were created for each scenario: one, two , and three features changed. Table 4 shows the total number of intended changes made by the study participants. For scenario 1, 36.4% of participa nts intended to alter one handwriting feature, 39% intended to change two features, and 24.7% intended to mak e changes to three features . A similar pattern emerged for scenario 2 where 39.9% made one change, 39.3% made two changes, and 20.8% made three cha nges. For scenario 3, 40.6% of the participants altered one feature, 4 3.2% altered two, and 16.2% altered three features . Table 4 . Number of handwriting features changed by scenario Number of Changes 1 2 3 Total Responses Frequency Share N (%) N (%) N (%) Time (Scenario) #1 112 (36.4) 120 (39.0) 76 (24.7) 308 #2 123 (39.9) 121 (39.3) 64 (20.8) 308 #3 125 (40.6) 133 (43.2) 50 (16.2) 308 61 A chi - square test of independence was used to assess the relationship between time (i.e. , scenarios) and the number of handwriting features that the study participants attempted to change. As shown in Table 5 , the chi - square test indicated that the relationship between the number of handw riting features changed and scenario , wa s not statistically significant, 2 = 7.00 4 , p < 0.05 . Al though these results we re statistically insignificant, the trend wa s in the expected direction: fewer changes we re made as the level of guardianship increase d . Table 5 . Results of Chi - square Test for handwriting features changed Value df Asymp. Sig. (2 - sided) 15 Pearson Chi - Square 7 . 004 4 0.1357 Likelihood Ratio 7.071 4 0.1322 N o. of Valid Cases 924 Research Question 3: What are the unintentional (i.e. , secondary) handwriting changes that occur as a direct res baseline (i.e., normal handwriting without disguise) , and also in each scenario condition in order to determine if (and how) featu res changed under the different scenarios. Changes from baseline that participants intended to make were recorded as intentional changes. All other deviations from the baseline condition were re corded as unintentional. Table 6 shows the handwriting feature s that were not changed, changed intentionally , and unintentionally changed by the study participants. When participants altered their signatures, the top five most common unintentional changes for scenario 1 were: lateral expansion (57%), the direction of the initial stroke (48%), the number of pen lifts (46%), the spacing of letters (45%), and connecting strokes (43%). 15 The asymptotic test and the p - value both represent the probability for the test statistic. In SPSS, t he Chi - square test result will return "Asymp. Sig (2 - tailed) , " "Exact (2 - tailed) , " and "E xact (1 - tailed) . " When the chi - square test was run , it returned th e Asymp. Sig. (2 - sided), which was used instead of the p - value. 62 Similarly, for scenario 2, lateral expansion (53%) was the most common unintentional change, followed by the number of pen lifts (44%), th e spacing of letters (40%), the direction of the initial stroke (40%), and letter form of the last name (40%). For scenario 3 , lateral expansion (30%), letter form of the last name (28%), the number of pen lifts (24%), the direction of the initial stroke ( 23%), and letter form of the first name (22%) were the most common unintentional changes. Additionally, when letter form of the first and last name were combined into a single letter form category, this handwriting feature became the most prominent uninten tional change at 74% for scenario 1, 75% for scenario 2, and 50% for scenario 3. Table 6 . Counts of Handwriting features not changed, intentionally changed, and unintentionally changed across each scenario Variable N Scenario #1 Scenario #2 Scenario #3 No Change Int Unint No Change Int Unint No Change Int Unint Lat Expansion 308 131 (43%) 0 (0%) 177 (57%) 146 (47%) 0 (0%) 162 (53%) 217 (70%) 0 (0%) 91 (30%) Instroke_Dir 308 159 (52%) 1 (0.3%) 148 (48%) 182 (59%) 4 (1%) 122 (40%) 237 (77%) 0 (0%) 71 (2 3%) Pen_Lifts 308 164 (53%) 1 (0.3%) 143 (46%) 174 (56%) 0 (0%) 134 (44%) 232 (75%) 1 (0.3%) 75 (24%) Spacing_ Letters 308 152 (49%) 17 (6%) 139 (45%) 174 (56%) 10 (3%) 124 (40%) 238 (77%) 17 (6%) 53 (17%) Connecting Strokes 308 175 (57%) 1 (0.3%) 132 ( 43%) 196 (64%) 2 (1%) 110 (36%) 261 (85%) 1 (0.3%) 46 (15%) Tstroke_Dir 308 171 (56%) 9 (3%) 128 (42%) 194 (63%) 8 (3%) 106 (34%) 240 (78%) 12 (4%) 56 (18%) Letter Form Last 308 23 (7%) 168 (55%) 117 (38%) 43 (14%) 142 (46%) 123 (40%) 106 (34%) 117 (38%) 85 (28%) Line Quality 308 189 (61%) 3 (1%) 116 (38%) 189 (61%) 4 (1%) 115 (37%) 257 (83%) 5 (2%) 46 (15%) Letter Form First 308 35 (11%) 163 (53%) 110 (36%) 62 (20%) 137 (44%) 109 (35%) 141 (46%) 99 (32%) 68 (22%) Size Absolute 308 151 (49%) 47 (15%) 1 10 (36%) 157 (51%) 51 (17%) 100 (32%) 218 (71%) 32 (10%) 58 (19%) Legible Sig 308 111 (36%) 101 (33%) 96 (31%) 139 (45%) 82 (27%) 87 (28%) 211 (69%) 51 (17%) 46 (15%) Slant 308 209 (68%) 9 (3%) 90 (29%) 233 (76%) 7 (2%) 68 (22%) 278 (90%) 5 (2%) 25 (8%) 63 Table 6 Tstroke_End 308 222 (72%) 6 (2%) 80 (26%) 228 (74%) 2 (1%) 78 (25%) 264 (86%) 3 (1%) 41 (13%) t - bar_Form 308 241 (78%) 0 (0%) 67 (22%) 244 (79%) 0 (0%) 64 (21%) 270 (88%) 0 (0%) 38 (12%) Pun_i/j - dot Form 308 233 (76%) 8 (3%) 67 (22%) 24 8 (81%) 9 (3%) 51 (17%) 264 (86%) 9 (3%) 35 (11%) Pun_i/j - dot Loc 308 235 (76%) 7 (2%) 66 (21%) 255 (83%) 7 (2%) 46 (15%) 266 (86%) 6 (2%) 36 (12%) Pressure 308 244 (79%) 2 (1%) 62 (20%) 241 (78%) 7 (2%) 60 (19%) 275 (89%) 3 (1%) 30 (10%) Align_Spec 2 3 08 240 (78%) 1 (0.3%) 67 (22%) 269 (87%) 1 (0.3%) 38 (12%) 287 (93%) 5 (2%) 16 (5%) t - bar_Loc 308 255 (83%) 0 (0%) 53 (17%) 267 (87%) 0 (0%) 41 (13%) 279 (91%) 0 (0%) 29 (9%) Align_Spec 1 308 273 (89%) 0 (0%) 35 (11%) 272 (88%) 1 (0.3%) 35 (11%) 296 (96% ) 2 (1%) 10 (3%) Order 308 272 (88%) 2 (1%) 34 (11%) 283 (92%) 1 (0.3%) 24 (8%) 288 (94%) 0 (0%) 20 (6%) Style of Allograph 308 270 (88%) 18 (6%) 20 (6%) 263 (85%) 15 (5%) 30 (10%) 293 (95%) 5 (2%) 10 (3%) Pun_period 308 286 (93%) 2 (1%) 20 (6%) 293 (95 %) 2 (1%) 13 (4%) 294 (95%) 2 (1%) 12 (4%) Spelling 308 237 (77%) 56 (18%) 15 (5%) 250 (81%) 44 (14%) 14 (5%) 262 (85%) 36 (12%) 10 (3%) Hiatuses 308 294 (95%) 0 (0%) 14 (5%) 293 (95%) 0 (0%) 15 (5%) 300 (97%) 0 (0%) 8 (3%) Abbreviation 308 273 (89%) 24 (8%) 11 (3%) 282 (91%) 18 (6%) 8 (3%) 290 (94%) 13 (4%) 5 (2%) Embellish 308 290 (94%) 9 (3%) 9 (3%) 285 (93%) 10 (3%) 13 (4%) 294 (95%) 7 (2%) 7 (2%) Instroke_ Start 308 298 (97%) 0 (0%) 10 (3%) 300 (97%) 0 (0%) 8 (3%) 305 (99%) 0 (0%) 3 (1%) Hand 308 299 (97%) 9 (3%) 0 (0%) 305 (99%) 3 (1%) 0 (0%) 303 (98%) 5 (2%) 0 (0%) * Note: percentages may not sum to 100 due to rounding. * * See Appendix A for handwriting variable definitions. In add ition to the information presented above, when the total number of unintentional changes for all of the handwriting features were summarized (Table 7 ), there wa s a noticeable drop in the number of unintentional changes in scenario 3. In fact, the total number of unintentional changes between scenario 1 and scenario 3 drop p ed by more than 50 percent. As 64 shown in Table 7 , this decreasing pattern wa s consistent with the total number of intentional changes. As the number of intentional changes decreased , the number of unintentional changes also decreased . Additionally, whe n the total numbers of changes across the scenarios we re compared, there we re three times as many unintentional changes as there we re intentional changes. However, when the total number s of changes in each scenario we re compared, the results illustrated th at for scenarios 1 and 2 , every intended change resulted in three unintended changes, but for scenario 3 , every intentional change resulted in two unintentional changes. Table 7 . Summary of handwriting features changed by scenario Scenario 1 Scenario 2 S cenario 3 Total Changes N (%) N (%) N (%) Number of Intentional Changes 664 (40) 567 (34) 436 (26) 1,667 Number of Unintentional Changes 2,136 (42) 1,898 (38) 1,030 (20) 5,064 Exploratory Factor Analysis First, a factor analysis was completed o n the baseline condition (the condition under which the study participants naturally signed their name without a disguise). This information identified the handwriting features that cluste red together und er normal conditions. Second, the guardianship condi tion (the condition under which the study participants intentionally altered their signatures) was analyzed . This information identified the handwriting features that cluste red together when the signatures we re intentionally changed. Finally, the results f or the baseline condition w ere compared to the results for the guardianship condition. I t was determined that five factors gave the most interpretable solution. Results of the orthogonal rotation of the baseline conditi on solution are shown in Table 8 . 65 Tab le 8 . Orthogonally rotated component loadings for the baseline condition 1 The Five Topic Factors When loadings less than 0.40 were excluded, the analysis yielded a five - factor solution. Five variables were loaded onto Factor 1, which explained almost 1 1% of the variance. All five of these variables relate d to handwriting features that ga ve t he writing a certain appearance , 66 including style of allograph, letter form, spacing, and slant. This factor was labeled "writing style." Four variables we re loaded o nto Factor 2, which wa s related to elements of execution (i.e. , the less obvious, more subt le elements of writing), which we re related to pen lifts , legibility, and i - dot form and location. This factor The two items that lo aded onto Factor 3 relate d to t - bar form and location. This factor was labeled - The four items that were loaded onto Factor 4 relate d to period marks, order of name, connecting strokes, and writing hand. This factor was labeled . Finally, the three items that were loaded onto Factor 5 relate d to alignment and line quality. This factor was labeled "writing fluidity." Overall, 39.9% of the variance in the baseline condition was explained by these five fa c tors and can be seen in Tabl e 9 . Table 9 . Total variance explained for the baseline Component Rotation Sums of Squared Loadings Total % of Variance Cumulative % 1 2.930 10.851 10.851 2 2.361 8.745 19.596 3 1.945 7.204 26.801 4 1.825 6.760 33.561 5 1.703 6.306 39.867 A subs equent factor analysis was performed on the combined guardianship condition 16 , and the results were compared to the baseline condition. This comparison identified the difference in loadings between the participants normal handwriting and their disguised ha ndwriting. Results of the rotation of the guardianship conditio n solution are shown in Table 10 . 16 The procedure used to complete the factor analysis for the baseline condition was the same procedure used for the combined g uardianship condition. 67 Table 10 . Orthogonally rotated component loadings for the guardianship condition 1 For the combined guardianship condition, there was a slight decrease in t he overall explained variability, indicating that the factor model d id not fit the data as well as t he baseline. As seen in Table 11 , the five factors account ed for 37.9% of the total variation . Slight changes in individual factor variances were also obser ved. 68 Table 11 . Total variance explained for the guardianship condition Component Rotation Sums of Squared Loadings Total % of Variance Cumulative % 1 2.597 9.617 9.617 2 2.335 8.648 18.265 3 2.001 7.412 25.677 4 1.740 6.443 32.120 5 1.585 5.869 37 .989 Table 12 illustrates the variables that we re more strongly correlated remain ed grouped across the baseline and guardianship conditions , whereas the variables that we re not as strong ly correlated did not appear in the guardianship condition. For exa mple, when Factor 1 from the baseline condition wa s compared to Factor 1 in the guardianship condition, the first three variables (LetterForm_Last, LetterForm_First, and Spacing_Letters ) we re consistent with the first three variables in the guardianship co ndition. However, the last two baseline variables for Factor 1 (Slant and Style of Allograph) dropped off Factor 1 for the guardianship condition. This could be the result of what is shown in Table 5 , where the first three variables in Factor 1 had high co rrelation scores of .941, .941, and .788 , indicating that they we re strongly correla ted. The last two variables had correlation scores of .427 and .414. These low correlation scores indicate d that these variables we re less related to the factor , and they explai ned less of the variation. Therefore, when the study participants start ed changing handwriting features under the guardianship condition, these weakly correlated variables drop ped off the factor. Additionally, when study participants start ed changing different aspects of their signatures, some variables appear ed on a factor in the guardianship condition that did not show up in the baseline condition. This is true for the variable " size." In general, when people sign their name normally, their handwrit ing size tends to be consistent (Huber & Headrick, 1999). Thus, the variable "size" did not appear in Factor 1 for the baseline condition, but it did come up in Factor 1 for the 69 guardianship condition. This finding suggested that when study participants be gan changing handwriting features related to letter form and spacing, they unintentionally altered the size of their writing. Table 12 also shows that while pen lifts and i - dot form and location remain ed grouped across the baseline and guardianship co ndition in Factor 2, handwriting legibility dropped off the factor in the guardianship condition. This finding suggested that pen lifts and i - dots we re features of handwriting that we re more stable and difficult to change , and that legibility wa s weakly co rrelated with the other handwriting variables in Factor 2. Therefore, when study participants start ed to change their i - dot form and location, they had less control over handwriting legibility, which became less important in explaining the variation. The t wo variables in Factor 3, which we re related to the t - bar (t - bar form and t - bar location), we re strongly correlated across the baseline and guardianship conditions. However, Factors 4 and 5 contain ed variables that we re more weakly correlated. In Factor 4, ed grouped across the baseline and guardianship conditions. However, the variables "ConnectStrokes" and " Hand" did not appear in the guardianship condition, suggesting that when study participants change d the o rder of their name, and they either add ed or remove d a period, the connecting strokes and writing hand beca me less important in explaining the variation. C onversely , when study participants change d the order of their name and the punctuation period, then i nitial strokes and abbreviations beca me more important. I n other words, when the order of the name and period marks we re changed, the study participants unintentionally change d their initial strokes and abbreviations. Factor 5 revealed that when the variab le "LineQuality" wa s changed by the study participants, the variables related to alignment (Alignment _Spec1, Alignment_Spec2) 70 beca me less important, and the variables "Hand , " "TStroke_Dir , " and "Slant" we re unintentionally changed. Table 12 . Comparison of baseline and guardianship factor solutions Factors Baseline Guardianship (Scenarios 1 - 3) 1 Letter Form_Last Letter Form_First Spacing Style of Allograph Slant Letter Form_Last Letter Form_First Spacing Size 2 Punctuation i - dot/jdot Form Punctuation i - dot/jdot Location Pen Lifts Legibility Punctuation i - dot/jdot Form Punctuation i - dot/jdot Location Pen Lifts 3 t - bar_Location t - bar_Form t - bar_Location t - bar_Form 4 Puncuation_Period Order of name Connecting Strokes Hand Order of name Initial Strok e Puncuation_Period Abbreviations 5 Alignment_Spec 1 Alignment_Spec 2 Line Quality Line Quality Hand Terminal Stroke_Direction Slant Research Question 4: How successful are writers in executing their intended disguise strategy? Overall, study partic ipants had a 61.8% success rate (the rate for each subject was based on the number of their attempted handwritin g changes). As shown in Table 13 , the success rate decreased as the level of guardianship increased for scenario 1, 70% for scenario 2, and 43% for scenario 3. 71 Table 13 . Table of success rates and 95% confidence intervals for each scenario Scenario Number Mean Lower 95% Upper 95% #1 308 73.1% 69.0% 77.2% #2 308 69.6% 65.3% 73.9% #3 308 42.9% 38.5% 4 7.3% Ta ble 14 illustrates the difference between scenario 1 and scenario 3 wa s significant ( p < 0.001), as wa s the difference between scenario 2 an d scenario 3 ( p < 0.001). However, the difference between scenario 1 and scenario 2 wa s not significant. T hese findings suggest ed that the study participants may have consider ed scenario 1 and scenario 2 to be si milar, whereas scenario 3 appeared to disrupt the concentration participants need ed in order to achieve their intended disguise strategy. This findin g wa s consistent with the pattern observed in Table 6 , where the intentional and unintentional changes in the participants' signature decrease d by more than 50 percent from scenario 1 to scenario 3, whereas the difference between scenario 1 and scenario 2 wa s approximately 15 percent. Table 14 . Table of pairwise comparisons with 95% confidence intervals for each scenario Scenario - Scenario Difference Lower CL Upper CL p - Value * #1 #3 30.2% 23.0% 37.4% <.0001* * #2 #3 26.7% 19.5% 33.9% <.0001* * #1 #2 3.5 % - 3.7% 10.7% 0.4871 * p - value from Tukey - comparison test ** p < 0.0001 Research Q uestions 5 and 6 level of guardianship? If yes, how does it change? T he chi - square test resul t that a ppears in Table 15 show ed no significant difference in disguise strategy across the three scenarios , 2 = 39.9778, p < 0 .05 . 72 Table 15 . Results of the chi - square test for assessing the relationship between guardianship and disguise strategies us ed Value df Asymp. Sig. (2 - sided) Pearson Chi - Square 39.9778 4 0.5601 Likelihood Ratio 43.2769 4 0.4166 N o. of Valid Cases 1678 The additional chi - square test of independence ( Table 16 ) used to assess the relationship between the level of guardianship and the top fi ve handwriting features ( Table 3 ) altered by the study participants showed no relationship (i.e., no change in distribution of the features that were changed based on the level of guardianship) 2 = 4.9421 , p < 0 .05 . 17 Table 16 . Results of chi - square test for assessing the relationship between guardianship and the top five disguise strategies used Value df Asymp. Sig. (2 - sided) Pearson Chi - Square 4.9421 8 0.7637 Likelihood Ra tio 5.0098 8 0.7565 N o. of Valid Cases 1256 Research Q uestion 7: - process when choosing a specific disguise strategy? As shown in Table 17 , the scenarios had - process es when cho osing a disguise strategy. Under the lowest level of guardianship (scenario 1), 85.1% of the study participants claimed to make major changes to their signatures. Conversely, under the highest level of guardianship (scenario 3) , only 2.6% of participants r eported making major changes to their signatures , and 93.2% claimed to make only one or two minor changes to their signatures. Under scenario 2, 67.5% of the study participants tried to keep 17 Separate chi - square tests were completed of scenario 3 compared to scenario 2, and scenario 3 compared to scenario 1. This supplemental analysis was first completed for all of the handwriting features, and then again for the top f ive handwriting features. The results of these chi - square tests revealed that there were no significant changes between these scenarios. 73 their signatur es similar to their original s ignatures , whereas 17 .5% chose to make major changes. Overall, a small number of participants indicated that the scenarios had no influence on their disguise strategy. For instance, only 1.3% of all participants stated that the scenario with the highest level of guardianship h ad no influence on their disguise strategy. Consequently, even when the participants were aware that their signatures would no t be compared, the scenario had an effect on their decision - making. Table 17 . Ways in which the scenarios influenced the particip Frequency Share No Influence Only 1 or 2 Minor Changes Kept Similar Major Changes Total Responses Time (Scenario) #1 27 (8.8%) 16 (5.2%) 3 (1.0%) 262 (85.1%) 308 Time (Scenario) #2 17 (5.5%) 29 (9.4%) 208 (67.5%) 54 (17.5%) 308 T ime (Scenario) #3 4 (1.3%) 286 (93.2%) 9 (2.9%) 8 (2.6%) 307 strategy would be less prominent. Supportin g the trend observed in Table 17 , which shows that partici pants changed their disguise strategy based on the level of guardianship, t he chi - square test result ( Table 18 ) was significant, 2 = 1147.48, p < 0.0001. Table 18 . C hi - square test for how participants were influenced by the scenarios Value df Asymp. Sig. (2 - sided) Pearson Chi - Square 1147 . 48 6 0.000* Likelihood Ratio 1149.38 6 0.000* N o. of Valid Cases 923 *p < 0.0001 Research Q uestion 8: Why do writers choose a specific disguise strategy ? To understand why participants select ed a particular disguise strategy, a frequency distribu tion table was created. The results reported in Table 19 show that the partic ipants devised 74 a disguise strategy based on the level of guardianship. Under the lowest level of guardianship , 88.3% of the study participants developed a disguise strategy based on the premise that the guardian would not evaluate or compare their signatur e to a known handwrit ing sample . Therefore, they could make any change without fear of being caught. In contrast, under the highest level of guardianship , 87% of participants devised a disguise strategy out of fear of being caught. Similarly, 8.5% were car eful when disguising their signature so that if the guardian questioned them, they could explain the difference between their disguised and known signatures. However, under scenario 2 , there were two competing reasons for choosing a particular disguise str ategy. On one side, 37.3% of participants felt that if the guardian evaluated their signature, he or she would not be very thorough. This made participants feel as though they could make any change and still get their signature past the guardian. Conversel y, 40.6% of participants were careful in devising their disguise just in case the guardian evaluated their signatures . The idea that the guardian could evaluate their signatures was enough of a reason to make them pause and develop a disguise strategy that they felt they could explain if asked. Table 19 . strategy 18 Frequency Share No I nfluence No Comparison Not Thorough Careful Feared Being Caught Total Responses 18 Part II of the su rvey questionnaire consisted of six short - answer open - ended questions; two for each of the three stu dy scenarios. These questions allowed participants to identify and describe the handwriting changes that they made when asked to disguise their signature, and explain why they chose a specific disguise strategy. The answers to these open - ended questions we re analyzed and coded as suggested by Creswell (2007). The participants' answers to the questions were segmented and categorized based on common themes. After completing the coding stage, five main categories were created : (1) the scenario did not influenc e the participants' disguise strategy (no influence), (2) the participants knew that the guardian would not evaluate their signature, thus they could do whatever they wanted and get away with it (no comparison), (3) if the guardian evaluated the signature, he/she would not be thorough (not thorough), (4) the participants wanted to be careful in case the guardian evaluated their signature (careful), and (5) the participants were concerned about being caught (feared being caught). 75 Table 19 Time (Sc enario) #1 27 (8.8%) 272 (88.3%) 2 (0.6%) 4 (1.3%) 3 (1.0%) 308 Time (Scenario) #2 14 (4.5%) 7 (2.3%) 115 (37.3%) 125 (40.6%) 47 (15.3%) 308 Time (Scenario) #3 3 (1.0%) 1 (0.3%) 10 (3.3%) 26 (8.5%) 267 (87.0%) 307 The chi - square test assess ing the rela tionship between the level of guardianship and the , why) for choosing a particular disguise strategy was significant, 2 = 1260.11, p < 0 . 0 001 ( Table 20 ) . This finding support ed the trend observed in Table 20 . As the level of gu resulted in a less prominent disguise strategy. Table 20 . C hi - square test for why participants were influenced by the scenario Value df Asymp. Sig. (2 - sided) Pearson Chi - Squ are 1260 . 11 8 0.000* Likelihood Ratio 1309.62 8 0.000* N o. of Valid Cases 923 *p < 0.0001 76 CHAPTER 6: DISCUSSION AND CONCLUSION Environmental criminologist s argue that because crime will persist rather than be eliminated from society , prevention is the best approach (Clarke, 1980; Eck & Weisburd, 1995; Felson & Boba, 2010). T he tenets of routine activity theory and situational crime prevention are all about focusing on specific types of crime. In regard to this study, signature forge ry is a crime that can be broken down into five different types: (1) simple, (2) simulation, (3) tracing, (4) cut - and - paste, and (5) auto - forgery. E ach type of signature forgery should be studied independently as each will have different patterns. The same is true when studying guardianship. There are different kinds of guardians and their effect s on the offender may vary depending on the type of crim e, where it takes place, and the capability (real or perceived) of the guardian. Since a focused approach on crime prevention requires a comprehensive understanding of the specific type of crime being studied, it is crucial to understand both the technical aspects of the particular criminal act and the impact that a specific kind of guardian has on the crime pro blem. Part I Disguised Handwriting (R esearch Questions 1 4) Although disguised handwriting is regularly encountered in the field of forensic document examination, distinguishing disguise from different authorship remains a problem area for many inves t igators (Bird et al., 2010; Bird et al. , 2012), which may be partly due to a serious lack of empirical work on this subject area (NRC/NAS, 2009). Two of the fundamental principles that underlie handwriting identification are: (1) no two writers share the same combination of handwriting characteristics, and (2) an individual's normal writing always has some degree of natural variation which prevents the writer from signing his or her name the exact same way twice. Although these principles are lacking in em pirical support (NRC/NAS, 2009) , currently they are the foundation that the field of questioned document examination is 77 built upon. It is beyond the scope of this study to address the need for such support. Thus, the methodology used to evaluate the signat ures in this study was based on standards (ASTM International, 2007) developed by the questioned document examination community. Primary Disguise S trategies The findings from the analysis used to identify the primary disguise strategies re vealed that alt hough there were 21 handwriting features that were purposely changed by study participants, only a few of these handwriting features were frequently changed. Letter form was the most common handwriting feature altered by the study participan ts. This findin g wa s consistent with previous research that also identified letter form as being the most frequently altered handwriting feature (Alford, 1970; Harris, 1953; Herkt, 1986; Konstantinidis, 1987; Michel, 1978; Mohammed, 1993). T he four other most frequently altered handwriting features identified in this study include d legibility, size, spelling 19 , and spacing. This finding wa s different from previous research studies that did not report legibility and spelling as frequently used disguise strategies (Herkt, 19 86; Konstantinidis, 1987; Michel, 1978; Mohammed, 1993 ) . Although previous research has reported signature size and spacing as commonly used disguise strategies (Herkt, 1986; Michel, 1978; Mohammed, 1993), neither of these tw o handwriting features were ide ntified as one of the top five handwriting features altered in these studies. The reason for the inconsistent findings may be due to differences in study design and the interpretation of the results. In this study , signature forgery was investigated exclus ively , there was a large sample size , and a survey questionnaire that solicited information about the participants intentional changes was incorporated . Pre vious studies, however, investigated disguised handwriting in signatures and handwritten notes (Har ris, 1953 ; Konstantinidis, 1987 ) , 19 Spelling refer red to the way the letters we re grouped to form a word. A list of definitions for the handwriting variables used in this study can be found in Appendix G. 78 and these earlier studies often utilized small sample sizes (e.g., Konstantinidis, 1987) with study populations that pose d a variety of threats to validity. For example, one study had a sample population composed entirely of police officers (Herkt, 1986) . Considering that some police officers receive training in forgery detection and investigate forgery crimes, their approach to signature disguise may be much different than the strategy used by a novice. Additionally, the authors of these p revious studies did not ask their participants to report the intentional changes they made. Instead, the researchers collected the disguised signatures and handwriting , and then interpreted the results. Therefore, they did not differenti ate between the intended and unintended changes. Consequently , the authors may have interpreted som e unintended changes as being intentiona lly made by the study subjects. For example, slant was reported as one of the most common disguise strategies ( Harris , 1953; Herkt, 1986; Konstantinidis, 1987; Michel, 1978; Mohammed, 1993 ) . However, slant may be an unintentional change that wa s wrongly interpreted as purposeful and categorized as a common disguise technique. Unintended H andw riting C hanges Previous res earch on disguised handwriting has not explored or identified the unintentional changes that occur as a direct result of the author's intentional changes. However, this information is important to investigators when they evaluate a questioned writing sampl e . S ome of the differences that we re noted between an authentic and questioned body of writing may have be en misinterpreted as a sign of a simulated forgery , when in fact it may have be en an unintended change that resulted from an intentional alteration m ade by the author when disguising his or her signature. For example, when writers intentionally alter letter forms to make their signatures appear differently, they may also unintentionally alter other handwriting features , 79 such as the speed of the writing . When a person consciously alters his handwriting , he may have to slow down the speed of his writing to execute his intended change. As a result of the slower writing speed , the writing line may show signs of tremor (i.e., an unsteady wavering stroke prod uced by a vibratory motion of the writing hand), which is a typical sign of a simulated forgery (Osborn, 1929). Consider ing that examiners have a difficult time differentiating a disguised writing from a simulated forgery ( Bird et al., 2010; Bird et al., 2 012), t he consequence of this for the investigator and the victims involved is considerable. Osborn (1929) and Hilton (1982) , two of the leading authorities of questioned document examination in North America, and other experts have asserted that if there is at least one significant 20 unexplainable difference, it can be inferred that the authors of the questioned and the known writings are different. Therefore, a disguised signature will contain at least one difference that may be considered significant by t he investigator and wrongly identified as a simulated forgery. In fact, it seems likely that this misidentification problem occurs quite often . A list of the unintentional handwriting changes w as develop ed for each scenario. Lateral expansion was the mos t frequently occurring unintentional change across all three scenarios. Other commonly occurring unintentional changes include d initial strokes, pen lifts, the spacing of letters, and letter form. Although t he f actor analysis provide d a limited amount of a dd itional information, it revealed that when participants began changing handwriting features related to letter form and spacing, they unintentionally changed the size of the writing. Letter form, 20 Significant is a subjective and controversial term used in the field of questioned document examination. Wha t may be considered a significant difference by one examiner may be classified as insignificant by another examiner. When forensic document examiners evaluate a signature for authenticity they rely upon their education, training, and experience to make a s ubjective determination regarding the similarities and differences they identify. If a difference is considered to be significant, then the signature in question is classified as a forgery. If the differences are considered to be insignificant, then they a re attributed to the natural variation of the writer and the signature is authenticated. 80 spacing , and size are handwriting features that , when combi ned , give the writing a certain "look." Therefore, if a writer wanted to make his or her signature "look" different by changing the letter form or spacing, then the size of the writing would most likely increase or decrease due to the forced change of the other two closely related features. The results of the analysis also revealed that w hen participants change d their i - dot form and location, they had less control over the legibility of the writing. Considerin g that i - dot form and location wa s a rare disg uise strategy, t his finding probably had more to do with the legibility of the writing. If the author altered the legibility of his or her signature, then it makes sense that the i - dot form and location would change. By making the writing neater , the autho r was more likely to reduce the speed of the writing , which resulted in a more carefully placed i - dot above the i - stem. Conversely, when the author made the writing messier , he or she most likely increased the speed of the writing , which made the location and form of the i - dot more variable. Additionally, when the order of the name and period marks were altered, the initial strokes and abbreviations were unintentionally changed. This finding wa s not surprising given that different letters have initial stro kes that begin in different locations, and wh en the order of the name was altered , the starting strokes changed by n ecessity . In other words, letters have initial strokes that begin in different locations. Therefore, by changing the order of the name , the initial strokes also changed . Moreover, when an author abbreviated his or her name , a period may have be en added . Similarly, if the writer removed an abbreviation by deleting it or spelling out the full name, then the period would have been removed. Addit ionally, when the line quality of the writing changed, variables related to alignment, the direction of the terminal stroke, and the slant unintentionally changed. Line quality has to do with the smoothness and rhythm of the writing line. Therefore, when a writer naturally produces a signature , there is a fluid "look" to it. W riting 81 with good line quality has a regular slant with letters that are arra nged in a consistent way. However, if the line quality of the writing is reduced , then the letters and slant become inconsistent , giving the writing a clumsy "look." Thus, changes in alignment and slant are closely related to line quality. It is important to consider that there we re a number of limitations with the Exploratory Factor Analysis. First, there wa s a question as to the factorability of the data. While it was noted Tabachnick and Fidell (2001), other guidelines, such as those identified by Kaiser (1974), would determine this to be below average , and th erefore not adequately related for factor analysis. Thus, it wa s difficult to determine whether the observed changes between baseline and guardian conditions reflect ed a pattern of behavioral changes or variability that may have reflect ed the modest factor ability of the data. Number of Handwriting Changes Made by the Study P articipants Study participants were given the freedom to incorporate as many changes to their signatures as they wanted to make. Interestingly, the total number of intended changes, as reported by the study participants, ranged fr om 1 to 3. This finding suggested that as the number of intended changes increase d, it became more difficult for study participants to execute their disguise strategy. Because handwriting is habit - formed (Calig iuri & Mohammed, 2012), when an individual intends to alter his handwriting, he must do so in a way that allows him to maintain some pictorial similarity 21 to his normal signature. However, the altered signature must be dissimilar enough that at a later poi nt in time , he can cl aim that the signature is not h i s . Therefore, if a writer wanted to devise a successful disguise , it seems like the best strategy would 21 The handwriting features that give a signature a certain "look" (e.g., letter form, slant, size, spacing) are often targeted by forgers to make their fake signature "look like" the authentic signature. Pictorial similarity is the term used to describe two signatures that are visually similar. 82 be one where fewer changes we re made to the handwriting. Indeed, if a writer incorporated too many cha nges at one time, then he risked creating a grotesque signature that would be more difficult to get past a guardian. In other words, a s the number of handwriting changes incorporated into the disguise strategy increased , the likelihood of success decrea sed . Although the c hi - square results were statistically insignificant, the trend th at fewer handwriting changes were made as the level of guardianship increased supported the idea that fewer changes we re easier to execute while maintaining a pictorial simi larity to the natural signature. This wa s important if the participants goal wa s to get the disguised signature past the guardian. Conversely, under the lowest level of guardianship , an increased number of changes may have result ed in a more grotesque - loo king signature that would have be en easier to deny later. Another interesting aspect of this f inding wa s that even though the number of changes decrease d as the level of guardianship increased, one change could be much more effective than three or more. For example, under scenario 1 , a participant could make one change that would make the signat ure unidentifiable. As shown below, the writer made one change to his normal signatu re (Figure 5a) , which resulted in a straight line (Figure 5b). Although there w as only one change , the change could be cl assified as a forgery by a non - certified expert 22 . Figure 5a. Naturally written signature Figure 5b. Altered signature 22 The field of questioned document examination is one of the most fragmented areas of forensic science. Certified experts would most likely reach an inconclusive opinion regarding the signature shown in Figure 5b. However, non - certified forensic document examiners and some law enforcement professionals might reach a different conclusion. More specifically, the signature may be cla ssified as a forgery and the writer of this disguise eliminated as authoring the signature. 83 T he straight line disguise strategy illustrated above was used under scenario 1 , but i t was not employed under scenario 3 . When a participant made a change under scenario 3 , i n which a guardian was present, it was less prominent. For example, Figure 6a wa s the writer ' s normal signature under scenario 3 , and Figure 6b wa s the author ' s altered signature under scenario 3. T he writer's disguise strategy was the addition of a line under his last name. In this case, it would not be difficult to identify the author of the signature. Figure 6a. Natural signature created Figure 6b. Altered signature created for scenario 3 for scenario 3 Considering these findings , it appe ared as though the study partici pants had a couple of constraints . First, because writing is habit formed , they were cognitively unable to concent rate on more than three changes at a time. If they tried to change too many handwriting features at the same time , then they risked failing their disguise attempt. Second, although one change could be much more effective than three, the kind of change empl oyed was determined by the level of guardianship. As illustrated in Figures 5a and 5b, one major change wa s only via ble when a guardian wa s ab sent. Therefore, more changes did not necessarily equal a more effective disguise strategy. Finally, the number o f inten tional changes was limited to 1 3 , but the numb er of unintentional changes was much higher. For example, the author of t he signature in Figure 5b made one intentional change that resulted in numerous unintentional changes such as letter form, size , initial strokes, terminal strokes, slant, and alignment. Although this example is a unique case because most study participants did not employ this strategy when disg uising their signatures, it did show th at an intentional change resulted in several unin tentional changes. 84 T he fact that one change generally resulted in two or three times as many unintentional changes illuminate d the problems examiners encounter when authenticating a signature. Therefore, examiners should consider disguise when evaluating all questioned signatu res , and use caution when weighing observed differences based on what is commonly believed to be the tell - tale signs of forgery. Indeed, these forgery indicators may simply be the unintentional changes that resulted from a specific di sguise strategy. Participants Success in E xecuting Their Intended Disguise S trategies Although the study participants intended to change their signatures in a variety of ways, they were not always successful in executing their disguise . Overall, 62% of the study participants success fully executed their disguise strategies. Interestingly , the participants success rate decreased as the level of guardianship increased. However, when evaluating the differences in success rates across scenarios , a pattern e merged. The difference between scenario 1 and 3 and scenario 2 an d 3 wa s significant, but the difference between scenario 1 and 2 was not . This suggested that the study participants may have viewed the guardianship under scenario 1 and scenario 2 in a sim ilar way. This helped explain why the success rates for scenarios 1 and 2 were similar, but different from scenario 3. Indeed, u nder the hig hest level of guardianship guardianship di srupted the concentration participants need ed in order to achieve their intended disguise strategy. Another possible explanation for the similar success rates between scenarios 1 and 2 was that under these scenarios , the participants intended changes we re more pronounced than under scenario 3 , which resulted in a higher success rate. For example, under scenarios 1 and 2 , the writer may have intended to change the writing size by making it larger. Because of the 85 decreased level of guardianship , the writer may have subconsciously made the writing much larger than he would have under scena rio 3. This makes sense when thinking about the effect that guardianship has on the writer. Not only did the number of changes decrease , but perhaps the quality of the inte nded change s also decreased as guardianship improved . Additionally, the overall success rate of 62% suggested that some participants may have thought that they executed th eir intended change when they had not. This finding wa s consistent with previous res earch that reported study participants failing to execute their intended disguise strategy (Harris, 1953; Konstantinidis, 1987). Habit may provide a good explanation as to why participants thought they made a change when they ha d not . Because a signature i s habit formed , it is unlikely that a person could explain how he ma d e his signature. In other words, when asked , most people would have a difficult time explai ning how they sign their name . As a result , when a person intentionally alters his signature , he has to guess what is normal and then deviate from that. Of course, this is hard to do because of natural variation , which creates a range of normally occurring differences. However, if h is attempted disguise strategy wa s more prominent , the chance of gett ing it right wa s improved. This is where guardianship really ha d an effect on the writer. Under the highest level of guardianship , the writer wa s more careful , and the chances of him creating a prominent change wer e less likely than when guardianship wa s a bsent. By altering a signature in a less prominent way , the writer increased his chance of making a ch ange that fell within his range of natural variation. Part II Guardianship (Research Q uestions 5 8) The Guardian 's E ffect on the P articipants D isgui se S trategy The participants' disguise strategy did not change depending on the level of guardianship. The reason for this may be due to the study participants limited understanding of handwriting 86 identification , and all of the different aspects involve d in the handwriting process. While certified experts are educated in all of the different handwriting features , most laypeople are not. Therefore, individuals who purposely change d their handwriting cho se to focus on those handwriting features that ga ve t he handwriting a certain "look" (Huber & Headrick, 1999). It is unlikely that a person who wa s not trained in handwriting identification would think to change such handwriting features as i - dot form and location, t - bar form and location , or initial and co nnecting strokes. Instead, l etter form, legibility, size, and spacing are four handwriting features that when changed make the handwriting "look" different , and they we re the handwriting features most frequently changed in this study . Thus, when participan t s alter ed their handwriting , they did not target different handwriting features based on the scenario. Instead, they modified the number of changes they mad e to the s ignature. hought - process B ehind a S pecific D isguise S trategy Research q uestions 7 and 8 were used to explore the thought - process behind the participants choice of disguise strategy , and tested how the level of guardianship influenced the participants decision - making. The analysis revealed that the participants were influenc ed by the scenario. The s eventh research question addressed how the participants were influenced by the level of guardianship , and the e ighth research question addressed why the level of guardianship influenced their decision to use a specific disguise str ategy. Under the lowest le vel of guardianship , part icipants knew that they could change whatever they wanted in their signature and get away with the disguise. Hence, most of the participants made major changes to their signatures. Some of these included t he writer signing a completely different name or drawing a straight line ( Appendix H). Although several participants viewed scenario 2 as having ineffective guardianship, many of them still wanted to be cautious in case the guardian did check their signatu re. As a result, a little over half of the participants kept their disguised signature similar to 87 their normal signature. This finding wa s interesting b ecause it demonstrated how a guardian that may be viewed as being ineffective can still affect the behav ior of the forger. Unlike scenarios 1 and 2, u nder the highest le vel of guardianship , the vast majority of participants only made one or two minor changes. These findings show ed that when participant s perceive d the guardian as capable, they change d their s trategy to ensure that the signature would be accepted by the guardian, while also leaving themselves with a point of difference to deny the signature at a later date. Of course, a large number of these minor changes fell within the writer's natural range of variation , resulting in a failed disguise strategy . As discussed above , forgery and document fraud crimes are different from the crimes that are generally studied by environmental criminologists. Unlike crimes of burglary and auto - theft where the offend er does not have to interact with a guardian, in crimes of forger y the offender must interact with the guardian. A guardian can help deter forgery crime. However, a guardian will not deter all forgery crimes, but he or she can impact the manner in which th e forgery is written, to the point where the disguise strategy fails. A fail ed disguise attempt is the next best thing to deterr ence since it will ensure that evidence that can help solve the crime is left behind by the forger . Research question 8 helped explain why the increased level of guardianship resulted in only minor changes to the handwriting. Under scenario 3 , the study participants reported a fear of being caught as the primary reason for only making one or two minor changes. U nder scenario 1 , t he study participants made what they consider ed to be major changes to t heir signature because they knew that no comparison would be made , which increase d the likelihood of getting away with the crime. Scenario 2 wa s interesting in the sense that participa nts had two competing reasons for their disguise strategy. One group who perceived the gu ardian as being incapable reported that they could do anything to their signature because the guardian would not be very 88 thorough in their comparison of the signatures . These participants also stated that even if the guardian did question them about the differences, they felt confident they could talk their way out of the situation 23 . The o ther group stated that they wanted to be careful in case their signature was compa red. As a result, they attemp ted to keep their signature similar to how they normally sign ed their name. These findings suggest ed that the p erceived level of guardianship wa s more important than the actual guardianship. If the participant perceived the g uardian as being capable, and feared being caught, they made minor changes to their signature. However, if the participant perceived the guardian as being incapable , then they made, what they claimed to be, major changes to their signature. This finding su pported Tilly's (2009) observation that guardian capability may be less important than guard ianship credibility. The result s from this study suggest ed that having a guardian present in crimes related to for gery and document fraud wa s not enough to preven t many of the crimes . E ven when a guardian was present , a third of the st udy participants made major c hanges to their signatures. Similarly, under scenario 3, the participants' disguise strategy was based on their perception of he guardian only had to be perceived as being capable, which created caution 24 in a majority of the participants, resulting in a considerable modification in their dis guise strategy. This change ultimately resulted in two outcomes. First, many of the disgui se attempts failed. Second, the successful disguise strategies resulted in only a few minor alterations 23 These thoughts were drawn and paraphrased from Part II of the survey questionnaire where the study participants answered open ended questions about how and why th e level of guardianship influenced their disguise strategy. 24 Participants used phrases such as "I feared being caught" and "I was scared that I would get caught" to explain why they chose a certain disguise strategy under scenario 3. 89 that mad e a future identification highly probable . 25 Figure 7a is an example of the writer's natural signature. Under scenario 3 , the writer successfully changed the ending stroke in his signature (Figure 7b). However, the alteration was minor and a certified expert would evaluate this signature and most likely conclude that the writer of the Figure 7a signature was the writer of the Figure 7b signature. T herefore , if the guardian wa s unable to prevent a crime from taking place , the results of this study illustrated that she c ould affect the forgers behavi or in a way that increased his chances of being identified after the crime ha d been completed , or the guardian could force a behavior that result ed in a failed attempt by the forger. Figure 7a. Natural signature created Figure 7b. Altered signature created for scenario 3 for scenario 3 Limitations of Study All research has lim itations and this study wa s no exception. Due to the nature of this project and the research questions that were internal validity. Therefore, in order to ensure that the change in the disguise strategy was produced solely b y the level of guardianship and not by an extraneous factor, the dat a collection phase of this res earch took place in a controlled environment . Of course, by focusing on the gs were necessarily weakened (Shadish, Coo k, & Campbell, 2002). However, this was considered to be an acceptable trade - off. 25 When a certified ha ndwriting expert evaluates a signature for authenticity, he compares the questioned writing to the known writing sample and notes all similarities and differences. If the questioned writing does not have any significant unexplainable differences, then the questioned signature is authenticated. In the current study, a participant's disguise strategy may have been successful, but when evaluated by a certified expert, the alterations may be categorized as insignificant due to natural variation. This would resu lt in a positive identification. 90 The integration of new technology must also be considered. I t has become common for retailers to use d igital tablets to capture c signatures, which varies from ink - to - paper signatures in a variety of ways. For example, the writing position and the writing instrument s used to sign digital tablet s are different from those used to sign a piece of paper. T his study did not expl behavior might have been affected by a digital tablet , or how their disguise strategy might have changed. Therefore, the findings can be applied to auto - forgery on paper, but not necessarily auto - forgery on tablets. E ach pa rticipant was given a limited amount of time to prepare for the handwriting exercise. Although the participants were initially "primed" to think like someone who wanted to disgu ise their handwriting, and were allowed to practice disguising their handwritin g for a few minutes before executing their disguise strategy, it is possible that some of the participants might not have been fully committed to the exercise. In other words, their effort to disguise their handwriting may have been different if they had b een in the real world. The participants in this study knew that they were in a controlled environment , and that it was an experiment with no consequences . As a result , they had a chance to be as daring as they wanted without worr y of being caught and punis hed. Moreover, in a real world setting , forgers may spend several hours thinking about the disguise strategy they want to use; this was not possible in a controlled environment. Final ly, although the measures used and the coding scheme completed were rel iable , there may be some questions regarding the coding of the handwriting data collected or the ability and interpretations of the coder . Due to resource limitations , a Board Certified Forensic Document Examiner completed all of the coding. Although the c oding scheme used wa s based 91 on established standards (ASTM International, 2007 ) , it wa s possible that other certified document examiners may h ave obtained different results. Despite the concerns identified above, t his wa s the first study that used a large sample size to investigate disguised signatures in a compreh ensive way. Furthermore, it wa s the only research that of capable guardianship affected the disguise strategy use d. The findings from this study can help scholars draw inferences from which more specific studies can be designed. For example, studies involving how manager - guardians interact with and affect the behavior of offenders can be developed. Additionally , the primary (intended) and secondary ( uninte nd ed) handwriting changes that have been identified in this study can be used to further our knowledge of the differences between forgery versus disguise, and forgery versus simulation. Directions for Future Research Future research could focus on signa ture forgery and the guardianship aspect of routine activity theory. Signature For gery It was beyond the scope of th is study to test whether the disguise strategies used b y the study participants would actually be accepted by a guardian or trick an exp ert into think ing that the disguised signature was not authored by them, but forged by someone else. Therefore, one direction for future research is to conduct a similar study to test the proficiency of handwriting experts and non - experts such as cashiers and bank tellers . Such a study could potentially be redesigned in the following way. First, a pool of study participants would be selected and then tol d to disguise their handwriting under conditions similar to the ones used in this study. Next, a group of people would be instructed to simulate the study participants ' normal signatur es. After 92 both of these steps wer e completed, the signatures would be assembled for the proficiency test which would consist of ten to fifteen normal signatures for the test tak ers to use as their comparison sample. Then a group of questioned signa tures would be created from the disguised signatures, the signature simulations, and the writers ' natural signatures. The experts, cashiers, and bank tellers would determine if the ques tioned signatures were authored by the known writer, disguised by the known writer, or simulated by somebody else. This kind of exercise would identify which disguise stra tegies we re the most effective in deceiving the experts. Another direction for future research relates to the importance of different handwriting features and the number of changes that are intentionally made to a signature. It is unknown what altered handwriting features are most likely to fool the experts when disguising a signature , and how many changes are necessary to convince handwriting experts that it is not a disguised signature. While it may seem tha t the more changes someone made to his/her signature , the more succe ssful the disguise strategy would be, this m ight not be the case. For example, a single intended change can be a more successful disguise ( able to get past the guardian and fool the experts into thinking it was not authored by the claimant) than a disguise strategy that incorporat es three or four intentional changes ( Fi gure 5a; b) . Therefore, this aspect of signature disguise , relating to number and type of features, needs to be investigated further. In this study , unintentional handwriting changes that occurred as a result of the writer's intentional changes were ide ntified. However, because each writer created up to three changes to his or her signature , it was not possible to identify which intentional changes cause d other handwriting features to unintentionally change. Future research may be able to identify the co rrelations between the intended and unintended handwriting changes by focusing on one change at a time. For example, participants could be asked to change a handwriting feature , such 93 as letter form . After making the intended change , the unintended change c ould be identified. Participants could then be asked to change another handwrit ing feature, such as size. This process could be repeated until the different handwriting features have been isolated and correla ted to the unintended changes. Another area in need of future research involves signatures on electronic devices. The use of signature produced on a digital tablet is different in a variety of ways from an ink - to - paper signature. For example, the writing position and the writing instrument used to sign vary. Future research c ould investigate disguise strategies might be affected when signing on a digital tablet , and how they might be different fro m the on es used on a paper document. Moreover, digital tablets provide additional information on handwriting features (e.g., speed) that cannot be accurately determined from a paper signature. Therefore, a study similar to this one, but using digital table ts , could be a contribution to the current literature on signature disguise. Finally, further investigations on other types of forgery such as simple and simulation would provide valuable insight into the differences betw een disguised signatures ( auto - fo rgery) and signatur es created by someone else ( simple and simulated forgeries). Perhaps, the handwriting features chosen by a person who is trying to simulate a signature to make it look authentic are the same features chosen by a disguiser to make the sig nature look different from the authentic. This kind of study would be helpful in understanding the different type s of document forgeries that exist and the thought process followed by the individuals creating the forgeries. 94 Environment and Guardianship Clarke ( 1997 ), Felson and Boba ( 2010 ) , and other crime prevention scholars clai med that focusing on crime s ituations can reduce crime. In this study , auto - forgery was investigated. However, if a similar study looked separately at other types of forgery ( si mple, simulation, tracing , and cut - and - paste), different patterns could emerge. Thus, a future study c ould be conducted to explore how guardian(s) affect different kinds of forgers. Other directions for future research include conducting observational st udies of manager - guardians in their workplace to examine how they interact with customers and evaluate documents. For example, what forms of identification do manager - guardians request from customers , and how long do they spend evaluating them? Similarly, studies c ould be designed to investigate how di fferent types of guardians ( handler - guardians , manager - guardians, target - guardians) influence forgers ' decision - making , and if their view of authority affects their guardian capability assessment. In other wor ds, are po sitional authority figures such as manager - guardians viewed as being more capable than target - guardians (e.g ., pedestrian, shopper, etc . )? These kinds of research questions may be answered through interviews and ethnographic studies. Finally, a n alternative approach to the present study could include an experiment within this design. Random assign ment of one - third of the study participants to each one of the three scenarios could strengthen the experiment . Adding this kind of randomization to th e research design would eliminate any potential biasing effects resulting from how the participants interpre ted the handwriting disguise exercise in relation to the increasing levels of guardianship. Policy Implications Guardianship mattered in forgery a nd document fr aud crimes. A s the level of guardianship increased , the study participants' disguise strategies changed. In terms of policy 95 implications, this suggested that prevention should be focused on reducing opportunities for signature disguise , and e ncouraging those closest to the crime problem to become more active in increasing the ri sks for the offender. Clarke (1997) developed sixteen situational crime prevention techniques that could be used by practitioners to reduce criminal opportunities. Sinc e 1997 the number of these preventative techniques has grown to twenty - five 26 (Cornish & Clarke, 2003; Pop Center, 2015), and are based on five main mechanisms: (1) increase the effort required to commit the crime , (2) increase the risks the offender must m ake to carry out the crime , (3) reduce the rewards obtained from the crime , (4) reduce provocations that may tempt offenders into criminal acts , and (5) remove excuses for the criminal offense . However, Clarke (2005) acknowledged that not all of the techni ques are equally suitable for all crimes. Thus, understanding a specific crime problem will enable practitioners to employ the most effective preventative technique. Based on the findings of this study 27 , the most effective situational preventative techni ques for forgery and document fraud crimes are those that increase risks. The following are some suggested actions that business owners and security professionals may take to prevent forgery and document fraud. First, people who alter their signature devis e a disguise strategy based on their assessment of the guardian s capability. Therefore, business owners could place a sign on their entrance door that says something along the lines of, "our employees are trained in forgery detection," "we check every ID and signature , " or "all forgers will be prosecuted." If the 26 For a complete list of Clarke's classification of the 25 situational crime prevention techniques visit: http://www.popcenter.org/25techniques/. 27 The possibility of getting caught was enough to impact the study partic ipants' disguise strategies. Thus, situational crime prevention techniques that make offenders think they have a higher risk of getting caught are enough in preventing or altering a disguise strategy. 96 offender believes that her signature will be checked, and that the risk of getting caught is higher, then she will either abandon the forgery attempt or change the way her signature is disguised. Another strategy would be to strengthen formal surveillance. For example, the business owner could employ loss prevention professionals who are visible and active around the checkout counters . These employees could actually be trained in forgery detection or they could act in a way that makes the customers believe that they are. In other words, the loss prevention staff could be seen checking IDs and signatures. This same effect could also be accomplished with the other store employees. The managers could have their cashiers take more time comparing signatures and checking IDs. Again, the important thing is to make the offender believe the guardians know what they are doing and that their risk of getting caught is higher. Similarly, utilizing additional p lace managers and rewarding vigilance would improve surveillance and send a message to the public that the business is cracking down on forgery and document fraud crimes. The additional place managers could assist other store employees by checking IDs and signatures more thoroughly while not increas ing the wait time for customers. Furthermore, rewarding store employees who report signatures that do not "look" right to store management could increase their interest in identifying forgeries. Finally, under s cenario 1 , most of the study participants made major changes to their signatures. Therefore, restaurant owners could institute a policy that all signatures are to be compared and that the wait staff will not leave the cus b een signed. Another option would be to have the customers pay for their meal at a checkout counter. This would require the customer to sign the document in front of a guardian , and would allow other customers to observe the cashier checking signatures and IDs . 97 Conclusion The current study expand ed the existing knowledge on signatur e disguise in two ways. First, it identified the handwriting features that were intentionally altered by the study participants when disguising their signatures . Second , it ide ntified the unintentional handwriting changes that resulted from the intentional changes. While some of the findings were consistent with previous research on disguised handwriting, other findings were not . For example, spelling was one of the most frequen t handwriting changes made by the participants in this study, but it was not identified as a commonly occurring change in previous studies. The results also showed that participants we re not always successful in executing their intended handwriting changes . The guardianship concept was explor ed by examining the influence that manager - guardians have on individuals who must interact with a guardian in order to accomplish their crime. Considering the findings of this study, and how the g uardians' role has bee n defined , it is clear that current definitions do not capture guardians and guardianship in a way that includes all types of crime. Forgery is a crime that often requires the offender and the guardian to interact. Moreover, considering the societal impact of forgery, it is clear that many of these crimes are not being prevented. Thus, in addition to preventing crime, the guardians' role should also include a behavior alteration component. In other words, if the guardian does not prevent the crime from occ urring, then the guardian should alter the behavior of the offender in such a way that she can be identified after the crime has occurred. The tenets of routine activity theory and situational crime prevention emphasize the importance of focusing on parti cular types of crime so that specific crime prevention strategies can be employed. Guardianship is one of the main concepts in routine activity theory, and one of the key elements in the criminal event model. This study has shown that guardianship can affe ct 98 how a forger disguises her handwriting . It appears that when the offender and guardian interact, the offender makes decisions on how best to carry out the crime based on his or her assessment erefore , in order to study the guardianship concept and apply it to different crime problems , the defi nition needs to be expanded to not only include crime deterrence, but to also include behavior alteration. As a result, the following def inition of a capa ble guardian i s proposed: a capable guardian can be defined as an individual(s) whose presence either prevents a crime from occurring or alters the crime scen e , which enables investigators to identify the offender at a later date. In other words, task is to prevent the crime from taking place, and the secondary task is to alter the behavior of the undeterred offender. Further, this study has shown that by combining forensic science with social science, a specific crime problem, such as forgery, can be studied and understood in a more holistic way. 99 APPENDICES 100 APPENDIX A : Definitions and Measurements of Guardian(ship) Table 2 1. Definitions and Measurements of Guardian(ship) Author and Publication Date Definition of Guardian(ship) Measurement of Guardian(ship) Cohen & Felson (1979) Labor force participation and single - adult households . Co hen & Cantor (1980) Guardianship is the condition in which specific individuals, be they law enforcement agents or the common citizenry, are actual or potential protectors of the offender's target (formal organizations specifically designe d for protective purposes, individuals in the proximity of persons or property). Guardians may be persons who are potentially able to prevent the successful completion of the crime, either by their physical presence alone or by some form of direct action ( companions, pedestrians, private security guards, law enforcement officers). The number of people in the household and major activity (work, keeping house, or being unemployed). 101 Table 21 (cont'd) Author and Publication Date Definition of Guardian(sh ip) Measurement of Guardian(ship) Cohen, Kluegel, & Land (1981) The effectiveness of persons (housewives, neighbors, pedestrians, private security guards, law enforcement officers) or objects ( burglar alarms, locks, barred windows) in preventing vio lations from occurring, either by their presence alone or by some sort of direct or indirect action. Household composition and labor - force statuses. Lynch (1987) The presence of persons or devices that can prevent or inhibit victimization ( professi onal guards, laymen with an interest in preventing victimization, alarms, etc.) . Frequency of local travel on the job and frequency of overnight trips (guardianship on the job). Stahura & Sloan III (1988) Police employment, police expenditure, fe male labor force nonparticipation. Miethe, Stafford , & Sloane (1990) Number of people in the household and daytime and nighttime activity. Bennett (1991) Female labor force participation. 102 Table 21 (cont'd) Author and Publication Date Defi nition of Guardian(ship) Measurement of Guardian(ship) Garofalo & Clark (1992) Household occupancy and the presence of lighting, dog s , alarm system s , neighborhood watch es . Eck & Weisburd (1995) People who can protect targets (friends, private security guards, public police). Felson (1995) Any person who serves by simple presence to prevent crime and by absence to make crime more likely. Rice & Smith (2002) The number of owner - occupied places, multifamily buildings, single - parent homes, commercial places. Tewksbury & Mustaine (2003) Self - protective behaviors (the possession and carrying of self - protective measures such as a weapon). 103 Table 21 (cont'd) Author and Publication Date Definition of Guardian(ship) Measur ement of Guardian(ship) Tseloni, Wittebrood, Farrell, & Pease (2004) Refers to the capability of persons and objects to prevent crime from occurring. Household composition and occupancy, neighbors agree to watch homes, use of self - protection measure s, participation in collective crime prevention enterprises. Coupe & Blake (2006) Occupancy of housing at time of burglary, detection of burglar (reporting), security devices. Felson (2006) Anybody passing by or assigned to look after people or property who keep s an eye on potential targets of crime. This usually refers to ordinary citizens, not police or private guards. Wilcox, Madensen, & Tillyer (2007) Individual - level guardianship is possessing qualities that relate to social ties and interpersonal control. Environmental - level guardianship is the collective degree to which individuals or objects in a bounded locale possess qualities related to social ties and social control. Individual - level target hardening, neighborhood - level targe t hardening, home occupancy, informal social control, place management and surveillance, natural surveillance. 104 Table 21 (cont'd) Author and Publication Date Definition of Guardian(ship) Measurement of Guardian(ship) Reynald (2009) Occupation of property, visibility of occupants, occupant(s) monitor property and streets, the willingness of occupant(s) to intervene. Felson & Boba (2010) Someone whose presence serves as a gentle reminder that someone is looking or those who engage in natural surv eillance, including ordinary citizens going about their daily lives but providing by their presence some degree of security. Reynald (2010) Residential guardians are the crime control agents who are most likely to assume primary responsibility for the ir residential space, including the people and property contained therein. Monitoring and intervention. Sampson, Eck , & Dunham (2010) Individuals who have the goal of protecting targets. Guardians are highly varied and include normal citizen behavior when they look out for each other and their property, groups of strangers who as a byproduct of their numbers guard each other , and individuals who are specifically employed to protect people and things. 105 Table 21 (cont'd) Author and Publication Date Definition of Guardian(ship) Measurement of Guardian(ship) Hollis - Peel, Reynald, van Bavel, Elffers, & Welsh (2011) Any person and every person on the scene of a potential crime that may notice and intervene (whether they intend to or not). Re ynald (2011a) Visible occupancy (available guardian), monitoring, intervention. D'Alessio, Eitle , & Stolzenberg (2012) Burglary rate (weekday, weeknight, weekend), unemployment rate. Hollis - Peel, Reynald, & Welsh (2012) Availability, mo nitoring, intervention. Breetzke & Cohn (2013) Population density, road density, community is partially or fully gated. Hollis, Felson, & Welsh (2013) The presence of a human element which acts, whether intentional ly or not, to deter the would - be offender from committing a crime against an available target. 106 Table 21 (cont'd) Author and Publication Date Definition of Guardian(ship) Measurement of Guardian(ship) Hollis - Peel & Welsh (2014) Availability of guardians, whether or not th e guardians are visibly monitoring and engaging in surveillance, and whether or not the guardians intervene. Uittenbogaard (2014) Visibility and surveillance. Coupe & Fox (2015) House (target) occupancy, neighbors' occupancy, number and types of all properties, number of roads and paths, security patrol, guardianship objects (CCTV, burglar alarms, dogs, other security devices). Note: Journal articles using another author s definition of guardianship w ere not included. Only original defi nitions were included. 107 APPENDIX B : Survey Questionnaire Survey Questionnaire on Intentionally Disguised Handwriting Instructions : Please answer each of the survey questions and statements below; your responses will remain anonymous. Thank you for your participation. Part I - General Questions Please answer the following questions . 1. Do you think the cashiers at your local retail and grocery stores are trained in forgery detection (i.e., trained to identify fake or altered signatures that may be presented to them)? (a) Yes (b) No 2. , , ined . Poorly Trained Very Well Trained 1 2 3 4 5 3. What do you think cashiers at stores, restaurants , and banks look for when they compare the signature on your personal identification (i.e. , drivers license, passport, ID card, etc.), or on the back of your credit card, to a check or credit card receipt that you have signed? For the statements below, please select from the following answer choices: Strongly Disagre e Disagree Neutral Agree Strongly Agree 4. It would be likely for a cashier at Walmart to accept a check or credit card receipt that contains a disguised signature. 1 2 3 4 5 5. It would be likely for a cashier at Target to accept a c heck or credit card receipt that contains a disguised signature. 1 2 3 4 5 108 Strongly Disagree Disagree Neutral Agree Strongly Agree 6. It would be likely for a cashier at Meijer to accept a check or credit card receipt that contains a disguised signature. 1 2 3 4 5 7. It would be likely for a cashier at Kroger (or similar grocery store) to accept a check or credit card receipt that contains a disguised signature. 1 2 3 4 5 8. It would be likely for a cashier a t K - Mart to accept a check or credit card receipt that contains a disguised signature. 1 2 3 4 5 9. It would be likely for a cashier at MSU Credit Union (or other bank) to accept a check or credit card receipt that contains a disguised signature. 1 2 3 4 5 10. It would be likely for a cashier at Michigan Department of Motor Vehicles (DMV) to accept a check or credit card receipt that contains a disguised signature. 1 2 3 4 5 11. It would be likely for a cashier at Olive Garden to accept a check or credit card receipt that contains a disguised signature. 1 2 3 4 5 12. It would be likely for a cashier at Old Chicago (or similar restaurant) to accept a check or credit card receipt that con tains a disguised signature. 1 2 3 4 5 Part II Scenarios Scenario #1 Restaurant Setting (signatures will not be compared) 13. How did you change or alter your original signature to make it appear different? Please provide as much detail as you can. 14. Did knowing that your signatures were not going to be compared influence your disguise strategy? How did it influence your strategy? Why? 109 Scenario #2 Retail Setting (signatures may or may not be compared) 15. How did you change or alter your origin al signature to make it appear different? Please provide as much detail as you can. 16. Did thinking that someone might compare your signatures influence your disguise strategy? How did it influence your strategy? Why? Scenario #3 Bank Setting (signatu res will be compared) 17. How did you change or alter your original signature to make it appear different? Please provide as much detail as you can. 18. Did k nowing that somebody for sure would be comparing your signatures influence your disguise strategy? How did it influence your strategy? Why? Part III Classification Please answer the following questions. 19. What is your age (in years)? __________________ 20. What is your gender? (a) Male (b) Female 21. What is your race/ethnicity? (a) Am erican Indian or Alaskan Native (b) Asian (c) Black or African American (d) Native Hawaiian or Other Pacific Islander (e) White Non Hispanic (f) White Hispanic (g) Other: _________________________ 22. What country are you from? _________________________ ____ 110 23. What are you currently majoring in? _______________________________ 24. What is your highest degree earned? (a) High School Diploma (b) Some college (c) College graduate - Associate degree Major: _____________________________ (d) College graduate Bachelor degree Major: _____________________________ (e) College graduate Master degree Major: _____________________________ What was your B.A. or B.S. degree in? __________ _________________ (f) College graduate Doctorate degree Major: _____________________________ (g) Professional school graduate (e.g. , DVM, MD, Law, etc . ) 25. Which hand do you write with? (a) Right (b) Left (c) Ambidextrous (write with both h ands) 26. Have you studied or received any training on forgery detection? (a) Yes Specify: __________________________ (b) No 27. Have you ever been a victim of forgery? (a) Yes Specify: __________________________ (b) No 28. Have you ever been a victim of identi ty theft? (a) Yes Specify: __________________________ (b) No 111 APPENDIX C : Consent Form Consent form for participation in the study entitled: The research study in which you are being asked to participate i nvestigates the kinds of disguise techniques used by individuals who sign a document with the intention of later denying ever signing it. The purpose of this study is to identify the thought process and techniques used by individuals who intentionally disg uise their handwriting, and offer some suggestions that will provide forensic document examiners with the appropriate examination techniques in relation to intentionally disguised signatures. You will be asked to sign your name several times and then ask ed to sign your name again , but with an attempt to disguise the writing. Following the writing activity you will be asked to take a brief survey regarding your disguise technique and strategy. We estimate that this process will consume no more than 30 minu tes of your time. Every effort will be made to maintain the privacy of your participation. The survey questionnaire and handwriting samples received for this project will be kept confidential and stored in a locked file cabinet in a locked office. Data from this research will be stored on a password protected computer in a locked office, and only research staff and the Michigan State University Human Research Protection Program will have access to the information over the course of the study (approximate ly 2 years). Your handwriting and any other identifying information will be kept confidential and all information received will be held in strict confidence , and your privacy will be protected to the maximum extent allowable by law. Upon conclusion of the study all handwriting samples will be destroyed. The data we collect may be used for publication or presentation, but your comments and identity will remain anonymous. The sole exception to this is if you agree to allow your handwriting to be used as exam ples in presentations and/or publications (below). There is no known risk or direct benefits to you when participating in this study. Your participation in this study is completely voluntary, and you may choose to refuse or discontinue participation in an y part of the process, or altogether, without penalty. The investigator(s) will be present when you are reading this form and will clearly answer any questions you have. If you have any concerns or questions about this research study, such as scientific i ssues, how to do any part of it, or to report an injury, please contact the researcher David Foran, Ph.D., by phone (517) 432 - 5439, email: foran@msu.edu, or regular mail: 560 Baker Hall, East Lansing, MI 48824. If you have questions or concerns about your role and rights as a research participant, would like to obtain information or offer input, or would like to register a complaint about this study, you may contact, anonymously if you wish, the Michigan State University's Human Research Protection Program at 517 - 355 - 2180 , Fax 517 - 432 - 4503 , e - mail irb@msu.edu , or regular mail at 207 Olds Hall, MSU, East Lansing, MI 48824 . Your signature below indicates your voluntary agreement to parti cipate in this study. Signature _________________________________ Date _______________ __________ Print Name_________________________________ 112 APPENDIX D : Documents used for Handwriting Exercise Figure 8a. Scenario 1 - Restaurant Setting Figure 8b. Scenario 2 - Retail Setting Figure 8c. Scenario 3 - Ban k Setting 113 APPENDIX E : Case Scenarios Signature will not be compared Restaurant Setting Scenario 1. It is Friday evening and you decide to go out and eat at a restaurant of your choice. After you finish your meal, the waitress leaves the check at yo ur table and tells you she will pick it up when you are ready to pay. After a few minutes she comes back and you give her the check and your credit card. When the waitress returns , she thanks you for dining with them and tells you to have a nice evening. Y ou know that the waitress will not compare your signature on the sales receipt to the decide to disguise (alter) your signature with the intention of later denying ever signing the sales receipt or eating at this restaurant on this date. After deciding on a disguise strategy, go ahead and sign the sales receipt in front of you. Signature may or may not be compared Retail setting Scenario 2. It is a Saturday afternoon and you decide to go to Walmart [or your choice of retail store] and do some shopping. After you have completed your shopping and are making your way to the cash register you start to think about whether or not the cashier will ask for or not. You know from past experience that sometimes the cashier will take a look at the signature on they will not ask for any identification at all. After thinking it over you decide that you will disguise your signature with the intention of later denying ever signing the sales receipt and purchasing the items. After deciding on a disguise strategy, go ahead and sign the sales receipt in front of you. Signature will be compared Bank setting Scenario 3. Wednesday morning on your way to work you decide to stop off at your bank and make a withdrawal. Thinking about all of the bills you have to pay, you consider trying to d isguise your signature with the intention of later denying ever making the cash withdrawal. However, you know that the bank teller will definitely ask for identification and make a comparison of your signatures. Although you know that your ID will be check ed and your signatures will be compared, you decide to try and disguise your signature anyway. After deciding on a disguise strategy, go ahead and sign the sales receipt in front of you. 114 APPENDIX F : Handwriting Coding Scheme Handwriting Study Coding Sc heme Elements of Style (ES) Elements of Execution (EE) Other (O) Participant Self - Report (Changes made, Yes or No) Handwriting Elements Scenario 1 - Restaurant Scenario 2 - Retail Scenario 3 - Bank Arrangement (ES) Yes/No Yes/No Yes/No Style of a llograph (ES) Yes/No Yes/No Yes/No Lower case letter form (ES) Yes/No Yes/No Yes/No Capital letter form (ES) Yes/No Yes/No Yes/No Spacing (ES) Yes/No Yes/No Yes/No Connecting Strokes (ES) Yes/No Yes/No Yes/No Slant (ES) Yes/No Yes/No Yes/No Size (ES) Yes/No Yes/No Yes/No Proportions (ES) Yes/No Yes/No Yes/No Lateral expansion (ES) Yes/No Yes/No Yes/No Alignment (EE) Yes/No Yes/No Yes/No Pressure (EE) Yes/No Yes/No Yes/No Speed (EE) Yes/No Yes/No Yes/No Hiatuses (EE) Yes/No Yes/No Yes/No Pen lif ts (EE) Yes/No Yes/No Yes/No Line Quality / Tremor (EE) Yes/No Yes/No Yes/No Initial Strokes (EE) Yes/No Yes/No Yes/No Terminal Strokes (EE) Yes/No Yes/No Yes/No Embellishments (EE) Yes/No Yes/No Yes/No Simplification (EE) Yes/No Yes/No Yes/No Abbrev iations (EE) Yes/No Yes/No Yes/No Punctuation/Diacritics (EE) Yes/No Yes/No Yes/No Legible signature (EE) Yes/No Yes/No Yes/No Illegible signature (EE) Yes/No Yes/No Yes/No Handedness (EE) Yes/No Yes/No Yes/No Writing movement (EE) Yes/No Yes/No Yes/N o Spelling (O) Yes/No Yes/No Yes/No Order of name (O) Yes/No Yes/No Yes/No 115 Scientific Detailed Analysis (What changes did the participants actually make) Handwriting Elements Scenario 1 - Restaurant Scenario 2 - Retail Scenario 3 - Bank Arrangeme nt (ES) Yes/No Yes/No Yes/No (Combo of many other elements) Style of allograph (ES) Yes/No Yes/No Yes/No 1. Print 2. Cursive 3. Script 4. Combo Lower case letter form (ES) Yes/No Yes/No Yes/No Capital letter form (ES) Yes/No Yes/No Yes/No Spacing (ES) Yes/No Yes/No Yes/No Letter : Word : 1. Narrow 1. Narrow 2. Moderate 2. Moderate 3. Wide 3. Wide Connecting Strokes (ES) Yes/No Yes/No Yes/No 1. Garland 2. Arcade 3. Angular 4. Thread 5. Indeterminable 6. Combo Slant (ES) Yes/No Yes/No Yes/No 1. Backward 2. Forward 3. Vertical 4. Combo 5. Indeterminable Size (ES) Yes/No Yes/No Yes/No 1. Small 2. Medium 3. Large 4. Extra - Large Proportions (ES) Yes/No Yes/No Ye s/No 1. Increase slightly 2. Increase greatly 3. Decrease slightly 4. Decrease greatly 116 Lateral expansion (ES) Yes/No Yes/No Yes/No 1. Narrow 2. Moderate 3. Wide Alignment (EE) Yes/No Yes/No Yes/No Signature to Baseline: 1. As cending 1. Above 2. Descending 2. Sitting On 3. Level 3. Below 4. Combo Pressure (EE) Yes/No Yes/No Yes/No 1. Heavy 2. Medium 3. Light Speed (EE) Yes/No Yes/No Yes/No 1. Fast 2. Medium 3. Slow Hiatuses (EE) Yes/No Yes/No Yes/No 1. One 2. Two 3. Three 4. Four or more Pen lifts (EE) Yes/No Yes/No Yes/No 1. One 2. Two 3. Three 4. Four or more Line Quality / Tremor (EE) Yes/No Yes/No Yes/No 1. Poor w/tremor 2. Poor without tre mor 3. Good without tremor 4. Excellent without tremor Initial Strokes (EE) Yes/No Yes/No Yes/No Start Location 1. Blunt 1. Top (12pm) 2. Feathery 2. Bottom(6pm) 3. Right (3pm) 4. Left (9pm) 5. Upper right 6. Upper left 7. Lower right 8. Lower left 117 Terminal Strokes (EE) Yes/No Yes/No Yes/No Start Location 1. Blunt 1. Top (12pm) 2. Feathery 2. Bottom(6pm) 3. Right (3pm) 4. Left (9pm) 5. Upper right 6. Uppe r left 7. Lower right 8. Lower left Embellishments (EE) Yes/No Yes/No Yes/No 1. Capital letters 2. Lower case letters 3. Beginning stroke 4. Ending stroke 5. Under signature 6. Combo Simplifica tion (EE) Yes/No Yes/No Yes/No 1. Capital letters 2. Lower case letters 3. Beginning stroke 4. Ending stroke 5. Under signature 6. Combo Abbreviations (EE) Yes/No Yes/No Yes/No 1. Eliminate letters 2. Letter combos that sacrifice form for speed 3. Mi ddle initial instead of full name. 4. Change number of letters in the abbreviation. Punctuation/Diacritics (EE) Yes/No Yes/No Yes/No Form Location 1. i - dot i - dot/j - dot 2. Period 1. Above stem 3. t - bar 2. Close to stem 4. Combo 3. Right of stem 4. Left of stem t - bar 1. Top of stem 2. 3/4 of stem 3. Middle - stem 4. Low - stem Legible signature (EE) Yes/No Yes/No Yes/No 1. All letters well formed and readable. 2. Most letters well formed and readable. 118 Illegible signature (EE) Yes/No Yes/No Yes/No 1. Letters are not well forme d and un - readable. 2. Only one or two letters are readable, the rest are loops and lines. 3. Writing consists of loops and lines only. Handedness (EE) Yes/No Yes/No Yes/No 1. Right hand 2. Left hand 3. Combo Writing movement (EE) Yes/No Yes/No Ye s/No 1. Garlanded 2. Arched 3. Angular 4. Indeterminable Spelling (O) Yes/No Yes/No Yes/No 1. Add letters 2. Delete letters 3. Change vowels (e.g., e to i) 4. Change consonants (e.g., s to c) 5. Combo Order of name (O) Yes/No Yes/No Yes/No 1. First, middle, last 2. First, last, middle 3. Last, first, middle 4. Last, middle, first 5. Middle, last, first 6. Middle first, last 7. First, last 8. Last, first 119 APPENDIX G : L ist of Handwriting V ariables List of Handwriting Variables 28 Abbreviations ( Abbreviations ): A shortened form of a word or title. Alignment (Align _Spec1 , Align _Spec2 ): The spatial organization of the writing pattern, its linear arrangement of words and intervening spaces , and their accommodation on the page. Connec ting Strokes (Connect Strokes ): The fusion of the terminal stroke of one lowercase cursive letter and the initial stroke of another. Embellishment (Embell ): Unnecessary o rnamentation or elaborative additions to the design of the letters or words. Handedne ss (Hand ): T he tendency of the writer to use either the right or the left hand more naturally than the other. Hiatuses ( Hiatuses ): A gap in a writing stroke of a letter formed when the writing instrument leaves the paper; an opening; an interruption in t he continuity of a line. Initial Stroke ( Instroke_Start , Instroke_Dir ): The first stroke of a letter or word. Lateral Expansion (Lat Expansion ): The horizontal dimension of writing produced by the width of letters and the space between letters and words. Legibility ( Legible Sig ): The ease with which a reader recognizes individual letter and character shapes. Letter Form (Letter Form_First , Letter Form_Last ): T he graphic form of a ny lowercase or uppercase letter of the alphabet. Line Quality ( Line Quality ) : The degree of regularity (i.e., smoothness or gradation) in the written stroke as may be judged from the consistency of its path in a prescribed direction. It varies from smooth and controlled to tremulou s and erratic. Order of Name ( Order ): The arrange ment of the first, middle , and last name. 28 Huber, R. A., & Headrick, A. M. (1999). Handwriting i dentification: Facts and fundamentals . Boca Raton, FL: CRC Press. 120 Pen Lifts ( Pen Lifts ): An interruption in a stroke due to the removal of the writing instrument from the paper. Pressure ( Pressure ): The amount of force exerted on the point of the writing instrument. It c an be de termined by the amount of indentation in the paper surface. Punctuation (Pun _i/j - dot Form , Pun_i/j - dot Loc , Pun_period ): The marks used in writing to separate sentences and their elements , and to clarify meaning (e.g., period, comma ). An i - dot ( and a peri od ) is a minute, roundish, solid mark. Size (Size _Absolute ): The absolute writing size is the distance between the top of the upper zone and the bottom of the lower zone. Slant ( Slant ): The angle of the axis of letters relative to the baseline of the wr iting. Spacing ( Spacing_Letter s): The distance between letters or words. Speed of Writing ( Speed ): The rate of line generation. Spelling (Spelling ): The way a word is spelled; the way letters are grouped to form a word. Style of A llograph (StyleofAllog raph) : A style (block capital, print script, or cursive form) of one of the 26 letters of the English alphabet. t - bar ( t - bar_Loc , t - bar_Form ): A stroke intersecting the stem or main portion of the lowercase or uppercase letter "T". Terminal Stroke ( Tstro ke_End , Tstroke_Dir ) : The last stroke of a word; endstroke. 121 APPENDIX H : Images Showing Intentional Handwriting Changes Figure 9. Examples of altered signatures in scenario 1 Genuine Disguised 122 Figure 10. Examples of altered signatures in scenario 2 Genuine Disguised 123 Figure 11. Examples of altered signatures in scenario 3 Genuine Disguised 124 Examples of altered signature s across scenarios Figure 12. Example #1 Genuine Disguised Scenario 1 Scenario 2 Scenario 3 125 Figure 13. Example #2 Genuine Disguised Scenario 1 Scenario 2 Scenario 3 126 Figure 14. Example #3 Genuine Disguised Scenario 1 Scenario 2 Scenario 3 127 Figure 15. Example #4 Genuine Disguised Scenario 1 Scenario 2 Scenario 3 128 Figure 16. Example #5 Genuine Disguised Scenario 1 Scenario 2 Scenario 3 129 REFERENCES 130 REFERENCES Abdi, H., & Williams, L. J. (2010). Newman - Keuls Test and Tukey Test. In N. Salkind (Ed.), Encyclopedia of Research Design . (1st ed., Vol. 2, pp. 897 902). Thousand Oaks, CA: Sage Publications. Agar, J. (2011). Recent arrests show fake documents are lucrativ e, sophisticated business . Retrieved February 13, 2013 from http://www.mlive.com/news/grand rapids/index.ssf/2011/01/recent_arrests_show_fake_docum.html Alford, E. F. (1970). Disguised handwriting: A statistical survey of how handwriting is most frequentl y disguised. Journal of Forensic Sciences, 15 (4), 476 488. American Bankers Association. (2013). 2013 Deposit account fraud survey report . Retrieved August 23, 2014 from http://www.aba.com/Products/Surveys/Pages/2013DepositAccount.aspx American Society of Crime Laboratory Directors. (2014) . Accredited Laboratory Index . Retrieved October 5, 2014 from http://www.ascld - lab.org/accredited - laboratory - index/ Asch, S. E. (1951). Effects of group pressure upon the modification and distortion of judgments. In H. Guetzkow (Ed.) Groups, leadership and men: Research in human relations . Oxford, England: Carnegie Press. Association of Certified Fraud Examiners (ACFE). (2014). Report to the nations on occupational fraud and abuse: 2014 global fraud study . Austin, T X: ACFE, Inc. Association of Financial Professionals. (2014). 2014 AFP payments fraud and control survey . Bethesda, MD: J. P. Morgan. ASTM International. (2007). Standard guide for examination of handwritten items . West Conshohocken, PA: ASTM Internatio nal. ASTM International. (2012). Standard guide for minimum training requirements for forensic document examiners . West Conshohocken, PA: ASTM International. Baker, J. N. (1955). The law of disputed and forged documents . The Mitchie Company: Charlottesvi lle, VA. Bandalos, D. L., & Finney, S. J. (2010). Factor analysis: Exploratory and confirmatory. In G. Hancock and R. Mueller (Eds), sciences (pp. 93 114). New York, NY: Routledge. Barclay, E., & Donnermeyer, J. F. (2011). Crime and security on agricultural operations. Security Journal, 24 (1), 1 18. 131 Barclay, P., Buckley, J., Brantingham, P. J., Brantingham, P. L., & Whinn - Yates, T. (1996). Preventing auto theft in suburban Vancouver commuter l ots: Effects of a bike patrol. In R. V. Clarke (Ed.), Preventing mass transit crime. Crime prevention studies (Vol. 6, pp. 133 161). Monsey: Criminal Justice. Bargh, J. A., & Chartrand, T. L. (2000). The mind in the middle: A practical guide to priming and automaticity research. In H. T. Reis and C. M. Judd (Eds.) Handbook of research methods in social and personality psychology . Cambridge, UK: Cambridge University Press. Bargh, J. A., Chen, M., & Burrows, L. (1996). Automaticity of social behavior: Dir ect effects of trait construct and stereotype activation on action. Journal of Personality and Social Psychology, 71 (2), 230 244. Barnett, C. (2000). The measurement of white - collar crime using Uniform Crime Reporting (UCR) Data . Washington, D.C.: Burea u of Justice Statistics. Bennett, R. R. (1991). Routine activities: A cross - national assessment of a criminological perspective. Social Forces, 70 (1), 147 163. Bennett, T. H., Holloway, K., & Farrington, D. P. (2006). Does neighborhood watch reduce crim e? A systematic review and meta - analysis. Journal of Experimental Criminology, 2 (4), 437 458. Bernasco, W., & Block, R. (2009). Where offenders choose to attack: A discrete choice model of robberies in Chicago. Criminology, 47 , 93 130. Bernasco, W., & Block, R. (2011). Robberies in Chicago: A block - level analysis of the influence of crime generators, crime attractors, and offender anchor points. Journal of Research in Crime and Delinquency, 48 , 33 57. Bernasco, W., & Luykx, F. (2003). Effects of attr activeness, opportunity , and accessibility to 83 burglars on residential burglary rates of urban neighborhoods, Criminology, 41 , 981 1002. Bird, C., Found, B., Ballantyne, K., & Rogers, D. K. (2010). Forensic handwriting examiners' opinions on the pr ocess of production of disguised and simulated signatures. Forensic Science International, 195 (1 3), 103 107. Bird, C; Found, B; & Rogers, D. (2010). Forensic handwriting examiners' skill in distinguishing between natural and disguised handwriting behavi ors. Journal of Forensic Sciences, 55 (5), 1291 1295. disguise behavior from handwritten text samples. Journal of Forensic Document Examination, 22 . 132 Boakye, K. A. (2009). The concept of capable guardianship: Exploring police views on tourist protection in Ghana. Crime Prevention and Community Safety, 11 (2), 104 123. Bossler, A. M., & Holt, T. J. (2009). On - line activities, guardianship, and malware infection: An examination of routine activities theory. International Journal of Cyber Criminology, 3 (1), 400 420. Brady, G. (2007). Crime lab struggles to keep up . Retrieved February 10, 2013 from http://www.newwest.net/city/article/crime_lab_struggles_to_keep _up/C101/L101/ Braga, A. A., Hureau, D. M., & Papachristos, A. V. (2011). The relevance of micro places to citywide robbery trends: A longitudinal analysis of robbery incidents at street corners and block faces in Boston. Journal of Research in Crime an d Delinquency, 48 , 7 32. Brantingham, P. J., & Brantingham, P. L. (1991). Environmental Criminology . Prospect Heights, IL: Waveland Press. Brantingham, P. L., & Brantingham, P. J. (2005). Situational crime prevention as a key component in embedded crim e prevention. Canadian Journal of Criminology and Criminal Justice, 47 (2), 271 292. Breetzke, G. D., & Cohn, E. G. (2013). Burglary in gated communities: An empirical analysis using routine activities theory. International Criminal Justice Review, 23 (1) , 56 74. Brown, B. B., & Bentley, D. L. (1993). Residential burglars judge risk: The role of territoriality. Journal of Environmental Psychology, 13 , 51 61. Brown, J. L. (2014). Friendly fraud spikes during security breaches. Retrieved October 16, 2014 from http://www.bizjournals.com/tampabay/news/2014/02/06/friendly - fraud - spikes - during - security.html?page=all Buckner, R. L., & Carroll, D. C. (2006). Self - projection and the brain. Trends in Cognitive Sciences, 11 (2), 49 57. Caligiuri, M. P., & Mohammed , L. A. (2012). The neuroscience of handwriting: Applications for forensic document examination . Boca Raton, FL: CRC Press. Carroll, J., & Weaver, F. (1986) . Shoplifters' perceptions of crime opportunities: A process tracing study. In, D. R. Cornish and R .V. Clarke (Eds.), The reasoning criminal: Rational choice perspectives on offending. New York, NY: Springer - Verlag. Chartrand, T. L., & Bargh, J. A. (1999). The chameleon effect: The perception - behavior link and social interaction. Journal of Personality and Social Psychology, 76 (6), 893 910. Clarke, R. V. (1980). Situational crime prevention: Theory and practice. British Journal of Criminology, 20 , 136 147. 133 Clarke, R. V. (1983). Situational crime prevention: Its theoretical basis and practical scope. Crime and Justice, 4 , 225 256. Clarke, R. V. (1992). Situational crime prevention: Successful case studies . Albany, NY: Harrow and Heston. Clarke, R.V. (1995). Situational crime prevention. In M. Tonry and D. P. Farrington (Eds.), Building a Safer Societ y: Strategic Approaches to Crime Prevention , Crime and Justice: A Review of Research, (Vol. 19, pp. 91 150). Chicago, IL: University of Chicago Press. Clarke, R.V. (1997). Introduction. In: R.V. Clarke (Ed.), Situational Crime Prevention: Successful Case Studies. Guilderland, NY: Harrow and Heston. Clarke, R. V. (2005). Seven Misconceptions of Situational Crime Prevention . In N. Tilley (Ed.), Handbook of Crime Prevention and Community Safety . Cullompton, Devon: Willan. Clarke, R. V., & Eck, J. E. (2009 ). Crime analysis for problem solvers: In 60 small steps . Retrieved on August 20, 2014 from http://www.cops.usdoj.gov/Publications/CrimeAnalysis60Steps.pdf Cohen, L. E., & Cantor, D. (1980). The determinants of larceny: An empirical and theoretical stu dy. Journal of Research in Crime and Delinquency, 17 (2), 140 159. Cohen, L. E., & Felson, M. (1979). Social change and crime rate trends: A routine activity approach. American Sociological Review, 44 , 588 608. Cohen, L. E., Kluegel, J. R., & Land, K. C . (1981). Social inequality and predatory criminal victimization: An exposition and test of a formal theory. American Sociological Review, 46 (5), 505 524. Community Oriented Policing Services. (2011). The impact of economic downturn on American police a gencies . Retrieved on February 20, 2013 from http://www.cops.usdoj.gov/files/RIC/Publications/e101113406_Economic%20Impact.pdf Compeau, D., Marcolin, B., Kelley, H., & Higgins, C. (2012). Generalizability of information systems research using student sub jects: A reflection on our practices and recommendations for future research. Information Systems Research, 23 (4), 1093 1109. Conway, J.V.P. (1959). Evidential Documents . Illinois: Charles C Thomas. Cornish, D. B., & Clarke, R. V. (1986). The reasoning criminal . New York, NY: Springer - Verlag. Cornish, D. B., & Clarke, R. V. (1987). Understanding crime displacement: An application of rational choice theory. Criminology, 254 , 933 947. 134 Cornish, D.B., & Clarke, R.V. (2003). Opportunities, precipitators , and criminal decisions: Cornish (Eds.), Theory for Practice in Situational Crime Prevention , Crime Prevention Studies, (Vol. 16). Monsey, NY: Criminal Justice Press. Coupe, T., & Blake, L. (2006). Daylight and darkness targeting strategies and the risks of being seen at residential burglaries. Criminology, 44 , 431 464. Coupe, T., & Blake, L. (2011). The effects of target characteristics on the sighting and arrest of offenders at burglary emergencies. Security Journal, 24 (2), 157 178. Coupe, T., & Fox, B. H. (2015). A risky business: How do access, exposure , and guardians affect the chances of non - residential burglars being seen? Security Journal, 28 (1), 71 92. Cre swell, J. W. (2007). Qualitative inquiry & research design: Choosing among five approaches. Thousand Oaks, CA: Sage Publications, Inc. Cullen, F. T., & Agnew, R. (2006). Criminological theory: Past to present . Los Angeles, CA: Roxbury Publishing Company . CyberSource. (2012). 2012 Online Fraud Report: Online Payment Fraud Trends, Merchant Practices , and Benchmarks . San Francisco, CA: CyberSource Corporation. D'Alessio, S. J., Eitle, D., & Stolzenberg, L. (2012). Unemployment, guardianship , and weekday residential burglary. Justice Quarterly, 29 (6), 919 932. Darley, J. M., & Batson, D. C. (1973). From Jerusalem to Jericho: A study of situational and dispositional variables in helping behavior. Journal of Personality and Social Psychology, 27 (1), 100 10 8. Darley, J. M., & Latane, B. (1968). Bystander intervention in emergencies: Diffusion of responsibility. Journal of Personality and Social Psychology, 8 (4), 377 383. Department of Justice (DOJ). (2010). The Department of Justice's Efforts to Combat Ide ntity Theft . Retrieved February 20, 2013 from http://www.justice.gov/oig/reports/plus/a1021.pdf Deutsch, M., & Krauss, R. M. (1962). Studies of interpersonal bargaining. The Journal of Conflict Resolution, 6 (1), 52 76. Dyer, A.G., Found, B. & Rogers, D . (2008). An insight into forensic document examiner expertise for discriminating between forged and disguised signatures. Journal of Forensic Sciences , 53 (5), 1154 1159. 135 Eck, J. E. (1994). Drug Markets and Drug Places: A Case - Control Study of the Spatial Structure of Illicit Drug Dealing. Unpublished Ph.D. dissertation. College Park, MD: University of Maryland. Eck, J. E. (2003). Police problems: The complexity of problem theory, research , and evaluation. In J. Knutsson (Ed.), Problem - Oriented Policin g: From Innovation to Mainstream, Crime Preventions Studies, (Vol. 15, pp. 79 113) . Monsey, NY: Criminal Justice Press. Eck, J. E., & Weisburd, D. (1995). Crime places in crime theory. In J. E. Eck and D. Weisburd (Eds), Crime and place, Crime Preventi on Studies, (Vol. 4, pp. 1 34). Monsey, NY: Willow Tree Press. Electronic Signatures in Global and National Commerce Act of 2000, Pub. L. No. 106 229, 114 Stat. 464 (2000). Ellen, D. (2006). Scientific examination of documents: Methods and techniques . Boca Raton, FL: CRC Press. Fabrigar, L. R., Wegener, D. T., MacCallum, R. C., & Strahan, E. J. (1999).Evaluating the use of exploratory factor analysis in psychological research. Psychological Methods, 4 (3), 272 299. Farrington, D. P., Bowen, S., Buc kle, A., Burns - Howell, T., Burrows, J., & Speed, M. (1993). Experiment on the prevention of shoplifting. In R. V. Clarke (Ed.), Crime Prevention Studies , (Vol. 1, pp. 93 119). Monsey, NY: Willow Tree Press. Federal Bureau of Investigation (FBI). (2008). T he Federal Bureau of Investigation financial institution fraud and failure report: Fiscal years 2006 and 2007. Retrieved February 14, 2013 from http://www.fbi.gov/stats - services/publications/fiff_06 - 07/fiff_06 - 07 Federal Bureau of Investigation (FBI). (20 13). Crime in the United States, 2012 . Retrieved June 10, 2014, from http://www.fbi.gov/about - us/cjis/ucr/crime - in - the - u.s/2012/crime - in - the - u.s. - 2012/offenses - known - to - law - enforcement/clearances Federal Bureau of Investigation (FBI). (2014). Uniform Cr ime Reports: UCR Publications . Retrieved June 10, 2014, from http://www.fbi.gov/about - us/cjis/ucr/ucr - publications#Crime Federal Reserve Bank of Boston. (2012). 2012 payments fraud survey summary of results: First district. Retrieved February 13, 2013 f rom http://www.bostonfed.org/bankinfo/payment strategies Federal Reserve Bank of Dallas. (2012). 2012 payments fraud survey summary of results: Eleventh district. Retrieved February 13, 2013 http://www.dallasfed.org/assets/documents/banking/firm/fi/fraud _survey.pdf 136 Federal Reserve Bank of Minneapolis. (2012). 2012 payments fraud survey summary of results: Ninth district. Retrieved February 13, 2013 from http://www.minneapolisfed.org/about/whatwedo/payments/2012_Payments_Fraud_Surv y_Sumary.pd Federal Tr ade Commission (FTC). (2014a). FTC releases top 10 complaint categories for 2012: Identity theft tops list for 13th consecutive year in report of national consumer complaints. Retrieved October 13, 2014 from http://www.ftc.gov/news - events/press - releases /2013/02/ftc - releases - top - 10 - complaint - categories - 2012 Federal Trade Commission (FTC). (2014b). Consumer Sentinel Network: Data book for January - December 2012. Retrieved October 13, 2014 from http://www.ftc.gov/sites/default/files/documents/reports/cons umer - sentinel - network - data - book - january/sentinel - cy2012.pdf Felson, M. (1986). Routine activities, social controls, rational decisions, and criminal outcomes. In D. Cornish and R.V. Clarke (Eds.), The Reasoning Criminal . New York, NY: Springer - Verlag . Felson, M. (1987). Routine activities and crime prevention in the developing metropolis. Criminology, 25 (4), 911 931. Felson, M. (1995). Those who discourage crime, In J. E. Eck & D. Weisburd (Eds.), Crime and place: Crime prevention studies , (Vol. 4, pp. 53 66). Monsey: Criminal Justice. Felson, M. (2006). Crime and nature . Thousand Oaks, CA: Sage Publications. Felson, M., & Boba, R. (2010). Crime and everyday life . Thousand Oaks, CA: Sage Publications. Felson, M., & Clarke, R. V. (1998). Oppor tunity makes the thief: Practical theory for crime prevention. Police Research Series, Paper No. 98. London, UK: Home Office. Felson, M., & Cohen, L. E. (1980). Human ecology and crime: a routine activity approach. Human Ecology, 8 (4), 389 406. Fenoff, R., (2013). Forgery. In L. M. Salinger (Ed.), Encyclopedia of White - Collar and Corporate Crime . (2nd ed., Vol. 6, pp. 360 364). Thousand Oaks, CA: Sage Publications. Fenoff, R., & Spink, J . (2013). Counterfeiting. In J. Albanese (Eds .). Encyclopedia of Criminology and Criminal Justice . New Jersey: Wiley - Blackwell. Festinger, L. & Carlsmith, J. M. (1959). Cognitive consequences of forced compliance. The Journal of Abnormal and Social Psychology, 58 (2), 203 210. 137 Field, A. (2000). Discovering Statistics using SPSS for Windows . Thousand Oaks , CA : Sage publications. Financial Crimes Enforcement Network. (2010). The SAR Activity Review: By the numbers . Retrieved on February 13, 2014 from http://www.fincen.gov/news_room/rp/files/sar_by_numb_14.pdf Finkl ea, K. M. (2012). Organized retail crime . Washington, D.C.: Congressional Research Service. Finney, S. J., & DiStefano, C. (2013). Non - normal and categorical data in structural equation modeling. In G. R. Hancock and R. O. Mueller (Eds), Structural equ ation modeling: A second course , (2nd ed., pp. 269 314). Charlotte, NC: Information Age Publishing, Inc. Flere, S., & Lavric, M. (2008). On the validity of cross - cultural social studies using student samples. Field Methods, 20 (4), 399 412. Foster, N. B. (2012). Stolen and forged checks on the rise. Retrieved February 13, 2013 from http://www.unionleader.com/article/20120201/NEWS03/702019961 Found, B., & Rogers, D. (2008). The probative character of forensic handwriting examiners' identification and eli mination opinions on questioned signatures. Forensic Science International, 178 , 54 60. Garcia - Retamero, R., & Dhami, M. K. (2009). Take - the - best in expert - novice decision strategies for residential burglary. Psychonomic Bulletin & Review, 16 (1), 163 16 9. Garofalo, J., & Clark, D. (1992). Guardianship and residential burglary. Justice Quarterly, 9 (3), 443 463. Gibbs, C., Gore, M. L., McGarrell, E. F., & Rivers III, L. (2010). Introducing conservation criminology. British Journal of Criminology, 50 , 12 4 144. Gilbert, D. T., Tafarodi, R. W., & Malone, P. S. (1993). You can't not believe everything you read. The Journal of Abnormal and Social Psychology, 65 (2), 221 233. Giles, L. (2013). PG police work to end internet crimes against children. Retrieved February 18, 2013 from http://www.heraldextra.com/news/local/crime - and - courts/pg - police - work - to end - internet - crimes - against - children/article_7b3fb9d9 - 52cf - 5868 - 8f64 55 81e2714c68.html Gravetter, F. J., & Forzano, L. B. (2012). Research methods for the behavioral sciences . Belmont, CA: Wadsworth, Cengage Learning. Griggs, B. (2012). When is social - media use a crime? Retrieved February 18, 2013 from http://www.cnn.com/20 12/12/18/tech/social - media/newtown - social - media - crime 138 Guy, S. (2014). New study says Chicagoans among biggest victims of data fraud . Retrieved October 13, 2014 from http://www.suntimes.com/news/metro/28694030 - 418/new - study - says - chicagoans - among - biggest - v ictims - of - data - fraud.html#.VD2cwRZCxQI Hamari, A. (2010). Identity theft investigations stretch police resources: Cases are time consuming to unravel. Retrieved on February 20, 2013 from http://www.brookfieldnow.com/news/93324139.html Harris, J. J. (1953 ). Disguised handwriting. Journal of Criminal Law, Criminology & Police Science, 43 , 685 689. Harrison, W.R. (1958). Suspect Documents: Their Scientific Examination . New York: Praeger. Harrison, D., Burkes, T. M., & Seiger, D. P. (2009). Handwriting exam ination: Meeting the challenges of science and the law. Forensic Science Communication, 11 (4), 1 16. Herkt, A. (1986). Signature disguise or signature forgery? Journal of the Forensic Science Society, 26 , 257 266. Hilton, O. (1982). Scientific Examinati on of Questioned Documents . New York, NY: Elsevier Science Publishing Co. Hirschi, T., & Gottfredson, M. (1983). Age and the explanation of crime. American Journal of Sociology, 89 , 552 584. Hollis, M. E., Fejes, L., Fenoff, R., & Wilson, J. M. ( 2014 ). Routine activities and product counterfeiting: A research note. International Journal of Comparative and Applied Criminal Justice . 1 16. Hollis, M. E., Felson, M., & Welsh, B. C. (2013). The capable guardian in routine activities theory: A theoretical and conceptual reappraisal. Crime Prevention and Community Safety, 15 , 65 79. Hollis, M. E., & Wilson, J. (2014). Who are the guardians in product counterfeiting? A theoretical application of routine activities theory. Crime Prevention and Community Sa fety, 16 , 169 188. Hollis - Peel, M. E., Reynald, D. M., van Bavel, M., Elffers, H., & Welsh, B. C. (2011). Guardianship for crime prevention: A critical review of the literature. Crime, Law and Social Change, 56 , 53 70. Hollis - Peel, M. E., Reynald, D. M., & Welsh, B. C. (2012). Guardianship and crime: An international comparative study of guardianship in action. Crime, Law and Social Change, 58 , 1 14. Hollis - Peel, M. E., & Welsh, B. C. (2014). What makes a guardian capable? A test of guardianship in action. Security Journal, 27 , 320 337. 139 Howard, H. M. (2005). The negligent enablement of imposter fraud: A common - sense common law claim. Duke Law Journal, 54 , 1263 1294. Huber, R. A., & Headrick, A. M. (1999). Handwriting identification: Facts and funda mentals . Boca Raton, FL: CRC Press. Iannacci, J., & Morris, R. (2000). Access device fraud and related financial crimes . Boca Raton, FL: CRC Press. IC3. (2011). IC3 2011 internet crime report released: More than 300,000 complaints of online criminal act ivity reported in 2011. Retrieved February 18, 2013 from http://www.ic3.gov/media/2012/120511.aspx IC3. (2013). 2011 Internet Crime Report. Retrieved February 13, 2013 from http://www.ic3.gov/media/annualreports.aspx Information Security Media Group (IS MG). (2014). The faces of fraud: The impact of retail breaches . Princeton, NJ: ISMG Corp. Jacobs, B. A. (2010). Serendipity in robbery target selection. British Journal of Criminology, 50 , 514 529. Jacobs, B. A., & Wright, R. (2010). Bounded rationalit y, retaliation, and the spread of urban violence. Journal of Interpersonal Violence, 25 (10), 1739 1766. Javelin Strategy & Research. (2012). 2012 2017: Retail point of sale forecast: Cash is no longer king. Retrieved February 13, 2013 from https://www.j avelinstrategy.com/brochure/251 Jiang, S., Lambert, E. G., and Wang, J. (2007). Capital punishment views in China and the United States: A preliminary study among college students. International Journal of Offender Therapy and Comparative Criminology, 51 (1), 84 97. Kaiser, H. F. (1974). An index of factorial simplicity. Psychometrika, 39 (1), 31 36. Keckler, J. A. (1997). Felonious disguise: A study of the most commonly used modes of disguise adopted by convicted felons. International Journal of Forensi c Document Examiners, 3 (2), 154 158. Kenney, D. J. (1986). Crime on the subways: Measuring the effectiveness of the guardian angels. Justice Quarterly, 3 , 481 496. Kline, P. (1994). An easy guide to factor analysis . New York, NY: Routledge Klynveld Pe at Marwick Goerdeler. (2004). 2003 fraud survey . U.S.A.: KPMG Forensic. 140 Knutsson, J. (2006). What is there to gain? A case study in implementing without self - interest. In J. Knutsson & R. V. Clarke (Eds.), Putting theory to work: Implementing situationa l crime prevention and problem - oriented policing, crime prevention studies (Vol. 20, 89 110). Monsey: Criminal Justice. Knutsson, J., & Clarke, R. V. (2006). Putting theory to work: Implementing situational crime prevention and problem - oriented policin g, crime prevention studies (Vol. 20). Monsey: Criminal Justice. Knutsson, J., & Kuhlhorn, E. (1992). Macro measures against crime. The example of check forgeries. In R. V. Clarke (Ed), Situational crime prevention: Successful case studies . Albany, NY: Harrow and Heston. evidence from multiple - opportunity searches. Law and Human Behavior, 30 (4), 455 468. Konstantinidis, S. (1987). Disguised handwriting. Jou rnal of the Forensic Science Society, 27 , 383 392. Kroneberg, C., Heintze, I., & Mehlkop, G. (2010). The interplay of moral norms and instrumental incentives in crime causation. Criminology, 48 (1), 259 294. Lacoste, J., & Tremblay, P. (2003). Crime inno vation: A script analysis of patterns in check forgery. Crime Prevention Studies, 16 , 171 198. Leung, S. C., Chung, M. W. L., Tsui, C. K., & Cheung, Y. L. (1988). A comparative approach to the examination of Chinese handwriting Part 3: Disguise. Journal of the Forensic Science Society, 28 , 149 165. Levi, M. (1998). Organising plastic fraud: Enterprise criminals and the sidestepping of fraud prevention. The Howard Journal of Criminal Justice, 37 (4), 423 438. Levi, M., Bissell, P., & Richardson, T. (1 991). The prevention of cheque and credit card fraud. Crime Prevention Unit Paper 26. London: Home Office. Levinson, J. (2001). Questioned documents: A lawyer's handbook . San Diego, CA: Academic Press. LexisNexis. (2013). True Cost of Fraud Study: Merc hants Struggle Against an Onslaught of High - Cost Identity Fraud and Online Fraud . New York, NY: LexisNexis, Inc. Liesik, G. (2010). financial information. Retrieved February 20, 201 3 from http://www.deseretnews.com/article/700009845/Card - fraud - turns - up - at - pumps - in - Sandy Police - se e - increase - of - skimming - financial - information.html 141 Luzer, D. (2010). When Cops Go to College . Retrieved February 20, 2013 from http://www.washingtonmonthly.com/college_guide/blog/when_cops_go_to_college.php Lynch, J. P. (1987). Routine activity and vict imization at work. Journal of Quantitative Criminology, 3 (4), 283 300. Lynch, J. P., & Cantor, D. (1992). Ecological and behavioral influences on property victimization at home: implications for opportunity theory. Journal of Research in Crime and Delinq uency, 29 (3), 335 362. Lynch , M. J., & Stretesky, P. B. (2003). The meaning of green: Contrasting criminological perspectives, Theoretical Criminology , 7 , 217 238. Madensen, T. D., & Eck, J. E. (2008). Violence in bars: Exploring the impact of place ma nager decision - making. Crime Prevention and Community Safety, 10 , 111 125. Masuda, B. (1996). An alternative approach to the credit card fraud problem. Security Journal, 7 (1), 15 21. Mativat, F., & Tremblay, P. (1997). Counterfeiting credit cards. Brit ish Journal of Criminology, 37 (2), 165 183. Michel, L. (1978). Disguised signatures. Journal of the Forensic Science Society, 18 , 25 29. Miethe, T. D., Stafford, M. C., & Sloane, D. (1990). Lifestyle changes and risks of criminal victimization. Journal of Quantitative Criminology, 6 (4), 357 376. Milgram, S. (1963). Behavioral study of obedience. The Journal of Abnormal and Social Psychology, 67 (4), 371 378. Miller, L. S. (1984). Bias among forensic document examiners: A need for procedural changes. J ournal of Police Science and Administration, 12 (4), 407 411. Miller, L. S. (1987). Procedural bias in forensic examinations of human hair. Law and Human Behavior, 11 , 157. Mohammed, L. A. (1993). Signature disguise in Trinidad and Tobago. Journal of the Forensic Science Society, 33 , 21 24. Musen, K. & Zimbardo, P. G. (1991). Quiet rage: The Stanford prison study. Video recording. Stanford, CA: Psychology Dept., Stanford University. National Check Fraud Center. (2013). Information about check fraud . Ret rieved February 20, 2013 from http://www.stopcheckfraud.com/ 142 National Consumers League. (2010). . Retrieved February 20, 2013 from http://www.nclnet.org/images/PDF/2009_top_scams.pdf National Institute of Justice. (2014). All Programs & Practices . Retrieved October 5, 2014 from http://www.crimesolutions.gov/programs.aspx National Research Council, Na tional Academy of Sciences (NRC/NAS). (2009). Strengthening forensic science in the United States: A path forward . Washington, D.C.: The National Academies Press. National White Collar Crime Center. (2008). Credit card fraud: White paper . Retrieved Febr uary 13, 2013 from http://www.nw3c.org/services/research/whitepapers/page/3/ National White Collar Crime Center. (2009). Check fraud: White paper. Retrieved February 13, 2013 from http://www.nw3c.org/services/research/whitepapers/page/4/ NBC News. (2013 ). ID theft on the rise again: 12.6 million victims in 2012, study shows. Retrieved on February 28, 2013 from http://www.wpsdlocal6.com/news/local/ID - theft on - the - rise - again - 126 - million - victims - in - 2012 - study - shows - 191998091.html New York Post. (2010). Check fraud victims defenseless . Retrieved February 11, 2011 from http://www.nypost.com/p/news/business/check_fraud_victims_defenseless_OWBipWf w7Gafxmt8Zh0tI Newman, G. R. (2003). Check and card fraud . Retrieved December 18, 2014 from http://www.popcenter.org/problems/credit_card_fraud/print/ Nisbett, R. E., and Cohen, D. (1996). Culture of honor: The psychology of violence in the South. Boulder, CO: Westview Press. O'Block, R. L. (1991). Metho ds of disguise in handwriting. IAQDE Journal, 4 (1), 5 18. Police structure of the United States . Retrieved February 10, 2013 from http://www.drtomoconnor.com/megapolice.htm le against identity theft . Retrieved on February 7 2013 from http://www.identitytheftcouncil.org/12 - reasons - why - were - losing the - battle - against - identity - theft Osborn, A.S. (1929 ). Questioned Documents . 2nd ed., Chicago, IL: Nelson - Hall Co. Palmer, E. J., Holmes, A., & Rollins, C. R. (2002). Investigating burglars' decisions: Factors influencing target choice, method of entry, reasons for offending, repeat victimization of a pr operty and victim awareness. Security Journal, 15 , 7 8. 143 Payne, B. K., & Chappell, A. (2008). Using student samples in criminological research. Journal of Criminal Justice Education, 19 , 175 192. Peachey, P. (2012). Complaints of crimes involving Facebook and Twitter increases 780% in just four years. Retrieved February 18, 2013 from http://www.independent.co.uk/news/uk/crime/complaints - of - crimes - involving - fac ebook and - twitter - increases - 780 - in - just - four - years - 8431881.html Pennell, S., Curtis, C., Henderson, J., et al. (1989). Guardian angels: A unique approach to crime prevention. Crime and Delinquency, 35 (3), 378 400. Peterson, R. A. (2001). On the use of college students in social science research: Insights from a second - order meta - analysis. Journal of Consumer Research, 28 , 450 461. Phillips, C. (1999). A review of CCTV evaluations: Crime reduction effects and attitudes towards its use. In K. Painter and N. Tilley (Eds.), Crime Prevention Studies , (Vol. 10, pp.123 155). Monsey, NY: Willow Tree Press. Pires, S., & Clarke, R.V. (2012). Are Parrots CRAVED? An Analysis of Parrot Poaching in Mexico. Journal of Research in Crime and Delinquency, 49 (1), 122 14 6. Pop Center. (2015). Twenty five techniques of situational prevention . Retrieved on July, 22 2015 from http://www.popcenter.org/25techniques/ Preacher, K. J., Zhang, G., Kim, C., & Mels, G. (2013). Choosing the optimal number of factors in explorato ry factor analysis: A model selection perspective. Multivariate Behavioral Research, 48 , 28 56. Privacy Rights Clearinghouse. (2014). Chronology of Data Breaches: Security Breaches 2005 Present . Retrieved August 23, 2014 from https://www.privacyrights .org/data - breach Quattrone, G. A., & Tversky, A. (1984). Causal versus diagnostic contingencies: On self deception and the voter's illusion. Journal of Personality and Social Psychology, 46 (2), 237 248. Regent, J. (1977). Changing slant: Is it the only c hange? Journal of Forensic Science, 22 (1), 216 221. Reynald, D. M. (2009). Guardianship in action: Developing a new tool for measurement. Crime Prevention and Community Safety: An International Journal, 11 , 1 20. Reynald, D. M. (2010). Guardians on guar dianship: Factors affecting the willingness to supervise, the ability to detect potential offenders, and the willingness to intervene. Journal of Research in Crime and Delinquency, 47 (3), 358 390. 144 Reynald, D. M. (2011a). Factors associated with the guar dianship of places: Assessing the relative importance of the spatio - physical and sociodemographic contexts in generating opportunities for capable guardianship. Journal of Research in Crime and Delinquency, 48 (1), 110 142. Reynald, D. M. (2011b). Guardin g against crime: Measuring guardianship within routine activity theory. Burlington, VT: Ashgate Publishing Company. Reynald, D. M. (2011c). Translating CPTED into crime preventive action: A critical examination of CPTED as a tool for active guardianship. European Journal on Crime Policy and Research, 17 , 69 81. Reyns, B. W. (2013). Online routines and identity theft victimization: Further expanding routine activity theory beyond direct - contact offenses. Journal of Research in Crime and Delinquency, 50 (2), 216 238. Rice, K. J., & Smith, W. R. (2002). Socioecological models of automotive theft: Integrating routine activity and social disorganization approaches. Journal of Research in Crime and Delinquency, 39 (3), 304 336. Robertson, E. W. (1991). Fun damentals of document examination . Chicago, IL: Nelson - Hall Inc. Ross, L., Greene, D., & House, P. (1977). The false consensus effect: An egocentric bias in social perception and attribution processes. Journal of Experimental Social Psychology, 13 (3), 27 9 301. Rossmo, D. K. (2013). Failures in criminal investigation. The Police Chief , February 2013. Sampson, R., Eck, J. E., & Dunham, J. (2010). Super controllers and crime prevention: A routine activity explanation of crime prevention success and failur e. Security Journal, 23 (1), 37 51. Scientific Working Group for Forensic Document Examination (SWGDOC). (2013). Standard for Examination of Handwritten Items. Retrieved March15, 2015 from http://www.swgdoc.org/images/documents/standards/SWGDOC%20Standa rd%20for%2 0Examination%20of%20Handwritten%20Items.pdf Seaman - Kelly, J. (1999). The examination of disguised numbers. Journal of Forensic Science, 44 (5), 1027 1028. Sears, D. O. (1986). College sophomores in the laboratory: Influences of a narrow data b ase on social psychology's view of human nature. Journal of Personality and Social Psychology, 51 (3), 515 530. Shadish, W. R., Cook, T. D., & Campbell, D. T. (2002). Experimental and quasi - experimental designs for generalized causal inference . Belmont, CA: Wadsworth. 145 Sherman, L. W., Farrington, D. P., Welsh, B. C., & MacKenzie, D. L. (2002). Evidence - based crime prevention . New York, NY: Taylor & Francis Group. Sherman, L. W., Gottfredson, D., Mackenzie, D., Eck, J., Reuter, P., & Bushway, S. D. (1998 ). Preventing crime: What works, what doesn't, what's promising . United States Department of Justice. Office of Justice Programs. National Institute of Justice (Research in Brief). Washington, DC: National Institute of Justice. Shreiner, K. (2013). Pr eventing corporate check fraud. Retrieved February 13, 2013 from http://www.atlcapbank.com/pages/knowledge - center/preventing - corporate - check fraud.php Smith, R. G. (1997). Plastic card fraud. Trends and Issues. Crime and Criminal Justice, No. 71. Canberr a, Australia: Australian Institute of Criminology. Snyder, M., Tanke, E. D., & Berscheid, E. (1977). Social perception and interpersonal behavior: On the self - fulfilling nature of social stereotypes. Journal of Experimental Social Psychology, 35 (9), 656 666. Sparacello, M. (2012). Undocumented foreigner accused of selling forged tax documents. Retrieved February 13, 2013 from http://www.nola.com/crime/index.ssf/2012/03/undocumented_foreigner_accused.html Stahura, J. M., & Sloan III, J. J. (1988). Urban stratification of places, routine activities and suburban crime rates. Social Forces, 66 (4), 1102 1118. Stasser, G., & Titus, W. (1985). Pooling of unshared information in group decision making: Biased information sampling during discussion. Journal of Experimental Social Psychology, 48 (6), 1467 1478. Tabachnick, B. G., & Fidell, L. S. (2001). Multivariate statistics . Needham Heights, MA: Allyn & Bacon . Tajfel, H., Billig, M. G., Bundy, R. P., & Flament, C. (1971). Social categorization and intergroup behaviour. European Journal of Social Psychology, 1 (2), 149 178. Tewksbury, R., & Mustaine, E. E. (2003). College students' lifestyles and self - protective behaviors: Further considerations of the guardianship concept in routine activity theory. Crimina l Justice and Behavior, 30 (3), 302 327. Tilley, N. (2009). Crime Prevention . Cullompton, UK: Willan. Tilley, N., Tseloni, A., & Farrell, G. (2011). Income disparities of burglary risk. British Journal of Criminology, 51 , 296 313. 146 Tilley, N., & Webb, J . (1994). Burglary reduction: Findings from safer cities scheme . Crime Prevention Unit Paper 51. London: Home Office. Tillyer, M. S., & Eck, J. E. (2011). Getting a handle on crime: A further extension of routine activities theory. Security Journal, 24 ( 2), 179 193. Tillyer, M. S., & Kennedy, D. M. (2008). Locating focused deterrence approaches within a situational crime prevention framework. Crime Prevention and Community Safety, 10 , 75 84. Tremblay, P. (1986). Designing crime. British Journal of Cri minology, 26 (3), 234 253. Truman, J. L., & Langton, L. (2014). Criminal victimization, 2013 . Washington, D.C.: Bureau of Justice Statistics. Tseloni, A., Wittebrood, K., Farrell, G., & Pease, K. (2004). Burglary victimization in England and Wales, the United States and the Netherlands: A cross - national comparative test of routine activities and lifestyle theories. British Journal of Criminology, 44 , 66 91. Tsui, D. C. K. (1997). Disguised Chinese handwriting by Chinese - Canadian writers. Canadian Soci ety Forensic Science Journal, 30 (3), 137 150. Tukey, J. (1949). Comparing Individual Means in the Analysis of Variance . Biometrics, 5 (2), 99 114. Tuohy, L. (2013). NH convict gets over 3 years in count erfeit plot. Retrieved February 13, 2013 from http://www.boston.com/news/local/new - hampshire/2013/01/31/convict - gets - over years - counterfeit - plot/4R80yKffC2VYuqFg2IX3uL/story.html Uittenbogaard, A. C. (2014). Assessing guardianship opportunities at undergr ound stations. Security Journal, 27 , 147 163. U.S. Department of Labor. (2012). Occupational Outlook Handbook, 2012 2013 Edition, Police and Detectives . Bureau of Labor Statistics. Retrieved February 20, 2013 from http://www.bls.gov/ooh/protective - servic e/police - and - detectives.htm U.S. Government Accountability Office (GAO). (2003). Counterfeit documents used to enter the United States from certain western hemisphere countries not detected . (Publication No. GAO 03 713T). Retrieved from GAO Reports Main P age via GPO Access database: http://www.gao.gov/products/GAO - 03 - 713T U.S. Government Accountability Office (GAO). (2008). Summary of covert tests and security assessments for the Senate Committee on Finance, 2003 - 2007 . (Publication No. GAO 08 757). Retr ieved from GAO Reports Main Page via GPO Access database: http://www.gao.gov/products/GAO - 08 - 757 147 Vogt, W. P., & Johnson, R. B. (2011). Dictionary of statistics & methodology: A nontechnical guide for the social sciences . Thousand Oaks, CA: SAGE Publicat ions, Inc. Walker, S., & Katz, C. M. (2005). The police in America . New York, NY: McGraw - Hill. Warchol, G. (2011). South African conservation crime and routine activities theory: A causal nexus? South African Journal of Criminology, 24 (2), 1 16. Webb, B. (1996). Preventing plastic card fraud in the UK. Security Journal, 7 (1), 23. Wegner, D. M., Schneider, D. J., Carter, S. R., & White, T. L. (1987). Paradoxical effects of thought suppression. Journal of Experimental Social Psychology, 53 (1), 5 13. We lsh, B. C., & Farrington, D. P. (2009). Public area CCTV and crime prevention: An updated systematic review and meta - analysis. Justice Quarterly, 26 (4), 716 745. Welsh, B. C., Mudge, M. E., & Farrington, D. P. (2010). Reconceptualizing public area survei llance and crime prevention: Security guards, place managers and defensible space. Security Journal, 23 (4), 299 319. White, M. C. (2012). Study: 10,000 identity theft rings in U.S. Retrieved February 14, 2013 from http://business.time.com/2012/11/20/st udy - 10000 - identity - theft - rings - in - u - s/ Whitt, T. (2008). You may be paying $400 to $600 a year to offset shoplifting costs. Retrieved February 14, 2013 from http://www.heraldtribune.com/article/20080310/BUSINESS/803100640 Wiecko, F. M. (2010). Research note: Assessing the validity of college samples: Are students really that different? Journal of Criminal Justice, 38 , 1186 1190. Wilcox, P., Madensen, T. D., & Tillyer, M. S. (2007). Guardianship in context: Implications for burglary victimization risk an d prevention. Criminology, 45 , 771 803. Willard, V. L. (1988). A study in hand printing. Journal of Forensic Document Examination, 2 , 1 12. Williams, M. M., Fletcher, R. B., and Ronan, K. R. (2007). Investigating the theoretical construct and invariance of the Self - Control Scale using confirmatory factor analysis. Journal of Criminal Justice, 35 (2), 205 218. World Bank. (2014). GDP Ranking. Retrieved August 23, 2014 from http://data.worldbank.org/country/united - states Wortley, R., & McFarlane, M. (201 1). The role of territoriality in crime prevention: A field experiment. Security Journal, 24 (2), 149 156. 148 Wright, R., Logie, R. H., & Decker, S. H. (1995). Criminal expertise and offender decision making: An experimental study of the target selection pr ocess in residential burglary. Journal of Research in Crime and Delinquency, 32 (1), 39 53. Wright, R. T., & Decker, S. H. (1996). Burglars on the job: Streetlife and residential break - ins . Boston, MA: Northeastern University Press. Wright, R. T., & Deck er, S. H. (1997). Armed robbers in action: Stickups and street culture . Lebanon, NH: Northeastern University Press. Yong, A. G., & Pearce, S. (2013). A beginner's guide to factor analysis: Focusing on exploratory factor analysis. Tutorials in Quantitati ve Methods for Psychology, 9 (2), 79 94.