Abstract
The financial exploitation of older adults has garnered the attention of society as well as state and federal governments in a way that elder abuse has never been able to achieve. It is frequently asserted that financial exploitation deserves this attention in part because it is the most prevalent form of elder abuse. This article systematically reviews the measurement of financial exploitation in comparison with other forms of elder abuse and concludes that its measurement is considerably more variable than other forms of abuse. Consequently, improvements in the measurement of financial exploitation are warranted.
Elder abuse is frequently defined as a single or repeated act, or lack of appropriate action, occurring within any relationship where there is an expectation of trust which causes harm or distress to an older person and typically encompasses five types of abuse: physical abuse, caregiver neglect (including abandonment), financial exploitation, psychological abuse, and sexual abuse (World Health Organization [WHO], 2002). Nationally representative studies find that overall, one in approximately 11 older adults in the United States experience some type of elder abuse in a given year, although prevalence varies by the type of abuse involved: financial exploitation (5.2%), caregiver neglect (5.1%), psychological abuse (4.6%), physical abuse (1.6%), and sexual abuse (<1%; Acierno et al., 2010).
The conceptualization of elder abuse has shifted over time, with some suggesting that the criminalization of elder abuse is responsible for the increased attention the social problem has received (Jackson, 2016). Criminalization may have helped the field gain some recognition in the 1990s, but by the early 2000s, it was unequivocally the emphasis on financial exploitation that elevated—catapulted—“elder abuse” into the sphere of a social problem. In reviewing the Senate Special Committee on Aging, Jackson (2017) concluded that financial fraud and abuse hearings have been more numerous than hearings on elder abuse generally. Arguably, financial exploitation is responsible for the increased state and federal attention targeting elder abuse. Efforts to preserve assets are justified by our belief that independence is based on financial security and that financial exploitation can lead to a diminished quality of life for older adults (Button, Lewis, & Tapley, 2014).
It has been asserted that this increased attention is warranted given that financial exploitation is the most prevalent form of elder abuse in the United States. In fact, two often-cited prevalence studies conducted in the United States conclude that financial exploitation is the most prevalent form of elder abuse (Acierno et al., 2010; Lachs & Berman, 2011). However, at least one national prevalence study in the United States finds that psychological abuse is the most prevalent form of elder abuse (Laumann, Leitsch, & Waite, 2008), consistent with Sooryanarayana, Choo, and Hairi’s (2013) review of 26 prevalence studies from around the world. In their reviews of financial exploitation, Lindert et al. (2013) and Fealy, Donnelly, Bergin, Treacy, and Phelan (2012) concluded that prevalence rates of financial exploitation differ between countries, with De Donder et al.’s (2011) review of the European prevalence studies concluding that “ . . . financial abuse can be considered only as number four of the most often occurring forms of abuse” (p. 137).
Which form of elder abuse is the most frequently occurring remains an unsettled question due in part to variations in methodology across studies. Four reviews of the prevalence literature—Cooper, Selwood, and Livingston (2008); Sooryanarayana, Choo, and Hairi (2013); Yan, Chan, and Tiwari (2015); and Dong (2015)—find tremendous variability in prevalence across studies, due in part to differences among study methodologies. However, none of these reviews have focused on financial exploitation. This study fills this gap.
Method
Search Parameters
PubMed was searched using “prevalence” in the title/abstract and “elder abuse” or “elder mistreatment” in the title (without regard to date). This search yielded 177 citations. Articles were reviewed for inclusion eligibility: The study measured the prevalence of financial exploitation (although it could also measure other forms of abuse); the article presented original research (not a review or making reference to prevalence rates); respondents were residing in the community (not long-term care residents or other congregate living situations); and the journal was in English (Spanish language journals were excluded). Exclusion criteria included studies on the development of screening or measurement instruments (e.g., Jervis, Fickenscher, & Beals, 2014); special populations were excluded (Moon, Lawson, Carpiac, & Spaziano, 2006); and dissertations and conference presentations were excluded. This review resulted in 13 studies meeting eligibility and exclusion criteria. A second search was conducted using the search terms “population-based” in the title/abstract and “elder abuse” or “elder mistreatment” in the title. This search resulted in 211 citations. Review of the citations for distinct (additional) citations that met criteria described above resulted in five additional citations. A search in PsychINFO with “elder abuse” in the title and “prevalence” in the abstract yielded 80 hits, and a search using “elder mistreatment” in the title and “prevalence” in the abstract yielded 18 hits, neither of which resulted in any new articles. Thus, 18 published studies were included in the sample of published studies in this review of financial exploitation measures (original unpublished final reports were not included in the review). Two additional unpublished studies were included. A prevalence study by Lachs and Berman (2011) was included because it is a widely cited study and upon which subsequent research is based, and Podnieks, Pillemer, Phillip, Shillington, and Frizzel (1990) was included because Lachs and Berman is in part predicated on Podnieks’s work. Appendix A contains a list of the 20 studies included.
The author is aware of three additional prevalence studies that were not detected by the above search: Kivela et al. (1993); Jordanova, Markovik, Sethi, Serafimovska, and Jordanova (2014); and Chompunud et al. (2010). Kivela et al. (1993) and Jordanova et al. (2014) reported the prevalence of financial exploitation. Chompunud et al. (2010), however, did not present the prevalence of financial exploitation, although a description of how financial exploitation was measured is provided. To maintain replicability, however, these studies were excluded from this review.
Extraction
Twenty studies were reviewed for the following characteristics: (a) the reported prevalence (by type of abuse); (b) the manner in which the instrument(s) used to measure prevalence was developed; (c) financial exploitation items used to measure financial exploitation; (d) any reported psychometrics associated with the measurement of financial exploitation; and (e) eight aspects of methodology.
Results
The Prevalence of Financial Exploitation
Appendix A provides a summary of 20 prevalence studies across types of abuse. As summarized in Table 1, of the 19 prevalence studies that compare types of abuse (Peterson et al., 2014, only measured financial exploitation), financial exploitation is the most common form of elder abuse in six prevalence studies, the second most common in eight prevalence studies, and the third most common in three prevalence studies, and fourth in two prevalence studies. The prevalence of financial exploitation is graphed in Figure 1.
Comparison of Number of Studies Within Prevalence Rank by Type of Abuse.
Note. Table based on Appendix A. Physical abuse and sexual abuse were never the most prevalent form of elder abuse in any study.
Peterson et al. (2014) only measured financial exploitation so there was no comparison.
Peterson et al. (2014) did not measure psychological abuse.
Peterson et al. (2014), Beach, Schulz, Castle, and Rosen (2010), and Laumann, Leitsch, and Waite (2008) did not measure caregiver neglect.
Financial Exploitation Instrument Development
Appendix B describes the ways in which prevalence instruments have been developed, separately for each type of abuse. Based on a review of this appendix, four categories were created. As summarized in Table 2, financial exploitation instruments were most frequently the result of “Author-created” or “Based on previous research.” Neglect instruments were most frequently based on the “Adaptation of a standardized measure” followed by “Author-created” and “Based on previous research.” In contrast, psychological abuse and physical abuse were most frequently based on the “Adaptation of a standardized measure.”
A Summary of the Nature of How Instruments Were Developed (Four Categories) by Type of Abuse.
Note. Table based on Appendix B.
Peterson et al. (2014), Beach, Schulz, Castle, and Rosen (2010), and Laumann, Leitsch, and Waite (2008) did not measure caregiver neglect.
Peterson et al. (2014) did not measure psychological abuse.
Peterson et al. (2014) and Beach et al. (2010) did not measure physical abuse.
Nine studies did not measure sexual abuse and two studies combined sexual abuse with physical abuse.
Prevalence of financial exploitation by country (in percentages).
Items in Financial Exploitation Instruments
The various ways in which financial exploitation measures have been developed suggest that there might be considerable variability in the items used to measure financial exploitation. Appendix C provides the financial exploitation items associated with each author’s measurement of financial exploitation and where available, the prevalence for each item. The measures range in the number of items from one item to nine across studies.
Reliability and Validity of Financial Exploitation Instruments
Appendix D provides a review of the reliability and validity measures reported in prevalence studies. As can easily be observed, only five studies report any type of reliability (and all five only report Cronbach’s alphas), and only three of those provide alphas by type of abuse. What is abundantly clear is that the alphas are consistently lower for the financial exploitation scale compared with the alpha for other types of abuse.
A similar situation regards the reporting of validity. Only eight studies describe any validity efforts, most frequently in the form of pre-testing the instrument, a sort of face validity. No other psychometrics (reliability or validity) were mentioned.
Methodology Across Studies
Appendix E provides a review of nine aspects of the methodology across these prevalence studies (all of which are cross-sectional). Of the 20 studies reviewed, 14 studies defined elder abuse, types of elder abuse, or both. The samples ranged from 200 to more than 15,000, although the majority of studies include a sample of 1,000s (only one study described the approach taken to determine the needed sample size). Response rates also varied, although the mode was in the 80% range (two studies reported cooperation rates as well). Only three studies had response rates in the 50% range or lower. Samples consisted primarily of cognitively intact respondents (n = 14 studies). Data collection primarily took the approach of a face-to-face interview in the home (n = 12), followed by telephone interviews (n = 6) and self-administered surveys (n = 2). And age was rather consistent, with 10 studies using age 60 and eight using age 65 or 66.
The two areas that showed some variability concerned prevalence period and abuser category. However, 12 of the studies used a past 12-month prevalence period (three studies also included since turning age 60). Two studies, however, used different time periods for different types of abuse. Abusers were primarily either anyone (n = 6) or a trusted person (n = 8). However, three studies used different abuser types for different types of abuse.
Discussion
Consistent with previous reviews (Cooper et al., 2008; Dong, 2015; Sooryanarayana et al., 2013; Yan et al., 2015), there was tremendous variability among the 20 prevalence studies reviewed in this article, with nine studies finding psychological abuse the most prevalent form of elder abuse (see Conrad, Iris, Ridings, Langley, & Anetzberger, 2011; Conrad, Iris, Ridings, Rosen, et al., 2011). However, this study concluded that financial exploitation prevalence rates varied from less than 1% to more than 16%. Furthermore, this study found that financial exploitation is the most common form of elder abuse in six prevalence studies, the second most common in eight prevalence studies, and the third most common in three prevalence studies, and fourth in two prevalence studies (see Table 1). Variability may be due in part to the manner in which instruments have been developed as well as methodological differences across studies. What is consistent within studies is the exclusion of older adults with diminished capacity across subtypes, which suggests that prevalence rates are likely higher than most studies indicate.
In stark contrast to the fairly consistent measurement of physical, psychological, and sexual abuse, and to a degree caregiver neglect, there is less consistency in how financial exploitation instruments have been developed as alluded to by Lowenstein, Eisikovits, Band-Winterstein, and Enosh (2009). For example, this review found that instruments were developed based on (a) “A review of the literature,” (b) Author-created, (c) Adaptation of standardized measure, and (d) Based on previous research. As summarized in Table 2, financial exploitation instruments were most frequently the result of “Author-created” or “Based on previous research.” Neglect instruments were most frequently based on the “Adaptation of a standardized measure” followed by “Author-created” and “Based on previous research.” In contrast, psychological abuse, physical abuse, and sexual abuse were most frequently based on the “Adaptation of standardized measure.” Pillemer and Finkelhor (1988) were the first to measure three forms of elder abuse (physical, verbal, and caregiver neglect) by adapting the Conflict Tactics Scale to measure physical and verbal abuse (citing Straus, 1979). Many authors have followed suit.
Authors who chose to use existing standardized instruments frequently have “adapted” the instrument in some way (see Appendix B). For example, authors would write “We selected and modified items from two well validated instruments for elder abuse: the Hwalek–Sengstock Elder Abuse Screening Test and the Vulnerability to Abuse Screening Scale” (Laumann et al., 2008, p. 4). However, such modifications tamper with the original psychometrics and undermine our confidence that the original psychometrics still ring true. Similarly, some authors would refer readers to the original study. For example, O’Keeffe et al. (2007) wrote that “behaviourally specific measures have a stable factor structure, moderate to high reliability, and considerable evidence of construct validity,” citing Straus (2007), but invariably modified the instruments. In either scenario, authors who relied on existing screens failed to re-evaluate the psychometrics for their sample. A model for future studies was conducted by Yan and Tang (2001) who reported the psychometrics for the original scale (Straus, Hamby, Boney-McCoy, & Sugarman, 1996), modified the Revised Conflict Tactics Scales (CTS2) for their study, and then assessed and reported the alphas that pertained to their study (physical abuse: .79, .82, respectively, and verbal abuse: .86, .81, respectively), providing the information readers need to be able to assess the quality of the instrument. However, they did not measure financial exploitation.
A standardized measure, the Vulnerability to Abuse Screening Scale (VASS; Schofield & Mishra, 2003), contains a coercion factor that is used as a measure of financial exploitation. However, a closer look at the coercion factor raises some concern. The only item suggestive of financial exploitation is “Has anyone taken things that belong to you without your OK?”, an item that has a factor loading of .35, unusually low for inclusion in a factor. The other two items in this factor were “Does someone in your family make you stay in bed or tell you you’re sick when you know you’re not?”, with a factor loading of .82, and “Has anyone forced you to do things you didn’t want to do?”, with a factor loading of .61. The use of this instrument as a measure of financial exploitation may be problematic.
Another concern with these modified instruments is the sample upon which the original instrument was tested. For example, the VASS was developed on a sample of older women (Schofield & Mishra, 2003) but is being used on samples of older men and women.
Furthermore, the U.S. Preventive Services Task Force has concluded that there are “ . . . no valid, reliable screening tools to identify abuse [or financial exploitation] of elderly or vulnerable adults in the primary care setting . . . ” (Moyer, 2013, p. 480), including those identified above, suggesting that the use of these screens (as screens are used for other purposes such as measuring prevalence) may be unsuitable, and particularly unsuitable when they are modified.
Given the differences in the manner in which financial exploitation instruments have been developed, it is unsurprising that there is tremendous variability in the items included in instruments developed to measure financial exploitation. When financial exploitation is measured, respondents are asked about one to nine types of financial exploitation when there are conceivably 30 or more ways in which a person might be financially exploited (Black, 2008; Blunt, 1996; Federal Trade Commission [FTC], 2015; Fuentes, 2009; Jackson & Hafemeister, 2012; Johnson, 2003; Lewis, 2001; Office for Victims of Crime, 1998; Sklar, 2000; Stiegel & VanCleave Klem, 2008; Thilges, 2000; U.S. Department of Justice [U.S. DOJ], 2015; U.S. Department of Housing and Urban Development [U.S. HUD], 2013). Of course, this suggests that prevalence studies are underestimating the prevalence of financial exploitation. However, generally there are only one or two items in the measure that account for the majority of the prevalence (see Appendix C).
The absence of psychometrics across studies is stunning. Typically, the psychometrics associated with an instrument are presented to allow scientists to assess the instrument’s validity and reliability (in all its forms). However, consistent with our findings, other reviews of elder abuse instrument development conclude that “ . . . none report on reliability and validity . . . ” (De Donder et al., 2011, p. 137) or there are “ . . . no measures with established psychometrics . . . ” (Cooper et al., 2008, p. 159). In fact, there are only three studies (Chokkanathan & Lee, 2006; DeLiema, Gassoumis, Homeier, & Wilber, 2012; Giraldo-Rodríguez & Rosas-Carrasco, 2013) which report Cronbach’s alpha for each type of abuse; no other type of reliability is even mentioned. What is particularly revealing in Appendix D is that Cronbach’s alphas are consistently lower for financial exploitation than for other forms of abuse, suggesting that the financial exploitation construct is not as internally consistent as other types of abuse. Only three studies mentioned validity, most commonly in the form of pre-testing the instrument on a group of individuals.
In some ways, these studies are remarkably methodologically similar. Most use a particular lower-boundary age (60 or 65), face-to-face interviews, with cognitively intact respondents, and most typically measuring abuse that occurred within the past 12 months. Sample size differed considerably, and only one study presented a power analysis, but most were robust in sample size. Recruitment methods were generally of a “random selection” nature but did differ considerably in terms of geography. Response rates also varied somewhat, although the mode was in the 80% range (two studies reported cooperation rates as well).
The two areas where some differences emerged concerned the prevalence time period and abuser category. A consistent time period across types of abuse is critically important. However, Podnieks et al. (1990) used a different time period for financial exploitation than for other forms of abuse. This choice was likely based on Pillemer and Finkelhor (1988) who measured physical abuse using “at least once since turning age 65” while all other forms of abuse were measured “in the past year,” which likely explains at least in part why physical abuse was the most prevalent form of elder abuse in their study. The difference this makes is illuminated by Lachs and Berman (2011) who found that financial exploitation was the most common form of abuse cited by respondents as having occurred in the 12 months (incidence), but emotional abuse was the most prevalent when calculating lifetime prevalence (since turning 60). This approach was relatively rare, but important to identify as types of elder abuse vie for “most prevalent” status.
Abuser category should likewise be consistent across types of abuse within a prevalence study. Who is defined as the offender has huge implications for prevalence rates (De Donder et al., 2011). Restrictions on offender varied tremendously across studies, but within several studies, offenders ranged from family member for some types of abuse to “anyone” for financial exploitation, or vice versa. Podnieks et al. (1990), for example, used “anyone” for financial exploitation, while requiring a close relationship for other types of abuse. In contrast, Acierno et al. (2010) used “anyone” for all types of abuse except financial exploitation, which required family members. Methodological differences across types of abuse within the same study impede the ability to make comparisons even within a study.
Finally, if definitions were provided, most scholars adopted either the Bonnie and Wallace (2003) or the WHO (2002) definition. However, the absence of definitions of elder abuse—or the subtypes or both—across studies was concerning. De Donder et al. (2011) asserted that depending on the author’s concept and definition of violence or abuse, the items included in the survey instruments vary greatly. It may be that because definitions of types of abuse differ, their instruments differ, making comparisons across types of abuse a fool’s errand.
New Directions
The results of this article underscore the consensus among researchers that improvements in definition and measurement is their number one priority (Stahl, 2015). However, prior to measurement development is the need to grapple with theoretical, conceptual, and definitional issues. For example, the most pervasive theory used to explain financial exploitation is the criminological Routine Activities Theory (RAT; Payne, 2011; Setterlund, Tilse, Wilson, McCawley, & Rosenman, 2007), and yet RAT fails to incorporate known risk factors associated with financial exploitation (e.g., cognitive impairment; see Dong, Simon, Rajan, & Evans, 2011). New theories unbeholden to existing theory remain a need (Jackson & Hafemeister, 2013).
The concept may benefit from placing boundaries around what constitutes financial exploitation. As one example, financial fraud (involving deception) could be distinguished from financial exploitation as some have argued (DeLiema, 2015; Jackson, 2015; Payne, 2011; Roubicek, 2008). The painstaking analysis in three conceptual papers on elder abuse (Goergon & Beaulieu, 2013; Harbison et al., 2012; Policastro, Gainey, & Payne, 2015) should be emulated in the context of financial exploitation. An empirical approach is to use concept mapping. Conrad, Iris, Ridings, Fairman, et al. (2011) invited professionals in the field of elder abuse to write descriptions of anything related to financial exploitation. While innovative and useful, the items were too expansive (e.g., offender motivation, financial management problems). A similar exercise restricted to types of financial exploitation would be a further advancement.
Based on theoretical and conceptual development, definitions of financial exploitation then could be derived. Jackson (2015) has argued that the definition of financial exploitation has expanded over time. While acceptable for public awareness purposes, greater precision is required in the context of research (see Centers for Disease Control and Prevention [CDC], 2016).
Finally, once the theoretical, conceptual, and definitional issues are resolved, researchers could turn their attention to measurement development. The difficult task of designing the instrument and gathering and presenting accompanying psychometrics to enable the field to assess the quality of the instruments is needed. Conrad, Iris, Ridings, Langley, and Wilber (2010) developed a self-report measure of financial exploitation that may be useful for determining whether someone is a victim of financial exploitation, but may be less useful for measuring prevalence. Instruments are designed for specific tasks (prevalence, screening, diagnosis, research, policy) and the field must cease cross-purposing instruments. With that said, there may be ways to repurpose Conrad et al.’s (2010) foundational work to meet the needs of those measuring prevalence.
Implications for Practitioners
This seemingly abstract issue of measurement has direct implications for practitioners. If our theories, conceptualizations, definitions, and instruments are wrong, then our prevention and intervention efforts are going to be wrong. Without accurate measurement, including those with diminished capacity, attempts to remedy the condition of older adults will be hampered. Four elder abuse intervention reviews, admittedly not exclusively focused on financial exploitation, are dismally discouraging (Ayalon, Lev, Green, & Nevo, 2016; Daly, 2011; O’Donnell, Phelan, & Fealy, 2015; Ploeg et al., 2009). It is our primary goal to prevent older adults from experiencing financial ruin due to financial exploitation, and for those who do experience it, to assist in their recovery, possibly aided by individualized goal attainment (Burnes & Lachs, 2017). And yet, without a greater understanding of the phenomenon and the cultural variation within, attempts to prevent and intervene will continue to be discouraging, ultimately harming the older adults we seek to protect.
Summary
Although a few methodological differences emerged, the real story appears to be the variable manner in which financial exploitation instruments have been developed. Taken together, this review concludes that the instruments used to measure financial exploitation are more variable and less psychometrically robust compared with the measurement of other forms of abuse. Prevalence is unequivocally important in attracting the attention of policymakers who fund programs for older adults, which has been in part the impetus for undertaking these national prevalence studies. And yet, with only one exception, the field has disregarded the tedious task of measurement development, a science unto itself. Given the concerns raised in this review, investment in financial exploitation measurement development is clearly warranted.
Footnotes
Appendix
Nine Aspects of Methodology Across Studies.
| Author | Location of study | Defines elder abuse/types of abuse | Sample size (Response rate) | Cognitively intact sample | Data collection format | Age of victim | Prevalence period | Abuser category | Sample selection method |
|---|---|---|---|---|---|---|---|---|---|
| Rahman and El Gaafary (2012; Egypt) | Rural area of Mansoura city, Egypt | Yes/No | 1,106 (95%) | No | Face-to-face interview in home | 60 + | Past 12 months | Caregivers | All elderly residing in the area were invited to participate |
| Chokkanathan and Lee (2006; India) | Chennai (urban), India | Yes/Yes | 400 (80%) | Yes | Face-to-face interview in home | + | Past 12 months | (seems to be expectation of trust) | Randomly selected |
| Oh, Kim, Martins, and Kim (2006; South Korea) | Seoul, Korea | No/No | 15,230 (53%) | No | Face-to-face interview in home | 65 + | Past 1 month | (seems to be family) | The entire population . . . living in Songpa Gu |
| Dong, Simon, and Gorbien (2007; China) | Urban China, NanJing Drum Tower Hospital | No/Yes | 412 (82%) | Yes | Self-administered survey | 60 + | — | Anyone (for financial exploitation only) | Subjects were identified when they registered |
| Wu et al. (2012; China) | Macheng (rural), China | Yes/Yes | 2,000 (89%) | Yes | Face-to-face interview in home | 60 + | Past 12 months | Anyone (for financial exploitation only) | Randomly selected rural villages within Macheng |
| Giraldo-Rodríguez and Rosas-Carrasco (2013; Mexico) | Mexico City, Mexico | No/No | 613 ( a ) | No | Face-to-face interview in home | 60 + | Past 12 mo | Anyone | Probabilistic sample (recruitment method unspecified) |
| Garre-Olmo et al. (2009; Spain) | Angle’s Primary Healthcare Area, Girona (rural) | Yes/No | 875 (82%) | Yes | Face-to-face interview in home | 75 + | Past 12 months | Anyone | Randomly selected from the municipal census, stratified by age (75-84 and 85+) |
| Gil et al. (2015; Portugal) | Portugal | Yes/Yes | 1,123 (74%) | Yes | CATI | 60 + | Past 12 mo | Anyone | Randomly selected telephone (land and cell) lines from population, stratified by seven geographic regions |
| Lowenstein, Eisikovits, Band-Winterstein, and Enosh (2009; Israel) | Urban areas in Israel | Yes/No | 1,045 (75%) | Yes | Face-to-face interview in home | 65 + | Past 12 months (past 3 and 6 months for neglect) | Family or care workers | Unclear (addresses of the older person’s households were obtained) |
| Comijs, Pot, Smit, Bouter, and Jonker (1998; The Netherlands) | Amsterdam | Yes/Yes | 1,797 (*) | No | Face-to-face interview in home | 65 + | Past 12 months | Personal or professional relationship | Randomly selected (stratified by age) from Amsterdam Study of the Elderly |
| Naughton et al. (2012; Ireland) | Ireland | Yes/No | 2,021 (83%) | Yes | Face-to-face interview in home | 65 + | Past 12 mo | Family, in-laws, close friends, care workers | Eligible participants within each cluster (stratified by region) were identified using a random route finding approach. |
| Biggs, Manthorpe, Tinker, Doyle, and Erens (2009; the United Kingdom) | United Kingdom | Yes/No | 2,111 (65%) | — | Face-to-face interview in home; CAPI in home | 66 + | Past 12 months | Expectation of trust (but asks about anyone) | Random probability sample of participants in the Health Survey; Wales used a random selection of addresses |
| Beach, Schulz, Castle, and Rosen (2010; the United States) | Allegheny County, PA | No/No | 903 (37.7%) | Yes | Four methods | 60 + | Past 6 months; since turning 60 | Trusted other (but asks about anyone for financial exploitation) | Random-digit dialing telephone (landline) sampling with screening for age |
| Peterson et al. (2014; the United States) | 10 regions in New York | No/Yes | 4,156 (67.4%) | Yes | Telephone interview | 60 + | Past 12 months; since turning 60 | Anyone | Random-digit dial sample; tested for selection bias |
| Laumann, Leitsch, and Waite (2008; the United States) | Across the United States | Yes/No | 3,005 (75.5%) | Yes | Face-to-face interview in home; self-administered survey | 57-85 | Past 12 months | Anyone (although prevalence was based on a family member) | Probability design; data from the National Social Life, Health and Aging Project (NSHAP) |
| Acierno et al. (2010; the United States) | Continental United States | No/No | 5,777 (69%) | Yes | CATI | 60 + | Past 12 months | Family for financial exploitation; other forms by anyone | Stratified random-digit-dialing |
| Amstadter et al. (2011; the United States) | South Carolina | Yes/No | 902 (*) | Yes | CATI | 60 + | Past 12 months | Family member for financial exploitation (other forms are variable) | Stratified random-digit dialing with an area probability sample based on census-defined “size of place” parameters (e.g., rural, urban). |
| DeLiema, Gassoumis, Homeier, and Wilber (2012; the United States) | Los Angeles, CA | No/No | 200 (65%) | Yes | Face-to-face interview in home | 66 + | Past 12 months | — | Randomly selected census tracts in Los Angeles; door-to-door recruitment |
| Lachs and Berman (2011; the United States) | All regions of New York | Yes/Yes | 4,156 (18.6%) | Yes | Telephone interview | 60 + | Past 12 months; since turning 60 | Anyone | Random-digit dialing strategy derived from census tracts |
| Podnieks, Pillemer, Phillip, Shillington, and Frizzel (1990; Canada) | Five regions in Canada | No/No | 2,008 (90%) | — | Telephone interview | 65 + | Varies by type of abuse | Anyone they know for financial exploitation; (other forms are variable) | Modified random-digit procedure |
Note. CATI = computer assisted telephone interview; CAPI = computer assisted personal interview.
Indicates unspecified or unavailable in text.
Declaration of Conflicting Interests
The author declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding
The author received no financial support for the research, authorship, and/or publication of this article.