Technology Use in Everyday Financial Activities: Evidence from Online and Offline Survey Data

Abstract

Objective

Internet use and mobile devices permeate every aspect of our lives and are changing our financial habits. Assessment of financial decision-making (FDM) has not yet caught up to apparent changes in financial behavior. To modernize assessment methods and create current and comprehensive FDM frameworks, we first need to establish the most commonly used and most preferred methods of performing specific financial activities.

Method

Cross-sectional survey data were collected using an online platform and offline approaches (in person and by mail) (N = 234). The frequency of using technological (e.g., laptop) and non-technological (e.g., in-person banking) means of completing seven financial activities was assessed first, including Depositing checks, Reviewing bank statements, Keeping track of money spent, Transferring funds, Withdrawing cash, Paying bills, and Purchasing products online. Second, preference for technological versus non-technological methods was assessed. Finally, linear regression models examined associations between demographics and preference for technological methods for each financial activity.

Results

The majority of respondents (77% online, 74% offline) used technology to perform various financial activities and preferred technological to non-technological methods for completing five out of the six financial activities. Increased preference for technological methods was associated with younger age for all the financial activities, and higher education was associated with reviewing bank statement and transferring funds.

Conclusions

Our survey findings provide empirical evidence for the changing nature of our financial habits. We discuss the implications of this change for researchers, clinicians, and the individuals themselves and emphasize the importance of modernizing FDM tools.

Introduction

Financial decision-making (FDM) is an important functional construct with relevance across the adult lifespan in both healthy and clinical populations. A gold-standard definition of FDM is lacking given the multifaceted nature of this construct as well as the many ways that different fields of study research this construct (e.g., psychology and economics) (Lichtenberg et al., 2015; Lusardi & Mitchelli, 2007; Marson, 2016; National Academies of Sciences, Engineering, & Medicine, 2016; Sunderaraman et al., 2019). In the current study, based on previous research, FDM is defined as the ability to autonomously conduct financial tasks in order to manage one’s finances without error or preventable financial loss (Marson, 2016; Lichtenberg et al., 2015; National Academies of Sciences, Engineering, & Medicine, 2016).

The study of FDM and its constituent abilities is important for establishing optimal functioning among adults of all ages, enhancing individuals’ abilities to accrue wealth in the form of savings for current and future expenses, such as retirement, and for enabling individuals to make financial decisions in line with their values. Additionally, studying FDM is critical in clinical populations for whom it becomes compromised. For example, in young adults with acquired brain injuries, impaired FDM has been shown to have significant ramifications for patients’ ability to regain independent functioning (Hoofien et al., 2001; Olver et al., 1996). Accurate assessment and ongoing monitoring of patients’ status becomes imperative as they transition to managing financial responsibilities (Martin et al., 2012; Ponsford et al., 2000; Sunderaraman et al., 2019). Another challenge arises in the growing frequency of older adults with neurodegenerative diseases, many of whom will at some point demonstrate decreased FDM (Gamble et al., 2015), and be at risk for making suboptimal financial decisions, such as poor investment choices (Korniotis & Kumar, 2011). Additionally, older adults are increasingly scammed by family members, friends, and third-party individuals; estimates of resultant financial loss range from a net loss of $2.9 billion to $36.5 billion (Consumer Financial Protection Bureau, 2016). Older adults are living longer and are estimated to hold about 70% of all disposable income (Taha et al., 2009; U.S. News and World Report, 2015). It is therefore critical that we develop more objective and more current methods of evaluating FDM across the adult lifespan.

Neuropsychologists, neurologists, gerontologists, and social workers, among other practitioners, often encounter the challenge of determining a person’s functional status in matters related to FDM. The determination of FDM relies on instruments that assess performance in a controlled setting such as an office (e.g., writing a check) or real-life behaviors that are examined via semi-structured interviews using self- and/or informant-reports (Griffith et al., 2003; Lichtenberg et al., 2015; Spreng et al., 2017). However, such methods have questionable ecological validity or rely on subjective report, which may not be entirely accurate and is often difficult to objectively verify (Engel et al., 2016; Germine et al., 2018; Miller & Barr, 2017; National Academies of Sciences, Engineering & Medicine, 2016; Sunderaraman et al., 2018).

The internet revolution and other rapid technological advances are dramatically changing the ways in which individuals handle their finances (Borkar & Pande, 2016; Choudrie & Vyas, 2014; Jiménez & Díaz, 2019). As of December 2017, the internet penetration rate (the percentage of the total population of a given country or region that uses the internet) for North America ranged between 84%–95%, with about 82% of individuals in New York City using the internet (Internet World Statistics, 2017). The use of online banking, internet-based shopping, electronic transfer of funds, and so forth are fast becoming common ways of conducting financial transactions. According to the Bank of America (2017) Trends in Consumer Mobility Report, about 62% of Americans (or more than three in five) use a mobile banking app, an increase by 8% compared to 2016. Of the individuals using mobile banking apps, 47% were aged 53 to 71 years and 40% were aged 72 years or above. Per Bank of America’s records, 23 million clients actively use the bank’s mobile banking platform with each user having logged onto their online account approximately 47 times in the second quarter (May to June) of 2017. The popularity of such online financial platforms is increasing given the ease of accessing any kind of financial information (e.g., account balance and spending activity). Moreover, use of technological platforms such as internet banking helps banks save costs and experiment with various strategies to attract clients (Chang, 2006; White & Nteli, 2004; Yiu et al., 2007). In another series of Bank of America (2017) surveys, the most commonly cited reason for downloading the bank’s mobile app was having 24/7 access to financial information (55%), followed by ease of depositing checks (44%), reducing trips to the bank (29%), and paying bills (28%). In another survey, about 38% of individuals ranked the provision of online banking services as more important than the proximity of branch location or low fees (Accenture North America Consumer Banking Survey, 2015).

Despite technology’s ubiquitous influence on our everyday financial transactions, practitioners and researchers continue to measure FDM using relatively traditional instruments (Miller & Barr, 2017). Among others, one critical reason for the lack of a more modern assessment approach is the dearth of information regarding which technologies are the most frequently used and are therefore important to examine in the context of FDM evaluations. Establishing current financial behaviors and modernizing assessment approaches become imminent as tasks such as balancing checkbooks, writing checks, or even using cash to buy and sell products may become extinct altogether. Indeed, 51% individuals find writing checks to be the most frustrating payment habit (Bank of America, 2017a).

In order to modernize the assessment of FDM, we first need to collect information on people’s financial habits and understand their preference for different types of technological products to conduct various financial tasks. Although the Bank of America survey was conducted on about 1,000 people, it was conducted online and consisted of individuals who owned a smartphone and who had a savings or checking account with the bank. In order to collect data from a more heterogeneous and widely representative sample, we administered a comprehensive survey to individuals using both online (Amazon Turk) and offline formats (in person or by mail). The overarching goal of the survey was to gather information about individuals’ current financial habits and their preferences for conducting various financial activities through their use of both technological and non-technological methods. Based on previous research, we anticipated that the majority of the respondents would use, and also prefer to use, technological methods to perform various financial tasks (Accenture North America Consumer Banking Survey, 2015; Bank of America, 2017a; Pew Research Center, 2017). A secondary goal of the current study was to determine if demographic characteristics were associated with preference for using technology. Previous research has found that being older, being female, and having fewer years of education were all associated with less technology use (Choudrie & Vyas, 2014; Fox, 2013; Pew Research Center, 2017). We thus anticipated similar associations in the current study after adjusting for method of data collection (online versus offline).

Materials and Methods

Participants

The inclusion criteria were that individuals be between 30 and 90 years old and living in New York City for at least 5 years to ensure that participants were acclimated with the financial institutions and more generally with the monetary system.

Measures

The Financial Habits and Technology Use survey

The survey, specifically developed for this study, consisted of two main sections. The first section collected demographic information, whereas the second section pertained to financial habits across seven broad financial activities relevant to daily life. These activities were Depositing Checks, Reviewing Bank Statements, Keeping Track of Spending Money, Transferring Funds, Withdrawing Cash, Paying Bills, and Purchasing Products Online. We first queried whether or not individuals conducted each of these financial activities and, if so, the frequency with which the activity was conducted. For those activities that were conducted, we queried the method(s) used to conduct the financial activity (e.g., in person or mobile app), the duration of time that each method has been used, and the top (most) preferred method for conducting the activity.

Procedures

The survey was distributed both offline and online. On average, participants took about 20–40 mins to complete the survey. All procedures were approved by the Columbia University Medical Center's Institutional Review Board.

1) Offline: Offline surveys were collected in person or via mail. A total of 30 participants completed a paper version of the survey at the end of an ongoing study of cognitive aging between March and November 2018. For mailed surveys, a mass mailing company was used to send surveys to 1,000 people who had been living in any of the five boroughs of New York City for at least 5 years. Individuals were randomly selected to represent New York City census data in terms of age, gender, race, ethnicity, socioeconomic status, and level of education. Participants were paid $10 for completing and returning the mailed surveys.

2) Online: A parallel copy of the survey was designed on Qualtrics and distributed online through Amazon’s Mechanical Turk. It was distributed two times in March 2018, for 2 days each. The goal of the second distribution was to increase the representation of older adults. A filter was added at the second distribution to ensure that previous respondents could not retake the survey. Online participants had to first respond to two screener questions pertaining to their age and number of years lived in New York City. An attention check question was embedded within the task to ensure that respondents were deliberatively responding. If they passed the screener and attention check question, they were then able to complete the survey. Upon completion of the survey, they were paid $5.

The offline versus online compensation amounts were based on an analysis of the typical pay rates for the specific length of questionnaire administered using the two methods.

Data Analysis

Statistical analyses were performed using IBM SPSS v.25. The in-person and mailed survey data responses were combined to represent the offline group of respondents, whereas those who completed the survey via M-Turk represented the online respondents. For each group of offline and online respondents, we characterized the frequency of engaging in each financial activity, the frequency of different methods used to perform the activity, and the duration of using each method. We also identified the most preferred method for completing each financial activity and characterized these methods as technological versus non-technological. We then examined whether preference for technological versus non-technological methods differed across demographic characteristics including age, gender, and education level using separate logistic regression models for each financial activity (except online purchases, an activity that was coded as Yes/No). The predictors in the model were age, gender, and years of education, with survey completion method (online versus offline) as a covariate. For two financial activity domains (transferring funds and reviewing bank statements) the assumption of linearity was not met for the education variable, and therefore, log-transformed data were used for these two analyses. Collinearity diagnostics were included to test for multicollinearity. Tolerance value of <0.1 (Menard, 1995) and variance inflation factor value of >10 (Myers, 1990) were used as cutoffs for multicollinearity.

Results

Survey Completion

With regard to offline survey data, 59 out of 1,000 surveys were returned via mail, of which 6 were excluded due to incomplete information. Together with the 30 surveys administered in person, offline survey data were thus available for 83 participants. Compared to the typical 10% anticipated rate of return for the mailed surveys, one likely factor contributing to a relatively lower return rate (i.e., 6%) could be the length of our survey, which was 35 pages long.

For online survey data, in the first distribution, 316 individuals initiated the survey; however, 203 did not pass the inclusion criteria for age and geographic location. Of the remaining 113, 13 failed the attention check question that was embedded within the survey to ensure that respondents were carefully reading the questions and then filling out the survey. A total of 100 respondents’ data were thus retained for analysis. The second time the survey was distributed, 189 individuals began the survey, with 135 not meeting screening criteria. Of the remaining 54, 3 failed the attention check question. Online data were retained for 51 participants in the second distribution. Overall, out of the 505 respondents who began the online survey, data were available for 151 participants.

Demographic and Overall Financial Habits

Demographic information for offline and online participants is presented in Table 1. In brief, the online group was younger, had fewer years of education, had a higher proportion of male respondents, had a different distribution of race, and had members that were more likely to be self-employed and had a full-time job. No differences between the groups were found for ethnicity, student status, or socioeconomic status. Supplementary material online, Table S1 provides demographic information on only offline respondents (In-Person and Mailed-in).

Table 1.

Summary of demographic characteristics for all the survey respondents

	Online	Offline
	Mean (SD); or frequency (%)		Statistics
N	151	83
Age (in years)	43.38 (12.12)	58.07 (15.68)	t = −7.98, p < .001
Education (in years)	15.03 (3.81)	16.12 (3.09)	t = −2.16, p = .03
Sex
Males	120 (79.5%)	32 (38.6%)	χ2 = 39.39, p < .001
Racea
Black / African American	14 (9.3%)	21 (25.3%)	χ2 = 26.70, p < .001
Caucasian	81 (53.6%)	43 (51.8%)
Asian	46 (30.5%)	6 (7.2%)
American Indian or Alaska Native	8 (5.3%)	1 (1.2%)
Hawaiian or Other Pacific Islander	0 (0)	2 (2.4%)
Other	2 (1.32%)	10 (12.05%)
Ethnicity (%)
Hispanic	30 (19.9%)	11 (13.3%)	χ2 = 1.34, p = .25
Non-Hispanic	121 (80.1%)	69 (83.1%)
Student Status
Non-students	142 (94%)	75 (90.4%)	χ2 = .09, p = .93
Students	9 (6%)	5 (6%)
Students enrolled full-time	6 (4%)	2 (2.4%)
Students enrolled part-time	3 (2%)	3 (3.6%)
Students receiving stipends	7 (4.6%)	2(2.4%)
Students not receiving stipends	2 (1.3%)	3(3.6%)
Employment
Self-Employed	34 (22.7%)	9(11.8%)	Fisher’s Exact = 57.13, p < .001
Full-Time	99 (66%)	26(34.2%)
Part-Time	8 (5.3%)	5(6.6%)
Homemaker	1 (0.7%)	-
Permanently sick, disable, or unable to work	0 (0)	3(3.9%)
Unemployed or temporarily laid off	0 (0)	6(7.9%)
Retired	8 (5.3%)	27(35.5%)
Managed Income
$0–$4,999	6 (4%)	3(3.6%)	Fisher’s Exact = 18.15, p = .09
$5,000–$9,999	8 (5.3%)	1(1.2%)
$10,000–$14,999	4 (2.6%)	4(4.8%)
$15,000–$19,999	5 (3.3%)	3(3.6%)
$20,000–$24,999	9 (6%)	3(3.6%)
$25,000–$29,999	4 (2.6%)	5(6%)
$30,000–$34,999	15 (9.9%)	7(8.4%)
$35,000–$49,000	22 (14.6%)	10(12%)
$50,000–$74,999	34 (22.5%)	16(19.3%)
$75,000–$99,999	15 (9.9%)	7(8.4%)
$100,000–$124,999	17 (11.3%)	4(4.8%)
$125,000–$149,999	5 (3.3%)	2(2.4%)
$150,000 and over	7 (4.6%)	15(18.1%)

Respondents include those who completed the survey online (using M-Turk) and offline (either in-person or mailed in the survey).

^a Note: Some respondents have checked multiple options (as it applied to them), and therefore, the total frequencies may be more than 151 (or more than 100%) in some cases.

As can be seen from Table 2, the majority of participants affirmed that they conducted the surveyed financial tasks (range of 69%–99% across tasks). Table 3 reveals that across six out of seven financial activities, respondents engaged in a given activity at least one to three times a month.

Table 2.

Percentage of respondents conducting specific financial activities

	Online (n = 151)	Offline (n = 83)
	N (%)
Depositing checks	136 (90%)	68 (82%)
Reviewing bank statements	139 (92%)	73 (89%)
Keeping track of spending money	129 (85%)	78 (95%)
Transferring funds	104 (69%)	66 (80%)
Withdrawing cash	128 (85%)	80 (96%)
Paying bills	135 (89%)	82 (99%)
Purchasing products online	132 (87%)	72 (88%)

Respondents include those who completed the survey online (using M-Turk) and offline (either in-person or by mail).

Note: Percentages have been rounded up.

Table 3.

Percentage of respondents conducting each financial activity at specified frequency

	Frequency	Online	Offline
Depositing checks
	Daily	1 (0.7%)	—
	1–3 times a week	32 (23.5%)	5 (7.4%)
	1–3 times a month	77 (56.6%)	41 (60.3%)
	4 times a year	20 (14.7%)	18 (26.5%)
	Twice a year	5 (3.7%)	3 (4.4%)
	Annually	1 (0.7%)	1 (1.5%)
	n	136	68
Reviewing bank statements
	Daily	9 (6.5%)	8 (11%)
	1–3 times a week	38 (27.3%)	7 (9.6%)
	1–3 times a month	72 (51.8%)	49 (67.1%)
	4 times a year	16 (11.5%)	9 (12.3%)
	Twice a year	3 (2.2%)	—
	Annually	1 (0.7%)	—
	n	139	73
Keeping track of spending money
	Daily	43 (33.3%)	29 (37.2%)
	1–3 times a week	50 (38.8%)	25 (32.1%)
	1–3 times a month	29 (22.5%)	24 (30.8%)
	4 times a year	6 (4.7%)	—
	Twice a year	1 (0.8%)	—
	n	129	78
Transferring fundsa
	Daily	3 (2.9%)	—
	1–3 times a week	21 (20.2%)	8 (12.1%)
	1–3 times a month	56 (53.8%)	42 (63.6%)
	4 times a year	21 (20.2%)	12 (18.2%)
	Twice a year	2 (1.9%)	2 (3%)
	Annually	1 (1%)	1 (1.5%)
	n	104	65
Withdrawing casha
	Daily	5 (3.9%)	—
	1–3 times a week	53 (41.4%)	24 (30%)
	1–3 times a month	58 (45.3%)	47 (58.8%)
	Four times a year	10 (7.8%)	5 (6.3%)
	Twice a year	2 (1.6%)	2 (2.5%)
	n	128	78
Paying billsa
	Daily	2 (1.5%)	—
	1–3 times a week	21 (15.6%)	14 (17.1%)
	1–3 times a month	106 (78.5%)	67 (81.7%)
	Four times a year	5 (3.7%)	—
	Annually	1 (0.7%)	—
	n	135	81
Purchasing products online
	1–3 times a week	33 (25%)	12 (16.7%)
	1–3 times a month	75 (56.8%)	40 (55.6%)
	Four times a year	24 (18.2%)	18 (25%)
	Annually	-	2 (2.8%)
	n	132	72

Respondents include those who completed the survey online (using M-Turk) and offline (either in-person or by mail). Note: Frequency percent are displayed in the cells, and the total number of respondents for each specific financial activity is represented as n.

^aPercentages do not add up to 100 because of missing data.

Interestingly, 74% (n = 61) of offline and 77% (n = 116) of online respondents use some form of technology to conduct their financial transactions, for example, online services accessed via laptop or computer and mobile phone apps (data not shown). Among the 23% (n = 35) of online respondents who did not use technology, 74% (n = 26) chose not to use such services, whereas 14% (n = 5) found them difficult to use. Overall, most online respondents (73%, n = 110) were very or quite comfortable using technology to perform everyday money management tasks. Among the 21% (n = 17) of offline respondents not using technology, about 53% (n = 9) chose not to use such services, whereas 18% (n = 3) did not have easy access to technology or found them difficult to use (12%, n = 2). Compared to online respondents, about half of offline respondents (48%) were very to quite comfortable using technology, with about 28% endorsing a medium level of comfort.

As seen in Table 4, during the past decade, both online and offline respondents affirmed using technology such as mobile phone apps and online services to conduct all of the surveyed financial activities, with technology use being most prominent for transferring funds (average 87%). Except for the offline group’s somewhat limited use of technology to withdraw money (48%), all other financial activities revealed technology usage above 54%.

Table 4.

Percentage of respondents using technology to conduct specific financial activities in the past decade

	Online	Offline
	N (%)
Depositing checks	109 (80%)	37 (54%)
Reviewing bank statements	125 (90%)	53 (75%)
Transferring funds	95 (91%)	53 (83%)
Keeping track of spending money	109 (85%)	56 (73%)
Withdrawing cash	113 (88%)	37 (48%)
Paying bills	107 (79%)	57 (72%)
Purchasing products online	132 (87%)	72 (88%)

Respondents include those who completed the survey online (using M-Turk) and offline (either in-person or by mail). Note. Both number and frequency percent are displayed in the cells.

Note: Percentages have been rounded up.

Specific Financial Habits

Tables 5 and 6 help visualize three characteristics of the data for each financial activity—the nature of the method used to conduct the activity, the duration of use, and the preferred method (split into technological and non-technological methods). Across the six financial activities (note: did not include Purchasing products online), the majority of respondents endorsed the use of technological products since the past year, with such products ranked as their most preferred method.

Table 5.

Frequency of respondents conducting financial activities at specified durations

			Duration (in %)
	Na		< 1 year	1–2 years	3–4 years	5–9 years	> 10 years
Depositing checks
Non-technological
Go to bank	135 (99.3%)	Online	48	19	10	5	19
	60 (88.2%)	Offline	5	0	3	10	82
Technological
Use bank’s mobile phone app	117 (86%)	Online	66	17	12	4	1
	28 (41.2%)	Offline	36	25	18	18	4
Reviewing bank statements
Non-technological
Go to bank	98 (70.5%)	Online	35	18	17	7	22
	8 (11%)	Offline	25	0	0	13	63
Receive via mail	118 (84.9%)	Online	34	25	14	9	18
	43 (58.9%)	Offline	2	7	2	7	81
Technological
Use bank’s online services via laptop or computer	134 (96.4%)	Online	55	10	14	15	5
	48 (65.8%)	Offline	4	17	17	38	25
Use bank’s mobile phone app	125 (89.9%)	Online	57	15	14	10	5
	28 (38.4%)	Offline	29	25	36	11	0
Receive via email	115 (82.7%)	Online	43	23	23	7	4
	20 (27.4%)	Offline	10	15	35	20	20
Keeping track of spending money
Non-technological
Keep paper records	94 (72.9%)	Online	18	13	30	15	24
	35 (44.9%)	Offline	0	0	0	0	100
Receive statements in the mail	103 (79.8%)	Online	27	21	20	13	18
	38 (48.7%)	Offline	0	0	0	0	100
Technological
Use bank’s online services via a laptop or computer	121 (93.8%)	Online	60	14	12	11	2
	51 (65.4%)	Offline	4	18	16	29	33
Use a mobile app	105 (81.4%)	Online	52	13	25	7	3
	26 (33.3%)	Offline	15	23	46	15	0
Use another online web-based program, or have purchased a computer program	75 (58.1%)	Online	41	39	12	5	3
	11 (14.1%)	Offline	9	9	45	18	18
Use spreadsheets	82 (63.6%)	Online	28	18	21	22	11
	13 (16.7%)	Offline	23	8	15	0	54
Transferring funds
Non-technological
Go to bank	81 (77.9%)	Online	32	14	16	15	23
	21 (31.8%)	Offline	5	10	5	24	57
Transfer funds by phone	61 (58.7%)	Online	23	41	18	8	10
	10 (15.2%)	Offline	0	20	20	10	50
Technological
Use bank’s online services via laptop or computer	102 (98.1%)	Online	52	13	22	13	1
	45 (68.2%)	Offline	9	7	24	40	20
Use bank’s mobile phone app	93 (89.4%)	Online	53	17	15	14	1
	31 (47%)	Offline	6	23	39	23	10
Withdrawing money
Non-technological
Go to bank teller	118 (92.2%)	Online	36	24	11	9	20
	36 (45%)	Offline	0	0	3	17	81
Use cashback option	102 (79.7%)	Online	39	25	15	14	8
	37 (46.3%)	Offline	11	8	32	24	24
Technological
Use ATM	126 (98.4%)	Online	59	8	2	8	24
	71 (88.8%)	Offline	3	4	4	11	77
Paying bills
Non-technological
Call company to pay by phone	89 (65.9%)	Online	30	33	15	12	10
	26 (31.7%)	Offline	8	12	19	27	35
Write and mail a check	93 (68.9%)	Online	24	17	20	14	25
	46 (56.1%)	Offline	4	4	0	9	83
Send a check regularly	73 (54.1%)	Online	16	22	23	18	21
	2 (2.45)	Offline	0	0	0	0	100
Technological
Use auto pay	122 (90.4%)	Online	61	11	13	7	8
	44 (53.7%)	Offline	7	18	18	30	27
Use online services via computer	110 (81.5%)	Online	52	14	15	11	8
	42 (51.2%)	Offline	5	7	19	36	33
Use mobile app	93 (68.9%)	Online	57	10	23	9	2
	25 (30.5%)	Offline	12	20	36	24	8

Respondents include those who completed the survey online (using M-Turk) and offline (in person or by mail).

^aThe N represents the number of respondents who responded to that option; the parentheses percentages represent the ratio of the number of those who responded to that option divided by the number of respondents who endorsed doing that specific activity (Table 2). The rows representing the duration add up to 100% based on the number of respondents who selected that option. Across all financial categories, the options “Ask someone to help” and “Use Other method” were grouped as non-technological. The numbers are not shown given the relatively small percentage of respondents endorsing these items.

Table 6.

Respondents’ most preferred way of conducting specific financial activities

	Online	Offline
Depositing check
Non-technological*	70 (53%)	51 (75%)
Go to the bank	59 (45%)	45 (66.2%)
Technological
Use bank’s mobile phone app	62 (47%)	17 (25%)
n	132	68
Reviewing bank statements
Non-technological	37 (27%)	31 (42.5%)
Go to the bank	22 (15.8%)	0 (0%)
Receive it via mail	13 (9.4%)	31 (42.5%)
Technological	100 (73%)	42 (57.5%)
Use bank’s online services via a laptop or computer	51 (36.7%)	27 (37%)
Use bank’s mobile phone app	29 (20.9%)	12 (16.4%)
I receive it via email	20 (14.4%)	4 (5.5%)
n	137	73
Keeping track of spending money
Non-technological	13 (10.2%)	29 (37.7%)
Keep paper records	7 (5.4%)	16 (20.5%)
Receive statements in the mail	2 (1.6%)	11 (14.1%)
Technological	115 (89.8%)	48 (62.3%)
Use bank’s online services via a laptop or computer	57 (44.2%)	26 (33.3%)
Use a mobile app	42 (32.6%)	17 (21.8%)
Use another online web-based program, or have purchased a computer program	8 (6.2%)	3 (3.8%)
Use spreadsheets	8 (6.2%)	4 (5.1%)
n	128	77
Transferring funds
Non-technological	19 (18.4%)	14 (21.5%)
Go to the bank	12 (11.5%)	9 (13.6%)
Transfer funds on the phone	3 (2.9%)	3 (4.5%)
Technological	84 (81.6%)	51 (78.5%)
Use bank’s online services via a laptop or computer	44 (42.3%)	34 (51.5%)
Use bank’s mobile phone app	40 (38.5%)	17 (25.8%)
n	103	65
Withdrawing money
Non-technological	32 (26.2%)	16 (20.3%)
Go to the bank teller	18 (14.1%)	12 (15%)
Use the cashback option	12 (9.4%)	4 (5%)
Technological	90 (73.8%)	63 (79.7%)
Use an ATM
n	122	79
Paying bills
Non-technological	19 (14.3%)	31 (37.8%)
Call the company and pay by phone	5 (3.7%)	5 (6.1%)
Write check mail	9 (6.7%)	22 (26.8%)
Technological	114 (85.7%)	51 (62.2%)
Use auto pay	64 (47.4%)	25 (30.5%)
Use online services via a computer or laptop	29 (21.5%)	21 (25.6%)
Use a mobile app	21 (15.6%)	8 (9.8%)
n	133	82

Respondents include those who completed the survey online (using M-Turk) and offline (in-person or by mail).

^aFew respondents ranked two or more options as 1. The options “Ask someone to help” and “Use Other method” were grouped as non-technological. The numbers are not shown given the relatively small percentage of respondents endorsing these items.

Logistic Regression Results

As shown in Table 7, across financial activities, among various demographic variables, age significantly contributed to all the models, with younger adults having greater preference for using technology. However, when stratified by age, both older (55 years and above) and young or middle-aged adults ranked technological methods as their top preference across all financial categories (except for depositing checks in older adults; see Supplementary material online, Table S1). Compared to all the models using each financial activity as the outcome, the largest amount of variance accounted for by the demographics was for the activity of keeping track of spending money. For this activity, the only significant demographic predictor was age, with young or middle-aged adults being more likely to use technology to keep track of spending than older adults. The explained variance in other activities from the combination of predictors ranged from 11% (Withdraw cash and Purchase products online) to 19% (Review bank statement). With regard to individual predictors, education significantly contributed to two financial activities (Transferring funds and Reviewing bank statements) with higher education related to greater preference for using technology. Gender did not significantly contribute to any model.

Table 7.

Logistic regression coefficients for variables predicting preference for using technological versus non-technological methods across specific financial categories

		B	S.E.	Wald χ2	p	OR	95% CI for odds ratio
							[Lower, Upper]
Depositing Checks
Model χ2 = 18.03***
R2 = 0.12
Age		0.03	0.01	7.92	0.01	1.04	[1.01,1.06]
Education		−0.07	0.04	2.96	0.09	0.93	[0.86,1.01]
Gender		0.13	0.35	0.14	0.71	1.14	[0.57,2.27]
Method		−0.64	0.41	2.48	0.12	0.53	[0.24,1.17]
Constant		0.31	0.94	0.11	0.74	1.36
Reviewing bank statement
Model χ2 = 28.65***
R2 = 0.19
Age		0.05	0.01	12.15	0.00	1.05	[1.02,1.07]
Education		−1.94	0.52	13.82	0.00	0.14	[0.05,0.40]
Gender		0.24	0.38	0.40	0.53	1.27	[0.60,2.69]
Method		−0.26	0.42	0.37	0.54	0.77	[0.34,1.77]
Constant		2.19	1.51	2.11	0.15	8.93
Keeping track of spending money
Model χ2 = 40.49***
R2 = 0.29
Age		0.07	0.02	16.48	0.00	1.07	[1.04,1.11]
Education		−0.05	0.06	0.75	0.39	0.95	[0.85,1.07]
Gender		0.32	0.46	0.49	0.49	1.38	[0.56,3.39]
Method		−1.00	0.49	4.18	0.04	0.37	[0.14,0.96]
Constant		−3.88	1.33	8.57	0.00	0.02
Transferring funds
Model χ2 = 18.46***
R2 = 0.17
Age		0.06	0.02	9.28	0.00	1.06	[1.02,1.10]
Education		−1.84	0.55	11.30	0.00	0.16	[0.05,0.46]
Gender		0.13	0.49	0.07	0.79	1.14	[0.44,3.01]
Method		0.33	0.56	0.36	0.55	1.39	[0.47,4.14]
Constant		0.31	1.62	0.04	0.85	1.36
Withdrawing cash
Model χ2 = 14.76**
R2 = 0.11
Age		0.05	0.01	9.49	0.00	1.05	[1.02,1.08]
Education		−0.08	0.05	2.66	0.10	0.92	[0.84,1.02]
Gender		−0.41	0.40	1.04	0.31	0.67	[0.30,1.45]
Method		1.27	0.49	6.70	0.01	3.56	[1.36,9.31]
Constant		−2.81	1.13	6.13	0.01	0.06
Paying bills
Model χ2 = 24.12***
R2 = 0.17
Age		0.04	0.01	8.84	0.00	1.04	[1.01,1.07]
Education		−0.08	0.05	2.43	0.12	0.92	[0.84,1.02]
Gender		0.21	0.41	0.25	0.62	1.23	[0.55,2.74]
Method		−0.85	0.45	3.60	0.06	0.43	[0.18,1.03]
Constant		−1.76	1.12	2.47	0.12	0.17
Purchasing products
Model χ2 = 13.82**
R2 = 0.11
Age		0.06	0.02	9.86	0.00	1.06	[1.02,1.10]
Education		0.07	0.06	1.19	0.28	1.07	[0.95,1.21]
Gender		−0.09	0.50	0.03	0.86	0.92	[0.34,2.45]
Method		1.17	0.60	3.89	0.05	3.23	[1.01,10.38]
Constant		−6.73	1.54	19.06	0.00	0.00

Note: R² = Nagelkerke R Square value; for all models, df = 4; ^***p < .001, ^**p < .01; Coding for the outcome variable was as follows: technological method = 0, non-technological method = 1; coding for method as a predictor included online respondents = 1, offline respondents = 2; coding for gender as a predictor included female = 0, male = 1.

Discussion

The Financial Habits and Technology survey used in the current study is a first step toward informing researchers and clinicians about the ways in which individuals use technology to conduct financial activities and thus the types of activities that should be considered in assessments of FDM moving forward. Findings from this study will enable the development of assessment tools, such as simulated apps, that are directly related to common financial behaviors as determined through empirical data. Overall results revealed that the majority of the respondents across various ages use technological methods to conduct each of the financial activities surveyed (Tables 4 and 5).

Overall Trends across Financial Activities

Our findings showed that the majority of respondents endorsed engaging in each of the surveyed financial activities, with shopping for products, depositing checks and reviewing bank statements being endorsed by relatively more people, and transferring of funds endorsed by relatively fewer people (Table 2). With regard to the frequency with which individuals engaged in various financial activities (Table 3), 33% (online participants) and 37% (offline participants) kept track of their money on a daily basis. All other surveyed financial activities, such as depositing checks, reviewing bank statements, transferring funds between accounts, withdrawing cash, paying bills, and buying products in store, were performed somewhat less frequently, at least one to three times a month (online and offline participants). Fewer than 4% of the respondents did any task less than one to two times a year. These findings show the recurring but multifaceted schedule of financial activities and reinforce the value of considering the various nuances of FDM in the development of assessment tools.

A common theme related to technology use and its preference emerged from a closer examination of the methods respondents used, their duration of use, and the rank assignments, thus strengthening the idea that researchers and clinicians need to rethink how to examine FDM (Tables 5 and 6). Below, we discuss the results for each financial activity.

Of those who endorsed depositing checks, 99% of online respondents and 88% of offline respondents reported going to the bank to deposit checks. However, 86% (online) and 41% (offline) respondents also endorsed using their bank’s mobile phone app. Regarding duration, the trends reveal that 95% of the online respondents and 79% of offline respondents began using their bank’s mobile phone app in the past 4 years. Regarding preferences, overall, both online (53%) and offline (75%) groups ranked using non-technological methods for depositing checks as the most preferred method. Given the relative ease with which checks can be deposited using a mobile app, this finding is somewhat surprising, especially in the case of the online survey respondents. However, some banks may not offer clients the option of mobile deposit and, if such an option is available, the amount that can be deposited may be limited, making it inconvenient for clients. Future work is needed to examine the reasons underlying the preferences reported in this study.

Among individuals who reported reviewing their bank statements, the bulk of respondents use their bank’s online services (96% of online survey respondents and 66% of offline survey respondents) to do so. Comparatively, only 83% (online survey) and 27% (offline survey) review statements received via email. A glance at the trends for the duration of method used suggests that an increasing number of individuals may be switching to online technology to review bank statements. Specifically, more than half of the online respondents began using their bank’s online services via a laptop or computer (55%) or mobile phone app (57%) to review their bank statements in the past year. For offline respondents, despite 81% receiving bank statements in their mail for more than a decade, a high number reported using technology (mobile app, 89%; laptop or computer, 38%) for the past 4 years. Regarding rank preferences, both groups (73% online; 58% offline) preferred using technological methods to review bank statements. A finer breakdown of the technological methods reveals that both groups (37% each) favored using the bank’s online services via a laptop or computer followed by using the mobile app (21% online; 16% offline) and finally by email (15% online; 5% offline). While speculative, the relatively low preference to review statements via email may reflect attitudes towards web security, wherein bank-based online services may be perceived to be more secure than email, a more generic platform that does not guarantee identity protection or information security. The interactive element of the online services platform, which allows one to navigate and access various kinds of transactions and information, may also appeal to clients.

Keeping track of money was another financial activity examined in the current study. Most respondents endorsed engaging in this activity on a daily or weekly basis and using some form of online technology to do so. For example, banks’ online services were used on a computer by 94% of online respondents and by 65% of offline respondents to keep track of money. While the majority of those completing the survey online also endorsed using the mobile app (81%), the offline respondents also reported using other methods such as relying on receiving statements in the mail. With regards to the duration, mimicking previous trends, a high number of online respondents have begun tracking their money using their bank’s online service (60%) or their mobile app (52%) during the past year compared to using other methods. For the offline respondents, the use of online services has been more prominent with about 63% using it anywhere from 1 to 9 years. Regarding preferences, most respondents (90% online; 62% offline) preferred using technological methods to keep track of money spent. Across both groups, accessing online services either via the computer or their mobile app is ranked as the most preferred method. Thus, this financial activity in particular seems to increasingly rely on technological methods as compared to other financial activities. The finding is not surprising given the quick access to immediate financial activities (e.g., money debited or credited can be viewed within a couple of hours) and interactive services (e.g., filter options, pie charts, categorizations, etc.) afforded by online platforms. Indeed, current findings corroborate results of the Bank of America survey reporting that clients prefer 24/7 access to financial information over proximity to a branch (Accenture North America Consumer Banking Survey, 2015).

An overwhelming majority of online respondents transfer funds. Of those who conduct this activity, the vast majority use their bank’s online services (98%) or their bank’s mobile phone app (89%). Although lower in absolute percentages, a parallel trend was observed among offline respondents, as follows: bank’s online services (68%), bank’s mobile phone app (47%), going to the bank (32%), and transferring via phone (15%). Regarding duration, 52%–53% of online respondents began using their bank’s online service via computer or mobile app in the past year. Among offline respondents, 64% reported using online services via computer for the past 3 to 9 years, and 85% reported using the mobile app for about less than a decade. Regarding rank preferences, overall, most respondents (82% online; 79% offline) prefer to use technological methods. Specifically, the use of bank’s online services (43% online; 53% offline) is the most popular choice.

To withdraw money, compared to other methods, almost all online (98%) and offline (89%) respondents endorsed using the ATM, with the majority of respondents rating it as the most preferred method (74%, online; 80%, offline). To pay bills, most of the online respondents reported using autopay (90%) and/or their bank’s online service via laptop or computer (81%), whereas most offline respondents either write a check and mail it (56%) and/or use autopay (54%). Regarding duration, autopay was reported as a relatively recent method for online respondents (61% started within the past year), whereas this method was adopted earlier among offline respondents (57% beginning at least 5 years ago to more than a decade). Regarding rank preferences, 86% (online) and 62% (offline) preferred using technological methods. Finally, for the final financial activity of purchasing products, approximately 87% (online) and 88% (offline) respondents shop for products online.

Use of Technological Methods as a Function of Demographic Variables

Overall, our findings provide evidence that technology is quickly transforming our financial landscape across multiple financial activities, and people are changing the way in which they manage their money. However, it is important to understand the extent to which trends observed in the current study vary as a function of key demographic features. Current findings suggest that the trends observed for the total sample of respondents did not differ as a function of gender and, for the majority of financial activities, did not differ as a function of education. There were, however, two areas in which education predicted the use of technology. Specifically, education impacted preference for using technology to transfer funds and review bank statements. The association between higher education and increased technology use has been documented previously and is tied closely to higher socioeconomic status, greater need for innovative learning, and higher level of access to resources (Giordani et al., 2014; Jiménez & Díaz, 2019; Polasik & Piotr Wisniewski, 2009). It is interesting that education had selective effects on technology use in the current study depending on the specific financial activity. It may be that for certain activities, individuals with fewer years of education may be more resistant to using technology and/or may find it harder to set it up and use that specific technology for certain activities. This interpretation is highly speculative, however; future studies are needed to further test and understand this dissociation.

In the current study, there was no effect of gender on technology use. Findings are in contrast to some previous papers, which found that the adoption of technology, such as internet banking, was relatively lower in women compared to men (Flavián et al., 2006; Jiménez & Díaz, 2019; Mottola, 2013; Polasik & Piotr Wisniewski, 2009; Yiu et al., 2007). However, it has been hypothesized that previously observed gender effects may reflect an artifact of study design, an interaction with other variables such as education, or historical forces such as salary gaps (Bimber, 2000; Jiménez & Díaz, 2019; Lera-López et al., 2011). It may also be the case that over the years, the gender gap for technology use may be diminishing.

As expected, age significantly affected technology use. Of note, however, when stratified by age, both young or middle-aged and older adults ranked technological methods as the most preferred across most financial activities. Specifically, although the relative frequency of technological method ranked as the top choice was higher in young or middle-aged as compared to older adults, the patterns remained generally comparable (see Supplementary material online, Table S2). Overall, this finding aligns with previous studies (Fox, 2013; Laukkanen et al., 2007; Passyn et al., 2011; Selwyn, 2004; Wagner et al., 2010). Digital divide—the perception of technology such as the internet as being non-essential—is proposed as one of the reasons for use of non-technological methods among older adults (Choudrie et al., 2018). Other reasons include barriers such as limited access, perceived financial risk, monetary costs associated with owning and using technology, sensorimotor problems, decreased motivation, and resistance to breaking away from traditions (Kwon & Noh, 2010; Molesworth & Suortti, 2002). While there are various theories addressing this phenomenon, including the Unified Theory of Acceptance and Use of Technology, Theory of Reasoned Action, Innovation Resistance Theory, the Theory of Planned Behavior, Diffusion of Innovation Theory, and Technology Acceptance Model (Chen & Chan, 2011; Kleijnen et al., 2009; Lian & Yen, 2014; Ram, 1987; Venkatesh et al., 2003), these theories examine the general use of technology across various contexts. It will be useful to develop multiple frameworks regarding the barriers and facilitators to technology use among older adults that are specific to individuals’ financial activities. It can be speculated that unique variables such as the level of cognitive or physical functioning and degree of concern about financial safety may contribute differentially to each framework, along with attitudes or perceptions about technology or preference for human contact when dealing with money (Chiou & Shen, 2012; Mansumitrchai & Chiu, 2012).

Clearly much more research into the role that demographic characteristics play in the context of various financial activities is needed. Such studies will help to tailor the assessment of technology use for specific financial services in the context of the demographic profile of an individual, in turn providing more relevant information to both researchers and clinicians. In addition, the recency of technology adoption and the frequency of use (e.g., using online banking once a month versus several times a week) will also potentially impact an individual’s fluency with the technology.

Conclusions

The current findings give researchers specific directions to develop technological assessments of specific financial activities that are part of everyday FDM. Developing computerized tasks for assessing financial activities such as keeping track of money spent and reviewing bank statements (both frequent and technology-preferred activities) may be critical. For example, an ATM is the most frequently way used to withdraw money, and therefore, researchers can develop specific assessment tools and activities to examine various aspects of an individual’s ability to use the machine and withdraw money. In fact, Czaja, Loewenstein, Sabbag, Curiel, Crocco, and Harvey (2017) developed a simulated ATM task and found that those with mild cognitive impairment (MCI) perform worse on this task (reaction time and accuracy) compared to healthy older adults. Clinicians can thus begin to use sensitive assessments of specific aspects of functioning and to then make referrals to specialists for a comprehensive evaluation to confirm their initial impressions.

Implications

For Researchers

• (1)Currently, given the heterogeneity of financial tasks—both in terms of quantity and quality—no unanimously agreed upon working model of FDM exists. By developing current and ecologically valid computerized tasks of specific abilities comprising FDM, people can be tested across a wide array of different contexts, making the development of such a theory more feasible. For example, people can be examined on their ability to navigate a simulated online banking website (performing a general procedure), their ability to decide from among different investment options (financial judgment), and their confidence across tasks and contexts (financial awareness). Akin to Finucane, Mertz, Slovic, and Schmidt (2005)’s Person–Task Fit Framework and Lichtenberg’s Financial Decisional Abilities Model, an FDM model can be conceptualized as consisting of factors that are task-specific (e.g., transferring money), decision maker-specific (e.g., demographic, intellectual, and experiential) and context-specific (e.g., geographic, economic, and political), which can then be developed and validated on an ongoing basis.
• (2)A standardized and relatively brief computer-based task will allow researchers to collect a relatively large quantity of data from diverse demographic groups and geographic regions, thus helping to comprehensively characterize financial ability across heterogeneous populations (Germine et al., 2018; Miller & Barr, 2017). By applying sophisticated statistical methods such as structural equation modeling to such data, both latent and observed variables can be identified.
• (3)By tracking individuals longitudinally, discriminant function analysis can be performed to identify which predictors classify those at risk for developing neurological conditions (e.g., classify healthy older adults versus those with Alzheimer’s disease [AD]) or impaired FDM.

For Clinicians Including Neuropsychologists and Evaluators in Various Settings

• (1)Neuropsychologists often use a battery of tests to assess various cognitive abilities such as attention, memory, language, executive function, spatial abilities, and processing speed. The interpretation of performance is based on standardized scores (t-scores, z-scores) generally derived from age, education, and other demographically adjusted norms. Although diagnosis of AD versus MCI is based on functional status, few measures exist to objectively and comprehensively measure the integrity of functioning with regard to financial capability (Appelbaum et al., 2016; Sperling et al., 2011). A computerized financial tool, once appropriate normative data are available, can be used by clinicians to objectively measure functioning in this area, as well as to address specific referral questions related to financial management.
• (2)Such a tool could also serve as a resource for identifying training needs and developing training activities, based on the performance tracked by the computer-based simulation. Depending on the profile of performance, clinicians and rehabilitation specialists could work to identify problem areas and then teach effective strategies to combat them. For instance, if an individual were to make errors while setting up a recurring auto-payment schedule, the rehabilitation specialist could work on correcting the errors by using various learning strategies such as shaping, superimposition, and so forth.
• (3)Such a tool could be conveniently used as a screening instrument by banks, adult protective service workers, and others to detect those at risk for making poor financial decisions. While the development and implementation of such tools for clinical use may take a few years, evaluators can begin to benefit by enquiring about an individual’s technology use for financial activities during their current clinical interviews. Table 8 provides some suggestions for questions that can be asked, in the form of a semi-structured interview, during the clinical enquiry of an individual’s instrumental activities of daily living (iADLs).

Table 8.

Suggestions for the kind of questions that can be asked during the clinical interview with the patient

• Do you use technology-based services such as online banking to conduct financial tasks?

• For what kind of activities do you use such technology (i.e., depositing checks, reviewing back statements, keeping track of spending money, transferring funds, withdrawing cash, paying bills and purchasing products)?

• Please explain the kind of technologies you use to conduct the task (i.e., website, app or ATM).

• When did you begin to use this technology?

• How often do you use it?

• Do you feel comfortable using the technology?

• Do you experience any difficulty while using this technology? If yes, please explain.

Other consideration during the interview:

It will be important to determine how technology is used in terms of:

• the amount of money (e.g., $100 versus $100,000) used for different types of financial activities, and

• the frequency and types of transactions (e.g., weekly, low stakes transactions such as using Paypal to pay a friend ~$20, for example, versus high stakes transaction such as paying multiple, monthly bills involving mortgages, utility bills, etc).

Note that the questions should be tailored depending on the specific patient responses.

For Individuals and their Caregivers

(1) After obtaining their performance report from the clinician, individuals can gain knowledge about their strengths and weaknesses. Assessments can also serve to track FDM over time.

(2) Depending on the level of performance by the individual, caregivers in collaboration with that individual, financial institutes, and attorneys could then make an informed determination about how money should be handled in the future, including deciding on stipulations of power of attorney and the creation of a will.

In summary, our survey data provide empirical evidence for the changing nature of our financial behavior and the need to upgrade methods of evaluating FDM. Based on current data, researchers and clinicians may wish to conceptualize FDM while considering the frequency with which certain financial tasks are conducted and the implications of what it may mean for an individual to decrease the frequency of or cease that activity altogether. From the current study, researchers may begin to distinguish between everyday financial tasks that require relatively straightforward but frequent monitoring by individuals (keeping track of money via online banking services) versus those done less routinely but which require decision-making and relatively more complex technical skills (buying products online by comparing them across websites). Given the changing pattern of financial habits, clinicians may benefit not only from asking if patients perform critical functions like managing finances but querying the specific functions that are performed (i.e., keeping track of money spent) and how they are performed (e.g., mobile app). The need to incorporate changes to keep up with the fluid nature of everyday FDM is now more apparent than ever.

Funding

This work was supported by the National Institutes of Health’s Ruth L. Kirschstein Predoctoral Individual National Research Service Award (5F32AG053035–02) and by the National Academy of Neuropsychology Clinical Research Grant.