DIFFERENCES IN U.S. MOBILE PHONE USE TO ACCESS ONLINE HEALTH - - PowerPoint PPT Presentation

THE DIGITAL DIVIDE: RACIAL AND ETHNIC DIFFERENCES IN U.S. MOBILE PHONE USE TO ACCESS ONLINE HEALTH INFORMATION.

Nancy Pontes, PhD, Rutgers University, Camden Manuel Pontes, PhD, Rowan University

The Digital Divide: Access to Internet

 The Digital Divide:

 Minorities and Persons with Lower Income are Less Likely to

Have Internet Access

 As A Result, they Have Less Access to Information and

Knowledge

 Minorities and Use of Mobile Devices

 Univariate Analyses Showed that Minorities are More Likely

to Use Mobile Devices to Access Online Health Information (Fox, 2012).

 Note: Hispanics are Younger than Non Hispanic Whites  Greater Mobile Use Among Hispanics May be Confounded

with Age (Hispanics More Likely to Use Mobile Devices Because they are Younger)

Pew Internet and Health Survey

 Pew Internet and American Life Project  A series of nationally representative (USA) telephone

surveys sponsored by the Pew Foundation.

 The goal of these surveys is to understand how US adults

use the Internet

 Pew Internet and Health Survey (n=3,014)

 Nationally representative (USA) survey of adults (18 years

• r older)

 Conducted by Princeton Associates  Sampling weights provided.  Need to use software that incorporates sampling weights for

analyses.

Use of R and survey package for analyses

 R (R Core Team 2015)

 Open source software for statistical analyses.  Thousands of add on packages for specialized

analyses

 Survey package (Lumley 2004, 2014)

 Used for analyses of survey data.  Can be used for data with sampling weights  Can be used for data from complex (multistage)

sampling designs.

 R and survey package were used for all analyses

Smartphones and mobile health

 Smartphones are changing the face of mobile

and participatory healthcare (Boulos, M. N., Wheeler, S., Tavares, C., & Jones, R. ,2011).

 Racial and ethnic minorities have smartphone

• wnership rates that are comparable to whites.

 Mobile devices may help bridge the digital

divide.

Logistic Regression Model: Independent variables



Race/Ethnicity

 Non-Hispanic Black  Hispanic  Non-Hispanic All Other  Non-Hispanic White (Ref)



Gender

 Male  Female (Ref)



Househod Income

 $50,000 or more  $ 0 – 49,000 (Ref)



Education

 Yes  No (Ref)



Age

 50 years or older  35 – 49 years  18 – 34 years (Ref)

Logistic Regression Model: Dependent variables



Used Smart Phone

 Yes  No (Ref)



Accessed Online Health Information (Any Device)

 Yes  No (Ref)



Accessed Online Health Information (Mobile Device)

 Yes  No (Ref)



Accessed Online Health Information (No Use of Mobile Device)

 Yes  No (Ref)



Had Health Apps on Phone

 Yes  No (Ref)

Table 1: Age Distribution of US Cell Phone Users by Race/Ethnicity (2012) NH= Non Hispanic, %=percentage of adult cell phone owners within race/ethnicity, SE=standard error of estimate, significance: *=p<0.05, **=p<0.01.

Table 2A: Logistic Regression – Used Smart Phone

% of adult cell phone users who used a smartphone, SE=standard error, t (M)= t statistic, OR=odds ratio, 95% CI=95% confidence interval, significance: *=p<0.05, **=p<0.01

Table 2B: Logistic Regression – Used Smart Phone.

 % of adult cell phone users who used a

smartphone, SE=standard error, t (M)= t statistic, OR=odds ratio, 95% CI=95% confidence interval, significance: *=p<0.05, **=p<0.01

% of adult cell phone users who used a smartphone, SE=standard error, t (M)= t statistic, OR=odds ratio, 95% CI=95% confidence interval, significance: *=p<0.05, **=p<0.01

Table 3A: Logistic Regression – Accessed Online Health Info: Any Device

% =percentage of adult cell phone users who accessed online health information, significance: *=p<0.05, **=p<0.01.

Table 3B: Logistic Regression – Accessed Online Health Info: Any Device

Table 4A: Logistic Regression – Accessed Online Health Info: Mobile Device

Table 4B: Logistic Regression – Accessed Online Health Info: Mobile Device

Table 5A: Logistic Regression – Accessed Online Health Info: No Mobile Device Use

Table 5B: Logistic Regression – Accessed Online Health Info: No Mobile Device Use

Table 6A: Logistic Regression – Had Health Apps on Mobile Device

Table 6B: Logistic Regression – Had Health Apps on Mobile Device

Key Findings (Univariate)

1.

Hispanics, and other minorites significantly more likely than non Hispanic whites to use smartphones.

2.

Hispanics, non-Hispanic blacks, and other minorities significantly more likely than non-Hispanic whites to access online health information through mobile devices.

3.

Hispanics, non-Hispanic blacks, and other minorities significantly less likely than non-Hispanic whites to access online health information with no use of mobile devices.

4.

Significant age difference (especially 50+ years) in the likelihood that persons access online health info with a mobile device

5.

No significant age difference in the likelihood that persons access online health info without any use of mobile deviceHispanics significantly more likely than non Hispanic whites to use smartphones. Note: Hispanics, and other minorities, significantly younger than Non-Hispanic whites

Key Findings (Multivariate)

After controlling for age, income and education, and gender

1.

Hispanics, non-Hispanic blacks, and other minorites significantly more likely than non Hispanic whites to use smartphones.

2.

Hispanics, non-Hispanic blacks, and other minorities significantly more likely than non-Hispanic whites to access online health information through mobile devices.

3.

Hispanics, non-Hispanic blacks, and other minorities significantly less likely than non-Hispanic whites to access online health information without no use of mobile devices.

4.

Younger adults (18-34 years), significantly more likely than other adults access online health info with a mobile device

5.

No significant age difference in the likelihood that persons access

• nline health info without any use of mobile device

Implications for Practice & Future Research

Implications for Practice

 Create Mobile-Friendly Web Sites  Websites that recognize whether viewer is using a

smart phone (or other mobile device) and format webpage for device.

 Width of webpage fits device

 Further research

 Race/Ethnicity Differences  Type of Information Search  Extent of Information Search  Quality of Information Search

Limitations of Study

 Self Report Measures of Online Health

Information Search

 Dependent variables were based on self-report by

respondents

 Do not measure amount of information search  Do not measure quality of information search

References

Boulos, M. N., Wheeler, S., Tavares, C., & Jones, R. (2011). K, et al. “How smartphones are changing the face of mobile and participatory healthcare: an overview, with example from eCAALYX. Biomedical engineering online, 10 (1), 24. Fox, S., and Duggan, M. (2012). Mobile health 2012. Washington, DC: Pew Internet A American Life Project. Lumley, T. (2004) Analysis of complex survey samples. Journal of Statistical Software 9(1): 1-19 Lumley, T. (2014) “Survey: analysis of complex survey samples". R package version 3.30. R Core Team (2015). R: A language and environment for statistical computing. R Foundation for Statistical Computing, Vienna, Austria. URL http://www.R-project.org/.

QUESTIONS?