Page 1 of 6

Archives of Business Research – Vol. 9, No. 7

Publication Date: July 25, 2021

DOI:10.14738/abr.97.10549. Belhadjali, M., Abbasi, S., & Whaley, G. (2021). Personal Information Privacy: Some Findings on Gender Difference. Archives of

Business Research, 9(7). 95-100.

Services for Science and Education – United Kingdom

Personal Information Privacy: Some Findings on Gender

Difference

Moncef Belhadjali

Norfolk State University

Sami Abbasi

Norfolk State University

Gary Whaley

Norfolk State University

ABSTRACT

The implementation of effective cybersecurity by organizations is a prerequisite to

privacy protection for personal information collected, used, stored, and shared

online. The trend for the potential of online privacy breaches has been moving

upward with our daily reliance on the Internet and cloud computing. While online,

individuals may choose to use a credit card to complete a transaction, access email,

access social media sites, and store pictures through a cloud storage. In some cases,

law enforcement agencies may access and use personal information stored online.

Do individuals approve of the usage of their personal information by these agencies

to solve crimes? Do demographic characteristics such a gender, education, and age

provide a reliable set of predictors for the probability of approval? Do females and

males differ with respect to the decision to approve information usage to solve

crimes? This study reports on the analysis of data from a 2019 Pew Research Center

survey of 1,365 individuals in the USA. Most respondents (63%) approve of

personal information usage by law enforcement agencies to solve crimes. The

purpose of this study is to determine the trend in the citizens’ approval for personal

information usage by law enforcement agencies, especially distinguishing the

genders. The results of a regression analysis showed that the demographic

variables -gender, education, and age- provide a statistically significant power to

predict the probability for information usage approval. A t-Test revealed that there

is a statistically significant difference between genders. Females are more likely to

offer the approval.

Keywords: Privacy; personal information; gender; law enforcement; Pew research

INTRODUCTION

Information privacy is about citizens’ right to be aware of how their personal information is

collected, used, stored, and shared. McKinsey conducted a survey of 1,000 consumers in North

America to find out what they think about their privacy and the collection of consumer data by

companies. The results uncovered that, consumers are becoming increasingly conscious about

what types of data they share—and with whom [1]. Also, that consumers are most comfortable

sharing data with healthcare providers and financial services -44% score-, compared to the

Page 2 of 6

96

Archives of Business Research (ABR) Vol. 9, Issue 7, July-2021

Services for Science and Education – United Kingdom

public sector and the government -11% score-. In addition, 71% of the respondents would stop

doing business with companies that share their data without their permission.

Although individuals are concerned about their privacy, they still share a great deal of personal

information online, especially through social media. To investigate this phenomenon, a study

proposed the “rational fatalism theory” [7]. They found that individuals with higher level of

fatalistic belief about business and technology are less likely to protect their privacy. Basically,

since they believe that they have no control over the outcome, therefore, risk becomes rational.

In a study published by the Pew Research Center, the authors reported that only 24% of the

American citizens surveyed know that private browsing only hides online activities from the

users of the same computers, and not from the website and the internet provider [5].

For any citizen, a major concern about unauthorized access to data may be summarized in the

following statement “Today, however, the biggest risk to our privacy and our security has

become the threat of unintended inferences, due to the power of increasingly widespread

machine learning techniques. Once we generate data, anyone who possesses enough of it can

be a threat, posing new dangers to both our privacy and our security.” [2]. One example of such

unintended inferences would be using online search history regarding a medical condition to

conclude that that a person might be suffering from a certain disease.

On social network sites, females displayed higher privacy concerns and behaviour than did

males [4]. Studies on gender and age for Facebook users found that females and younger users

were more concerned about the privacy of sharing their photos [3]. In a study of students in

higher education to elicit their opinion about privacy and personal data collection in online

courses, the authors found that students were aware that their learning behaviours were being

recorded. However, students were not concerned that the data collected about them were used

to improve learning. In addition, the results of the study revealed that there was no significant

difference between the students’ gender with respect to their concern about privacy [6]. The

purpose of this current study is to utilize data from a survey of American citizens to compare

the decision of females and males to accept that their personal information is accessed and used

by law enforcement agencies to solve crimes.

METHODOLOGY

Data

The data used in the study was obtained from the Pew Research Center. The data was collected

via an online survey developed by the Pew researchers and administered during the period of

June 3, 2019 to June 17, 2019. The respondents are members of the American Trends Panel

(ATP), a pool of US citizens of age 18 and older. The total responses were 4,272, however, this

current study used 1365 cases only. This is due to eliminating records with missing or

incomplete data (i.e. respondent refused to answer) for the variables selected. For the purpose

of this paper, four variables were selected as being of interest to the researchers. One outcome

variable relates to the approval for usage of personal information by law enforcement agencies

to help solve crimes. Three demographic variables were also selected, they are gender,

education, and age. Many researchers consider these three variables as relevant in a personal

decision context. Next, all records with missing values were eliminated. Finally, the dataset was

reduced to 1365 records. Table 1 provides a snapshot of the final dataset used in this study.

Page 3 of 6

97

Belhadjali, M., Abbasi, S., & Whaley, G. (2021). Personal Information Privacy: Some Findings on Gender Difference. Archives of Business Research,

9(7). 95-100.

URL: http://dx.doi.org/10.14738/abr.97.10549

Table 1. Data by gender, education, and age

College Graduate (CLG) Some College (SCL) High School or Less

(HSL)

Totals

18to2

9

30to4

9

50plu

s

18to2

9

30to4

9

50plu

s

18to2

9

30to4

9

50plu

s

Femal

e

41 85 117 39 63 98 49 83 160 735

(54%)

Male 33 73 165 18 54 99 24 39 125 630

(46%)

Totals 74

(5%)

158

(12%)

282

(21%)

57

(4%)

117

(9%)

197

(14%)

73

(5%)

122

(9%)

285

(21%)

1365

(100%

)

Age brackets: 18to29; 30to49; 50plus

Table 1 above shows that females make up the majority (54%) of the respondents in this

sample, with 46% males. For education, college graduates (38%), some college (27%), and high

school or less (35%). For age, 18to29 (14%), 30to49 (30%), and 50plus (56%). Although for

education the three categories are almost equally represented, for the variable age, the majority

of respondents are 50 or older.

Analysis and Results

Table 2 below reports on the number of respondents (n out N) who approve of the law

enforcement agencies’ usage of their personal information to help solve crimes. That is, 63%

consider this type of information usage as acceptable (n = 857 out of N = 1365). For each

combination of education, gender, and age, two quantities –the number of respondents (N) and

the frequency (n)- are needed to estimate the population proportion of those who consider this

type of usage as acceptable, in each category. For example, among the respondents (Female,

CLG, 18to29), 61% (25 out of 41) consider this type of information usage as acceptable.

However, for (Male, CLG, 18to29), 39% (13 out of 33) consider this type of information usage

as acceptable.

Table 2. Usage of personal information by law enforcement agencies to help solve crimes:

Acceptable

Female Male

18to29 30to49 50plus 18to29 30to49 50plus

n N n N n N n N n N n N

CLG 25

(61%)

41 53

(62%)

85 81

(69%)

117 13

(39%)

33 32

(44%)

73 96

(58%)

165

SCL 17

(44%)

39 40

(63%)

63 73

(74%)

98 9

(50%)

18 22

(41%)

54 62

(63%)

99

HSL 30

(61%)

49 61

(73%)

83 126

(79%)

160 10

(42%)

24 26

(67%)

39 81

(65%)

125

The dimensions of Tables 2 are determined by the number of independent variables (or factors)

and the number of categories (or factor levels) per variable. With three categories of education

(CLG, SCL, HSL), three categories of age brackets (18to29, 30to49, 50plus), and two categories

of gender (Female, Male), we obtain 3 x 3 x 2 = 18 cells. A general observation from Table 2

Page 4 of 6

98

Archives of Business Research (ABR) Vol. 9, Issue 7, July-2021

Services for Science and Education – United Kingdom

above is that when comparing the proportions in the 9 cells under Female to those under Male,

all proportions under Female are higher than those under Male, except one for (SCL, 18to29).

To analyse the data, we built a regression model with categorical independent variables. Also,

we used a dummy-variable coding for the categories corresponding to education, age, and

gender. The reference categories are education (HSL), age bracket (50plus), and gender (Male).

The resulting model can be written as:

P = b0 + b1CLG + b2SCL + b318to29 + b430to49 + b5Female

Where P is the probability of approving of personal information usage controlling for education,

age, and gender. The parameter b0 is the intercept, and the remaining coefficients (b1... b5)

represent the effects of education, age, and gender.

A regression analysis run on the data of Table 2 using the model above, provided the following:

P = 0.674 - 0.090*CLG - 0.087*SCL - 0.185*18to29 - 0.097*30to49 + 0.130*Female

SPSS Output

Model Summary

Model R R Square

Adjusted R

Square

Std. Error of

the Estimate

1 .897a .804 .722 .06531

a. Predictors: (Constant), Female, A30to49, SCL, A18to29, CLG

ANOVAa

Model Sum of Squares df Mean Square F Sig.

1 Regression .210 5 .042 9.848 .001b

Residual .051 12 .004

Total .261 17

a. Dependent Variable: PL

b. Predictors: (Constant), Female, A30to49, SCL, A18to29, CLG

Coefficientsa

Model

Unstandardized Coefficients

Standardized

Coefficients

B Std. Error Beta t Sig.

1 (Constant) .674 .038 17.871 .000

CLG -.090 .038 -.352 -2.387 .034

SCL -.087 .038 -.339 -2.298 .040

A18to29 -.185 .038 -.724 -4.906 .000

A30to49 -.097 .038 -.378 -2.564 .025

Female .130 .031 .540 4.222 .001

a. Dependent Variable: PL

DISCUSSION

The results shown above reflect the marginal effects of the categorical variables representing

education, age, and gender on the dependent variable P representing the probability of