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