International Journal of Technology and Management

Cognitive Factors and Health Behaviors of Social Media IJOTM

Users In Sub-Saharan Africa ISSN 2518-8623

Volume 6. Issue I
Kituyi G. Mayoka pp. 1-21, July 2021
Makerere University Business School ijotm.utamu.ac.ug
Email: [email protected] email: [email protected]

Clive K. Tsuma
ICT University
Email: [email protected]

Victor Mbarika
East Carolina University
Email: [email protected]

Abstract
This study investigated the mediation effect of Locus of Control, Self-Regulation, Age sensitivity and
Behavioral Intention in the relationship between Cognitive Factors and Health Behaviors of social media
users in Sub-Sahara Africa.

A positivist epistemology was adopted in trying to create knowledge from empirical and observational
evidences across Sub-Saharan Africa. Data were collected from 450 social media users in Uganda, Cameroon
and Nigeria. Descriptive statistics, correlation and regression analysis methods were used to analyze the data.
Baron and Kenny procedure and Medgraphs by Jose were used to test for mediation.

Findings reveal that Internal Locus of Control and Behavioral Intention mediate the relationship between
Cognitive Factors and Health Behavior; Self-Regulation mediates the relationship between Cognitive Factors
and External Locus of Control; External Locus of Control mediates the relationship between Self-Regulation
and Health Behavior; Self-Regulation mediates the relationship between Cognitive Factors and Health
Behavior; Self-Regulation mediates the relationship between Age Sensitivity and Health Behavior. The
findings however reveal that External Locus of Control does not mediates the relationship between Cognitive
Factors and Health Behavior.

In order to promote the learning of positive health behaviors amongst social media users in Sub-Saharan
Africa, Self-Regulation and External Locus of Control should be enhanced.
1
Key words
Internal Locus of Control; External Locus of Control; Self-Regulation; Age Sensitivity; Behavioral Intention;
Cognitive Factors; Health Behaviors; Social Media

Introduction
Technology has become a driving force in almost every aspect of life. Innovations that address the challenges
and the wellbeing of man are sprouting up – many of these in the area of Information Communication
Technology (ICT). ICT is the umbrella word for all elements of communication equipment (Rouse, 2014). It
includes hardware, software and liveware that operate to accept and process data, store and disseminate
information to the users. These include – but not limited to computer hardware and software, radios, phones,
televisions and communications networks. ICTs are being used in all business disciplines across the globe.
Different ICT applications have been developed for use in areas such as education - for the case of e-learning
and education management systems, healthcare such as hospital information management systems,
telemedicine applications, and general communications such as Social Media (SM) and social networks,
telecommunications, among others (Hilliard, 2012).

There are different categories of users engaged on these media including young and old professionals,
marketers, politicians, spiritual leaders, and educationists. Anyone can create or consume information as long
as they have access to the internet. Abbasi et al. (2016) posits that social media platforms have transformed
information consumers into information producers. A lot of the social media information is health-related.
The PWC Health Research Institute (2012) posits that people in the United Stated of America (USA) are
increasingly using social media to exchange health-related information. Patients in most developed countries
and some developing countries are able to access health information, share their health problems and get
assistance via some kind of social media. The active engagement, instant sharing and consumption of
information via social media is influencing behavior in various forms. Even those behaviors that are not
necessarily originating from a health perspective may in the long run have health bearings on the consumers.
Overtime, behaviors are being learned and unlearned (Abbasi et al. 2016; Cui, 2016). Some of these behaviors
influence the health of consumers which this study proposes to examine. This study sought to investigate the
mediation effect of External Locus of Control, Internal Locus of Control, Self-Regulation, and Behavioral
Intention in the relationship between Cognitive Factors and Health Behavior of social media users in Sub-
Sahara Africa.

1. Literature Review
1.2 Social media and health behavior in developed countries
Though relatively new compared to traditional approaches, social media usage in health communication is
being applied across the developed world to solve medical and public health problems. For example, a study
conducted by Wright (2016) on social networks and health support services revealed that social media was a
key tool in providing interpersonal social support for positive health outcomes in communities. Adewuyi and
Adefemi (2016) further show that social media were being used by health professionals to influence health
2
behaviors across nations. They however, caution about usage of social media without proper strategies as it
could lead to negative health behavioral outcomes and called for active user involvement of relevant
stakeholders in designing of health communication strategies for social media. Newbold and Campos (2011)
argue that whereas usage of traditional channels for communicating public health information had proved
effective in transforming health behavior, little empirical evidence existed for the case of social media, thereby
making it difficult to establish with certainty the efficacy of social media in health behavioral change.

Recent literature reveals that the impact of social media on health behavior in developed countries was
immense. It is both negative and positive and is most outstanding amongst teenagers and youth who were the
most vulnerable groups to social media. These groups were the biggest users of social media. Carroll and
Kirkpatrick (2011) reveal that whereas social media helped teenagers to create useful networks and develop
their identities, sexuality and self-esteem in the USA. The technology exposed teenagers to various risks.
Teens who regularly used social media stood high chances of developing mental health problems, due to peer
rejection and cyber bullying (Boyar, 2010) leading to depression and stress. According to (Salathé et al. 2017),
social media was a practical avenue where individuals got exposed to negative sentiments. This is supported by
Rideout (2010) who established that regular social media users exhibited signs of low esteem, dissatisfaction
with their physical and material abilities, unhappiness and boredom (Boyar, 2010).

Scholarly initiatives have been tried to address the challenges being brought in by social media in developed
countries. For example, studies were conducted to see how best social media could be used without causing
the negative effects to society (e.g., Adewuyi & Adefemi, 2016; Centola, 2013; Korda & Itani, 2011; Centola,
2010; Smith & Christakis, 2008). Newbold and Campos (2011) postulate that social was showing great
potential and could become a tool of choice for health communication in the near future given its growing
wide usage. Therefore, there was a need for development of standards, policies and legal frameworks to foster
best practices in the application of social media in health communication (Newbold & Campos, 2011).

1.3 Theoretical framework

This study was grounded on the Social cognitive theory of Albert Bandura which posits that Cognitive
Factors play a big role in causing behavioral change through the process of reciprocal determinism with other
constructs (Bandura, 1986). In addition, we borrow the constructs of Internal Locus of Control and External
Locus of Control that were identified as important factors in influencing the learning of new behaviors by
Rotter in his Social Learning Theory (Rotter, 1966). Another hypothetical addition to the study is the role of
age in moderating behavioral learning process as had been proposed by researchers in the field of psychology.
In this Age Sensitivity is hypothesized to moderate the relationship between Cognitive Factors and Self-
Regulation. Individuals’ level of age sensitivity either increases or reduces the influence of their Cognitive
Factors on Self-Regulation in the behavioral learning process. Further, studies by Bandura (1986; 1988), and
Rotter (1966) show that locus of control influence behavior through increasing or decreasing the learner’s
behavioral intention. However, Rotter (1966) cautions that locus of control is relevant only in a responsive
environment. Hence, in this study, we hypothesize that both internal and external locus of control influence

3
behavioral intention of social media users. Where there is locus of control, social media users are more open to
adopt new observed behaviors from elements in the environment and the reverse is true.

The other constructs hypothesized in this study are Self-Regulation (Blalock et al. 2016; Bandura 1988;
1986) and Behavioral Intention (Venkatesh el al. (2003). In this study, Cognitive Factors positively influence
Self-Regulation while Self-Regulation is also hypothesized to have a positive influence on the External Locus
of Control as well as Health Behavior. On the other hand, Behavioral Intention is hypothesized to positively
influence Health Behavior as well as mediate the relationship between Internal Locus of Control and Health
Behavior. Figure 1 presents the conceptual framework.

Figure 1: Conceptual framework (Source: Bandura, 1965; Rotter, 1966; Bandura, 1986;
Venkatesh et al. 2003)

2.3 Research hypotheses

The study tested six hypotheses stated below:
H1: Internal Locus of Control and Behavioral Intention mediate the relationship between Cognitive Factors and
Health Behavior of social media users in Sub-Sahara Africa.
H2: Self-Regulation mediates the relationship between Cognitive Factors and External Locus of Control of social
media users in Sub-Sahara Africa.
H3: External Locus of Control mediates the relationship between Self-Regulation and Health Behavior of social
media users in Sub-Sahara Africa.
H4: Self-Regulation mediates the relationship between Cognitive Factors and Health Behavior of social media users
in Sub-Sahara Africa.

4
H5: External Locus of Control mediates the relationship between Cognitive Factors and Health Behavior of social
media users in Sub-Sahara Africa.
H6: Self-Regulation mediates the relationship between Age Sensitivity and Health Behavior of social media users in
Sub-Sahara Africa.

2. Methodology
2.2 Research epistemological orientation and design
Tennis (2008) defines epistemology as “how we know”. It is a term used to describe the process of discovering
knowledge through scientific methods. It is used to validate justifications about truths and beliefs. It centers
on philosophical analysis, problems of skepticism, sources of information, and mechanisms for validating new
knowledge. Dretske (2016) argues that epistemology is one effective way of doing cognitive science since it
helps researchers to understand how knowledge is sourced, processed and extended. It encompasses critical
reasoning also known as critical realism (Hjørland, 2004).

Epistemology helps scholars to differentiate knowledge from beliefs, truths and justifications. There are cases
where truths held in society about something transform into knowledge. However, such knowledge is obscure
without justifications. This is because the beliefs may be mistaken and untrue. Therefore, for a belief to
become knowledge, it must be true and justifiable.

Currently, there are beliefs that social media influences Health Behavior of its users. It is also true that social
media are influence behavior of its users. However, these beliefs and truths are unsubstantiated and
unsupported because there is no empirical justification. Hence, in this study, the main aim will be to justify
these beliefs and truths about the influence of social media on Health Behavior. These justifications can then
help us build knowledge through development a theoretical model explaining the relevant factors and
constructs responsible for social media and health behavioral change.

Although little literature exists on the term, different schools of thought have proposed various types of
epistemology as; positivist epistemology – which posits that scientific knowledge should emerge from
observable scientific evidence (IGI, 2016); genetic epistemology – which is a study aimed at understanding
the original sources of knowledge (Wikipedia, 2016a), or the genesis of knowledge; feminist epistemology -
which is a study of knowledge from a feminist point of view (Wikipedia, 2016b); social epistemology - which
construes knowledge as a collective work of society (Wikipedia, 2016c); constructivist epistemology - which is
a study of natural science embedded in and can be explained by mental constructs of nature by measuring
them (Wikipedia, 2016d).

This study used a positivist epistemology in trying to create knowledge from empirical and observational
evidences across Sub-Saharan Africa region. The type of epistemology assumes that the researcher is
independent of the study. This helps to alleviate biases. Hence it is the most appropriate epistemological

5
orientation for the current study. Further, the study adopted a cross-sectional research design where data were
collected at one point in time.

2.3 Study population and sample

According to Internet World Statistics (2015), Sub-Saharan Africa has a population of 943,445,548 people.
However, given that this study is targeting SM users, the general population cannot work as our study
population. This is because out of the 943,445,548 people, only 234,342,776 people use the internet and
only 68,968,500 people use Facebook (Internet World Statistics, 2015). Further, given that Facebook alone
cannot represent SM (as had been defined in chapter two), we were unable to use the total number of
Facebook users as our study population. In this case therefore, the actual study population remained
unknown. Generally, however, as of the year 2013, the population of Cameroon was 22 million people; the
population of Nigeria was 174 million people and the population of Uganda was 38 million people. Hence, if
we were to look at the entire population with no regard to social media usage, the study population would
have been 234m people. However, we could not use country populations because this study targeted social
media users only.

Given that the study population was unknown, a formula for determining sample sizes using population
proportion (Hyde, 2017) was used with the following assumptions:

¾ Population proportion = 0.5

¾ Margin of error = 4%
¾ Confidence level = 95% (z=1.96)

Determining sample for each country

Where;
p = population proportion / percentage picking a choice of 50% (0.5)
z = standard normal deviation set at 95% confidence level (1.96)
E = Estimate / Margin of error

6
Hence total sample for all three countries = 152 x 3 = 453

2.4 Mediator variables

A mediator variable is one that connects a linkage between the independent and dependent variables. It is a
third variable influencing a zero-order correlation (Statistics Solution, 2016; Baron & Kenny, 1986).
Mediator variables in this study are; External Locus of Control, Internal Locus of Control, Self-Regulation
and Behavioral Intention.

2.5 Measurement and operationalization of variables

Each variable and construct were measured by a set of items evaluated on a 5-point likert scale. The
operationalization of each variable and the theories used to measure it are discussed below:

Cognitive Factors are attributes unique to an individual such as knowledge, beliefs and attitudes that help in
the learning process. This variable was structured into three constructs namely Knowledge, Beliefs and
Attitudes. Knowledge investigated whether respondents had the necessary knowledge, skills and experience for
using the technology as well as examining whether respondents knew the benefits of using social for health-
related purposes.

On the other hand, the construct of Beliefs was used to investigate respondents’ beliefs, cultural norms,
religion and traditions allowed them to use social media for purposes of sharing and consuming health related
information via social media.

Similarly, the construct of Attitudes was used to study the respondents’ attitudes towards using social media
to access and consume health related information. The relevant theories and literature that informed
Cognitive factors as a variable were Bayrón (2013) and Bandura (1986).

Internal Locus of Control is used to refer to an individual’s locus of control or state of being where one
controls the consequences of his / her behavior. This variable was used to investigate whether respondents
controlled the consequences of their actions while using social media. The variable also investigated whether
respondents maintained interpersonal relations, made greater efforts to learn, were in in charge of my
activities and assumed success or failure before taking action while using social media. These were proposed
by Boundless (2016) and Rotter (1966).

On the other hand, External Locus of Control is used to refer to an individual’s locus of control or state of
being where one is unable to controls the consequences of his / her behavior. Consistent with Boundless
(2016) and Rotter (1966), External Locus of Control variable was employed to examine whether respondents
were not in control of the consequences, achieve less, had low morale to learn, did not maintain good
7
relations, considered themselves lucky, were not responsible for the bad things that happen to them, did not
think about the consequences of their actions before doing them, and if the respondents were unable to help
themselves when faced with challenging situations while using social media.

Self-Regulation postulates the extent to which an individual internally controls himself or herself in the face of
imitation and observation of new behavior. Self-Regulation was measured by Blalock et al. (2016), Winters et
al. (2003) and Bandura (1988; 1986). The variable investigated among others whether respondents freely
brought up a health-related issue to their peers, friends and family, freely disclosed their health-related
problems to their peers, friends and family, freely challenged their peers, friends and family on health-related
matters via social media.

Age Sensitivity is used to refer to the feelings an individual has towards people of different age groups. This
variable was measured by literature from WHO (2000) and NIHCE (2007) and aimed to investigate whether
respondents freely interacted with people of different age groups via social media on health-related matters,
freely interacted with people of their age group via social media on health-related matters, and also whether
the respondents did do not mind learning new health behaviors from people of different ages as well as their
own age group via social media.

Behavioral Intention is used to show the likelihood that an individual or community will learn new behaviors.
It was measured by Venkatesh el al. (2003). This variable was employed to examine whether respondents
intended to acquire new health skills, knowledge and practices via social media.

Finally, the variable Health Behavior is used to refer to learned action, skills, practices an individual does that
influence his wellbeing in terms of health. It was measured by Bandura (1986); Blalock et al. (2016); Winett
et al. (1999); Bandura (1990); Blalock et al. (2016); Kane (2004). This variable was structures into four
constructs namely; Skills, Practice, Observational learning, and Moral degeneration.

Skills construct was used to study whether respondents learned new health related skills by using social media.
It also investigated whether respondents were able to treat diseases, manage chronic diseases and also whether
they learned how to look after patients using information obtained via social media.

The construct of Practice was used to study whether respondents learned new health related behavioral
practices through observing role models and training themselves via social media.

Observational learning construct on the other hand was used to establish whether respondents learned new
health related behaviors by observing other influential people in society such as celebrities, political leaders,
elders, religious do them.

Lastly, Moral degeneration construct was used to study whether respondents learned new health related
behaviors that decayed their morals by using social media. It investigated whether respondents learned how to
8
and actually smoked, used drugs, drunk alcohol, consumed pornography, became gay, and had sex with
multiple partners because of the information they consumed over time via social media.

2.6 Data collection and analysis

Data were collected using questionnaires and were analyzed using descriptive statistics, correlation, regression
and bootsrap analysis. Baron and Kenny (1986) mediation testing approach was used to test for mediation.
Further, the medgraph was used to graphically analyze mediation effects. The findings were confirmed using
structural equation modeling techniques.

3. FINDINGS
3.2 Level of education
Descriptive statistics were used to examine respondents’ level of education. Table 1 presents the results.

Table 1: Level of education

Frequency Percent Valid Cumulative
Percent Percent
Valid Primary 8 2.2 2.2 2.2
Secondary 12 3.4 3.4 5.6

Certificate 35 9.8 9.8 15.4

Diploma 40 11.2 11.2 26.5
Bachelors 120 33.5 33.5 60.1
Masters 132 36.9 36.9 96.9
PhD 11 3.1 3.1 100
Total 358 100 100

Results in Table 1 show that most respondents had master’s degrees (Freq=132, 37%), followed by those with
bachelors’ degrees (freq=120, 34%). Those with diplomas were 40 (11%), those with certificates were 35
(10%), while those with secondary level education were 12(3%). Only 11(3%) respondents were PhD level
while 8 (2%) were primary school level educated.

3.3 Marital status

Descriptive statistics were also used to examine the marital status of respondents as seen in Table 2.

Table 2: Marital status

Frequency Percent Valid Cumulative
Percent Percent
9
Valid Single 204 57 57 57
Married 143 39.9 39.9 96.9

Divorced 11 3.1 3.1 100

Total 358 100 100

Results in Table 2 show that most respondents were single (freq=204, 57%), followed by those who were
married (freq=143, 40%). Only 11(3%) respondents were divorced.

3.4 Testing for Mediation

The objective of this study was to examine the mediation effect of External Locus of Control, Internal Locus
of Control, Self-Regulation, and Behavioral Intention in the relation between Cognitive Factors and Health
Behavior of social media users in Sub-Sahara Africa. In order to investigate this objective, six hypotheses were
formulated. Baron and Kenny (1986) procedure was used to test for mediation. Thereafter, the results were
entered in Jose (2013) Medgraphs to graphically produce mediation by Sobel z values, P values, direct and
indirect effects. The following section presents mediation results:

H1: Internal Locus of Control and Behavioral Intention mediate the relationship between Cognitive Factors and
Health Behavior of social media users in Sub-Sahara Africa.

Since this hypothesis had two mediators, it was skipped because Baron and Kenny procedure does not cater
for two variable mediations (Jose, 2013; Baron & Kenny, 1986). The mediation effects of the hypothesis were
tested using SEM Bootstrap mediation effects as will be seen later under Confirmatory analysis using
structural equation modeling.

H2: Self-Regulation mediates the relationship between Cognitive Factors and External Locus of Control of social
media users in Sub-Sahara Africa.

Baron and Kenny (1986) procedure was used to test for mediation of effect of Self-Regulation in the
relationship between Cognitive Factors and External Locus of Control. Jose (2013) MedGraph was used to
graphically produce mediation by Sobel z values, P values, direct and indirect effects as seen in Figure 1
Medgraph.

10
Figure 1: MedGraph for Self-Regulation in Cognitive Factors and External Locus of Control

The results presented in Figure 1 above reveal that Self-Regulation partially mediates the relationship between
Cognitive Factors and External Locus of Control (Sobel z-value=2.05, P=0.040811). The mediation is
significant at 95% confidence level. For there to be mediation, P value must be less than 0.05 (Jose, 2013;
Baron and Kenny, 1986). Both lower and upper bound values are above zero, implying significance. Further,
the indirect path from Cognitive factors to External Locus of Control contributes 33.8% of the total
mediation effects. The variance explained by the Self-Regulation in the relationship between Cognitive
Factors and External Locus Control is 18.7% (R2=0.187). With these results therefore, H2 stating that Self-
Regulation mediates the relationship between Cognitive Factors and External Locus of Control of social
media users in Sub-Sahara Africa was supported.

H3: External Locus of Control mediates the relationship between Self-Regulation and Health Behavior of social
media users in Sub-Sahara Africa.

11
Baron and Kenny (1986) procedure together with Jose (2013) Medgraph were used to test for mediation of
effect of External Locus of Control in the relationship between Self-Regulation and Health Behavior as seen
in Figure 2 Medgraph.

Figure 2: External Locus of Control mediating Self-Regulation and Health Behavior

The results presented in Figure 2 above reveal that External Locus of Control partially mediates the
relationship between Self-Regulation and Health Behavior (Sobel z-value=3.062629, P=0.002194). Both
direct and indirect effects are significant. Also, both lower and upper bounds are above zero, indicating
significance. This mediation is significant at 95% confidence level. The indirect path from Self-Regulation to
Health Behavior contributes 17.5% of total effects. Further, the indirect path explains 46.4% of variance in

12
total effects (R2=0.464). According to Jose (2013) and Baron and Kenny (1986), there is mediation if P<=
0.05. Therefore, H3 stating that External Locus of Control mediates the relationship between Self-Regulation
and Health Behavior of social media users in Sub-Sahara Africa was supported.

H4: Self-Regulation mediates the relationship between Cognitive Factors and Health Behavior of social media users
in Sub-Sahara Africa.

Baron and Kenny (1986) procedure together with Jose (2013) Medgraph were used to test for mediation of
effect of Self-Regulation in the relationship between Cognitive Factors and Health Behavior as seen in Figure
3 Medgraph.

Figure 3: MedGraph Self-Regulation in Cognitive Factors and Health Behavior

The results presented in Figure 3 above reveal that Self-Regulation fully mediates the relationship between
Cognitive Factors and Health Behavior (Sobel z-value=4.1838678, P=0.000035). This is because the direct

13
effect is not significant (Beta=0.091). This mediation is significant at 95% confidence level. Both lower and
upper bounds are above zero, indicating significance. The indirect path from Cognitive factors to Health
Behavior contributes 53% of total effects. Further, the indirect path explains 78.9% of variance in total effects
(R2=0.789). For there to be mediation, P value must be less than 0.05 (Jose, 2013; Baron and Kenny, 1986).
Therefore, H4 was supported.

H5: External Locus of Control mediates the relationship between Cognitive Factors and Health Behavior of social
media users in Sub-Sahara Africa.

Baron and Kenny (1986) procedure was used to test for mediation of effect of External Locus of Control in
the relationship between Cognitive Factors and Health Behavior as seen in Figure 4 Medgraph.

Figure 4: Medgraph External Locus of Control in Cognitive Factors and Health Behavior

14
The results presented in Figure 4 above reveal that External Locus of Control does not mediate the
relationship between Cognitive Factors and Health Behavior (Sobel z-value=1.619432, P=0.105354). This is
because Sobel z-value P of 0.105354 is above 0.05 (Jose, 2013; Baron & Kenny, 1986). Therefore, H5 stating
that External Locus of Control mediates the relationship between Cognitive Factors and Health Behavior of
social media users in Sub-Sahara Africa was rejected.

H6: Self-Regulation mediates the relationship between Age Sensitivity and Health Behavior of social media users in
Sub-Sahara Africa.

Baron and Kenny (1986) procedure was used to test for mediation effect of Self-Regulation in the relationship
between Age Sensitivity and Health Behavior as seen in Figure 5 Medgraph.

Figure 5: Medgraph Self-Regulation in Age Sensitivity and Health Behavior

15
The results presented in Figure 5 above reveal that Self-Regulation fully mediates the relationship between
Age Sensitivity and Health Behavior (Sobel z-value=5.987725, P=0.000001). This is because Sobel z-value P
of 0.000001 is less than 0.05 (Jose, 2013; Baron and Kenny, 1986). The direct effect is not significant
(Beta=0.046). The path of Age sensitivity and Health behavior contributes 79.9% of total effects. Further,
variables along the indirect path explain 96.7% of variance in the total effects (R2=96.7%). Therefore, H6
stating that Self-Regulation mediates the relationship between Age Sensitivity and Health Behavior of social
media users in Sub-Sahara Africa was supported. Table 68 presents a summary of chapter five.

3.5 Confirmatory analysis using Structural equation modeling

In order to test for hypotheses 1 to 6, mediation tests were run using structural equation modeling techniques.
This is done through analyzing the direct and indirect effects and their significance levels by running the
AMOS bootstrap procedure. The mediation effects are as presented in Table 3.

16
 Table 3: Bootstrap Mediation effects
Dependent variable Mediating variable Independent variable DE P IE P Mediation effect Hypothesis
Health Behavior <--- Internal Locus of Control <--- Cognitive Factors .152 .019 .008 .026 Partial mediation H1
and Behavioral Intention
External Locus of Control <--- Self-Regulation <--- Cognitive Factors - .433 .121 .001 Full mediation H2
.060
Health Behavior <--- External Locus of Control <--- Self-Regulation .329 .001 .142 .001 Partial mediation H3
Health Behavior <--- Self-Regulation <--- Cognitive Factors .135 .069 .158 .001 Full mediation H4
Health Behavior <--- External Locus of Control <--- Cognitive Factors .262 .001 .031 .454 No mediation H5

17
H1: Internal Locus of Control and Behavioral Intention mediate the relationship between Cognitive Factors and
Health Behavior of social media users in Sub-Sahara Africa.

Results in Table 3 show that the direct effect of the relationship between Cognitive Factors and Health
Behavior was significant (Beta=.152, P=.019). At the same time, the indirect effect of the relationship between
Cognitive Factors and Health Behavior was also significant (Beta=.008, P=.026). Given that both the direct
and indirect effects were significant, there is partial. This implies that Internal Locus of Control and
Behavioral Intention mediate the relationship between Cognitive Factors and Health Behavior of social media
users in Sub-Sahara Africa. Hence H1 was supported.

H2: Self-Regulation mediates the relationship between Cognitive Factors and External Locus of Control of social
media users in Sub-Sahara Africa.

Results in Table 3 reveal that the direct effect of the relationship between Cognitive Factors and External
Locus of Control is not significant (Beta=-.060, P=.433). The results however show that the indirect
relationship between Cognitive Factors and External Locus of Control was significant (Beta=.121, P=.001).
This means that there is full mediation of Self-Regulation in the relationship between Cognitive Factors and
External Locus of Control. Since the introduction of Self-Regulation in the relationship renders the direct
effect insignificant, it can be concluded that Self-Regulation fully mediates the relationship between Cognitive
Factors and External Locus of Control. Therefore, H2 was supported.

H3: External Locus of Control mediates the relationship between Self-Regulation and Health Behavior of social
media users in Sub-Sahara Africa.

As seen in Table 3, the direct effect of the relationship between Self-Regulation and Health Behavior was
significant (Beta=.329, P=.001). The direct effect of the relationship between Self-Regulation and Health
Behavior via External Locus of Control was also significant (Beta=.142, P=.001). This is in-line with H3 that
External Locus of Control mediates the relationship between Self-Regulation and Health Behavior of social
media users in Sub-Sahara Africa.

H4: Self-Regulation mediates the relationship between Cognitive Factors and Health Behavior of social media users
in Sub-Sahara Africa.

Further, in Table 3 the results show that the relationship between Cognitive Factors and Health Behavior
direct effect was not significant (Beta=.135, P=.069). However, the relationship between Cognitive Factors
and Health Behavior via Self-Regulation was significant (Beta=.158, P=.001). This means that Self-
Regulation fully mediates the relationship between Cognitive Factors and Health Behavior. Thus, H4 that
stated that Self-Regulation mediates the relationship between Cognitive Factors and Health Behavior of social
media users in Sub-Sahara Africa was supported.

H5: External Locus of Control mediates the relationship between Cognitive Factors and Health Behavior of social
media users in Sub-Sahara Africa.

18
Volume 5. Issue II. December 2020
Results in Table 3 reveal that direct effect of the relationship between Cognitive Factors and Health Behavior
was significant (Beta=.262, P=.001). However, the indirect effect of the relationship between Cognitive
Factors and Health Behavior via External Locus of Control was not significant (Beta=.031, P=.454). This
means that External Locus of Control does not mediate the relationship between Cognitive Factors and
Health Behavior via External Locus of Control. Therefore, H5 which stated that External Locus of Control
mediates the relationship between Cognitive Factors and Health Behavior of social media users in Sub-Sahara
Africa was rejected.

4. DISCUSSION OF FINDINGS
4.2 The mediation effect of Internal Locus of Control and Behavioral Intention in the
relationship between Cognitive Factors and Health Behavior of social media users
Findings revealed that Internal Locus of Control and Behavioral Intention mediated the relationship between
Cognitive Factors and Health Behavior of social media users. Whereas there was a negative significant
relationship between Cognitive Factors and Health Behavior, the introduction of Internal Locus of Control
and Behavioral Intention as mediators also yielded a positive significant relationship between Cognitive
Factors and Health Behavior. This meant that Internal Locus of Control and Behavioral Intention enhanced
the relationship between Cognitive Factors and Health Behavior.

The above finding resonates to Rotter (1966) theory that argues that individuals who control the
consequences of their actions, exhibit high interpersonal relations, and make greater efforts to learn, if
knowledgeable and possesses strong beliefs, they are likely to learn new Health Behaviors via social through
observational learning and practice. Further, the finding is in agreement with Venkatesh et al. (2003) that
Behavioral Intention facilitates behavioral learning.

4.3 The mediation effect of Self-Regulation in the relationship between Cognitive Factors
and External Locus of Control of social media users
Sobel test z-value results revealed that Self-Regulation significantly mediated the relationship between
Cognitive Factors and External Locus of Control. SEM findings also indicated that Self-Regulation had a
partial positive mediation effect on the relationship between Cognitive Factors and External Locus of Control.
Meanwhile the relationship between Cognitive Factors and External Locus of Control was negative. The
introduction of Self-Regulation as a mediator created a positive relationship between Cognitive Factors and
External Locus of Control which was previously negative via direct path. Hence, where there is no Self-
Regulation, an increase in Cognitive Factors reduced External Locus of Control of social media users which is
in contravention with Boundless (2016) and Rotter (1966).

However, once Self-Regulation is introduced as a mediator, the indirect relationship between Cognitive
Factors and External Locus of Control through Self-Regulation becomes positive – implying that an increase
in Cognitive Factors as well as Self-Regulation also increases External Locus of Control of social media users.

19
Volume 5. Issue II. December 2020
This confirms the literature that argues that Cognitive Factors improve External Locus of Control (Boundless,
2016 & Rotter, 1966).

4.4 The mediation effect of External Locus of Control in the relationship between Self-
Regulation and Health Behavior of social media users
Findings divulged a partial mediation effect caused by External Locus of Control in the relationship between
Self-Regulation and Health Behavior implying that social media users’ tendency of relying on online
communities for learning new behaviors, facilitated the level of carefulness and control they had in the process
of learning new Health Behaviors via social media. This finding is in-line with literature of (Blalock et al.
2016; Boundless, 2016; Kane, 2004; Winett et al. 1999; Bandura, 1988; 1986; Rotter, 1966)

Individuals with high Self-Regulation learn selectively. They control the consequences of their actions and are
likely to access information from trusted sources. Therefore, introducing External Locus of Control helps to
ease up Self-Regulation in the learning. An individual becomes more open to various sources of information
via social media. According to Boundless (2016), an individual with External Locus of Control seeks solutions
to his / her problems from people surrounding him / her. In this case social media users with External Locus
of Control tend to ask for counseling and guidance from online communities about their problems. This
helps them get solutions which if practiced, gradually change their Health Behaviors.

4.5 The mediation effect of Self-Regulation in the relationship between Cognitive Factors
and Health Behavior of social media users
Sobel test z-value results revealed that Self-Regulation significantly mediated the relationship between
Cognitive Factors and Health Behavior. SEM findings also revealed that Self-Regulation mediated the
relationship between Cognitive Factors and Health Behavior. This finding helps to suggest that social media
users who set goals, freely discussed their health issues via social media, and were in control of their health
affairs were likely to learn new Health Behaviors if they were knowledgeable and had strong beliefs in their
culture and religious norms. This finding was in-line with literature (Blalock et al. 2016; Bayrón, 2013; Kane
2004; Bandura, 1988; 1986).

The learning of new Health Behavior through observation and practice can be attributed to social media
users’ knowledge, beliefs as well as their level of Self-Regulation. Individuals with strong beliefs, good
knowledge, and control their actions are likely to learn new useful health related behaviors. This can also
happen if an individual is freer to handle their health-related problems via social media.

4.6 The mediation effect of External Locus of Control in the relationship between Cognitive
Factors and Health Behavior of social media users
Sobel test z-value results revealed that External Locus of Control did not significantly mediate the relationship
between Cognitive Factors and Health Behavior. Whereas there was a significant relationship between
Cognitive Factors and Health Behavior in SEM, an introduction of External Locus of Control revealed that

20
Volume 5. Issue II. December 2020
the relationship became positive and significant. This finding indicated that External Locus of Control helped
to mediate the relationship between Cognitive Factors and Health Behavior of social media users. This
finding supports the literature arguing that knowledgeable individuals who were in control of their actions
during online engagements were better learners of new Health Behaviors (Blalock et al. 2016; Bayrón, 2013).
The findings also support an argument that knowledgeable individual with strong beliefs who seek solutions
to their health problems from online communities learned new Health Behaviors (Boundless, 2016; Rotter,
1966).

4.7 The mediation effect of Self-Regulation in the relationship between Age Sensitivity and
Health Behavior of social media users
Sobel test z-value results revealed that Self-Regulation significantly mediated the relationship between Age
Sensitivity and Health Behavior. Whereas it was revealed that Age Sensitivity had a negative significant
relationship with Health Behavior in SEM, an introduction of Self-Regulation as a mediator in this
relationship caused it to become positive and significant. This finding suggests that without Self-Regulation in
terms of self-control, self-management and being charge of one’s affairs, social media users who are age
sensitive were unlikely to learn. However, those social media users who minded the age groups of online
platforms where health related information emanated from and were in charge of their actions in online
engagements were likely to learn new Health Behaviors. This finding is in-line with NIHCE (2007) and
WHO (2000) who acclaim the role played by age in affecting the learning process. Age Sensitivity can delay
or speedup the learning process. In this case for example, we discover that Age Sensitivity delays learning
where there is no self- regulation (Blalock et al. 2016; Bayrón, 2013).

5. CONCLUSION AND RECOMMENDATIONS

The objective of this study sought to examine the mediation effect of External Locus of Control, Internal
Locus of Control, Self-Regulation, Age sensitivity and Behavioral Intention in the relationship between
Cognitive Factors and Health Behavior of social media users in Sub-Sahara Africa. This was digested into six
hypotheses including H1, H2, H3, H4, H5 and H6.

Findings supported H1 that Internal Locus of Control and Behavioral Intention partially and mediated the
relationship between Cognitive Factors and Health Behavior. Given this finding, we conclude that Internal
Locus of Control and Behavioral Intention mediate the relationship between Cognitive Factors and Health
Behavior of social media users in Sub-Saharan Africa.

The findings also supported H2 since Self-Regulation fully mediated the relationship between Cognitive
Factors and External Locus of Control. Therefore, we conclude that Self-Regulation fully mediates the
relationship between Cognitive Factors and External Locus of Control of social media users in Sub-Saharan
Africa.

On H3, the finding revealed that Self-Regulation and External Locus of Control mediated the relationship
between Self-Regulation and Health Behavior. this leads us to a conclusion that Self-Regulation and External

21
Volume 5. Issue II. December 2020
Locus of Control mediate the relationship between Self-Regulation and Health Behavior of social media users
in Sub-Saharan Africa.

Findings also supported H4 where Self-Regulation was found to mediate the relationship between Cognitive
Factors and Health Behavior. Therefore, we conclude that Self-Regulation mediates the relationship between
Cognitive Factors and Health Behavior of social media users in Sub-Saharan Africa.

Further, H5 was supported whereby External Locus of Control mediated the relationship between Cognitive
Factors and Health Behavior. This finding leads us to a conclusion that External Locus of Control mediates
the relationship between Cognitive Factors and Health Behavior of social media users in Sub-Saharan Africa.

Finally, on H6, although both direct and indirect effects were significant suggesting partial mediation, the
direct effect was negative while indirect effect was positive. This implied that Self-Regulation as a mediator
transformed a negative relationship to positive. Therefore, H6 was also supported that Self-Regulation
mediated the relationship between Age Sensitivity and Health Behavior of social media users in Sub-Sahara
Africa.

Lastly, given that both of the mediator variables mediated their respective relationships, it is pertinent that
these mediators including Self-Regulation and External Locus of Control are enhanced in order to promote
positive Health behavioral change via social media platforms.

22
Volume 5. Issue II. December 2020
References
Abbasi M., Chai S., Liu H., and Sagoo K. (2016) Real-World Behavior Analysis through a
SM Lens, Arizona State University. Retrieved on 26th March 2016 from:
http://www.public.asu.edu/~mabbasi2/papers/SBP12Lens.pdf
Adewuyi E. O. and Adefemi, K. (2016) Behavior Change Communication Using Social Media:
A Review, The International Journal of Communication and Health, 2016 / No. 9
Blalock S. J., Bone L., Brewer N. T., Butterfoss F. D., Champion V. L., Epstein R. E., …
(2016) Health Behavior and Health education: Theory, Research and Practice. Karen G., Rimer
B. K., Viswanath K. (Ed.), Perelman School of Medicine, University of Pennsylvania, Retrieved on
March 3rd 2016 from: http://www.med.upenn.edu/hbhe4
Bandura, A. (1990). Mechanisms of moral disengagement. In W. Reich (Ed.), Origins of
terrorism: Psychologies, ideologies, theologies, states of mind (pp. 161-191). Cambridge: Cambridge
University Press.
Bandura, A. (1986). Social foundations of thought and action: A social cognitive theory.
Englewood Cliffs, NJ: Prentice-Hall.
Bayrón, C.E (2013) Social Cognitive Theory, Entrepreneurial Self-Efficacy and Entrepreneurial
Intentions: Tools to Maximize the Effectiveness of Formal Entrepreneurship Education and Address
the Decline in Entrepreneurial Activity, Revista Griot. Vol. 6, 1, Pg. 66-77
Boundless (2016). Bandura's and Rotter's Social-Cognitive Theories of Personality, Boundless
Psychology, Boundless, 08 Jan. 2016. Retrieved 21 Mar. 2016
from https://www.boundless.com/psychology/textbooks/boundless-psychology-textbook/personality-
16/social-cognitive-perspectives-on-personality-81/bandura-s-and-rotter-s-social-cognitive-theories-
of-personality-314-12849/
Boyar, R., Levine, D., and Zensius, N. (2011) Tech SexUSA: Youth Sexuality and Reproductive
Health in the Digital Age, ISIS.
Carroll, J.A. and Kirkpatrick, R.L. (2011) Impact of social media on adolescent behavioral
health, Oakland, CA: California Adolescent Health Collaborative.
Centola, D. (2013) Social Media and the Science of Health Behavior, American Heart
Association
Centola D (2010) The spread of behavior in an online social network experiment, Science. 329:
1194–1197.
Cui P. (2016) Behavioral Analysis and Prediction in Social
Networks, Tsinghua University. Retrieved on 26th March 2016 from:
http://centres.smu.edu.sg/larc/files/2013/03/Peng_Cui_slides.pdf
Dretske F. (2016) Epistemology and Information, UVA, retrieved on 19th July 2016 from
http://www.illc.uva.nl/HPI/Draft_Epistemology_and_Information.pdf
Hilliard C. (2012). SM for Healthcare: A Content Analysis of M.D. Anderson’s
Facebook Presence and its Contribution to Cancer Support Systems, The Elon Journal of
Undergraduate Research in Communications, Vol. 3, No. 1
Hyde S. K. (2017) Math 221: Confidence Intervals for a Population Proportion Chapter 19,

23
Volume 5. Issue II. December 2020
 (Moore, 5th Ed.), Brigham Young University, retrieved on 30th April, 2017 from
http://jekyll.math.byuh.edu/courses/m221/handouts/handout19_prop_ci.pdf
IGI (2016) What is Positivist Epistemology, IGI, retrieved on August 7th 2016 from
http://www.igi-global.com/dictionary/positivist-epistemology/23062
Internet World Statistics (2015) Africa 2015 population and internet users’ statistics for 2015,
Internet World Statistics, Retrieved on 26 March, 2016 from:
http://www.internetworldstats.com/stats1.htm
Kane RL, Johnson PE, Town RJ, Butler M. A, (2004) Structured Review of the Effect of
Economic Incentives on Consumers ‘Preventive Behavior, American Journal of Preventive Medicine
27:4, 327-352
Korda, H., and Itani, Z. (2011) Harnessing Social Media for Health Promotion and Behavior
Change, Health Promotion Practice, Sage Journals, Vol 14, Issue 1, 2013
Newbold, K. B. and Campos. S. (2011) Media and Social Media in Public Health Messages: A
Systematic Review, McMaster Institute of Environment & Health
NIHCE (2007) Behaviour change at population, community and individual levels, National
Institute for Health and Clinical Excellence
PWC Health Research Institute (2012) SM “likes” healthcare from marketing to social
business, PWC
Rideout, V.J., Foehr, U.G., and Roberts, D.F. (2010) Generation M2: Media in the lives of 8-18
year olds, Kaiser Family Foundation. Retrieved on April 30th 2017 from
http://www.kff.org/entmedia/upload/8010.pdf
Rotter, J. B. (1993). Expectancies. In C. E. Walker (Ed.), The history of clinical psychology in
autobiography (vol. II) (pp. 273-284). Pacific Grove, CA: Brooks/Cole
Rouse M. (2014) ICT (Information and Communications Technology - Or Technologies),
TechTarget, Retrieved On 26 March, 2016 From: Http://Searchcio.Techtarget.Com/Definition/Ict-
Information-And-Communications-Technology-Or-Technologies
Salathé, M. Vu D. Q. Khandelwal, S. and Hunter. D. R. (2017) The Dynamics of Health
Behavior Sentiments on a Large Online Social Network, Penn State University
Smith K, Christakis N. (2008) Social networks and health. Annu Rev Sociol, 34:405–429.
Statistics solution (2016) Moderator variable, retrieved on 7th March 2016 at
https://www.statisticssolutions.com/directory-of-statistical-analyses-general-moderator-variable/
Tennis. J. T. (2008) Epistemology, Theory, and Methodology in Knowledge Organization:
Toward a Classification, Metatheory, and Research Framework, In Knowledge Organization, 35(2/3):
102-112.
Venkatesh V., Morris, M. G., Davis, G. B., and Davis. D. (2003) User Acceptance of
Information Technology: Toward a Unified View, MIS Quarterly. Vol. 27, No. 3, pp. 425-478
WHO (2000) Health and Health Behaviour among Young People, World Health Organization
Policy Series: Health policy for children and adolescents Issue 1; International Report
Wikipedia (2016a) Genetic epistemology, Wikipedia, retrieved on August 7th 2016 from
https://en.wikipedia.org/wiki/Genetic_epistemology
Wikipedia (2016b) Feminist epistemology, Wikipedia, retrieved on August 7th 2016 from
https://en.wikipedia.org/wiki/Feminist_epistemology

24
Volume 5. Issue II. December 2020
Wikipedia (2016c) Social epistemology, Wikipedia, retrieved on August 7th 2016 from
https://en.wikipedia.org/wiki/Social_epistemology
Wikipedia (2016d) Constructivist epistemology, Wikipedia, retrieved on August 7th 2016 from
https://en.wikipedia.org/wiki/Constructivist_epistemology
Winett R. A., Anderson E. S., Whiteley J. A., Wojcik J. R., Rovniak L. S., Graves K. D., Galper
D. I., Winett S. G. (1999) Church-based Health Behavior programs: Using Social Cognitive Theory
to formulate interventions for at-risk populations, Applied & Preventive Psychology 8:129-142.
Winters E, Petosa R, Charleton T. (2003) Using Social Cognitive Theory to explain discretionary
"Leisure-time" physical exercise among high school students. Journal of Adolescent Health Vol.
32:436-442.
Wright, K. (2016) Social Networks, Interpersonal Social Support, and Health Outcomes: A
Health Communication Perspective. Frontiers in Communication, 1:10. doi:
10.3389/fcomm.2016.00010

25
Volume 5. Issue II. December 2020