International Journal of Managing Information Technology (IJMIT) Vol.10, No.

2, May 2018

INVESTIGATING TANZANIA GOVERNMENT

EMPLOYEES’ ACCEPTANCE AND USE OF SOCIAL
MEDIA: AN EMPIRICAL VALIDATION AND
EXTENSION OF UTAUT
Julius Raphael Athuman Mhina , Md Gapar Md Johar

Management & Science University, Faculty of Information Sciences & Engineering,

Shah Alam, Selangor, Malaysia
ABSTRACT
The aim of this paper was to present factors which influence work-related adoption of social media among
government employees. The study adapted the Unified Theory of Acceptance and Use of Technology
(UTAUT) and integrated it with three constructs namely Hedonic Motivation (HM), Perceived Personal
Image (PI) and Attitude (AT). The sample of study consisted of 600 usable responses collected using survey
questionnaire from employees based in Tanzania Ministries, Departments and Agencies (MDAs) employing
convenience sampling technique. The data was analyzed using Structural Equation Modelling. The results
showed that, PI and HM influence attitude (AT), while PI, Social Influence (SI) and AT influence Behavior
Intention (BI). Moreover, SI, PI, AT and BI showed significant effect on actual Use Behavior. Furthermore,
the results showed that PI is affected positively by HM and SI. The added constructs proved to be important
modifications to the UTAUT model as suggested by the increase in explained variance by UTAUT from
70% to 73% on BI and from 53% to 73% on actual use behavior.

KEYWORDS
Web 2.0 and Social Media, Hedonic Motivation, Perceived Personal Image, Attitude, UTAUT, Government
Employee, Tanzania.

1. INTRODUCTION
There is a substantial evidence suggesting that social media are widely and increasingly being
used by people of all demographic classes [1][2]. A special report by Kemp [3] revealed that, as
of January, 2018, out of 7.593 billion total world population; 5.135 (68%) billion were unique
mobile phone users, 4.021 (53%) billion were internet users, 3.191 (42%) billion were active
social media users and 2.958 (39%) billion were active mobile social media users. The report
further reports that an average internet user spends, on average 6 hours per day online. Moreover,
the report reveals that about 200 million people used internet for the first time in 2017 with
Africa recording the fastest growth rate, in which the number of internet users across the
continent is increasing by 20% per year. Overall, the report shows that, the global number of
social media users has increased by 13% in the last the past year and this staggering increase in
internet users is associated with the increasingly affordable smartphones and mobile data plans.

Studies show that, in response to the large number of citizens using social media, government
organizations are increasingly adopting social media [4]-[7] hence making social media a key
component of electronic government (hereafter called e-government). Use of social media in e-

DOI : 10.5121/ijmit.2018.10205 75
 International Journal of Managing Information Technology (IJMIT) Vol.10, No.2, May 2018
government increases government-citizen engagement [8], fosters citizens’ electronic
participation [9], increases government transparency [10], strengthens the communication,
interaction and collaboration between government and stakeholders [11] and increases citizens’
trust in the government [12].

A study conducted by Dwivedi et al., [13] showed that there are many studies on use of social
media in electronic government. However, most of the studies did not employ well established
theories [13]. This may be due to a lack of well suited model or theory for acceptance and use of
social media in electronic government as social media is relatively a new phenomenon hence
calling for a new model to suit this kind of technologies. Moreover, for social media use in
government to be successful, the technologies should be accepted by the employees. However,
the factors influencing government employees to use social media for work-related purposes are
not well known [14].

Therefore, the main objective of this paper was to empirically validate UTAUT and integrate it
with Hedonic Motivation, Perceived Personal Image and Attitude in order to study the factors
that influence government employee’s adoption of web 2.0 and social media applications for
work-related purposes as a pre-step toward developing a model for acceptance and use of social
media among government employees. These three constructs emerged in the literature to be
important antecedents of technology acceptance [15]. The sample for the current study was
obtained from Tanzania government employees based in Ministries, Departments and Agencies
(MDAs).

This paper proceeds as follows; In section 2- literature review, section 3- Model development
and hypotheses formulation, section 4- research methodology, section 5- analysis of the results,
section 6 – discussion and conclusion, Section 7- further work, section 8- references and
Appendices in section 9.

2. LITERATURE REVIEW

UTAUT is a unified model for adoption of information systems which was developed by
Venkatesh et al. [16] from other eight technology adoption models. It was developed with the
aim of providing a standard theoretical framework for individual adoption of technology in
organizations by combining the following theories/models: Theory of Reasoned Action,
Technology Acceptance Model, Motivation Model, Theory of Planned Behavior , Combined
TAM and TPB, Model of Personal Computer Utilization, Innovation Diffusion Theory , Social
Cognitive Theory [16].The UTAUT showed highest total explained variance on Behavior
Intention (BI) (70%) and on Use Behavior (UB) (53%) among the eight from which it was
synthesized [16].

Since its development, UTAUT has been applied in a variety of contexts such as in mobile
learning [17]-[19], e-learning [20], consumer [21], e-government [22] [23], mobile commerce
and internet banking [24] [25], citizens [26], research [27], Citizens’ Facebook use [28] [29] and
small businesses [30].

However, Venkatesh, Thong and Xu [31] reviewed the use of UTAUT model from when it was
developed in 2003 to 2014 and concluded that the model needs refinement . Their study involved
1,267 research papers published in major information system journals and classified the papers
into four themes: 1) general citation of to the original UTAUT without any substantial use of the
theory (1,205), 2) an application of UTAUT as underlaying research theory in an empirical study
(12), 3) an integration of UTAUT with other theories (13), and 4) an extension to the UTAUT by
76
 International Journal of Managing Information Technology (IJMIT) Vol.10, No.2, May 2018
adding another construct (37). Looking at the number of studies which used UTAUT as their
base model excluding the general citations, it shows that that there are relatively small number of
studies i.e. 62 which used UTAUT as a base model contrary to the original expectation of the
model development as it was expected to be a standard theory for studying adoption of
information systems among researchers hence suggesting a need to refine the model [16] [31].

Figure 1 : Proposed Research model [Modified UTAUT]

In line with the Venkatesh et al., [31] call for need to refine the UTAUT theory, in this study
we therefore seek to validate and refine the theory by integrating it with three constructs
namely Hedonic Motivation, Perceived Personal Image and Attitude . There is substantial
evidence from the literature which suggest that these constructs are important antecedents of
intention and use of technology [14] [15].

3. MODEL DEVELOPMENT AND HYPOTHESES FORMULATION

This study adapts UTAUT model and integrate it with Hedonic Motivation (HM) and Attitude
(AT) (see Figure 1). The proposed research model consists of eight (8) constructs namely;
Performance Expectancy (PE), Effort Expectancy (EE), Social Influence (SI), Facilitating
Conditions (FC), Hedonic Motivation (HM), Perceived Personal Image (PI), Attitude (AT),
Behavior Intention (BI) and Use Behavior (UB). While the UTAUT relationships are shown
using thick lines, the added constructs are written in italic. In the proposed model, Behavior
Intention (BI) is postulated to be determined by PE, EE, SI, HM, PI and AT while AT is
determined by SI, FC, PI and HM. Furthermore, the UB is postulated to be determined by SI, FC,
AT, PI and BI and PI is determined by SI. While the UTAUT relationships are shown using thick
lines, the added constructs are written in italic. In the proposed model, Behavior Intention (BI) is
postulated to be determined by PE, EE, SI, HM, PI and AT. AT on the other hand is postulated to
be determined by SI, FC, PI and HM. Furthermore, the UB is postulated to be determined by SI,
FC, AT, PI and BI. Furthermore, the model posits that PI is determined by SI. In this study, the
moderating variables of gender, age, experience and voluntariness were not analyzed. The
hypotheses for this model are explained in the next subsections.

77
 International Journal of Managing Information Technology (IJMIT) Vol.10, No.2, May 2018

3.1 PERFORMANCE EXPECTANCY (PE)

In this study, PE is defined as the degree to which a government employee believes that using
social media for work-related purposes will enhance his/her job performance [16]. Apart from
Venkatesh et al., [16] who found that PE positively affects BI (see figure 1), other researchers
also confirmed influence of PE on acceptance of technology [14] [15] [18] [19] [29] [32] [33].
However, PE was not found to have any significant effect on BI in the following studies [17]
[22]. Hence this study empirically tests the following hypothesis:

H1 Performance Expectancy has significant and direct effect on intention of government

employee to adopt and use social media for work-related purposes.

3.2 EFFORT EXPECTANCY (EE)

In this study, Effort Expectancy (EE) is defined as the perception that using social media is ease
([16]). As shown in figure 1, EE was found to have significant and positive influence on
employee’s intention to adopt the new technology in [16]. Other researchers found similar
results [15] [18] [22] [33]. However, EE was not found to significantly effect BI in the
following studies [19] [29] [33]. Hence this study empirically tests the following hypothesis:

H2 Effort Expectancy has significant and direct effect on intention of employees to adopt
and use social media for work-related purposes.

3.3 SOCIAL INFLUENCE (SI)

In this study SI is defined as the degree to which government employee’s adoption of social
media is influenced by the pressure from other people he/she deems important [16]. Previous
studies show that SI has significant positive effect on employees’ intention to adopt technology
[14] [15] [16] [19] [29] [32] [33] and on actual use [29]. Also, SI was found to have significant
effect on attitude in the following studies [15] [19]. Furthermore, social norms (similar to social
influence) has shown significant positive effect on user’s image (here called perceived personal
image) [51]. However, SI was not found to have any significant effect on BI in the following
study [18] as suggested by the UTAUT findings. Nevertheless, based on the findings of majority
of studies, we hypothesize that:

H3 SI has positive significant effect on perceived personal image of a government employee to

use social media for work-related purposes.

H4 SI has positive significant effect on attitude of a government employee to use social media
for work-related purposes.

H5 SI has positive significant effect on intention of employees to use social media for work-
related purposes.

H6 SI has positive significant effect on government employee’s actual use of social media for
work-related purposes.

78
 International Journal of Managing Information Technology (IJMIT) Vol.10, No.2, May 2018
3.4 FACILITATING CONDITIONS (FC)

Facilitating Conditions (FC) is defined as the individual belief that an organizational and
Technical infrastructure exists to support the use of the system [16]. In this study, FC is defined
as the belief that there is enough organizational support toward use of social media for work-
related purposes. It includes supportive organizational policies, trainings and other technical
assistance that may be required by an employee in order to engage in work-related use of the
technologies. Previous studies show that FC has significant effect on actual use of technology
[15] [16] [21] [29] [33]. Also, FC is found to have significant direct effect on attitude [14] [15].
However, the following study did not find any significant effect of FC on UB [22]. Based on
majority of findings, we hypothesize that:

H7 FC have positive significant effect on Attitude of government employees to use social

media for work-related purposes.

H8 FC have significant and direct effect on government employee’s actual use of social
media for work-related purposes.

3.5 HEDONIC MOTIVATION (HM)

Hedonic Motivation Hedonic motivation is defined as the perceived enjoyment, fun or pleasure
experienced when using the technology [21]. In this study hedonic motivation is defined as
perceived enjoyment anticipated up on use of social media. Previous studies show that hedonic
motivation has positive significant influence on attitude to use a technology [34]. Also, HM was
found to have significant effect on intention to use technology in the following studies [21], [29],
[32]. Hence, we hypothesize that:

H9 HM has significant positive effect on the attitude of a government employee toward

use of social media for work-related purposes.

H10 HM has significant positive effect on the intention of a government employee to use
social media for work-related purposes.

3.6 Perceived Personal Image (PI)

This construct has its roots in the construct called self-identity expressiveness [36]. Personal
image or Self-image is defined as the degree to which use of an innovation is perceived to
enhance one's image or status in one's social system [52]. Social media enable employees to
engage rapidly and simultaneously with peers, colleagues, customers, and the general public
thereby offering possibilities to convey a favorable image of the self [36]. Studies show that, if
important members among an individual’s circle of friends believe that the person should use an
innovation, then using it will improves his/her standing within this social system [52]. Empirical
evidence show that self-identity has effect on both attitude and intention to engage in behavior
[14]. In another study, it was found that, self-expressiveness (here called perceived personal
image) has significant positive effect on actual use of social media for work-related purposes
[36]. Hence, we hypothesize that:

79
 International Journal of Managing Information Technology (IJMIT) Vol.10, No.2, May 2018

H11 PI has positive influence on the Attitude of a government employee toward intention to
engage in use of personal social media for work-related purposes.

H12 PI has positive influence on the government employee’s intention to engage in use of
personal social media for work-related purposes.

H13 PI has positive influence on the government employee’s intention to engage in use of
personal social media for work-related purposes

3.7 ATTITUDE (AT)

Attitude is defined as an individual’s positive or negative feelings (evaluative affect) about

performing the target behavior [35]. In this study, attitude is defined as the government
employee’s positive or negative feelings about using personal social media accounts for work-
related. Previous studies show that attitude is an important antecedent of intention to adopt new
technology [14] [15] [36]. Also, studies show that apart from having direct positive effect on BI,
also attitude has direct influence on actual use [14] [15]. Hence, we propose the following
hypotheses for this study:

H14 Attitude of a government employee toward social media has positive influence on
his/her intention to adopt them for work-related purposes.

H15 Attitude of government employee toward social media has positive influence on
his/her actual use of such technologies for work-related purposes.

3.8 Behavior Intention (BI)

BI is defined as a “degree to which a person has formulated conscious plans to perform or not
perform some specified future behavior’’ [37]. Studies show that BI is an immediate predictor
of actual use behavior [16] [35] [38] and that it has significant positive effect on actual use of
technology [15] [29] [33] [36]. Hence this study assesses the willingness of government
employees to engage in use of personal social media accounts for work-related purposes as an
important antecedent of actual use behavior.

H16 Behavior Intention has positive significant effect on government employees’ actual
use of personal social media for work-related purposes.

3.9 USE BEHAVIOR (UB)

This is the target behavior in this study. It is the actual use of the systems. In this study it is
defined as the actual use of personal social media accounts for work-related purposes. It involves
posting work-related messages on web 2.0 or social media platforms such as Twitter, Facebook,
YouTube, WhatsApp, LinkedIn, Instagram etc. [36]. In this study, use behavior is measured
through self-reported frequency of using personal social media accounts for work-related
purposes.

80
 International Journal of Managing Information Technology (IJMIT) Vol.10, No.2, May 2018

4. RESEARCH METHODOLOGY
4.1 Population, Sample Size and Sampling Technique

The population for this study is government employees based in Tanzania Ministries,
Departments and Agencies (MDAs). These employees were selected because they were
considered the key people toward implementation of government policies. Depending on the
findings of the study, the government policies can be developed or reviewed to enhance
beneficial usage of social media among government employees. Since the population size is
finite and known, the Yamane formula [39] was used to calculate the sample size for the study in
which 400 is considered adequate minimum sample. For this study, a Quantitative technique [40]
using self-administered questionnaire [41] was employed. The convenient sampling technique
was used to distribute questionnaire to 1061respondents in the following distribution: nine (9)
Ministries, eighteen (18) Government Departments and five (5) Agencies based in Dar es salaam
city, Tanzania. 675 (63.6%) questionnaires were returned. After data screening, 600 usable cases
were obtained. The data collection was conducted between January 2016 and March 2017.
Before the actual data collection, a pilot study was conducted to fine tune the research
instruments.

Table 1 : Respondents’ Social Media Usage (N = 675)

81
 International Journal of Managing Information Technology (IJMIT) Vol.10, No.2, May 2018

4.2 Instrument Design and Development

The initial measurement instrument used in the study consisted of 38 items (See Appendix A)
with all of the items adapted from the literature. While the items for the following constructs -
PE, EE, SI, FC, AT and BI were adapted from Venkatesh et al. [16]; the items for HM were
adapted from Venkatesh et al., [21]. Modification was done on the items to fit the area of study.
A 7-point Likert scale was adapted in this study with scores ranging from 1- Strongly Disagree
to 7-Strongly Agree. The validated questionnaire consisted of twenty-eight (28) items (see
Appendix A).

4.3 Study procedure

Table 1 shows the home or personal usage of social media by respondents. Respondents were
presented with a list of most popular web 2.0 and social media applications in Tanzania and were
asked to indicate the application (s) they use for personal purposes whereby the most popular
social media among the respondents is WhatsApp (21.1%) which is used by 555 (92.2%)
respondents, Facebook is the second most used application (16.4%) used by 432 (71.8%)
respondents, YouTube (14.2%) ranked third and is used by 408 (61.2%) respondents, Instagram
(12.4%) used by 356 (53.4%) respondents, Twitter (7.6) used by 219 (32.8%) respondents,
LinkedIn (6.4) used by 185 (27.7%) respondents, Google+ (6.9%) used by 199 (29.8%)
respondents, Blogs (6.2%) used by 179 (26.8%) respondents. An almost negligible number of
respondents 5 (0.2%) indicated that they don’t use any social media for any purpose. Those
respondents who used at least one application for home or personal purposes, were asked to
indicate their perception on the web 2.0 and social media applications and their willingness to
use them for professional or work-related purposes.

5. RESULTS & ANALYSIS

SEM technique was employed to conduct multivariate statistical analysis as it is a robust
technique for analyzing multi-interrelationships between latent variables [43]. A two-stage
analysis was adapted, starting with analysis of the measurement model through confirmatory
factor analysis (CFA) followed by analysis of the structural model [46]. AMOS version 21 and
SPSS version 21 tools were used to facilitate the analysis process.

5.1 Descriptive Analysis

Table 2shows demographic characteristics of the sample with 209 (31%) respondents coming
from Ministries, 331 (49%) respondents from government departments and 135 (20%) from
government agencies. The youth age group (18-35) contributes more than half of the sample 376
(55.7%). Majority of the respondents are males 338 (57.7%), More than half of the respondents
have a bachelor degree 369 (54.7) and 545 (80.8%) have high experience (≥ 3 years) in using
social media for personal purposes.

5.2 Missing Data, Normality and Outlier Assessment

Data screening was performed on the collected data for detection of missing values, normality
and outliers in order to produce valid and reliable results [42]. This process resulted in deletion
of 75 cases due to high missing rate (more than 5%) and suspicious pattern such as non-
engagement [43]. The remaining data with missing values (less than 5%) were imputed using

82
 International Journal of Managing Information Technology (IJMIT) Vol.10, No.2, May 2018
median [44]. The data screening process resulted into 600 (88.9%) usable cases which were used
in all subsequent analyses.

Univariate normality was tested by using Skewness and Kurtosis. The results of the analysis are
as shown in Table 3 in which the Skewness and Kurtosis values of all items were within the
acceptable value of ±2 for Skewness and ±3 for Kurtosis [45].

Table 2: Respondents demographic profile (N = 675)

5.3 Measurement Model

This study adapted the two-step approach in conducting analysis [46]. First, CFA was
conducted in order to assess the measurement model for reliability and validity of measured
items followed by analysis of structural equation model for analysis of causal relationships
between model constructs. To assess the model fitness, the following model fit indices were
used for both the measurement model and structural model for this study: A normed χ-square
(χ2/df), Root Mean Square Error of Approximation (RMSEA); Standardized Root Mean
Residual (SRMR), Comparative Fit Index (CFI) and Tuller-Lewis Index (TLI) [43] (see table 4).

A confirmatory factor analysis (CFA) was carried out to validate a pooled measurement model
with 38 items. In order to attain good model, fit, ten (10) items were deleted from the initial
measurement model in the following order: PE4, FC3, FC5, HM1, SI4, BI4, PI1, PI3, UB1 and
EE2 due to due to low factor loading (<0.5) and being redundant with other items from other
constructs [43]. Also, the pair of error terms for items S1 and S3, HM2 and HM3 were
constrained due to high modification index (MI>15) [45].

The resulting model after these modifications showed a good model fit (χ²/df =2.378; RMSEA =
0.048; SRMR = 0.0412; CFI = 0.966, TLI =0.959) (see table 4, column 4). Since all model fit
indices had been achieved, then, the psychometric features of the model i.e., convergent
validity, discriminant validity and constructs reliability were assessed.

Cronbach’s alpha (α), Composite Reliability (CR), Average Variance Extracted (AVE) and
Maximum Shared Variance (MSV) were all used to assess reliability, discriminant validity and
convergent validity of the measurement model. For reliability to be achieved, α >0.7 and CR
should be ≥0.6 or preferably >0.7, for convergent validity to be achieved; AVE>0.5 and for
discriminant validity to be achieved MSV<AVE [43]. Table 3 shows that all the CR for all
constructs ranged from 0.766 (FC) to 0.960 (BI), the AVE ranged from 0.77 (FC) to 0.96 (BI)
and all MSV<AVE. Furthermore, √AVE- which is shown as the diagonal value (in bold) is

83
 International Journal of Managing Information Technology (IJMIT) Vol.10, No.2, May 2018
greater than all values along the row and column for all constructs hence discriminant validity is
achieved [45]. In order to test for multicollinearity, SPSS was used to calculate tolerance and
Variance Inflation Factor (VIF). A value of tolerance value greater than 0.1 and VIF of less than
3.0 is considered good indication of absence of multicollinearity [47]. Since all VIFs were <3
and Tolerance >0.1, then it is concluded that there is no multicollinearity in the sample (see
table 3). Based on this assessment, it is evident that the convergent validity, reliability and
discriminant validity of the constructs have been achieved, therefore the assessment of the
structural model is done in the next section. The final measurement model consisted of 28 items
(See Appendix B).

Table 3 : Constructs Reliability, Convergent Validity and Discriminant Validity

CR AVE MSV PE EE SI FC HM PI AT BI UB

0.8
PE 0.67 0.33 0.82
6

0.7
EE 0.56 0.31 0.56 0.75
9

0.8
SI 0.59 0.3 0.5 0.24 0.77
1

0.7
FC 0.52 0.36 0.58 0.51 0.55 0.72
7

H 0.9
0.81 0.36 0.44 0.28 0.54 0.6 0.9
M 5

0.8
PI 0.54 0.51 0.34 0.11 0.53 0.41 0.49 0.73
2

0.8
AT 0.73 0.68 0.26 0.05 0.44 0.42 0.53 0.7 0.86
9

0.9
BI 0.89 0.68 0.27 0.03 0.49 0.39 0.45 0.72 0.83 0.94
6

0.9 0.9
UB 0.84 0.66 0.31 0.05 0.55 0.38 0.48 0.71 0.78 0.81
1 1

1.68 1.46 1.58 1.74 1.63 N/A N/A N/A N/A

VIF
4 8 6 0 0

0.85 0.75 0.75 0.94 0.86 0.88 0.95 0.9

Cronbach’s alpha 0.77
4 8 6 9 3 9 9 1

0.59 0.68 0.63 0.57 0.61 0.69 N/A N/A N/A

Tolerance
4 1 1 5 3 8

-.380 .17
Skewness -.592 -.620 -.201 -.374 -.585 -.300 -.029
2

1.05 -.174
Kurtosis .175 -.014 .151 .048 -.498 -.731 -
1
.65

84
 International Journal of Managing Information Technology (IJMIT) Vol.10, No.2, May 2018
1

Notes: CR=Composite Reliability, AVE = Average Variance Extracted, MSV = Maximum Shared Variance,
PE=Performance Expectancy, EE=Effort Expectancy, SI = Social Influence, FC= Facilitating Conditions; HM =
Hedonic Motivation, PI = Perceived Personal Image, AT: Attitude; BI= Behavior Intention, UB = Use Behavior, VIF=
Variance Inflation Factor.

5.4 Structural Model

The same criteria that were used to assess goodness of fit of the measurement model were used
to assess the goodness of fit (GOF) of the structural model. Both models showed similar model
fit results (see table 4). Since structural model showed good model fit (χ²/df =2.375,
RMSEA=0.048, SRMR=0.0446, CFI=0.965,TLI=0.959), therefore path analysis can be done.
The results of structural model analysis are shown in table 5 and figure 2. The structural model
results show that eleven (11) out of seventeen (17) standardized path coefficients were
significant.

Table 4. Model Fit indices for the final measurement and structural model
Name of Name
Recommended Measurement Structural
Category
Of Value a
Model Model
Index
Absolute χ² ns at p<0.05 739.626 759.909
Degree of
df n/a
freedom 311 320
Parsimonious χ ²/df 5< Preferable <3 2.378 2.375
Absolute RMSEA <0.08 0.048 0.048
Incremental SRMR <0.08 0.0412 0.0446
Incremental CFI >0.95 0.966 0.965
Incremental TLI >0.95 0.959 0.959
Notes: χ2 = Chi-square; df = degree of freedom; RMSEA = Root Mean Square Error of
Approximation; SRMR= Standardized Root Mean Residual; CFI = Comparative Fit Index.
Sources: a [43] [48] [49].

More specifically, the following factors showed significant effect on Attitude: Hedonic
Motivation (HM) (β = 0.231, p<0.001) and Perceived Personal Image (PI) (β = 0.581, p<0.001)
hence supporting hypotheses H9 and H11 respectively and contributed to 54% of explained
variance on attitude (see table 5). However, the following hypothesized effects on attitude were
not confirmed: SI (β = -0.02, p>0.05), Facilitating Conditions (FC) (β = 0.042, p>0.05) and hence
hypotheses H4 and H7 and H10 were not supported. Furthermore, the following factors showed
significant effects on Behavior Intention (BI): Social Influence (SI) (β = 0.126, p<0.01)

Perceived Personal Image (PI) (SI) (β = 0.234, p<0.001) and Attitude (AT) (β = 0.646, p<0.001)
hence showing support for the hypotheses H5, H12 and H14 and contributed to 73% of explained
variance on behaviour intention (see table 5) but the findings did not support the effect of
performance expectancy (PE) (β = 0.018, p>0.05), effort expectancy (EE) (β = -0.053, p>0.05)
85
 International Journal of Managing Information Technology (IJMIT) Vol.10, No.2, May 2018
and Hedonic Motivation (HM) (β = -0.07, p>0.05) on behaviour intention (BI) as hypothesized
hence hypotheses H1, H2 and H01 were contradicted (see table 5).

Table 5. Structural Path Results

P- Hypothesis
HNO Path (β) S.E. C.R.
Value Testing
H3 SI PI 0.385 0.044 6.566 *** Supported
E HM PI 0.281 0.036 5.29 *** Emerging
H11 PI AT 0.581 0.098 9.567 *** Supported
H9 HM AT 0.231 0.055 4.585 *** Supported
H4 SI AT -0.02 0.063 -0.373 0.709 Unsupported
H7 FC AT 0.042 0.065 0.798 0.425 Unsupported
H1 PE BI 0.018 0.055 0.44 0.66 Unsupported
H2 EE BI -0.053 0.065 -1.526 0.127 Unsupported
H5 SI BI 0.126 0.049 3.129 0.002 Supported
H12 PI BI 0.234 0.082 4.672 *** Supported
H14 AT BI 0.646 0.048 13.598 *** Supported
H10 HM BI -0.07 0.04 -1.959 0.05 Unsupported
H15 AT UB 0.289 0.058 4.89 *** Supported
H16 BI UB 0.387 0.056 6.733 *** Supported
H8 FC UB -0.046 0.045 -1.22 0.223 Unsupported
H6 SI UB 0.176 0.049 4.243 *** Supported
H13 PI UB 0.156 0.082 3.015 0.003 Supported
Notes: HNO: Hypothesis Number= Emerging, PE=Performance Expectancy,
EE=Effort Expectancy, SI = Social Influence, FC= Facilitating Conditions; HM:
Hedonic Motivation, I: Perceived Personal Image, AT: Attitude, BI=Behavior
Intention; UB = Use Behavior; Significance levels: **p<0.01, *** p < 0.001; ns:
Not Significant (p>0.05); β: Standardized Path Coefficients.

On the other hand, the following factors showed significant effects on actual Use Behaviour:
Social Influence (β = 0.176, p<0.001), Perceived Personal Image (β = 0.156, p<0.001), Attitude
(β = 0.289, p<0.001) and Behaviour Intention (β = 0.387, p<0.001) hence showing support for
the hypotheses H6, H13, H14 and H16 respectively and contributed to 73% of explained
variance on actual use behaviour (see table 5). However, the proposed effect of facilitating
conditions on actual use behaviour could not be confirmed hence hypothesis H7 was
contradicted. These factors contributed to 73% of explained variance on actual use behaviour
(see table 5).

Finally, the results showed that Social Influence has positive significant effect on Perceived
Personal Image (β = 0.385, p<0.001) hence showing support for hypothesis H3. Also,
unexpected significant relationship between hedonic motivation and perceived personal image
(β = 0.281, p<0.001) as suggested by the high modification index (HM) (HM = 16.485) [53].
These two factors explained up to 35% of variance on perceived personal image (see table 5).

86
 International Journal of Managing Information Technology (IJMIT) Vol.10, No.2, May 2018

Figure 2. Structural Model Results for Proposed Model

These results show that the proposed model showed higher explained variance on Behaviour
Intention (73%) and actual Use Behaviour (73%) than the 70% variance on BI and 53% on UB
which was obtained by Venkatesh et al., (2003). However, our model did not include interaction
terms as it was the case with the original UTAUT model. This is because the used software in
this case i.e. AMOS does not support including interaction terms in the analysis.The structural
model results as determined by AMOS are shown in figure 2.

6. DISCUSSION
The aim of this paper was to validate UTAUT and integrate it with Hedonic Motivation (HM),
Perceived Personal Image (PI) and Attitude (AT) in order to investigate the antecedents of
acceptance and work-related use of personal social media accounts among Tanzania government
employees. The results showed that, the added constructs improved the model performance as
shown by the explained variances on behavior intention and actual use.

The results showed that, attitude is the strongest determinant of intention and actual use of
personal social media for work-related purposes among the respondents consistent with other
studies [15] [36].

The second most important antecedent of acceptance and use social media among the
respondents is perceived personal image (PI) which was found to be highly significant on
attitude, behavior intention and actual use behavior . The positive effect of PI implies that ,

87
 International Journal of Managing Information Technology (IJMIT) Vol.10, No.2, May 2018
employees will engage in use work-related use of social media if they believe that , engagement
in this behavior will enhance their social image among their colleagues and society [36].

Regarding social influence, our results confirmed its effect on perceived personal image,
behavior intention and actual use consistent with other previous studies [14] [15] [16] [21] and
[51]. This suggests that the decision of respondents to use personal social media for work-related
purposes is highly influenced by others especially the supervisors and other colleagues deemed
important.

Regarding hedonic motivation, the results confirmed that perceived enjoyment i.e. hedonic
motivation has positive effect on perceived personal image and attitude of respondents toward
adoption of work-related use of social media. Hence employees who perceive social media as
enjoyable have positive attitude toward use the technologies for work-related purposes.

However, our results could not confirm effect of Performance Expectancy (PE) and Effort
Expectancy on behavior intention consistent with [17] [22] [19] [29] [33] [50] who obtained
similar results regarding PE and EE respectively.

Overall, our results show that the acceptance and use of personal social media accounts for work-
related purposes among Tanzania government employee is influenced by individual or
psychological factors (hedonic motivation, social influence, perceived personal image and
attitude) and not by the system attributes i.e. performance expectancy and effort expectancy
which in this case did not play any salient role in the intention of the employees to engage in the
behavior.

7. FURTHER WORK
The main objective of this study was to investigate the factors that influence government
employee’s use of personal social media accounts for work-related purposes. Although this
objective was achieved, nevertheless the model has a number of limitations that can limit its
applicability. Firstly, this research was based on non-probabilistic sampling technique which
could limit generalizability of the results hence in future a probabilistic sampling technique can
done to offset this limitation. Secondly, this study is a cross-sectional since it was conducted in
only in one point in time hence the results should be used with caution since changes in time and
technology may affect the attitude of employees toward use personal social media accounts for
work-related purposes.

Thirdly, the explained variance on perceived personal image (35%), attitude (54%), behavior
intention (73%) and use behavior (73%) shows that there is room for improvement of the model
by investigating the unidentified factors in order to achieve even higher explanatory power
especially on perceived personal image and attitude in which the contributing factors showed
relatively low explained variance. Fourthly, this study employed a quantitative approach in order
to achieve the objectives of the study, however in future a mixed method approach can be used to
provide in-depth explanation of the results.

88
 International Journal of Managing Information Technology (IJMIT) Vol.10, No.2, May 2018

REFERENCES

[1] Chui, M. et al. (2012) ‘The social economy: Unlocking value and productivity through social
technologies’, McKinsey Global Institute, (July), pp. 1–18. doi: 10.1007/s11002-013-9246-1.
[2] Goncalves, J. et al. (2015) ‘Increasing the Reach of Government Social Media: A Case Study in
Modeling Government-Citizen Interaction on Facebook’, Policy & Internet, 7(1), pp. 80–102. doi:
10.1002/poi3.81.
[3] Kemp, S. (2018) Digital in 2018: World’s internet users pass the 4 billion mark - We Are Social UK.
Available at: https://wearesocial.com/uk/blog/2018/01/global-digital-report-2018 (Accessed: 17
February 2018).
[4] Magro, M. J. (2012), “A Review of Social Media Use in E-Government,”Administrative Sciences,
Vol. 2 No. 4, pp.148–161.
[5] Banday,T, M., & Mattoo, M. M. (2013),“Social Media in e-Governance: A Study with Special
Reference to India”, Social Networking, pp.47–56.
[6] Al-badi, A. H. (2013),” The adoption of social media in government agencies: Gulf Cooperation
Council case study”, Journal of Technology Research, pp.1–26.
[7] Mainka, A. et al. (2014) ‘Government and social media: A case study of 31 informational world
cities’, Proceedings of the Annual Hawaii International Conference on System Sciences, (January),
pp. 1715–1724. doi: 10.1109/HICSS.2014.219.
[8] Mossberger, K., Wu, Y. and Crawford, J. (2013) ‘Connecting citizens and local governments? Social
media and interactivity in major U.S. cities’, Government Information Quarterly, 30(4), pp. 351–358.
doi: 10.1016/j.giq.2013.05.016.
[9] Boudjelida, A., Mellouli, S. and Lee, J. (2016) ‘Electronic Citizens Participation’, in Proceedings of
the 9th International Conference on Theory and Practice of Electronic Governance - ICEGOV ’15-16.
New York, New York, USA: ACM Press, pp. 31–39. doi: 10.1145/2910019.2910097.
[10] Carlo Bertot, J., Jaeger, P. T. and Grimes, J. M. (2012) ‘Promoting transparency and accountability
through ICTs, social media, and collaborative e‐government’, Transforming Government: People,
Process and Policy. Edited by S. Ae Chun. Emerald Group Publishing Limited, 6(1), pp. 78–91. doi:
10.1108/17506161211214831.
[11] Al-Jamal, M. and Shanab, E. A. (2016) ‘The influence of open government on e-government website:
the case of Jordan’, International Journal of Electronic Governance, 8(2), p. 159. doi:
10.1504/IJEG.2016.078131.
[12] Picazo-Vela, S., Gutiérrez-Martínez, I. and Felipe Luna-Reyes, L. (2012) ‘Understanding risks,
benefits, and strategic alternatives of social media applications in the public sector’. doi:
10.1016/j.giq.2012.07.002.
[13] Dwivedi, Y. K., Rana, N. P., Tajvidi, M., et al. (2017) ‘Exploring the Role of Social Media in e-
Government’, Proceedings of the 10th International Conference on Theory and Practice of Electronic
Governance - ICEGOV ’17, pp. 97–106. doi: 10.1145/3047273.3047374.
[14] Mhina, J.R.A and Johar,M.G.M (2017),“A conceptual Framework for Acceptance and Use of Social
Media among Government Employees”, International Journal of Emerging Trends & Technology in
Computer Science, Vol. 6 No. 6,pp.181- 186. [15] Dwivedi, Y. K., Rana, N. P., Jeyaraj, A., et al.
(2017) ‘Re-examining the Unified Theory of Acceptance and Use of Technology (UTAUT): Towards
a Revised Theoretical Model’, Information Systems Frontiers. Information Systems Frontiers, pp. 1–
16. doi: 10.1007/s10796-017-9774-y.
[16] Venkatesh, V., Morris, M. G., Davis, G. B., & Davis, F. D. (2003), “User Acceptance of Information
Technology: Toward a Unified View”, MIS Quarterly, Vol. 27 No. 3, pp.425–478.
[17] Jairak, K., Praneetpolgrang, P., & Mekhabunchakij, K. (December, 2009),”An Acceptance of Mobile
Learning for Higher Education Students in Thailand”, International Journal of the Computer, the
Internet and Management, Vol. 17(SP3),pp. 36.1-36.7.
[18] Nassuora, A. B. (2013),“Students Acceptance of Mobile Learning for Higher Education in Saudi
Arabia”, International Journal of Learning and Management Systems, Vol. 1 No. 1, pp.1–9.
[19] Thomas, T., Singh, L., & Gaffar, K. (2013),“The utility of the UTAUT model in explaining mobile
learning adoption in higher education in Guyana”, International Journal of Education and
Development using Information and Communication Technology,Vol. 9 No. 3, pp.71–85.

89
 International Journal of Managing Information Technology (IJMIT) Vol.10, No.2, May 2018
[20] Yoo, S. J., Han, S.-H. and Huang, W. (2012) ‘The roles of intrinsic motivators and extrinsic
motivators in promoting e-learning in the workplace: A case from South Korea’. doi:
10.1016/j.chb.2011.12.015.
[21] Venkatesh, V. (2012),” Consumer Acceptance and Use of Information Technology: Extending the
Unified Theory of Acceptance and Use of Technology,” MIS Quarterly, Vol. 36 No. 1, pp.157–178.
[22] Al-Shafi, S., & Weerakkody, V. (2010),” Factors affecting e-government adoption in the state of
Qatar”, Proceedings of the European and Mediterranean Conference on Information Systems Abu
Dhabi, pp.1–23.
[23] Alshehri, M. A. (2012) ‘Using the UTAUT Model to Determine Factors Affecting Acceptance and
Use of E-government Services in the Kingdom of Saudi Arabia’, pp. 1–264.
[24] Lai, D. C. F., Lai, I. K. W., & Jordan, E. (2009). An extended utaut model for the study of negative
user adoption behaviours of mobile commerce. Proceedings of the International Conference on
Electronic Business (ICEB), pp.720–728
[25] Tobergte, D. R. and Curtis, S. (2013) ‘Analyzing the Use of UTAUT Model in Explaining an Online
Behavior: Internet Banking Adoption’, Journal of Chemical Information and Modeling, 53(9), pp.
1689–1699. doi: 10.1017/CBO9781107415324.004.
[26] Workman, M. (2013) ‘New media and the changing face of information technology use: The
importance of task pursuit, social influence, and experience’. doi: 10.1016/j.chb.2013.10.008.
[27] Gruzd, A., Staves, K., & Wilk, A. (2012). Connected scholars: Examining the role of social media in
research practices of faculty using the UTAUT model. Computers in Human Behavior, 28(6), 2340–
2350. https://doi.org/10.1016/j.chb.2012.07.004
[28] Liew, E. J. Y., Vaithilingam, S. and Nair, M. (2014) ‘Facebook and socio-economic benefits in the
developing world’, Behaviour & Information Technology. Taylor & Francis, 33(4), pp. 345–360. doi:
10.1080/0144929X.2013.810775.
[29] Lallmahomed, M. Z. I., Nor, N. Z., Ibrahim, R., & Rahman, A. A. (2013). Predicting different
conceptualizations of system use: Acceptance in hedonic volitional context (Facebook). Computers in
Human Behavior, 29(6), 2776–2787. https://doi.org/10.1016/j.chb.2013.07.018
[30] Serben, D. F. (2014) ‘The Examination of Factors Influencing Social Media Usage by African
American Small Business Owners Using the UTAUT Model’, pp. 1–134.
[31] Venkatesh, V. et al. (2016) ‘Unified Theory of Acceptance and Use of Technology: A Synthesis and
the Road Ahead’, Jais, 17(5), pp. 328–376.
[32] Guo, Y. (2015) ‘Moderating Effects of Gender in the Acceptance of Mobile SNS Based on UTAUT
Model’, International Journal of Smart Home, 9(1), pp. 203–216. doi: 10.14257/ijsh.2015.9.1.22.
[33] Liu, L., & Chang, C. (2016),” Verification of Social Network Site Use Behavior of the University
Physical Education Students”, EURASIA Journal of Mathematics, Science & Technology Education,
Vol. 12 No. 7, pp.793–805.
[34] Tung, B. Y. E. W. et al. (2016) Determinants of Engagement in Virtual Social Networks in Singapore
a Thesis Submitted in Fulfillment of the. University of Canberra.
[35] Fishbein, M., & Ajzen, I. (1975). Belief, attitude, intention and behavior: An introduction to theory
and research. Reading, MA: Addision-Wasely.
[36] van Zoonen, W., Verhoeven, J. W. M., & Elving, W. J. L. (2014), “Understanding work-related social
media use: An extension of theory of planned behavior”, International Journal of Management
Economics and Social Sciences, 3(4), pp.164–183.
[37] Venkatesh, V. and Bala, H. (2008) ‘Technology Acceptance Model 3 and a Research Agenda on
Interventions’, Decision Sciences. 39(2), pp. 273–315.
[38] Ajzen, I. (1991), “The theory of planned behavior”, Organizational Behavior and Human Decision
Processes, Vol. 50 No. 2, pp. 179-211.
[39] Yamane, T. (1967). Statistics:AnIntroductoryAnalysis (2 ed.). New York: Harper&Row.
[40] Creswell, J. W. (2014). Qualitative,Quantitative and Mixed Methods Research. In J. W.Creswell,
Research Desig(4th ed.). Los Angeles,London,New Delhi,Singapore,Washington: SAGE.
[41] Kumar, R. (2011). Research Methodology:A step-by-step guide for beginners. London,New
Delhi,California,Singapore: SAGE Publications Ltd.
[42] Tabachnick, B.G. and Fidell, L.S. (2007), Using Multivariate Statistics, 5th ed., Allyn & Bacon,
Needham Height, MA.
[43] Hair, Joseph F; Black, William C; babin, Barry J; Anderson, Rolph E;. (2010). Multivariate Data
Analysis: A Global Perspective. Upper Saddle River, NJ: Person.

90
 International Journal of Managing Information Technology (IJMIT) Vol.10, No.2, May 2018
[44] Gaskin, J., (2016), "Data Screening", Gaskination's Statistics.
https://www.youtube.com/user/Gaskination
[45] Awang, Z. (2016), SEM Made Simple: A Gentle Approach to Learning Structural Equation Modeling
(2nd ed.). MPWS Rich Publication,Selangor,KualaLumpur.
[46] Schumacker, R. E., & Lomax, R. G. (2010). A Beginner’s Guide to Structural Equation Modeling.
York: Routledge, Lawrence Erlbaum.
[47] Pallant, J. (2010), SPSS Survival Manual: A Step by Step Guide to Data Analysis using SPSS, 4th ed.,
Open University Press, Berkshire.
[48] Hu, L. and Bentler, P.M. (1999), “Cutoff criteria for fit indexes in covariance structure analysis:
conventional criteria versus new alternatives”, Structural Equation Modeling: A Multidisciplinary
Journal, Vol. 6 No. 1, pp. 1-55.
[49] Kline, R. B. (2011). Principles and practice of structural equation modelling (3rd ed.). New York:
Guilford Press.
[50] Mandal, D. and Mcqueen, R. J. (2012) ‘Extending UTAUT to Explain Social Media Adoption by
Microbusinesses’, International Journal, 4(4), pp. 1–11. doi: 10.5121/ijmit.2012.4401
[51] Venkatesh, V. and Davis, F. D. (2000) ‘A Theoretical Extension of the Technology Acceptance
Model: Four Longitudinal Field Studies’, Management Science. INFORMS, 46(2), pp. 186–204. doi:
10.1287/mnsc.46.2.186.11926.
[52] Moore, G. C. and Benbasat, I. (1991) ‘Development of an Instrument to Measure the Perceptions of
Adopting an Information Technology Innovation Instrument development—Innovation diffusion—
Information technology adoption—Research methodology’, Information Systems Research, 2(3), pp.
192–222.
[53] Denison, D. R., Hart, S. L., & Kahn, J. A. (1996). From chimneys to cross-functional teams:
Developing and validating a diagnostic model. Academy of Management Journal, 39(4), 1005–1024

BIBLIOGRAPHY

Julius Raphael Athuman Mhina Received B.Sc. in Computer Science from The
University of Dar es Salaam in Tanzania in 2005 and M.Sc. In Advanced Computer
Science from University of Sheffield in UK in 2007. He is working as assistant lecturer
at the Institute of Finance management; one of the higher learning institutions in
Tanzania. He is currently undertaking PhD. studies in ICT at the Management and
Science University in Malaysia. His research areas of interest include: Electronic
Government, Cyber Security, Social Media, Internet Programming, Software
Engineering, Web Science and Operating Systems.

Prof. Dr. Md Gapar Md Johar He is Senior Vice President Research, Innovation,

Technology and System of Management and Science University, Malaysia. He is a
professor in Software Engineering. He holds PhD in Computer Science, MSc in Data
Engineering and BSc (Hons) in Computer Science. He is a Certified E-Commerce
Consultant. He has more 35 years of working experience and worked in various
organizations such as Ministry of Finance, Ministry of Public Enterprise, Public
Service Department, Glaxo Malaysia Sdn Bhd and Management & Science
University. His research interests include learning content management system, knowledge management
system, data mining, e-commerce, image processing, data science, character recognition and healthcare
management system. He has developed many computerized systems such as Housing Loan Accounting
System, Commercial Vehicle Licensing System, Campus Management System, Knowledge and Learning
Management System, Asset and Inventory Management System, Library Information System, RFID
application and Time Attendance System.

91
 International Journal of Managing Information Technology (IJMIT) Vol.10, No.2, May 2018
Appendices

Appendix A. Measurement Items

Item
Construct Statement
Code

PE1
I find social media to be useful in my job.
Performanc PE2 Using social media for work-related purposes will enable me to
e accomplish tasks more quickly.
Expectancy Using social media for work-related purposes will increase my
(PE) PE3
productivity.
PE4 If I use social media for work-related purposes, I will increase
chances of getting my salary raised (dropped).
EE1
My interaction with social media is clear and understandable.

Effort EE2
It is easy for me to become skillful using social media (dropped).
Expectancy
(EE) EE3
I find it easy to use social media.
EE4
Learning to operate social media is easy for me.

SI1 People who influence my behavior think that I should use personal
social media for work-related purposes.

Social SI2 People who are important to me think that I should use personal
Influence social media for work-related purposes.
(SI)
SI3
My supervisor uses personal social media for work-related purposes.

SI4 In general, my organization supports use of personal social media for

work-related purposes (dropped).
FC1 I have the resources necessary to use personal social media for work-
related purposes.
FC2 I have the knowledge necessary to use personal social media for
work-related purposes.
Facilitating
Conditions FC3 Personal Social media are compatible with other technologies I use
(FC) (dropped).
FC4 I can get help from others when I have difficulties using personal
social media for work-related purposes.
FC5 A specific person (or group) is available for assistance with using
personal social media for work-related purposes (dropped).
Hedonic Using personal social media for work-related purposes would be
HM1
Motivation interesting (dropped).
92
 International Journal of Managing Information Technology (IJMIT) Vol.10, No.2, May 2018
(HM) Using personal social media for work-related purposes would be
HM2
enjoyable.
Using personal social media for work-related purposes would be very
HM3
entertaining.
Using personal social media for work-related purposes would give me
HM4
pleasure.
Using personal social media for work-related purposes would be
HM5
exciting.

PI1 I use social media to express who I want to be.

PI2 I use social media to express my personal values.

People in my organization who use personal social media for work-

PI3
Perceived related purposes have more prestige.
Image (PI) People in my organization who use personal social media for work-
PI4
related purposes have high status.
I feel that Use of personal social media to share messages about my
PI5
organization will boost my image.
I feel that the Use of personal social media for sharing messages
PI6
about my organization will boost my status.

AT1 Using personal social media for work-related purposes is a good idea

Attitude
AT2 I would like to use personal social media for work-related purposes.
(AT)
I believe that using personal social media for work-related purposes
AT3
would be fun.
I intend to use personal social media for work-related purposes in the
BI1
next 6 months.
I predict I will use personal social media for work-related purposes in
Behavior BI2
the next 6 months.
Intention
(BI) I have a plan to use personal social media for work-related purposes
BI3
in the next 6 months.

BI4 I will always try to use personal social media for work-related
purposes in my daily life (dropped).

UB1 I frequently do use personal social media for work-related purposes in

my daily life (dropped).
Use
I intensively do use personal social media for work-related purposes
Behavior UB2
in my daily life.
(UB)
UB3 Overall, I use personal social media for work-related purposes a lot.

Notes: Scale: 1 = Strongly Disagree (SD); 2 = Moderately Disagree (MD); 3 = Disagree (D); 4
= Neutral (N) 5 = Agree (A); 6 = Moderately Agree (MA); 7 = Strongly Agree (SA).

93
International Journal of Managing Information Technology (IJMIT) Vol.10, No.2, May 2018

Appendix B: Final Measurement Model

94