Asma Mushtaq, Lecturer
(Corresponding Author), Institute of Management Sciences (IMS), University of
Baluchistan, Quetta. Email: asmakhan.sbk@gmail.com
Kamal Badar, Assistant
Professor, Institute of Management Sciences (IMS), University of Baluchistan,
Quetta. Email:kamal.badar1980@gmail.com
Muhammad Anwar, Professor
and Director, Institute of Biochemistry, University of Baluchistan, Quetta. Email: anwarpanezai@yahoo.com
Syed Gohar Abbas, Associate Professor,
Department of Business Administration, Sarhad University of Science &
Information Technology, Peshawar. Email: abbas.ba@suit.edu.pk
Abstract: This research endeavor aims to investigate the impact
of network centrality (degree, closeness and betweenness) on academic
performance (CGPA) of female students in an academic advice network. Where
degree centrality refers to the number of direct links that an actor has with
other actors, betweenness centrality refers to the degree to which an actor
lies on the geodesic paths between other actors and closeness centrality
focuses on how close a member is to all other members in network. Data was
collected from 182 female students enrolled in various programs at a public
sector University in Baluchistan through name generators for egocentric network
by Burt. Techniques of correlation and standard multiple regression analysis
are employed to test the hypotheses. The results of statistical analysis
revealed that high degree and betweenness centrality leads to increase academic
performance of students whereas higher closeness centrality leads to decrease
academic performance. The result of the study has practical implications for
students’ academic life: it will aid the female students to discover structural
pattern of social ties of their advice network and enhance their tendency for
forming more ties related with academic advices for the achievement of
excellent academic performance.
Key words: Network
centrality, academic performance, advice network
Introduction
According to the “social network approach”, the
performance of a person is affected by the types of ties, and structural
characteristics of network more than by qualities, traits, characteristics that
an individual owns (Yang &Tang, 2003).Structural characteristics refer to
the complete arrangement of relations among the system's actors (Tichy,
Tushman, & Fombrun, 1979), like, clustering, bridge, gatekeeper, density
and centrality etc. Structural characteristics (Granovetter, 1985) in social
network have the power to impact positively or negatively on performance of
entities (Yang & Tang, 2003). Structural characteristics (Granovetter,
1985) served as focal principle in social network studies. The unique
characteristic of this stream of investigation lies in how it pleas to the
structural properties of social networks in clarifying results (Sparrowe,
Liden, Wayne, & Kraimer, 2001). From this point of view, people enjoy gains
or suffer damages by virtue of social structure (their places inside social
networks) in any kind of situation. For example, organizational integration
(Sparrowe & Liden, 1997), advancement, progress, development or promotion
(Burt, 1992), authority, power, control (Brass, 1984), creativity (Ibarra,
1993), learning aftermaths (Baldwin, Bedell, & Johnson, 1997), and job
successes (Sparrowe, et al., 2001) all depends on actors position in social
network. In addition, individuals embedded in advice networks can benefit from
the ties by gaining structurally beneficial positions in the advice network to
derive positive outcomes for example enhanced research performance (Badar,
Hite, & Ashraf, 2015).
Several
studies were conducted to prove impact of structural social capital (Zheng,
2008) on performance but their unit of analysis was comprised of both genders,
disregarding (ignoring) individuals- just male or female students and their
position in social networks which influence their academic performance.
Researchers like, Guldner and Stone-Winestock (1995); Baldwin et al. (1997);
Sparrowe et al. (2001) conducted research to find the impact of structural
social capital on students’ performance and confirmed that student’s location
in social network enhance their academic performance.
Researchers
(particularly in education) have been quite interested in exploring the factors
which significantly contribute for performance improvement of students (Shah,
Rahman, & Abbas, 2015). Some scholars have conducted research on females’
centrality in network but their context were different from education like,
Burt (1998) tried to find the impact of structural social capital on women
manager’s advancement and concluded weaker relationship among these variables.
Hence, this research endeavor (unlike others) attempts to encompass the
knowledge of advice networks by concentrating on a social structure of female
student’s network and studying the influence of three traditional dimensions of
centrality (degree, closeness and betweenness) on female student’s academic
performance.
Centrality
Centrality is defined as the level to which a focal
actor is linked to other actors within identified network (Wasserman &
Faust, 1994). Freeman (1979) suggested three measured of centrality, commonly
used to show position of actor in network, that are degree centrality which is based on the number of direct links that
an actor has with other actors, betweenness
centrality that is the extent to which an actor lies on the geodesic paths
between other actors and closeness
centrality focuses on how ‘‘close’’ a member is to all other members in
network. This current research used these dimensions to predict impact of
centrality on academic performance of female students in advice network.
Centrality is one of the most significant notions in
social network analysis (Domhoff, 2013). Central position of individuals offers
them a lot of benefits (Tsai & Ghoshal, 1998) like firm’s innovation (Ahuja,
2000) and performance (Tsai 2001; Yang, 2007) and a person’s ingenuity (Smith & Shalley,
2003) and knowledge creation (Mcfadyen & Cannella, 2004). Baldwin et al.
(1997) indicated that network centrality has positive effect on MBA team
member’s success. Similarly, Hopp, Iravani, Liu and Stringer (2011) indicated
that individuals having central position in advice network show good job
performance. Likewise, Brass (1981) revealed that the flow of the work
represented by centrality of one’s position indirectly impact performance. Aktamov and Zhao (2014) concluded that firms
having higher degree centrality leads in the field of innovation. It is
affirmed by researchers e.g. Spillane and Kim (2012) that actor who occupies
central position in networks are better linked than others and are more
noticeable in the network.
Academic Performance
The utmost critical assets of educational institutes
are students. The development of any country depends on number of qualified
individuals, larger the number of quality graduates opens up endless ways for
country to run on the road of success (Mushtaq & Khan, 2012). The best and
easiest way to judge presence of quality graduates with in country is to check
the academic performance of students through academic reports. As suggested by Allen (2005) academic
performance of students are conversed through academic report that tells about
students’ progress/ achievement (in terms of CGPA/ GPA) established during
course of study in particular discipline. The foremost purpose for allocating
CGPA/GPA is to generate a public record of student’s academic success that can
perfectly and excellently communicate the level of mastery of a subject a
student has demonstrated (Airasian, 1991).
Conceptual
Framework
This conceptual framework is showing the relationship
between centrality measures (degree centrality, degree closeness, betweenness
centrality) and academic performance of female student’s in advice network.
Academic
Performance Degree centrality Betweenness centrality Closeness centrality
Figure 1 Conceptual Framework
Theory and
Hypothesis
Centrality and academic
performance
Advice
networks means set of ties through which individuals share information, data,
knowledge, facts and guidance that assist them in the completion of their
assigned task (Webster & Hackley, 1997) in
better way. Student’s position in advice network plays a dynamic role
(Picciano,2002) but if a student resides the central location s/he can get much
more academic benefits. Gaining advices from number of contacts, not only
develop (Tsai, 2001) intellectual ability of student but can aid him in
locating, griping, diffusing, and blending relevant advices into valuable
resources and ultimately enhancing performance (Borgatti, 2005; Tsai, 2001).
Student can also get assistance form advice network, if s/he became only source
of interaction among other two or more students. Asking for advices from each
other they have to depend on him/her (Spillane & Kim, 2012). While helping
other students in exchange of advices, s/he gain a lot of advices which can be
used by him/her in resolving his own study related problems (Hopp, Iravani,
Liu, & Stringer, 2011). Access to timely and dissimilar advice’s is also
possible for students only if they are at shortest geodesic distance from
others in network. By finding position near to others saves his cost and energy
too in searching for required academic advices. With the help of timely and
dissimilar academic advices he can solve his difficulties which impede his
performance instantly and hastily.
Thus, it was
proposed that:
H1: Female students with higher degree
centrality scores in advice network will have higher academic performance (in
terms of CGPA).
H2: Female
students with higher betweenness centrality scores in advice network will have
higher academic performance (in terms of CGPA).
H3: Female
students with higher closeness centrality scores in advice network will have higher
academic performance (in terms of CGPA).
Method
Sample
182 female
students enrolled in various programs at a public sector university served as
sample of this study. Convenience sampling technique was used to collect data.
Measures
Dependent variable: Academic performance. CGPA of each
student was used to measure academic performance. Which was taken from
university examination record but CGAP can be calculated with the help of
grades, grade weights (A=4, B=3, C=2, D=1), credit hours attempted, and number
of semesters / classes / subjects attended.
Independent variables: centrality measures. According to
Freeman, (1979), the indexes/measures of centrality .i.e. degree, closeness and
betweenness are used as a measure of closeness centrality, degree centrality,
and betweenness centrality of female student’s advice network. Where degree is
the total sum of links to other people in the network. Betweenness states the
level to which a person lies in the middle of other actors in the network.
Closeness is the level to which a person is positioned near all the other
individuals in a network (directly or indirectly).
Procedure
Data was
collected through network questionnaire based on name generators (these are
type of questions that ask central/ focal personal to list the names of those
individuals from whom s/he receive or to whom s/he provide resources)for
egocentric network. Name generator was adapted from standard method of name
generator for egocentric (personal) network as developed by Burt (1984) that is
also used in United States general social surveys (GSS) since
1984 (McPershon, Smith-Lovin, & Brashears, 2006). Name
generators asked the respondents to write the names of those learners/students
with whom the s/he (respondent) most frequently discuss their study related
problems and take academic advices. Reliability of questionnaire was determined
by Cronbach’s alpha to be 0.69.Participants were given questionnaire along with
informed consent form. Participants were requested to read inform consent form
carefully and if they agree to take part in this study then sign on form. After
that participants were requested to read the guidelines cautiously and sensibly
to fill out demographic and name generator sections of the questionnaire.
Ethical guidelines
It was
announced that participants are free to withdraw participation. When all
questionnaires were returned back participants were praised and let them know
that the study was in fact considering for association among social network
structure and academic performance of students. It was made sure to
participants that data would be used for research purpose only and their names
were kept anonymous. Participants were given the right to ask question if they
want. The data collection instrument by Burt (1984) is free to use for all
researchers so no permission was required.
Analysis
The study was
carried out using the statistical techniques of correlation and standard
multiple regression. Correlation is used to show the relationship of all dependent
and independent variables with each other while, standard multiple regression
is used to predict the effect of predictor variables on the dependent variable.
Analysis was performed using Statistical Package for the Social Sciences (SPSS)
20.0 to determine potential associations amongst the variables and effect of
independent variables on dependent variable.
Performance
of students was measured by CGPA that was taken from university examination
record. Data about academic advice was collected through questionnaire based on
name generator (Burt, 1984) for egocentric network. With the help of data
collected through name generator 182 X 182 matrix was constructed in MS Excel,
where zero (0) shows absence of advice tie and one (1) shows presence of advice
tie. This matrix was later on imported in UCINET VI (Borgatti, Everett, &
Freeman, 2002) where three centrality measures were calculated (degree,
closeness and betweenness). Data was coded, entered and analyzed through IBM
SPSS 20.0 to determine potential relationships among the variables.
Correlation
The means,
standard deviation and Spearman’s correlation coefficients can be seen in Table
1. Total number of students participated
in the study was 182. Table 1
also shows the data about the performance of students in exam as measured by
their CGPA. The highest possible CGPA was 3.85. The lowest possible CGPA was
0.24. The mean CGPA score of all students was 2.74 (S.D=0.74). The measures of centrality were likely to be correlated
with the academic performance but were also expected to be non-parametric in
nature (Yan & Ding, 2009). Therefore, Spearman’s correlation was calculated
to measure strength and direction of relationship among variables. The
correlation table disclosed that degree centrality have positively and
significantly correlated with CGPA (r =.60,
p < 0.05), closeness centrality
also shows positive and significant correlation with CGPA (r = .39, p< 0.05) and
betweenness centrality positively but non-significantly correlated with CGPA (r=.33, p>0.05).
Table 1: Summary Statistics and spearman’s
Correlation Matrix
|
Mean |
Std. Dev |
1 |
2 |
3 |
4 |
1. CGPA |
2.74 |
.74 |
___ |
|
|
|
2. degree_centra |
.01 |
.01 |
.60** |
___ |
|
|
3. betweenness_centra |
.03 |
.09 |
.39** |
.63** |
___ |
|
4. closness_centra |
18.33 |
79.17 |
.33** |
.57** |
.18* |
___ |
**. Correlation is significant at the 0.01 level
(2-tailed).
*. Correlation is significant at the 0.05 level
(2-tailed).
Regression Analysis
Before applying multiple regression some assumptions
related with multiple regression must be addresses like normality of residuals,
no outliers, independence of residuals, Multicollinearity.
The assumption of normality of residuals and
independent variable was checked by using histogram(figure 1), and normal P–P
(figure 2) and Q–Q plots (figure 3).The shape of histogram curve nearly
approaches a bell-curve, which means that residuals are approximately normally
distributed. All data points at normal P-P and Q-Q plot lies near or on the
straight line which also give indication of normality of data.
The assumption of no outlier can be checked through
histogram. Outlier is an observation that is different from general
configuration of distribution. Through graphical examination of histogram we
concluded that overall trend of data set is almost same, so there is no significant
outlier.
Figure 3 Normal P-P plot of CGPA
Figure 4. Normal Q-Q
plot of CGPA
To check that
the three measures of centrality(independent variables) can significantly
predict the academic performance (dependent variable) of female student’s,
standard multiple regression analysis was executed. Result of the standard
multiple regression is reported in Table 2. This table also shows the value of
Durbin-Watson statistic through which we can check the assumption of
independence of residuals (autocorrelation). Independence of residuals means
that error should not be interconnected to another error. The value of
Durbin-Watson statistic is always lies between 0 and 4. Zero (0) mean positive
autocorrelation, values near 4 mean direct negative autocorrelation and values
near 2 indicates no autocorrelation. In case of current study the value of
Durbin-Watson statistic was near 2 (1.63) indicating our data has almost no
autocorrelation.
Multicollinearity
is another problem associated with regression analysis. Therefore, data should
be checked for this problem before to apply regression. Multicollinearity means
that independent variables show high or moderate level of association with each
other to such an extent that they make it difficult to recognize the
contribution of single independent variable in predicting dependent variable.
The assumption of no multicollinearity was checked through the values of VIF.
Normally, it is believed that the value of VIF below 10 indicates no or low
level of multicollinearity (Myers, 1990). Table 2 specifies that there is no
problem of multicollinearity as all values of VIF were lowers than 10.
Table 2 shows
the result of the multiple regression i.e. positive and significant
relationship between degree centrality (β1=0.74,
p<0.05), betweenness centrality (β2=0.32, p < 0.05) and academic performance.
Table 2: Summary of Slandered Multiple Regression
Analysis
|
Unstandardized |
Standardized |
VIF |
||
Model |
Coefficients |
Coefficients |
|||
|
B |
Std Error |
β |
||
|
Degree_centrality |
28.22 |
3.79 |
.74* |
2.62 |
Betweenness_centrality |
267.1 |
57.34 |
.32* |
2.6 1 |
|
Closness_centrality |
0.00 |
0.00 |
-0.03 |
1 |
|
|
R2 |
|
0.32 |
|
|
|
F |
|
28.13** |
|
|
|
Durbin
Watson |
|
1.63 |
|
|
Independent variable:
Degree_centrality, Betweenness_centrality, Closness_centrality.
Dependent
variable: Academic performance.
Discussion
and Conclusion
Based on results, we find support for H1 and H2. This
means that female students having high degree centrality and betweenness
centrality in their advice network are expected to achieve better academic
performance. This study findings are supported by prior researchers like,
Aktamov and Zhao (2014); Tsai (2001); Cross and Cummings (2004).While closeness
centrality and academic performance shows negative and insignificant
relationship (β3= -0.03, p>0.05) finding no support for H3, that means
academic performance of female students is reduced with every increment in
closeness of female students in advice network. This finding challenges the
prior studies like, Claro, Neto, and Claro (2013) but conform the finding of
some researchers like, Aktamov and Zhao (2014); Perry-Smith and Shalley (2003).
Closeness centrality shows negative relationship with
female student’s academic performance this may be due to defining our limited
network boundary (students were restricted to give answer to questions, just
remembering to their ties with class fellows not individuals outside of their
class). It is argued that too high or too little network centrality may be
restraining. There should be an intermediary degree of network centrality which
wins the maximum incentives and/or rewards (Rotolo & Petruzzelli, 2013;
Mcfayden & Cannella, 2004; Badar et al., 2015).In case of current research
closeness centrality is high (all advisers are from the same class) due to
which students can get access to large but conflicting viewpoints (Podolny
& Baron, 1997). Which may be the cause of negative relationship between closeness
centrality and female student’s academic performance.
This study aimed to explore and investigate the effect
of structural social capital on female student’s academic performance in advice
network. Our findings of degree centrality and betweenness centrality are in
line with what was found in prior studies, indicating that higher degree
centrality and betweenness centrality higher would be academic performance of
female student’s. While, closeness centrality shows negative relationship with
female student’s academic performance which means that higher closeness
centrality lower would be academic performance of female student’s.
Limitations
and Recommendations
This study has certain limitations e.g. our study lack
generalization as female students of one university may not represent all. More
disseminated findings should be created through involvement of other university
female students and results of research can be strengthen by increasing sample
size.
The result of closeness centrality is different from scholar
expectations of current study. Therefore, it is recommended that more future
research should be done concerning relationship between closeness centrality
and female student’s academic performance with an enlarged network boundary.
References
Ahuja,
G. (2000). Collaboration networks, structural holes, and innovation: A
longitudinal study. Administrative
Science Quarterly, 45(3), 425–455.
Airasian,
P. W. (1991). Perspectives on measurement instruction. Educational Measurement: Issues and Practice, 10(1), 13–26.
Aktamov,
S., & Zhao, Y. (2014). Impact of network centrality positions on innovation
performance of the firm: evidence from china automobile industry. Business Management and Strategy, 5(1).
Allen,
D. J. (2005). Grades as valid measures of academic achievement of classroom
learning.A Journal of Educational Strategies, 78(5), 13.
Badar,
K., Hite, J. M., & Ashraf, N. (2015). Knowledge network centrality, formal
rank and research performance: evidence for curvilinear and interaction
effects. Scientometrics, 105(3),
1553-1576.
Badar,
K., Frantz, T. L., & Jabeen, M. (2016). Research performance and degree
centrality in co-authorship networks: The moderating role of homophily. Aslib
Journal of Information Management, 68(6), 756–771
Baldwin,
T. T., Bedell, M. D., & Johnson, J. L. (1997). The social fabric of a
team-based M.B.A. program: Network effects on student satisfaction and
performance. Academy of Management
Journal, 40(6), 1369-1397.
Borgatti,
S. P., Everett, M. G., & Freeman, L. C. (2002). UCINET for Windows: Software for Social Network Analysis. Harvard,
MA: Analytic Technologies.
Borgatti,
S. P. (2005). Centrality and network flow. Social
Networks, 27, 55–71.
Brass,
D. J. (1981). Structural relationships, job characteristics, and worker
satisfaction and performance. Administrative
Science Quarterly, 26, 331-348.
Brass,
D. J. (1984). Being in the right place: A structural analysis of individual
influence in an organization. Administrative
Science Quarterly, 29, 518-539.
Burt, R.
S. (1984). Network items and the general social survey. Social Networks, 6, 293-339.
Burt, R.
S. (1992). Structural Holes: The Social
Structure of Competition. Cambridge: Harvard University Press.
Burt, R.
S. (1998). The gender of social capital. Rationality
and Society, 10, 15–46.
Claro,
P.D., Neto, S. A., & Claro, P. B. (2013).The enhancing impact of friendship
networks on sales managers’ performance. BAR,
Rio de Janeiro, 10(2), 158-175.
Cross,
R., & Cummings, J. N. (2004). Tie and network correlates of individual
performance in knowledge-intensive work. Academy
of Management Journal, 47(6), 928–937.
Domhoff, W.G.
(2013) Centrality in networks, and how it is measured. http://whorulesamerica.net/power_elite/centrality.html.
Freeman,
L. C. (1979). Centrality in social networks. Conceptual clarification. Social Networks, 1, 215–239.
Granovetter,
M. S. (1985). Economic action and social structure: The problem of
embeddedness. American Journal of
Sociology, 91(3), 481-510.
Guldner,
C. E., & Stone-Winestock, P. (1995).The use of sociometry in teaching at
the university level.Journal of Group Psychotherapy, Psychodrama &
Sociometry, 47(4), 177-186.
Hopp, J.
W., Iravani, M. S., Liu, F., & Stringer, J. M. (2011). The impact of
discussion, awareness, and collaboration network position on research
performance of engineering school faculty.
Ibarra,
H. (1993). Network centrality, power and innovation involvement: Determinants
of technical and administrative roles. Academy
of Management Journal, 36, 471-501.
Mcfadyen,
A. M., & Cannella, J. A. (2004). Social capital and knowledge creation:
Diminishing returns of the number and strength of exchange relationships. Academy of Management Journal, 47(5),
735–746.
McPherson,
M.., Smith-Lovin, L., & Brashears, M. E. (2006). Social isolation in
America: Changes in core discussion networks over two decades. American Sociological Review,71(3),
353-375.
Mushtaq,
I., & Khan, S. H. (2012). Factors affecting students’ academic performance.
Global Journal of Management and Business
Research, 12 (9).
Myers, R
(1990). Classical and modern regression
with applications (2nd ed.). Boston, MA: Duxbury.
Perry-Smith,
J. E., & Shalley, C. E. (2003). The social side of creativity: A static and
dynamic social network perspective. The
Academy of Management Review, 28(1), 89–106.
Picciano,
A. G. (2002). Beyond student perceptions: Issues of interaction, presence, and
performance in an online course. Journal
of Asynchronous Learning Networks, 6(1), 21-40.
Podolny,
J. M., & Baron, J. N. (1997). Relationships and resources: Social networks
and mobility in the workplace. American
Sociological Review, 62, 673–693.
Rotolo,
D., &Petruzzelli, M. (2013). When does centrality matter? Scientific
productivity and the moderating role of research specialization and
cross-community ties. Journal of
Organizational Behavior, 34(5), 648–670.
Shah R.,
Rahman W., & Abbas S. G., (2015) An analysis of students’ academic
performance: A case study of Sarhad University, Peshawar, Pakistan. Sarhad Journal of Management Sciences, 1(1), 31-41.
Sparrowe,
T. R., & Liden, C. R. (1997). Process and structure in leader-member
exchange. The Academy of Management Review, 22(2), 522-552.
Sparrowe,
R. T., Liden, R. C., Wayne, S.J., & Kraimer, M. L. (2001). Social networks
and the performance of individuals and groups. Academy of Management Journal, 44(2), 316-325.
Spillane,
J. P., & Kim, C. M. (2012). An exploratory analysis of formal school
leaders’ positioning in instructional advice and information networks in
elementary schools.American Journal of
Education, 119 (1), 73-102.
Tichy,
N. M., Tushman, M. L., & Fombrun, C. (1979). Social network analysis for
organizations. Academy of Management
Review, 4(4), 507-519.
Tsai,
W., & Ghoshal, S. (1998). Social capital and value creation: The role of
intra-firm networks. Academy of
Management Journal, 41, 464–476.
Tsai, W.
P. (2001). Knowledge transfer in intra-organizational networks: Effects of
network position and absorptive capacity on business unit innovation and
performance. Academy of Management
Journal, 44(5), 996–1004.
Wasserman,
S., & Faust, K. (1994). Social
Networks Analysis: Methods and applications. Cambridge: Cambridge
University Press.
Webster,
J., & Hackley, P. (1997). Teaching effectiveness in technology-mediated
distance learning. Academy of Management
Journal, 40(6), 282-309.
Yang,
H., & Tang, J. (2003). Effects of social network on students’ performance:
A web-based forum study in Taiwan. JALN,
7(3).Retrieved from: http://www.adesignmedia.
com/onlineresearch/socialnetworksv7n3_yang.pdf.
Yang, K.
S. (2007).Firms’ network position,
industry hierarchy position and innovation and an additional examination on
structural equivalent block models and between-sector position. Academy of
Management Annual Meeting Proceeding, 1-6.
Yan, E.,
& Ding, Y. (2009). Applying centrality measures to impact analysis: A
co-authorship network analysis. Journal
of the American Society for Information Science and Technology, 60(10),
2107–2118.
Zheng,
W. (2008). A social capital perspective of innovation from individuals to nations:
Where is empirical literature directing us? International
Journal of Management Reviews, 10(4), 1–39.