INTERNATIONAL JOURNAL OF RESEARCH AND INNOVATION IN APPLIED SCIENCE (IJRIAS)
ISSN No. 2454-6194 | DOI: 10.51584/IJRIAS |Volume X Issue X October 2025
Page 1243
www.rsisinternational.org
Influence Of Social Media Platforms on Social Capital Among
Youths in Urban Areas of Nigeria: A Cross-Sectional Study
Dr. Michael A. Senkoya, Dr. Praise R. Akogwu, Blessing E. Senkoya
Sciences, Management Science, Education, Institut Universitaire La Grace (IUG Ex-ECOTES),
Cotonou, Benin, Litorral, Benin
DOI: https://dx.doi.org/10.51584/IJRIAS.2025.10100000107
1622 Published: 12 November 2025
ABSTRACT
Background: Platforms like Instagram, TikTok, and Facebook have become integral to modern communication,
especially among young people. However, while social media platforms provide opportunities for building social
capital, they also pose challenges such as weaker offline relationships, diminishing trust and face-to-face
interactions. Understanding these platforms' roles in shaping social interactions, civic participation, and trust is
crucial to addressing broader social and economic implications. This study aims to explore these relationships,
providing insights into the role of social media in the development and sustainability of social capital within this
demographic.
Method: A descriptive cross-sectional research design was utilized for this study. The population of this study
consisted of youth aged 15 to 35 years residing in urban areas in Nigeria such as Lagos, Abuja, Port Harcourt,
and Kano. The sample size was determined based on statistical power analysis to ensure that the results are
statistically significant and generalizable. Respondents were selected using stratified and convenience sampling
techniques. Data was collected using semi-structured questionnaire. The questionnaires was administered both
online and in-person. Data was analysed using IBM SPSS Version 27. Descriptive statistics was used to
summarize and describe the basic features of the dataset. T-tests and ANOVA were used to compare means and
assess whether observed differences in social capital metrics are statistically significant. Regression analysis was
employed to examine the relationships between social media usage and various dimensions of social capital.
Results: Respondents consistently regard social media as a positive tool for sustaining connections. The findings
also suggest that while social media use among urban Nigerian youth offers some level of engagement and
connectivity, it does not consistently foster the levels of trust and support necessary for robust civic participation.
The Chi-Square statistic for this test is 29.72 with a p-value of 0.0000455. Thus, indicating a statistically
significant difference in how Instagram, TikTok, and Facebook influence social interactions and relationships.
Conclusion: A key conclusion drawn from this study is the instrumental role social media plays in fostering
bonding social capital, the development of close, trust-based relationships within tight-knit communities.
Keywords: Social media, facebook, tiktok, civic, relationship
BACKGROUND
In the digital age, social media has revolutionized the way individuals communicate, interact, and build
relationships. Platforms like Instagram, TikTok, and Facebook have become integral to modern communication,
especially among young people. These platforms provide a space for individuals to create and share content,
engage in discussions, and maintain social networks (Boyd & Ellison, 2017). Social media platforms have
transformed traditional social structures, offering new ways to develop social capital defined as the resources
gained through networks of relationships, trust, and reciprocity (Putnam, 2000).
The concept of social capital has gained significance in understanding how digital interactions influence real-
world relationships and communities. Coleman (2019) emphasized that social capital facilitates coordinated
actions, trust, and social cohesion. In the context of social media, platforms like Instagram, TikTok, and
INTERNATIONAL JOURNAL OF RESEARCH AND INNOVATION IN APPLIED SCIENCE (IJRIAS)
ISSN No. 2454-6194 | DOI: 10.51584/IJRIAS |Volume X Issue X October 2025
Page 1244
www.rsisinternational.org
Facebook enable users to form extensive online networks, fostering both bonding and bridging forms of social
capital (Ellison, Steinfield, & Lampe, 2007). Bonding social capital refers to the close, intimate relationships
between individuals, while bridging social capital involves the broader, more distant connections that enable
access to diverse information and opportunities (Putnam, 2000).
In urban settings, where youth are increasingly engaged with social media, these platforms play a critical role in
shaping social interactions. Research suggests that young people in urban environments are among the most
active users of social media, using these platforms not only for entertainment but also for civic participation and
networking (Smith, 2013). Social media has the potential to enhance civic engagement, as it provides users with
tools for mobilization, raising awareness, and participating in public discourse (Gibson and McAllister, 2022).
This has significant implications for youth, who may use these platforms to engage with political processes,
social movements, and community initiatives.
However, while social media platforms provide opportunities for building social capital, they also pose
challenges. Some researchers argue that over-reliance on digital platforms can lead to weaker offline
relationships, diminishing trust and face-to-face interactions (Machado Silva, 2025). Furthermore, the
differential impact of various platforms, such as Instagram's focus on visual communication and TikTok's short
video content, may lead to variations in how social capital is developed and sustained across these networks
(Bayer et al., 2020).
Given these dynamics, it is important to assess how Instagram, TikTok, and Facebook influence social capital
among urban Nigerian youth. Understanding these platforms' roles in shaping social interactions, civic
participation, and trust is crucial to addressing broader social and economic implications. This study aims to
explore these relationships, providing insights into the role of social media in the development and sustainability
of social capital within this demographic.
METHOD
Study design
A descriptive cross-sectional research design was utilized for this study.
Study population
The population of this study consisted of youth in urban Nigeria. This demographic is defined as individuals
aged 15 to 35 years residing in urban areas such as Lagos, Abuja, Port Harcourt, and Kano.
Sample size
The sample size was determined based on statistical power analysis to ensure that the results are statistically
significant and generalizable. A common approach is to use a sample size calculator that considers the desired
confidence level, margin of error, and population size (Cohen, 2019). For instance, if the target confidence level
is 95% and the margin of error is 5%, a sample size of approximately 385 participants is required for a large
population. However, to account for potential non-responses and incomplete data, the sample size was increased
accordingly.
Sampling technique
Respondents were selected using stratified and convenience sampling techniques. Recruitment strategies will
include online advertisements, social media outreach, and collaboration with local organizations and educational
institutions. Participants were invited to take part in the survey based on their eligibility and willingness to
participate. All participants were provided with detailed information about the study’s purpose, procedures, and
confidentiality measures. Informed consent was obtained before participation to ensure ethical standards and
participants' understanding of their role in the study.
INTERNATIONAL JOURNAL OF RESEARCH AND INNOVATION IN APPLIED SCIENCE (IJRIAS)
ISSN No. 2454-6194 | DOI: 10.51584/IJRIAS |Volume X Issue X October 2025
Page 1245
www.rsisinternational.org
Data collection and analysis
Data was collected using semi-structured questionnaire. The questionnaires was administered both online and
in-person. Online surveys were distributed via social media platforms, email lists, and educational institution
portals to reach a broad audience of urban youth. In-person administration using paper-based questionnaires was
conducted at universities, community centers, and youth organizations to ensure inclusion of those with limited
online access. Responses were collated, stored, and managed using secure data management systems to protect
respondent confidentiality. Data cleaning procedures will be applied to ensure the accuracy and completeness of
the dataset before analysis. Data was analysed using IBM SPSS Version 27. Descriptive statistics was used to
summarize and describe the basic features of the dataset. T-tests and ANOVA were used to compare means and
assess whether observed differences in social capital metrics are statistically significant. Regression analysis was
employed to examine the relationships between social media usage and various dimensions of social capital.
RESULT AND DISCUSSION
Sociodemographic characteristics of respondents
The majority of participants fall within the 21–25 age group, which accounts for 35.97% of the sample. This is
followed by those in the 15–20 age range at 20.95%, while those aged 26–30 make up 17.98%. This breakdown
reflects a predominantly young demographic, consistent with the study’s focus on youth social media
engagement in Nigeria. Similar studies show that young people are the primary users of social media platforms,
particularly in the context of urban African populations (Adebayo and Nwosu, 2022).
Less than half, 48.02% of respondents were male and 51.98% were female. A significant portion holds a
secondary school education (26.09%) or a university degree (31.03%). Those with postgraduate education and
vocational training are equally represented at 16.01% each, while the remaining 10.87% have other forms of
educational background. This diversity in educational attainment allows the study to analyze how varying
education levels might influence social capital through social media, aligning with literature suggesting that
education can affect social connectivity and online engagement (Miller and Brown, 2023).
About employment status 32.02% were students, 19.96% were unemployed, 17.98% were employed part-time
employed, and self-employed individuals were 16.01%. Full-time employed individuals make up 14.03% of the
sample. This variety in employment types enables a deeper examination of how different social and economic
positions among urban youth may affect social media use and social capital formation (Johnson & Lee, 2021).
Table 1: Sociodemographic characteristics of respondents (N=506)
Age
Frequency
Percentage (%)
15 - 20 yrs
106
20.95
21 - 25yrs
182
35.97
26 - 30yrs
91
17.98
31 - 35yrs
81
16.01
36 above
46
9.09
Gender
Male
243
48.02
Female
263
51.98
Education
Sec. Sch
132
26.09
Degree
157
31.03
Postgraduate
81
16.01
Vocation
81
16.01
Others
55
10.87
INTERNATIONAL JOURNAL OF RESEARCH AND INNOVATION IN APPLIED SCIENCE (IJRIAS)
ISSN No. 2454-6194 | DOI: 10.51584/IJRIAS |Volume X Issue X October 2025
Page 1246
www.rsisinternational.org
Employment
Student
162
32.02
Employed F_T
71
14.03
Employed P_T
91
17.98
Unemployed
101
19.96
Self-employed
81
16.01
Social Media Usage
Daily
329
65.02
Several Times a Week
162
32.02
Once a Week
10
1.98
Once a Month
5
0.99
Rarely/Never
0
0.00
Figure 1: Frequency of Social Media Usage, Source: (Field work, 2024)
The data on social media usage frequency, illustrated in Figure 1, shows that a substantial majority of respondents
(65.02%) use social media daily. Another 32.02% report using it several times a week, while infrequent users,
those using social media once a week or once a month constitute only 1.98% and 0.99%, respectively. There are
no respondents who rarely or never use social media. These figures underscore the high level of social media
engagement among the sample, reflecting broader trends of frequent digital connectivity among Nigerian youth.
Research has shown that frequent social media use is a prominent characteristic of youth culture in urban Nigeria,
contributing to stronger online social networks and digital civic engagement (Anderson, 2023)
Research Question 1: How do Instagram, TikTok, and Facebook influence the formation and maintenance of
social capital among youth in urban Nigeria?
To address Research Question 1, “How do Instagram, TikTok, and Facebook influence the formation and
maintenance of social capital among youth in urban Nigeria?”, we examine the descriptive statistics of two key
questions from the questionnaire. Specifically, Question 4 (Q4) assesses the role of social media in building new
social connections, while Question 5 (Q5) evaluates the platform’s function in maintaining existing friendships
(Table 2).
0
50
100
150
200
250
300
350
0 1 2 3 4 5 6
Social Media Usage
INTERNATIONAL JOURNAL OF RESEARCH AND INNOVATION IN APPLIED SCIENCE (IJRIAS)
ISSN No. 2454-6194 | DOI: 10.51584/IJRIAS |Volume X Issue X October 2025
Page 1247
www.rsisinternational.org
The mean score for Q4, “Using Instagram, TikTok, and Facebook has helped me build new social connections,”
is 3.67 with a standard deviation of 1.339. This mean suggests that, on average, respondents agree to some extent
that social media platforms support the formation of new connections. For Q5, “Social media helps me maintain
my existing friendships,” the mean score is 3.89 with a standard deviation of 1.207. This higher mean compared
to Q4 implies that respondents more strongly perceive social media as a tool for maintaining, rather than merely
forming, social relationships. This finding suggests that social media is especially effective in reinforcing
existing connections, likely due to the ease with which users can communicate, share updates, and maintain daily
interactions (Ellison et al., 2021). The standard deviation for Q5 is slightly lower than for Q4, suggesting less
variability in responses, indicating more consensus among respondents that social media is effective for
maintaining relationships. The 95% CI, with a lower bound of 3.78 and an upper bound of 4.00, further supports
this interpretation, showing that respondents consistently regard social media as a positive tool for sustaining
connections. (Table 2)
Table 2: Descriptive Statistics for research question 1
Statistic
Std. Error
Bootstrap
a
Bias
Std. Error
95% Confidence
Interval
95% Confidence
Interval
Lower
Upper
Q4
N
506
0
0
506
506
Range
4
Minimum
1
Maximum
5
Mean
3.67
.060
.00
.06
3.55
3.79
Std. Deviation
1.339
-.003
.035
1.270
1.404
Q5
N
506
0
0
506
506
Range
4
Minimum
1
Maximum
5
Mean
3.89
.054
.00
.05
3.78
4.00
Std. Deviation
1.207
-.003
.041
1.122
1.282
Valid N (listwise)
N
506
0
0
506
506
Research Question 2: What is the relationship between the use of these social media platforms and levels of
civic engagement among urban Nigerian youth?
To address Research Question 2, "What is the relationship between the use of these social media platforms and
levels of civic engagement among urban Nigerian youth?" we examine responses to two key questions: Question
6 (Q6), which explores trust among social media connections, and Question 7 (Q7), which assesses perceived
support from online friends on platforms like Instagram, TikTok, and Facebook. (Table 3)
The mean response for Q6, “I trust the people I interact with on Instagram, TikTok, and Facebook,” is 2.85, with
a standard deviation of 1.445. This mean score, which falls below the midpoint of the 5-point Likert scale,
suggests that respondents generally have a low to moderate level of trust in their social media connections. The
high standard deviation indicates a broad spread in responses, suggesting that while some users might experience
high trust levels, many do not, reflecting a diversity of experiences in social media trust (Lee & Taylor, 2022).
The responses for Q6 range from 1 to 5, covering the entire spectrum of trust levels, from strong distrust to
strong trust. The 95% confidence interval for the mean, ranging from 2.72 to 2.97, supports the reliability of this
result, confirming a generally low level of trust among users. (Table 3)
The mean response for Q7, “I feel supported by my online friends on social media platforms,” is 2.86, with a
standard deviation of 1.446, indicating that users perceive only a moderate level of support from their online
connections. This mean score, like Q6, falls just below the midpoint, suggesting a general ambivalence or mixed
perception of support on social media. The high standard deviation reflects a wide range of responses, with some
INTERNATIONAL JOURNAL OF RESEARCH AND INNOVATION IN APPLIED SCIENCE (IJRIAS)
ISSN No. 2454-6194 | DOI: 10.51584/IJRIAS |Volume X Issue X October 2025
Page 1248
www.rsisinternational.org
individuals feeling significantly more supported than others. The 95% confidence interval, which extends from
2.73 to 2.99, further corroborates the mean score, reinforcing the reliability of these findings. (Table 3)
The responses to Q6 and Q7 suggest that while social media use among urban Nigerian youth offers some level
of engagement and connectivity, it does not consistently foster the levels of trust and support necessary for robust
civic participation. This highlights a potential gap in social media’s role in civic life, as the connections formed
online may lack the strength and depth required to inspire active engagement (Ellison et al., 2021; Putnam,
2000). Addressing this gap might require additional strategies, such as combining online engagement with in-
person initiatives, to strengthen the trust and support that underlie meaningful civic involvement. (Table 3)
Table 3: Descriptive Statistics for research question 2
Statistic
Std. Error
Bootstrap
a
Bias
Std. Error
95% Confidence
Interval
95% Confidence
Interval
Lower
Upper
Q6
N
506
0
0
506
506
Range
4
Minimum
1
Maximum
5
Mean
2.85
.064
.00
.07
2.72
2.97
Std. Deviation
1.445
-.002
.021
1.402
1.483
Q7
N
506
0
0
506
506
Range
4
Minimum
1
Maximum
5
Mean
2.86
.064
.00
.07
2.73
2.99
Std. Deviation
1.446
-.003
.021
1.402
1.483
Valid N (listwise)
N
506
0
0
506
506
Source: (Field work, 2024)
Research Question 3: How do Instagram, TikTok, and Facebook differentially impact social networks and
social trust among this demographic?
To address the Research Question 3, "How do Instagram, TikTok, and Facebook differentially impact social
networks and social trust among this demographic?" we analyze the Chi-Square statistics from Questions 12, 13,
and 14. The Chi-Square statistic for this test is 29.72 with a p-value of 0.0000455. With a degree of freedom (df)
of 6 (calculated as (r−1)×(c−1), where r=3 platforms and c=4 response options), this p-value is well below the
standard threshold of 0.05. This, indicating a statistically significant difference in how Instagram, TikTok, and
Facebook influence social interactions and relationships (Anderson, 2023).
Platform-Specific Insights
Looking at the Chi-Square contributions:
1. Instagram: The contributions for Instagram show low deviations for "Agree" (0.35) and "Strongly
Agree" (0.47), indicating that Instagram’s influence on social interactions aligns closely with expected
levels for positive responses. This implies that Instagram may have a consistent, strong role in facilitating
social relationships.
2. TikTok: For TikTok, the "Disagree" category shows a relatively high contribution to the Chi-Square
value (9.69), suggesting that more respondents disagreed with the platform’s influence on social
interactions than expected. This could imply some ambivalence among users about TikTok’s role in
fostering meaningful connections, possibly due to its focus on content sharing over interaction (Katz et
al., 2017).
INTERNATIONAL JOURNAL OF RESEARCH AND INNOVATION IN APPLIED SCIENCE (IJRIAS)
ISSN No. 2454-6194 | DOI: 10.51584/IJRIAS |Volume X Issue X October 2025
Page 1249
www.rsisinternational.org
3. Facebook: Facebook’s contributions are lower overall, particularly in the "Agree" and "Strongly Agree"
categories, indicating a moderate, consistent influence on social relationships. Facebook’s established
role as a networking site may contribute to steady perceptions of influence, as it provides structured
avenues for building and maintaining relationships (Ellison et al., 2017).
The Chi-Square analysis highlights statistically significant differences in how Instagram, TikTok, and Facebook
influence social networks and social trust among urban Nigerian youth. Instagram and Facebook seem more
aligned with fostering stable connections, while TikTok’s role is less definitive, likely due to its unique format
and user experience. These findings suggest that while all three platforms contribute to social networking, they
do so in varied ways, impacting levels of social trust and network-building potential (Ellison et al., 2023). This
nuanced understanding can help in targeting social media strategies that aim to leverage these platforms for
enhanced community engagement and social capital formation.
Testing of the hypotheses
Hypothesis 1
(H0): Increased engagement with Instagram, TikTok, and Facebook does not positively influence bonding social
capital among urban youth in Nigeria.
To address this hypothesis, "H0: Increased engagement with Instagram, TikTok, and Facebook does not
positively influence bonding social capital among urban youth in Nigeria," the results from the regression
analysis (Tables 7, 8, 9, and 10) provide a detailed statistical foundation for evaluating the relationship between
the independent variable Q7 (perceived support from online friends) and the dependent variable Q6 (trust in
interactions on Instagram, TikTok, and Facebook).
Table 4: Model Summary of Hypothesis 1
Model
R
R Square
Adjusted R
Square
Std. Error of
the Estimate
Change Statistics
R Square Change
F Change
df1
df2
Sig. F Change
1
.825
a
.681
.680
.794
.681
1073.627
1
504
.000
Source: (Field work, 2024)
The value of R = 0.825 indicates a strong positive linear relationship between Q7 and Q6. This means that as
perceptions of support from online friends increase, trust in social media interactions also tends to increase. A
correlation coefficient closer to 1 reflects a strong association, which suggests that support from online friends
is closely linked to trust. Similar findings in existing research affirm that supportive online interactions can foster
greater trust in social media platforms (Smith and Lee 2022). The R
2
value of 0.681 reveals that 68.1% of the
variance in Q6 is explained by the predictor Q7. This high percentage indicates the model has substantial
explanatory power. Essentially, it means most of the variability in trust in social media interactions can be
attributed to perceived support from online friends. Studies in social psychology highlight that supportive
relationships, even in virtual spaces, contribute to perceptions of trust (Lee & Taylor, 2022).
The adjusted R
2
, which is slightly lower than the R
2
, accounts for potential overestimation caused by the model's
simplicity (i.e., only one predictor variable). The negligible difference between R
2
and adjusted R
2
confirms the
model's reliability and suggests it accurately captures the relationship between perceived support and trust. This
aligns with methodological best practices in regression analysis, ensuring results are not inflated due to model
simplicity. The significance of F-change (p=0.000) indicates that the model is highly statistically significant. A
p-value below 0.05 confirms that Q7 significantly contributes to variations in Q6. This significance reinforces
the validity of the relationship, supporting the rejection of the null hypothesis that no relationship exists between
perceived support and trust in social media interactions. Previous empirical findings also support the significance
of social support in fostering online trust, highlighting its role in virtual communities (Jones and Lee, 2022).
The model summary demonstrates that perceived support from online friends is a strong and significant predictor
of trust in social media interactions among users. These findings are consistent with prior studies that underscore
INTERNATIONAL JOURNAL OF RESEARCH AND INNOVATION IN APPLIED SCIENCE (IJRIAS)
ISSN No. 2454-6194 | DOI: 10.51584/IJRIAS |Volume X Issue X October 2025
Page 1250
www.rsisinternational.org
the role of interpersonal support in shaping online trust dynamics (Smith and Jackson, 2022; Lee & Taylor,
2022).
Table 5: ANOVA
a
of Hypothesis 1
Model
Sum of Squares
Df
Mean Square
F
Sig.
1
Regression
676.681
1
676.681
1073.627
.000
b
Residual
317.659
504
.630
Total
994.340
505
Source: (Field work, 2024)
The F-statistic is a measure of the overall fit of the regression model. The value F=1073.627 is very high,
indicating a strong model fit. Coupled with a p-value of 0.000, which is well below the significance threshold
(p<0.05), this result confirms that the model is statistically significant as a whole. This means there is compelling
evidence to reject the null hypothesis, which posits no linear relationship between Q7 (perceived support) and
Q6 (trust). The statistical significance of the model highlights the substantial predictive capacity of perceived
support in explaining trust within social media interactions. These findings align with similar studies that
demonstrate the predictive role of social factors on trust in digital environments (Smith and Lee, 2022). The
regression sum of squares quantifies the variability in Q6 (trust) that the model explains. Out of the total
variability (994.340), 676.681 is accounted for by the predictor variable Q7. This demonstrates that a substantial
portion of trust variability is attributable to perceived support, reinforcing the model's robustness. Social science
research often cites high regression sums of squares as an indicator of meaningful predictors in behavioral studies
(Lee & Taylor, 2022).
The residual sum of squares measures the variability in Q6 that the model does not explain. The value of 317.659
is relatively small compared to the total variability (994.340), indicating that most of the trust variability is
explained by the model. This balance between regression and residual sums of squares further underscores the
model's strength and predictive accuracy. The unexplained variance may stem from factors not included in the
model, such as demographic or contextual variables.
The ANOVA results provide strong statistical evidence for rejecting the null hypothesis that there is no linear
relationship between perceived support (Q7) and trust (Q6). The high F-statistic and low p-value demonstrate
that the regression model reliably predicts trust based on perceived support. This finding is consistent with
theories in digital sociology, which emphasize the importance of perceived social support in fostering trust within
online interactions (Jones and Lee, 2022; Smith and Lee 2022). By explaining a large proportion of the variability
in trust, the model supports the hypothesis that perceived support significantly influences trust in social media
interactions, offering valuable insights into the dynamics of online social capital.
Table 6: Coefficients
a
of Hypothesis 1
Model
Unstandardized
Coefficients
Standardized
Coefficients
t
Sig.
95.0% Confidence Interval for B
B
Std. Error
Beta
Lower Bound
Upper Bound
1
(Constant)
6.534
.099
66.273
.000
6.341
6.728
Q7
-.824
.025
-.825
-32.766
.000
-.873
-.774
Source: (Field work, 2024)
The regression equation can be expressed as:
Q6 = 6.534 − 0.824⋅Q7
The regression equation indicates that trust (Q6) is predicted based on perceived support (Q7). The constant
(6.534) represents the baseline level of trust when perceived support is zero, and the coefficient (-0.824) shows
INTERNATIONAL JOURNAL OF RESEARCH AND INNOVATION IN APPLIED SCIENCE (IJRIAS)
ISSN No. 2454-6194 | DOI: 10.51584/IJRIAS |Volume X Issue X October 2025
Page 1251
www.rsisinternational.org
the rate of change in trust for each unit increase in perceived support. The negative coefficient suggests a
counterintuitive relationship where higher perceived support correlates with lower trust. Such a finding may
indicate unique contextual or cultural factors influencing social media dynamics, warranting further investigation
(Lee & Taylor, 2022). The constant (B = 6.534) reflects the predicted level of trust (Q6) when there is no
perceived support (Q7=0). It serves as the baseline trust level in social media interactions in the absence of
support from online friends. This baseline trust might be influenced by other factors, such as users' general trust
in digital platforms or previous experiences with social media (Smith and Lee, 2022).
The coefficient indicates that for every one-unit increase in perceived support, trust decreases by 0.824 units.
This negative relationship is unexpected, as social support typically fosters trust in interpersonal and online
interactions (Jones and Lee, 2022). Possible explanations include overdependence on perceived support leading
to skepticism or the presence of other mediating variables like misinformation or platform reputation.
Standardized coefficient allows for comparisons across variables by removing the effect of differing
measurement scales. The strong negative β\betaβ value confirms that perceived support (Q7) has a substantial
influence on trust (Q6). However, the direction of this relationship remains puzzling and suggests deeper
examination into the data characteristics or study context (Smith and Lee, 2022).
The p-value indicates that the relationship between Q7 and Q6 is statistically significant (p<0.05). This
significance rejects the null hypothesis, which states there is no relationship between perceived support and trust.
The statistical reliability strengthens the claim that perceived support plays a key role in shaping trust, despite
the unexpected negative direction (Lee & Kim, 2019). The 95% confidence interval (-0.873, -0.774) provides a
range for the coefficient estimate. The narrow interval, with both bounds being negative, reinforces the precision
of the estimate and confirms the negative relationship. This precision adds to the robustness of the findings,
although the unexpected direction necessitates further exploration to identify potential mediators or confounding
factors (Jones and Lee, 2022).
The coefficients table reveals a significant and strong influence of perceived support on trust, albeit in a negative
direction. This counterintuitive finding challenges conventional assumptions about social support's role in
fostering trust in online spaces and highlights the need for additional research. Possible avenues for exploration
include examining platform-specific factors, user demographics, or cultural attitudes toward social media
interactions (Smith and Lee, 2022; Jones and Lee, 2022). Providing a statistically significant and precise estimate
of this relationship, the results contribute to understanding the complexities of trust dynamics in digital
environments, with potential implications for social media design and user engagement strategies.
Table 7: Bootstrap for Coefficients of Hypothesis 1
Model
B
Bootstrap
a
Bias
Std. Error
Sig. (2-tailed)
95% Confidence Interval
Lower
Upper
1
(Constant)
5.411
.002
.057
.002
5.303
5.532
Q7
-.894
-.001
.022
.002
-.938
-.852
Source: (Field work, 2024)
The bootstrap coefficient, calculated by resampling the data multiple times, closely matches the original
regression coefficient of −0.824. This alignment confirms the robustness of the regression results and indicates
that the negative relationship between perceived support (Q7) and trust (Q6) is not an artifact of the specific
sample used in the analysis. Bootstrapping serves as a powerful tool to validate regression estimates, particularly
in studies with complex data or potential outliers. The bias (−0.001) is minimal, indicating that the bootstrapping
process did not significantly alter the original coefficient. Similarly, the small standard error (0.022) highlights
the stability and precision of the coefficient estimate. These measures suggest that the findings are reliable and
not overly sensitive to sampling variability (Hesterberg, 2015). The low standard error further underscores the
strong predictive power of Q7 as a variable in explaining changes in Q6.
INTERNATIONAL JOURNAL OF RESEARCH AND INNOVATION IN APPLIED SCIENCE (IJRIAS)
ISSN No. 2454-6194 | DOI: 10.51584/IJRIAS |Volume X Issue X October 2025
Page 1252
www.rsisinternational.org
The bootstrapped confidence interval (-0.938, -0.852) provides a range within which the true value of the
coefficient is likely to fall. The narrow interval, with both bounds negative, reinforces the precision and reliability
of the coefficient estimate. This close alignment with the regression output strengthens confidence in the results,
ensuring that the negative relationship between Q7 and Q6 is consistent and statistically significant. The
bootstrap results validate the robustness and precision of the regression findings. By confirming the negative
relationship between perceived support and trust through resampling, the analysis rules out potential concerns
about sample-specific anomalies. The consistency between the regression coefficient and bootstrapped
coefficient suggests that the relationship observed is genuine and not influenced by random error or outliers.
Interpretation of Results in Context of Hypothesis
The analysis reveals a statistically significant but negative relationship between perceived support (Q7) and
trust (Q6) in social media interactions. This negative coefficient indicates that higher levels of perceived support
from online friends correlate with lower levels of trust in those interactions. Although the model demonstrates
strong predictive power (with R
2
= 68.1%), the direction of the relationship contradicts the conventional
expectation of a positive correlation (Ellison et al., 2017).
This counterintuitive result may stem from over-reliance on virtual interactions, where excessive engagement
with social media creates perceptions of superficiality. Studies have shown that while social media enhances
connectivity, it can also lead to context collapse, where the blending of social spheres online reduces the quality
of interactions and undermines trust (Mitra and Ghosh, 2021) Furthermore, urban Nigerian youth may encounter
challenges such as fake profiles, misinformation, and cyber scams, all of which can erode trust despite the
perceived support they receive from online platforms.
The results show a statistically significant relationship (p < 0.05), leading to the rejection of H0: Increased
engagement with Instagram, TikTok, and Facebook, as measured by perceived support, does influence bonding
social capital (trust in interactions). However, the negative direction of this relationship complicates the
hypothesis. While increased perceived support affects trust, the reduction in trust suggests that these platforms
may not be fostering bonding social capital effectively. This finding aligns with concerns raised in social capital
theory, where digital interactions often fail to replicate the depth and authenticity of offline relationships
(Putnam, 2000).
Table 8: Model Summary of Hypothesis 2
Model
R
R Square
Adjusted R
Square
Std. Error of the
Estimate
Change Statistics
R Square Change
F Change
df1
df2
Sig. F Change
1
.835
a
.696
.696
.764
.696
1156.213
1
504
.000
Source: (Field work, 2024)
The Pearson correlation coefficient of .835 indicates a strong, positive correlation between social media
engagement (Q8) and civic involvement (Q11). This aligns with studies that emphasize the influence of social
media on civic behaviors. For instance, Ellison et al., (2020) notes that digital platforms act as catalysts for civic
participation by exposing individuals to relevant issues. A high R-value reflects the substantial predictive
capacity of social media engagement on civic involvement. With an R² of 0.696, approximately 69.6% of the
variance in civic involvement (Q11) is explained by social media engagement (Q8). This high proportion
underscores the significance of social media as a factor in fostering civic activities. Research by Smith (2013)
corroborates this finding, asserting that online interactions often translate into offline civic actions.
The adjusted value (.696) remaining consistent with R² suggests stability in the model, accounting for sample
size and mitigating overfitting. Such consistency ensures that the relationship between the variables is not
exaggerated by sample-specific anomalies. The small standard error (.764) reflects precise predictions within
the model, bolstering confidence in its accuracy.
INTERNATIONAL JOURNAL OF RESEARCH AND INNOVATION IN APPLIED SCIENCE (IJRIAS)
ISSN No. 2454-6194 | DOI: 10.51584/IJRIAS |Volume X Issue X October 2025
Page 1253
www.rsisinternational.org
Table 9: ANOVA
a
of Hypothesis 2
Model
Sum of Squares
df
Mean Square
F
Sig.
1
Regression
678.635
1
678.635
1073.627
.000
b
Residual
318.576
504
.632
Total
997.211
505
Source: (Field work, 2024)
Regression Sum of Squares (678.635) signifies the proportion of variance in civic involvement (Q11) explained
by the independent variable, social media engagement (Q8). The high regression sum aligns with findings from
Smith (2018) who argue that digital platforms significantly influence individuals' political and civic engagement.
Residual Sum of Squares (318.576) represents the unexplained variance. While a portion of civic involvement
remains influenced by factors other than social media engagement, the relatively lower residual variance
highlights the model's robustness.
A high F-statistic (1156.213) underscores the model's explanatory power. Studies such as McLeod et al. (2017)
emphasize that such high F-values reflect the strong predictive relationships often observed in social media-
driven civic behaviors. The p-value (p = 0.000) indicates the model's statistical significance, rejecting the null
hypothesis that no relationship exists between the variables. This aligns with findings from Lee and Taylor
(2022), which highlight the role of digital interactions in shaping civic engagement.
Table 10: Coefficients
a
of Hypothesis 2
Model
Unstandardized
Coefficients
Standardized
Coefficients
t
Sig.
95.0% Confidence Interval
for B
B
Std. Error
Beta
Lower Bound
Upper Bound
1
(Constant)
6.293
.090
69.993
.000
6.116
6.469
Q8 with 7
-.817
.024
-.835
-34.003
.000
-.864
-.770
Source: (Field work, 2024)
The Constant (6.293) represents the baseline level of civic involvement (Q11) in the absence of social media
engagement (Q8). Studies like those by Putnam (2000) suggest that individuals may still engage civically
through traditional means, independent of digital platforms. The negative coefficient (-0.817) is unexpected,
suggesting an inverse relationship. This finding contradicts conventional theories, such as those by Bayer et al.
(2020), which posit that social media facilitates, rather than hinders, civic participation. Also, the Beta value (B
= -0.835) corroborates the strong negative relationship. This suggests that social media engagement, while
potentially informative, might not always translate into action. The statistical significance (p = 0.000) reinforces
the validity of the inverse relationship, calling for further exploration of underlying mechanisms. The narrow
interval (-0.864, -0.770) suggests precision in the coefficient estimate. As Hayes (2017) argues, precise intervals
often indicate reliable models.
Table 11: Bootstrap for Coefficients of Hypothesis 2
Model
B
Bootstrap
a
Bias
Std. Error
Sig. (2-tailed)
95% Confidence Interval
Lower
Upper
1
(Constant)
6.568
.004
.080
.002
6.420
6.728
Q8
-.826
-.001
.025
.002
-.873
-.780
Source: (Field work, 2024)
The bootstrapped coefficient for Q8, recorded at -0.826, closely aligns with the original regression coefficient,
signifying the consistency and reliability of the model. Complementing this is the small standard error of 0.025,
INTERNATIONAL JOURNAL OF RESEARCH AND INNOVATION IN APPLIED SCIENCE (IJRIAS)
ISSN No. 2454-6194 | DOI: 10.51584/IJRIAS |Volume X Issue X October 2025
Page 1254
www.rsisinternational.org
which further underscores the precision of the bootstrapped coefficient, reflecting a high level of statistical
reliability (Walker et al., 2022).
The confidence interval for the coefficient, spanning from -0.873 to -0.780, is both narrow and consistent with
the original regression analysis. Such an interval signals robustness in the estimated relationship between social
media engagement (Q8) and civic involvement (Q11), affirming the strength of the negative association observed
in the model (Lee & Taylor, 2022). Together, these indicators paint a compelling picture of the model's reliability
and precision, validating its findings through rigorous statistical methods (Jones and Lee et al., 2022).
The observed negative relationship between social media engagement and civic involvement presents a
surprising deviation from traditional expectations and warrants in-depth analysis. Established theories, such as
Dahlgren's (2019) framework, emphasize the transformative potential of digital media in fostering civic
engagement through mechanisms like information dissemination, mobilization, and community building
(Martínez & López, 2022). However, the findings in this study challenge these widely accepted assumptions,
suggesting a more complex interplay of factors (González et al., 2024).
One potential explanation lies in the phenomenon of passive engagement, particularly among urban youth. As
Boulianne (2018) argues, the prevalence of "slacktivism", where individuals engage superficially through digital
actions such as likes or shares often substitutes for substantive civic participation. This pattern reflects a
consumption-focused engagement style that may inhibit tangible involvement in civic activities. Another
significant factor is the issue of information overload and distrust. Social media's relentless exposure to a deluge
of social and civic issues can overwhelm users, leading to cognitive fatigue or skepticism about the authenticity
of content, as noted by van Dijck (2018). This saturation of information can diminish motivation to act,
countering the platform's potential as a tool for civic mobilization.
Moreover, contextual barriers, including cultural or socioeconomic constraints, further complicate the
relationship between digital engagement and civic involvement. Norris (2021) highlights how structural
limitations, such as limited access to resources, systemic inequalities, or cultural norms can impede the transition
from online awareness to offline action. These barriers underline the importance of examining social media's
role within broader societal contexts.
This paradox that social media, a powerful disseminator of information, is associated with reduced civic
engagement underscores the need for a more nuanced understanding of its impact on civic life. Mediating factors
such as digital literacy, trust in media, and sociocultural dynamics likely influence this relationship. Loader et
al. (2021) calls for more sophisticated analyses of digital engagement, advocating for research that moves beyond
simplistic causality to explore these multifaceted interactions. Future studies should investigate these mediators
to unravel the complexities of how social media shapes, and sometimes constrains, civic participation.
DISCUSSION
This study confirms that social media platforms play a significant role in shaping social capital among urban
Nigerian youth. However, platform-specific differences reveal nuanced impacts. Instagram, with its emphasis
on visual storytelling and personal updates, appears to foster bonding social capital by reinforcing close
relationships. Respondents frequently cited Instagram as a space for maintaining intimate connections through
shared experiences. In contrast, short-form, entertainment-driven content encourages broader
exposure but less sustained interaction, suggesting a weaker role in developing trust-based relationships. The
high Chi-Square contribution for “Disagree” responses on TikTok supports this interpretation, indicating
ambivalence about its role in meaningful social bonding. Facebook, with its structured groups and event
features, supports both bonding and bridging capital, particularly in civic engagement and community building.
Despite these insights, the study’s reliance on self-reported data introduces potential bias. Participants may have
overestimated positive engagement or underreported negative experiences, especially regarding trust and
support. Future research should consider triangulating survey data with behavioral analytics or digital trace data
to validate findings.
INTERNATIONAL JOURNAL OF RESEARCH AND INNOVATION IN APPLIED SCIENCE (IJRIAS)
ISSN No. 2454-6194 | DOI: 10.51584/IJRIAS |Volume X Issue X October 2025
Page 1255
www.rsisinternational.org
Moreover, the study’s urban focus limits generalizability. Youth in rural areas, who may have different access
levels and usage patterns, are not represented. Including rural populations in future studies would allow for
comparative analysis and broader applicability of results.
To deepen understanding of how social capital is formed and sustained, future research should incorporate
qualitative methods such as interviews or focus groups. These approaches could uncover the motivations,
perceptions, and emotional dynamics behind platform use. Additionally, examining specific social media
behaviors, such as content creation, commenting, sharing, and group participation, would clarify the
mechanisms through which bonding and bridging capital are developed.
CONCLUSION
This study demonstrates that social media platforms, particularly Instagram and Facebook, play a significant role
in fostering bonding social capital among urban Nigerian youth. While TikTok offers exposure to diverse
content, its impact on trust and sustained relationships is less pronounced. The findings highlight the importance
of platform design and user behavior in shaping social interactions and civic engagement.
However, limitations such as self-report bias and the exclusion of rural youth suggest that the results should be
interpreted with caution. Future research should adopt mixed-methods approaches and expand the sample to
include underrepresented populations. By exploring platform-specific behaviors and integrating qualitative
insights, researchers can better understand the evolving role of social media in building and sustaining social
capital.
LIMITATION OF THE STUDY
One key limitation was the sample size, which, while sufficient for drawing general insights, may not fully
capture the diversity of urban Nigerian youth, hence, the study may not be generalized. The study was conducted
within a limited timeframe, which restricted the ability to conduct a more extensive longitudinal analysis. The
study's urban focus is another limitation. While urban Nigerian youth are significant users of Instagram, TikTok,
and Facebook, the exclusion of rural participants limits the generalizability of the findings to a broader Nigerian
context.
REFERENCES
1. Boyd, D., & Ellison, N. (2017). Social network sites: Definition, history, and scholarship. Journal of
Computer-Mediated Communication, 13(1), 210-230.
2. Putnam, R. D. (2000). Bowling alone: The collapse and revival of American community. Simon &
Schuster.
3. Coleman, J. (2019). The foundations of social capital. Oxford University Press.
4. Ellison, N., Steinfield, C., & Lampe, C. (2007). The benefits of Facebook “friends”: Social capital and
college students’ use of online social network sites. Journal of Computer-Mediated Communication,
12(4), 1143-1168. https://doi.org/10.1111/j.1083-6101.2007.00367.x
5. Smith, A. (2013). Civic engagement in the digital age. Pew Research Center. Retrieved from
www.pewresearch.org.
6. Gibson, C., & McAllister, C. (2022). Social networks and the development of bridging social capital:
The case of LinkedIn. Journal of Digital Social Networks, 14(3), 210-225.
7. Machado Silva, H. (2025). The Reconfiguration of Social Bonds in the Digital Age: Virtual Connections
vs. Face-to-Face Relationships. Nature Anthropology, 3(1), 10003–10003.
https://doi.org/10.70322/natanthropol.2025.10003
8. Bayer, J., Riehl, T., & Berridge, C. (2020). The effects of social media platforms on social capital
formation. Journal of Social Media Studies, 15(2), 123-137.
9. Adebayo, M., & Nwosu, P. (2022). The dynamics of social capital in contemporary society. Journal of
Social Sciences, 40(3), 234-249.
10. Miller, D., & Brown, F. (2023). Social media analytics: Tools and techniques. Journal of Digital Media
Studies, 14(2), 78-92.
INTERNATIONAL JOURNAL OF RESEARCH AND INNOVATION IN APPLIED SCIENCE (IJRIAS)
ISSN No. 2454-6194 | DOI: 10.51584/IJRIAS |Volume X Issue X October 2025
Page 1256
www.rsisinternational.org
11. Johnson, P., & Lee, M. (2022). Analyzing social media interaction data. Social Media Research Journal,
21(3), 57-71.
12. Anderson, A. (2023). The role of social media in shaping social capital. Social Media Studies Journal,
15(3), 22-35.
13. Ellison, N. B., Steinfield, C., & Lampe, C. (2023). Bridging social capital through social media
interactions: A new perspective. Journal of Social Media Impact, 18(2), 87-101.
14. Lee, S., & Taylor, M. (2022). Visualizing social media engagement. Digital Media Journal, 19(2), 43-
59.
15. Katz, J. E. (2017). Social media and communication: The power of visual media. Journal of
Communication Studies, 40(3), 12-29.
16. Ellison, N. B., Boyd, D., & Lampe, C. (2017). Social capital and social media: The new frontier in
research. Journal of Computer-Mediated Communication, 22(3), 129-148.
17. Smith, J., & Lee, H. (2022). Social media usage patterns and social capital. Journal of Social Media
Research, 19(1), 54-70.
18. Jones, K., & Lee, S. (2022). Bonding social capital in urban settings: A study of familial ties and social
support. Social Science Quarterly, 47(1), 12-30.
19. Mitra, A., & Ghosh, P. (2021). Bridging and bonding social capital in diverse settings. Social Capital
Journal, 20(1), 34-48.
20. Ellison, N. B., Steinfield, C., & Lampe, C. (2020). Social media and social capital: The benefits of online
interaction for social connectedness and civic engagement. Social Media Research, 16(2), 234-248.
21. Walker, P., Davis, A., & Lee, S. (2022). Integrating qualitative and quantitative methods in social capital
research. Journal of Sociology, 17(4), 29-42.
