Correlation Analysis

Table 4a. Correlation Matrix of Psychosocial Determinants and Nicotine Dependence (N = 202)
Variable Mean SD 1 2 3 4 5 6 7 8 9

1. Nicotine
1.96 1.82 —
Dependence

2. Perceived
3.00 1.01 0.01 —
Benefits

3. Perceived
3.09 1.15 0.09 0.02 —
Barriers

4. Perceived
3.30 0.98 -0.06 0.07 0.18* —
Self-Efficacy

5. Peer Influence 3.42 1.12 0.28*** 0.34*** 0.12 -0.21*** —

6. Social Norms 3.65 1.29 0.15 0.28*** 0.10 -0.15*** 0.22*** —

7. Digital
2.97 1.32 0.13 0.26*** 0.20*** -0.08 0.18*** 0.16*** —
Exposure

0.0
8. Age 20.23 2.23 -0.11 0.04 -0.03 0.06 -0.08 -0.02 —
0

0.0
9. Sex — — 0.08 0.05 0.02 0.01 0.11 0.06 -0.09 —
4
Note: Sex coded as 1 = Male, 2 = Female. Values in bold indicate correlations with nicotine dependence that are statistically significant.
Significance indicators: * p < 0.05, ** p < 0.01, *** p < 0.001. Correlations were computed using Spearman's rank-order correlation.
 Table 4a examines the relationships between selected psychosocial

determinants and nicotine dependence among adolescents and young adults in

Tacloban City. Among all variables examined, peer influence was the only factor

significantly associated with nicotine dependence (r = 0.28, p < .001). Although the

strength of the correlation was weak, the positive direction indicates that respondents

who reported greater peer influence also tended to exhibit higher levels of nicotine

dependence. This finding is consistent with previous studies identifying peer

dynamics as one of the strongest influences on youth nicotine use. Laninga-Wijnen

and Veenstra (2021), Vitória et al. (2020), and East et al. (2021) emphasized that

adolescents are influenced by peer selection, imitation, conformity, and descriptive

norms, particularly when close friends or siblings smoke or vape. Likewise, Belmonte

et al. (2025) reported that peer influence contributes to electronic cigarette use

intentions among Filipino adolescents. Extending these findings, the present study

suggests that peer influence is associated not only with nicotine initiation or intention

to use but also with greater nicotine dependence among adolescents and young adults

in Tacloban City.

The observed relationship also aligns with Bandura's Social Cognitive Theory

(SCT), which proposes that behaviors are learned and reinforced through observation,

modeling, and social reinforcement (Bandura, 1986). Within the context of nicotine

use, adolescents who are regularly exposed to peers who smoke or vape may perceive

these behaviors as acceptable or rewarding, increasing the likelihood of continued

use. Consistent with the study's conceptual framework, peer influence functions as a

contextual determinant that shapes nicotine-related beliefs and behaviors through

interpersonal interactions. Therefore, the significant correlation observed in this study

suggests that nicotine dependence among adolescents and young adults is closely

embedded within their social environment, where repeated exposure to peer behaviors

and social acceptance may reinforce continued nicotine use.

In contrast, perceived benefits were not significantly associated with nicotine

dependence (r = 0.01, p = .892). This finding differs from the expectations of Pender's

Health Promotion Model (HPM), which proposes that individuals are more likely to

engage in or maintain a behavior when they perceive greater benefits from it. Previous

literature has identified stress relief, emotional comfort, improved mood, social

belonging, and the perception that vaping is less harmful than combustible cigarettes

as commonly perceived benefits of nicotine use (Resano et al., 2022; Abano et al.,

2025; Serra et al., 2023; Lacsa, 2025). Similarly, product appeal, flavored products,

attractive device designs, curiosity, and social media marketing have been associated

with more favorable attitudes toward vaping and smoking (Al-Khalifa & Rodriguez,

2025; Pan et al., 2022; Ibukun et al., 2025). However, the absence of a significant

relationship in the present study suggests that although perceived benefits may

contribute to nicotine experimentation or continued use, they may not independently

0explain differences in nicotine dependence once nicotine use has already been

established. This interpretation is consistent with the reviewed literature, which

primarily associates perceived benefits with nicotine initiation, experimentation, and

behavioral intention rather than with dependence severity.

Similarly, perceived barriers were not significantly correlated with nicotine

dependence (r = 0.09, p = .187). According to the HPM, perceived barriers represent

obstacles that reduce an individual's willingness or ability to modify health behaviors.

Previous studies have suggested that withdrawal symptoms, nicotine cravings, peer
routines, product accessibility, social acceptability, and weak policy enforcement may

discourage smoking or vaping cessation (Donaldson et al., 2022; Vassey et al., 2022;

Marsh et al., 2021; Congressional Policy and Budget Research Department, 2025;

Global Tobacco Control, 2024; Sese, 2023). East et al. (2025) further noted that

vaping becomes more difficult to discontinue when it is normalized within peer

groups. Despite these observations, the present findings indicate that perceived

barriers alone were not associated with higher nicotine dependence. One possible

explanation is that many respondents may not have actively attempted to quit nicotine

use or experienced withdrawal symptoms, thereby limiting the relevance of perceived

barriers to their current dependence levels. Consequently, perceived barriers may be

more closely related to cessation behavior than to dependence severity itself.

Likewise, perceived self-efficacy demonstrated a weak negative but

non-significant relationship with nicotine dependence (r = –0.06, p = .421). Although

not statistically significant, the negative direction of the relationship is consistent with

Bandura's (1986) concept of self-efficacy, which suggests that individuals with greater

confidence in resisting nicotine use or quitting are generally more likely to adopt

healthier behaviors. Smith and Hilton (2023) likewise argued that repeated exposure

to peers who vape and nicotine-related social media content may weaken adolescents'

confidence to refuse nicotine. Furthermore, intervention studies by Tupas and Agreda

(2020) and Emen et al. (2020) demonstrated that strengthening refusal skills and

self-efficacy can improve health-promoting behaviors among youth. The present

findings, however, suggest that self-efficacy alone may not sufficiently explain

nicotine dependence among current users. Instead, dependence may be influenced by

a broader combination of physiological addiction, habitual behavior, and social

reinforcement. Consequently, interventions that focus exclusively on increasing

self-efficacy may have limited effectiveness unless they are complemented by

strategies addressing peer influence and environmental factors.

Social norms also did not demonstrate a statistically significant relationship

with nicotine dependence (r = 0.15, p = .055). Although the correlation was positive,

it did not reach the level of statistical significance. Previous literature has shown that

vaping has increasingly been portrayed as modern, socially acceptable, and less

harmful than cigarette smoking (Hoek et al., 2022), while Sanchez et al. (2021)

reported that adolescents often perceive vaping as socially integrated and acceptable

within their peer groups. The present findings suggest that general perceptions

regarding the social acceptability of nicotine use may not be sufficient to explain

dependence severity. Instead, direct interpersonal influences from close peers may

exert a stronger association with nicotine dependence than broader societal

perceptions.

Similarly, digital exposure was positively but not significantly correlated with

nicotine dependence (r = 0.13, p = .063). Previous research has consistently

demonstrated that exposure to nicotine-related content on social media platforms

contributes to experimentation, susceptibility, and current tobacco use among

adolescents (Donaldson et al., 2022; Vassey et al., 2022; Purba et al., 2023). However,

the findings of the present study indicate that digital exposure was not independently

associated with nicotine dependence. Rather than exerting a direct influence on

dependence severity, digital exposure may contribute indirectly by reinforcing

favorable beliefs regarding nicotine use, increasing perceived social acceptance, and

strengthening peer-related influences. This interpretation is further supported by the

significant correlations observed between digital exposure and perceived benefits,

perceived barriers, peer influence, and social norms.

The interrelationships among the psychosocial determinants further support

the study's conceptual framework. Peer influence was positively correlated with

perceived benefits (r = 0.34, p < .001) and negatively correlated with perceived

self-efficacy (r = –0.21, p < .001), indicating that respondents experiencing stronger

peer influence were more likely to perceive nicotine use favorably while reporting

lower confidence in resisting or quitting nicotine. These findings are consistent with

Laninga-Wijnen and Veenstra (2021), Vitória et al. (2020), East et al. (2021), and

Belmonte et al. (2025), who emphasized that peer interactions influence adolescents'

health beliefs and refusal behaviors through social learning and conformity. Likewise,

digital exposure demonstrated significant positive correlations with perceived

benefits, perceived barriers, peer influence, and social norms, further supporting the

study's theoretical position that contextual factors derived from Social Cognitive

Theory shape the core Health Promotion Model cognitions rather than acting as

independent predictors of nicotine dependence.

Finally, neither age (r = –0.11, p = .131) nor sex (r = 0.08, p = .279)

demonstrated statistically significant relationships with nicotine dependence. These

findings suggest that demographic characteristics alone were not associated with

dependence severity among adolescents and young adults in this study. Instead,

psychosocial and interpersonal factors, particularly peer influence, appeared to be

more closely related to nicotine dependence than demographic variables.

The findings both support and extend the existing literature. Consistent with

previous studies, peer influence emerged as the only psychosocial determinant

significantly associated with nicotine dependence, reinforcing the importance of the

immediate social environment in shaping nicotine-related behaviors among

adolescents and young adults. In contrast, the primary HPM constructs as well as

context determinants were not directly associated with nicotine dependence.

Nevertheless, the significant intercorrelations among these variables indicate that they

remain theoretically important because they influence one another in ways consistent

with the study's conceptual framework. Specifically, the findings support the role of

Social Cognitive Theory as a contextual framework, wherein peer influence, digital

exposure, and social norms shape the HPM behavior-specific cognitions of perceived

benefits, perceived barriers, and perceived self-efficacy.

Multiple Linear Regression Analysis

Table 4b. Predictors of Nicotine Dependence Among Adolescent and Young Adult Nicotine Users
Independent Variables B SE β t p 95% CI

(Constant) 1.420 0.450 3.16 0.002 0.533 to 2.307

Peer Influence 0.462 0.101 0.284*** 4.57 <0.001 0.263 to 0.661

Perceived Benefits 0.351 0.117 0.195*** 3.00 0.008 0.120 to 0.582

Perceived Self-Efficacy -0.268 0.098 -0.143*** -2.73 0.007 -0.461 to -0.075

Perceived Barriers 0.164 0.082 0.088*** 2.00 0.045 0.003 to 0.325

Social Norms 0.214 0.111 0.112 1.93 0.055 -0.005 to 0.433

Sex (Male = 1) 0.229 0.168 0.125 1.36 0.175 -0.103 to 0.561

Digital Exposure 0.059 0.072 0.042 0.82 0.414 -0.083 to 0.201

Age -0.025 0.034 -0.031 -0.74 0.462 -0.092 to 0.042

2
Note: 𝑅 = 0. 342, 𝐹(8, 193) = 12. 55, 𝑝 < 0. 001​
Abbreviations: β, standardized regression coefficient; CI, confidence interval; SE, standard error.​
Significance indicators: * p < 0.05, ** p < 0.01, *** p < 0.001.
 Table 4b shows that the overall model significantly predicted nicotine

dependence among adolescent and young adult nicotine users, R² = 0.342, F(8,193) =

12.55, p < 0.001. This means that the combined predictors explained 34.2% of the

variance in nicotine dependence scores. In practical terms, the model has meaningful

explanatory value because more than one-third of the differences in nicotine

dependence among respondents can be accounted for by the psychosocial and

demographic variables included in the analysis. This supports the study’s theoretical

direction, particularly the Health Promotion Model, which emphasizes that health-risk

behaviors are shaped by behavior-specific cognitions such as perceived benefits,

perceived barriers, and perceived self-efficacy, together with interpersonal and

situational influences.

Among all predictors, peer influence emerged as the strongest predictor of

nicotine dependence. It had a positive and statistically significant effect, B = 0.462, β

= 0.284, t = 4.57, p < 0.001, 95% CI [0.263, 0.661]. This indicates that for every

one-unit increase in peer influence, nicotine dependence increased by 0.462 points,

holding other variables constant. This result strongly agrees with the present

literature, which identifies peer dynamics as one of the most consistent drivers of

nicotine initiation and continuation. The existing literature notes that adolescents may

be influenced through peer selection, imitation, conformity, and descriptive norms,

especially when friends or siblings smoke or vape. Studies by Laninga-Wijnen and

Veenstra (2021), Vitória et al. (2020), East et al. (2021), and Belmonte et al. (2025)

support the role of peer influence in shaping nicotine-related behavior. Therefore, the

present finding extends the literature by showing that peer influence is not only
associated with experimentation or intention but also independently predicts nicotine

dependence severity among adolescents and young adults in Tacloban City.

Perceived benefits also significantly predicted nicotine dependence, B =

0.351, β = 0.195, t = 3.00, p = 0.008, 95% CI [0.120, 0.582]. This means that

respondents who perceived more benefits from nicotine use tended to report higher

nicotine dependence scores. This finding supports the existing literature, which

explains that perceived benefits include beliefs that nicotine provides stress relief,

improved mood, emotional comfort, social belonging, or reduced harm compared

with cigarettes. Studies by Resano et al. (2022) and Abano et al. (2025) linked

nicotine use with coping narratives, while Serra et al. (2023) and Lacsa (2025) noted

that misconceptions about vaping being safer than cigarettes may encourage

experimentation and continued use. Product appeal, flavors, attractive designs,

curiosity, and social media marketing were also discussed by Al-Khalifa and

Rodriguez (2025), Pan et al. (2022), and Ibukun et al. (2025) as factors that strengthen

favorable expectations toward vaping. Compared with the correlation result, where

perceived benefits were not significantly related to nicotine dependence, the

regression result suggests that perceived benefits become important when other

predictors are controlled. This implies that benefit beliefs may have a unique

predictive role in dependence when separated from overlapping influences such as

peer influence, social norms, and digital exposure.

Perceived self-efficacy was a significant negative predictor of nicotine

dependence, B = -0.268, β = -0.143, t = -2.73, p = 0.007, 95% CI [-0.461, -0.075].

This means that higher confidence in resisting, reducing, or quitting nicotine use was

associated with lower nicotine dependence. This finding is consistent with Bandura’s
Social Cognitive Theory and the Health Promotion Model, both of which emphasize

self-efficacy as a key determinant of behavior regulation and behavior change. The

RRL explains that self-efficacy refers to confidence in refusing nicotine, resisting peer

pressure, managing triggers, and quitting successfully. It also notes that repeated

exposure to peers who smoke or vape and nicotine-related content may weaken young

people’s refusal confidence over time, as discussed by Smith and Hilton (2023). Thus,

the present result confirms the protective role of self-efficacy: adolescents and young

adults with stronger confidence to resist nicotine are less likely to show higher

dependence.

Perceived barriers also significantly predicted nicotine dependence, although

its effect was smaller than peer influence, perceived benefits, and self-efficacy: B =

0.164, β = 0.088, t = 2.00, p = 0.045, 95% CI [0.003, 0.325]. This indicates that

higher perceived barriers to reducing or stopping nicotine use were associated with

higher nicotine dependence. This finding supports the Health Promotion Model,

which explains that barriers can sustain unhealthy behaviors when individuals

perceive quitting as difficult. The RRL identifies these barriers as cravings,

withdrawal experiences, social acceptability, peer routines, accessibility of nicotine

products, and weak policy enforcement. Evidence from East et al. (2025) shows that

social acceptability may make vaping cessation more difficult, while Donaldson et al.

(2022) and Vassey et al. (2022) emphasize how digital exposure may normalize

nicotine use and reinforce continued behavior. Structural barriers are also supported

by Marsh et al. (2021) and Philippine-focused sources such as Serra et al. (2023),

Congressional Policy and Budget Research Department (2025), Global Tobacco

Control (2024), and Sese (2023). The present result therefore aligns with the literature
by showing that perceived difficulty in quitting or reducing nicotine use contributes to

dependence severity.

On the other hand, social norms did not significantly predict nicotine

dependence, although the result approached significance, B = 0.214, β = 0.112, t =

1.93, p = 0.055, 95% CI [-0.005, 0.433]. The positive coefficient suggests that

respondents who perceived greater social acceptance of nicotine use tended to have

higher dependence scores, but the effect did not reach the 0.05 level. This partially

supports the literature, which states that vaping has been increasingly positioned as

modern, socially acceptable, and integrated into youth culture. Hoek et al. (2022) and

Sanchez et al. (2021) explain that these shifting norms may lower perceived

consequences and increase barriers to cessation. However, in the present model, social

norms were weaker than direct peer influence. This suggests that actual peer influence

may be more powerful than general perceptions of social acceptability.

Digital exposure was not a significant predictor of nicotine dependence, B =

0.059, β = 0.042, t = 0.82, p = 0.414, 95% CI [-0.083, 0.201]. This finding differs

from literature showing that social media exposure is associated with

experimentation, susceptibility, and current nicotine use. The RRL cites Donaldson et

al. (2022) and Vassey et al. (2022) as evidence that online tobacco and e-cigarette

content can amplify pro-vaping messages, while Purba et al. (2023) links social media

use with adolescent health-risk behaviors. The non-significant finding suggests that

digital exposure may not directly predict dependence severity once stronger predictors

such as peer influence, perceived benefits, barriers, and self-efficacy are included in

the model. It may instead operate indirectly by shaping attitudes, increasing perceived

benefits, normalizing use, or reinforcing peer-related exposure.

 Finally, age and sex were not significant predictors of nicotine dependence.

Age had B = -0.025, β = -0.031, p = 0.462, while sex had B = 0.229, β = 0.125, p =

0.175. These findings indicate that nicotine dependence in this sample was not

primarily explained by demographic characteristics. Instead, dependence was more

strongly predicted by psychosocial and cognitive factors. This supports the theoretical

framework of the study, which emphasizes that nicotine-related behaviors are shaped

by beliefs, emotions, peer groups, social media, accessibility, and policy environments

rather than by demographic profile alone.

The regression findings strengthen and refine the results of the correlation

analysis. While the correlation matrix showed that only peer influence was

significantly related to nicotine dependence, the regression model revealed that peer

influence, perceived benefits, perceived self-efficacy, and perceived barriers were all

significant predictors when analyzed together. This indicates that the Health

Promotion Model predictors have important unique contributions to nicotine

dependence after controlling for other variables. Compared with the present literature,

the study confirms that peer influence remains the strongest social predictor, while

perceived benefits and barriers increase dependence risk and self-efficacy serves as a

protective factor. These findings suggest that prevention and cessation programs

should combine peer-focused interventions, correction of perceived benefits,

barrier-reduction strategies, and self-efficacy-building activities such as refusal-skills

training and coping-skills development.

Table 4c. Diagnostic Tests for the Multiple Linear Regression Model

Assumption Test Result Remarks

Normality Shapiro-Wilk p = 0.248 Satisfied

No Multicollinearity VIF (Mean) 1.34 Satisfied

Homoscedasticity Breusch-Pagan p = 0.512 Satisfied

Independence of Errors Durbin-Watson 1.96 Satisfied

Table 4c presents the diagnostic tests conducted to determine whether the

multiple linear regression model met the basic assumptions required for valid

interpretation. These tests are important because assumption violations may affect the

accuracy, reliability, and interpretability of regression estimates, including the

regression coefficients, confidence intervals, and significance values (James et al.,

2021; Field, 2024).

The normality assumption was assessed using the Shapiro–Wilk test, which

produced a result of p = 0.248. Since this value is greater than the 0.05 level of

significance, the result indicates that the residuals were approximately normally

distributed. This means that the regression model did not violate the normality

assumption. In relation to the regression findings in Table 4b, this supports the

validity of the reported t-values, p-values, and confidence intervals. Field (2024)

emphasizes that normal residuals help support valid hypothesis testing in regression

analysis.

The assumption of no multicollinearity was assessed using the Variance

Inflation Factor (VIF), with a mean VIF value of 1.34. This value is far below the

commonly accepted thresholds of 5 or 10, indicating that the independent variables

were not excessively correlated with one another. This is important because high

multicollinearity can inflate standard errors and make regression coefficients unstable.

Since the VIF result was acceptable, the significant predictors in Table 4b—peer

influence, perceived benefits, perceived self-efficacy, and perceived barriers—can be

interpreted as having unique contributions to nicotine dependence. This supports

James et al. (2021), who noted that VIF is used to identify whether predictors are too

highly intercorrelated in regression models.

The homoscedasticity assumption was tested using the Breusch–Pagan test,

which yielded p = 0.512. Since the result is greater than 0.05, there is no evidence of

heteroscedasticity. This means that the variance of the residuals was relatively

constant across the predicted values of nicotine dependence. This strengthens the

reliability of the standard errors and significance tests reported in the regression

model. Field (2024) explains that satisfying homoscedasticity supports the accuracy

of statistical significance testing in regression.

The assumption of independence of errors was assessed using the

Durbin–Watson statistic, which produced a value of 1.96. Since values close to 2

generally indicate that residuals are independent, the result suggests that there was no

serious autocorrelation among the errors. This means that the observations in the

model were sufficiently independent, further supporting the validity of the regression

estimates.

Overall, the diagnostic results show that the regression model was statistically

appropriate for examining the predictors of nicotine dependence. The assumptions of

normality, no multicollinearity, homoscedasticity, and independence of errors were all

satisfied. Therefore, the findings in Table 4b may be interpreted with greater

confidence. This strengthens the conclusion that peer influence, perceived benefits,

perceived self-efficacy, and perceived barriers are meaningful predictors of nicotine

dependence among adolescent and young adult nicotine users in Tacloban City.