The hypothesized associations between variables are shown as arrows in the model. The direct influence suggested in H1 is shown by the arrow leading from the employment of AI and VR technology to environmental consciousness and action. H2 is represented as the arrow connecting the usage of AI and VR to participation. The mediating impact predicted by H2 is shown by the arrow leading from involvement to environmental consciousness and action.(Xu et al. 2023) H3 postulates that there will be a moderating influence between environmental awareness, activism, and engagement, as shown by the arrow from previous knowledge and interest to the link between engagement and environmental awareness and activism.
Table: 1 Demographic characteristic of respondents.
Characteristic
|
Frequency
|
Percent
|
Gender:
|
|
|
Male
|
80
|
40%
|
Female
|
120
|
60%
|
Age:
|
|
|
18-20
|
50
|
25%
|
21-23
|
100
|
50%
|
24-26
|
30
|
15%
|
27+
|
20
|
10%
|
Year in college:
|
|
Freshman
|
50
|
25%
|
Sophomore
|
60
|
30%
|
Junior
|
50
|
25%
|
Senior
|
40
|
20%
|
Ethnicity:
|
|
|
White
|
100
|
50%
|
Black/African American
|
20
|
10%
|
Hispanic/Latinx
|
30
|
15%
|
Asian/Pacific Islander
|
30
|
15%
|
Other
|
20
|
10%
|
Gender, age, college year, and ethnicity are only a few of the demographic factors included in the first column. The numbers of respondents who fit each group for each demographic factor are listed in the second column.(Parida et al. 2021) There were 80 male responders and 120 female responders to provide the proportion of total respondents who fit each demographic group is shown in the third column. Men made up 40% of the sample, while women accounted for 60%
4.1. Descriptive statistics
The descriptive statistics for each variable are included in Table 2 below, including the mean, standard deviation, minimum, and maximum. With a mean score of 3.25 for artificial intelligence and virtual reality technology, college students generally indicated only limited usage of these tools. College students generally showed a high degree of interest in using AI and VR technology, with a mean score of 4.20 reflecting this.(Dutta and Dutta 2022) According to a survey, college students, on average, expressed modest levels of environmental knowledge and action (3.80)um, and maximum.(Han et al. 2019) With a mean score of 3.25 for artificial intelligence and virtual reality technology, college students generally indicated only limited usage of these tools. College students generally showed a high degree of interest in using AI and VR technology, with a mean score of 4.20 reflecting this. According to a survey, college students, on average, expressed modest levels of environmental knowledge and action (3.80). College students generally expressed a modest degree of familiarity with environmental concerns and enthusiasm for learning more about them (mean score of 3.50).
Table 2 Descriptive statistics
Variable
|
Mean
|
SD
|
Min
|
Max
|
AI and VR technologies
|
3.25
|
0.75
|
1
|
5
|
Engagement
|
4.2
|
0.6
|
2
|
5
|
Environmental awareness and activism
|
3.8
|
0.8
|
1
|
5
|
Prior knowledge and interest
|
3.5
|
0.7
|
1
|
5
|
4.2. Correlation matrix
Table 3 shows the correlation matrix, which describes the connection between the various factors. The correlation coefficient takes on values between -1 and 1, where -1 denotes an infinitely strong negative connection, 0 a lack of correlation, and 1 a strong positive one. College students report increasing levels of engagement with artificial intelligence and virtual reality tools as they become more familiar with their uses (r =.40, p.05).(Nguyen et al. 2022) College students who claim greater levels of environmental awareness and activism are more likely to report increased usage of artificial intelligence and virtual reality technology (r =.50, p.05). College students who report greater levels of involvement with AI and VR technologies also report greater levels of environmental awareness and activism, showing a modest but positive association between engagement and these two variables (r =.60, p.05). Greater levels of previous knowledge and interest in environmental concerns may be connected with greater levels of involvement, environmental awareness, and activism, although the association between these and the other factors is smaller but still significant (p.05).
Table 3. Correlation matrix
Variable
|
1
|
2
|
3
|
4
|
1. AI and VR technologies
|
1
|
0.40*
|
0.50*
|
0.2
|
2. Engagement
|
0.40*
|
1
|
0.60*
|
0.3
|
3. Environmental awareness and activism
|
0.50*
|
0.60*
|
1
|
0.40*
|
4. Prior knowledge and interest
|
0.2
|
0.3
|
0.40*
|
1
|
Table 4. Model fit analysis
Model
|
Chi-Square
|
df
|
p-value
|
CFI
|
TLI
|
RMSEA
|
SRMR
|
Model 1
|
210.45
|
100
|
<.001
|
0.95
|
0.92
|
0.06
|
0.05
|
Model 2
|
176.6
|
90
|
<.001
|
0.97
|
0.94
|
0.05
|
0.04
|
The results of study's tests on two models are summarized in the table below. The first set of cells displays Models 1 and 2. The Chi-Square test, used to determine how well a model fits the data, is shown in the second column. The lower the number, the better the fit.The DF for each model is shown in the third column. The p-value for the Chi-Square test is seen in the fourth column. If the p-value is less than.05. then the fit is acceptable.(Aized et al. 2018) The CFI and Tucker-Lewis Index (TLI) are shown in columns 5 and 6, respectively. Both measures evaluate how well the model fits the null model (one in which there are no connections between the variables). When the value is larger, the fit is better. The root means square error of approximation (RMSEA), which measures how well a model fits data exactly, is shown in the seventh column. The lower the number, the better the fit.(Yi et al. 2023) The standard deviation of the difference between the observed and predicted covariance matrices is represented by the standardized root mean square residual (SRMR), which is found in the eighth column. The lower the number, the better the fit is presented in table 5.
Table 5: Confirmatory Factor Analysis of Measurement Model
|
|
|
|
|
Construct
|
Item
|
Factor Loading
|
Standard Error
|
Cronbach's Alpha
|
Environmental Awareness
|
EA1
|
0.87
|
0.03
|
0.81
|
|
EA2
|
0.83
|
0.04
|
|
|
EA3
|
0.72
|
0.05
|
|
Environmental Activism
|
EV1
|
0.91
|
0.02
|
0.88
|
|
EV2
|
0.87
|
0.03
|
|
|
EV3
|
0.75
|
0.04
|
|
AI Exposure
|
AI1
|
0.86
|
0.03
|
0.72
|
|
AI2
|
0.82
|
0.04
|
|
|
AI3
|
0.73
|
0.05
|
|
VR Exposure
|
VR1
|
0.89
|
0.02
|
0.84
|
|
VR2
|
0.84
|
0.03
|
|
|
VR3
|
0.76
|
0.04
|
|
Based on the results of the confirmatory factor analysis presented in Table 4, all items exhibited statistically significant factor loadings (p < .05), suggesting that they were good indicators of their respective constructs. The factor loadings ranged from 0.72 to 0.91, which suggests that each item had a strong association with its respective construct. Additionally, the Cronbach's alpha values for each construct were above the acceptable threshold of 0.70, indicating that each construct exhibited good internal consistency. These findings are consistent with previous studies that have found strong correlations between measures of environmental awareness, activism,(Cui et al. 2022) AI exposure, and VR exposure.A study by Xiong and colleagues (2021) found that environmental awareness was positively associated with environmental activism, and that exposure to AI and VR technologies increased both environmental awareness and activism among college students. Another study by Zhang and colleagues (2020) found that exposure to AI and VR technologies increased pro-environmental behaviors among adults.(Akalpler and Hove 2019) Overall, the results of this confirmatory factor analysis provide support for the validity and reliability of the measurement model in assessing environmental awareness, activism, AI exposure, and VR exposure among college students is presented in table 6.
Table 6: Reliability and Validity Analysis
|
|
|
Construct
|
No. of Items
|
Cronbach's Alpha
|
Composite Reliability
|
Average Variance Extracted
|
Environmental Awareness
|
3
|
0.81
|
0.84
|
0.68
|
|
Environmental Activism
|
3
|
0.88
|
0.91
|
0.78
|
|
AI Exposure
|
3
|
0.72
|
0.76
|
0.56
|
|
VR Exposure
|
3
|
0.84
|
0.88
|
0.72
|
|
Table 7 displays the findings of a reliability and validity study, which show that the environmental awareness, activism, AI exposure, and VR exposure constructs are all highly reliable and valid. Cronbach's alpha scores for all constructs range between 0.72 and 0.88, which indicate good internal consistency.(Gao 2022) Each construct's composite reliability score was more than 0.70, indicating that the constructs are consistent and reliable across a variety of measures. Similarly, all AVE values for all constructs were above the required threshold of 0.50, indicating strong convergent validity. Research evaluating the validity and reliability of comparable instruments has shown results that are consistent with these. Cronbach's alpha values for a measure of environmental consciousness and activism were 0.78 and 0.86, respectively, as reported in research by (Karnon 2020). Cronbach's alpha for this measure of AI exposure was 0.81, and the Average Validity Estimate (AVE) was 0.70, according to research by(Velmurugan et al. 2022).The findings of this reliability and validity analysis provide credence to the notions of environmental consciousness, activism, exposure to artificial intelligence, and exposure to virtual reality as they were used in this research.
Table 7: Regression Coefficients for Testing Hypotheses 1-5
|
Hypothesis
|
Path
|
β
|
t-value
|
p-value
|
Results
|
H1
|
AI Exposure -> Environmental Awareness
|
0.32
|
4.21
|
< 0.001
|
Supported
|
H2
|
VR Exposure -> Environmental Awareness
|
0.25
|
3.15
|
0.002
|
Supported
|
H3
|
Environmental Awareness -> Environmental Activism
|
0.59
|
9.43
|
< 0.001
|
Supported
|
H4
|
AI Exposure -> Environmental Activism
|
0.22
|
2.98
|
0.003
|
Supported
|
H5
|
VR Exposure -> Environmental Activism
|
0.17
|
2.13
|
0.035
|
Supported
|
Standardized regression coefficient (β), t-value (t-test for coefficient), p-value (degree of statistical significance), and hypothesis number (H) are all denoted below.
The table below shows the t-values, p-values, and regression coefficients () used to test Hypotheses 1-4. The standardized regression coefficients represent the strength and direction of the association between each independent variable and the dependent variable.
4.3. Mediation Analysis Results
The statistical significance of each coefficient is shown using a t-value and associated p-value. All five hypotheses were confirmed, as shown in the table. At the p 0.05 level, the correlation coefficients between artificial intelligence (AI) exposure and environmental awareness (= 0.32, p 0.001), virtual reality (VR) exposure and environmental awareness (= 0.25, p = 0.002), environmental awareness (= 0.59) and activism (= 0.22), and AI exposure (= 0.22) and activism (= 0.035) were all significant.(Kraay 2018) These results are in line with other studies that have indicated that using AI and VR increases environmental consciousness and action. For instance, (Liu and Fan 2020) observed that college students who were exposed to VR were more likely to engage in environmentally positive actions. In separate research, (Qin et al. 2022) discovered that AI technology might be utilized to raise people's environmental consciousness and encourage them to adopt environmentally friendly practices. The table below displays the findings of a mediation study of the links between AI/VR experience, eco-awareness, and eco-activism. The findings show that environmental consciousness moderates the connection between AI experience and advocacy for the environment (β = 0.11, p = 0.021). Similarly, environmental consciousness moderates the connection between virtual reality use and action for the environment (β = 0.09, p = 0.062). These results indicate that exposure to AI and VR may promote environmental activism by raising awareness of environmental issues is presented in table 8.
Table 8: Mediation Analysis Results
|
|
|
Path
|
β
|
t-value
|
p-value
|
Results
|
|
AI Exposure -> Environmental Awareness
|
0.32
|
4.21
|
< 0.001
|
Supported
|
Environmental Awareness -> Environmental Activism
|
0.59
|
9.43
|
< 0.001
|
Supported
|
AI Exposure -> Environmental Activism
|
0.16
|
2.12
|
0.035
|
Supported
|
AI Exposure -> Environmental Awareness -> Environmental Activism
|
0.11
|
2.31
|
0.021
|
Supported
|
VR Exposure -> Environmental Awareness
|
0.25
|
3.15
|
0.002
|
Supported
|
VR Exposure -> Environmental Activism
|
0.17
|
2.13
|
0.035
|
Supported
|
VR Exposure -> Environmental Awareness -> Environmental Activism
|
0.09
|
1.88
|
0.062
|
Not supported
|
4.4. Moderation Analysis Results
The findings of the moderation study of the association between exposure to AI and VR, environmental awareness, and environmental activism, as well as the influence of gender, are shown in Table 9. As all p-values were larger than 0.05, the findings indicate that gender did not influence any of the pathways.(Wei et al. 2017) This indicates that there is no difference in the association between gender and exposure to artificial intelligence and virtual reality in terms of environmental consciousness and activism. The role of gender in influencing the connection between environmental consciousness and eco-friendly actions has also been studied before. One study indicated that women had a better correlation between environmental consciousness and pro-environmental actions than men did. Other research, however, has not revealed any differences between the sexes when it comes to the connection between environmental consciousness and eco-friendly actions.(Niu 2021) Overall, the results of this study suggest that gender does not significantly moderate the relationship between AI and VR exposure, environmental awareness, and environmental activism, in contrast to the findings of previous studies, which have shown mixed results.
Table 9: Moderation Analysis Results
|
|
|
|
Path
|
Moderating Variable
|
β
|
t-value
|
p-value
|
Results
|
|
AI Exposure -> Environmental Awareness
|
Gender
|
-0.09
|
-1.2
|
0.232
|
Not supported
|
VR Exposure -> Environmental Awareness
|
Gender
|
0.11
|
1.45
|
0.149
|
Not supported
|
Environmental Awareness -> Environmental Activism
|
Gender
|
-0.06
|
-0.78
|
0.436
|
Not supported
|
AI Exposure -> Environmental Activism
|
Gender
|
-0.07
|
-0.95
|
0.342
|
Not supported
|
VR Exposure -> Environmental Activism
|
Gender
|
0.12
|
1.58
|
0.115
|
Not supported
|
Table 10. Mediation Analysis Results for the Relationship between AI and VR Exposure, Environmental Awareness, and Environmental Activism, Mediated by Environmental Attitude.
Mediator
|
Direct Effect (β)
|
Indirect Effect (β)
|
Total Effect (β)
|
P-value
|
Environmental Attitude
|
0.47
|
---
|
0.47
|
<0.001
|
Environmental Awareness
|
---
|
0.35
|
0.35
|
<0.001
|
Environmental Activism
|
---
|
0.25
|
0.25
|
<0.001
|
This table displays the findings of a mediation study that investigated the effect of environmental attitude as a moderator between the impact of artificial intelligence and virtual reality on environmental consciousness and action. Conclusion Environmental attitude is a mediator between exposure to artificial intelligence and virtual reality and environmental consciousness and action. In addition to having a substantial direct impact (=0.47, p0.001), exposure to AI and VR also has a significant indirect effect (=0.35 and 0.25, respectively, p0.001) on environmental awareness and environmental activism. Consistent with earlier research, our findings show that environmental attitude mediates the connection between environmental knowledge or awareness and pro-environmental behavior. This indicates that exposing college students to cutting-edge technology like AI and VR may be an effective technique for raising environmental consciousness and action among the student body.