Phase 1
The results of a thematic analysis revealed that participants who encountered travel adversities due to traffic congestion experienced negative affect. They manifested their negative affect through fight or flight responses. The results were negative road occurrences that intensified traffic congestion.
The theme of negative affect was organized from the basic themes of participant experiences that elicited stress, aggression, pressure, anxiety, tension, anger, irritation, and frustration during their regular commute through traffic congestion. They experienced mood swings, exhibited road rage, impatience, and resentment, felt sad, helpless, and tensed, and struggled to maintain composure during their commute. The commute rendered them emotionally distraught. They were irritated, anxious, and emotionally distressed. The disturbance of peace caused them to feel insane and disregard others.
The theme fight was organized from the basic themes of aggressive driving, horn honking, stress eating, vengeful driving, changing riding or driving style, verbal abuse and mob mentality. A flight response was also evident when some participants opted to stay inside to avoid traffic congestion.
The theme of negative road occurrences was organized from the basic themes such as engaging in verbal and physical altercations on the roads, physically damaging vehicles, showing a racing mentality, mild forms of vengeance, risky and unethical behaviour during rush hour, and reckless and negligent driving. The basic themes also illustrated the commuters' participation in thrill-seeking activities, such as speeding, performing stunts, displaying defiant behaviour, disobeying the authorities, refusing to follow the rules, and wilfully violating traffic regulations. Some commuters exhibited entitled behaviour, as evidenced by a sense of entitlement and the expectation of special treatment despite lawbreaking.
Following are a few examples from the excerpts of participant interview responses that led to the basic themes.
I think I get reckless because obviously someone's not giving me way; you try honking, you lower down your screen, it’s not working and then you start manoeuvring and get wherever possible and just go and.. you are helpless right… (M.L. December 21, 2019)
There have been times when I have overtaken cars and cut them off, just to get the point across that they were pissing me off, which has happened quite a few times (A.K. December 14, 2019)
I am one of them. I have no patience anymore after a point, I am like, okay, maximum, I have to pay a fine. So I’d rather wait for the letter or, maximum, I might get caught by the cops. I will work it out with them. If they ask me to pay a heavy fine if I have the money at that point, I will think about it, or I will say, as per the rules, you take a photograph, and you send me the thing, and I will make the payment that's it (A. R. December 12, 2019).
When you get irritated, it automatically shows on your accelerator, it shows for me, when I get irritated obviously it will show on my accelerator or my horn (R.M. December 7, 2019).
Based on these findings, investigators concluded Phase 1 with the following assumptions.
Travel adversities cause negative affect on commuters. Commuters would be compelled to fight or flee because they would be unable to see any other options. Because commuters avoid exposure to traffic congestion, flight actions have no repercussions. Fight actions will result in undesirable or negative road occurrences, which will further intensify travel adversity. This sequence may result in a vicious cycle as depicted in Fig. 2.
Phase 2
Structural equation modelling (SEM) (IBM AMOS 27) was employed to validate the model from phase 1 statistically. Initially, we analysed the correlation between the organizing themes (Fig. 3A & Table 2).
Table 2
Retained basic themes after removing those with high covariance.
Basic themes
|
|
Organizing themes
|
B
|
S.E.
|
β
|
RE1(Congestion is a part of the daily commute)
|
<---
|
Travel Adversity
|
1.00
|
--
|
0.38**
|
RE2(Being stuck in congestion feels chaotic)
|
<---
|
1.64
|
0.34
|
0.66**
|
RE3(Experiencing wastage of time in congestion)
|
<---
|
0.98
|
0.28
|
0.33**
|
RE5(Negative road experiences in the beginning of the day impacting interactions during the day)
|
<---
|
2.07
|
0.41
|
0.73**
|
RE6(Riding/Driving is influenced by the actions around in the street)
|
<---
|
1.64
|
0.35
|
0.60**
|
Af3(Pressure)
|
<---
|
Negative Affect
|
1.00
|
--
|
0.79**
|
Af5(Anxiety)
|
<---
|
1.13
|
0.09
|
0.84**
|
Af7(Negative Mood/Feelings)
|
<---
|
1.04
|
0.08
|
0.72**
|
Af9(Tension)
|
<---
|
1.11
|
0.08
|
0.80**
|
Af11(Resentfulness)
|
<---
|
0.96
|
0.09
|
0.88**
|
Af15(Frustration)
|
<---
|
0.98
|
0.08
|
0.72**
|
Af16(Emotional Distress)
|
<---
|
1.17
|
0.08
|
0.76**
|
Af18(Feeling Helpless)
|
<---
|
1.04
|
0.09
|
0.71**
|
Af19(Nervousness)
|
<---
|
0.94
|
0.08
|
0.74**
|
Af20(Annoyance)
|
<---
|
0.88
|
0.08
|
0.74**
|
Af23(Restlessness)
|
<---
|
0.97
|
0.08
|
0.65**
|
Af24(Overwhelming Emotions)
|
<---
|
0.98
|
0.08
|
0.72**
|
Af27(Feeling Edgy)
|
<---
|
0.91
|
0.08
|
0.70**
|
ARO1(Aggressive Driving)
|
<---
|
Fight
|
1.00
|
--
|
0.76**
|
ARO3(Mob-Mentality)
|
<---
|
0.87
|
0.10
|
0.78**
|
ARO4(Verbal Abuse )
|
<---
|
1.00
|
0.10
|
0.86**
|
ARO5(Change in Riding / Driving style)
|
<---
|
1.01
|
0.10
|
0.68**
|
ARO6(Persistent and Recurring Honking)
|
<---
|
1.10
|
0.10
|
0.84**
|
ARO11(Physical Altercations)
|
<---
|
Negative Road occurrences
|
1.00
|
--
|
0.68**
|
ARO14(Defiant Behaviour)
|
<---
|
1.31
|
0.10
|
0.75**
|
ARO15(Entitled Behaviour)
|
<---
|
1.03
|
0.10
|
0.77**
|
ARO16(Unethical Behaviour)
|
<---
|
1.11
|
0.09
|
0.77**
|
ARO17(Racing Mentality)
|
<---
|
1.24
|
0.12
|
0.81**
|
Note. Model fit indices: χ2 = 664.45, df = 344, p < .01; RMSEA = .066, LO90 = .058, HI90 = .073; CFI = .912; TLI = .903; IFI = .913; χ2/df = 1.932
RMSEA = root mean square error of approximation; LO = lower limit; HI = upper limit; CFI = Comparative Fit Index; TLI = Tucker Lewis index; IFI = Incremental Fit Index; χ2 = model chi square; df = degrees of freedom.
**p < .01
|
To attain model fitness, basic themes with a covariance greater than 15 with another basic theme were removed. The range of factor loadings (standardized estimates) of the retained items of travel adversity (β = .33 to .73), negative affect (β = .65 to .84), fight (β = .76 to .86), and negative road occurrences (β = .68 to .81) showed basic themes as a moderate to strong components of the organizing themes. There was a strong correlation between negative affect and travel adversities (r = .80, p < .01). Also, there was a strong correlation between negative road occurrences and the fight (r = .78, p < .01). Probably, the participants could not discriminate travel adversities from negative affect and negative road occurrences from the fight. There was a moderate correlation between fight and travel adversities (r = .56, p < .01), negative road occurrences and travel adversities (r = .43, p < .01), fight and negative affect (r = .55, p < .01), and negative road occurrences and negative affect (r = .41, p < .01). The root mean square error of approximation (RMSEA = .066, LO90 = .058, HI90 = .073), Comparative fit index (CFI = .912), Tucker Lewis index (TLI = .903), Incremental Fit Index (IFI = .913), and Standardized Root Mean Square Residuals (SRMR = .0634) showed that the model is of adequate fit.
Using SEM, we tested if travel adversities, negative affect, fights, and negative road occurrences, respectively, had a sequentially predictable relationship (Fig. 3B & Table 3).
Table 3
Standardized and unstandardized coefficients in the SEM that tested if travel adversities, negative affect, fights, and negative road occurrences, respectively, had a sequentially predictable relationship.
Endogenous
|
|
Exogenous
|
B
|
S.E.
|
β
|
Negative affect
|
<---
|
Travel adversities
|
1.65
|
0.33
|
0.80
|
Fight
|
<---
|
Negative affect
|
0.51
|
0.07
|
0.55
|
Negative road occurrences
|
<---
|
Fight
|
0.61
|
0.07
|
0.78
|
RE1(Congestion is a part of the daily commute)
|
<---
|
Travel adversities
|
1.00
|
--
|
0.39
|
RE2(Being stuck in congestion feels chaotic)
|
<---
|
1.65
|
0.33
|
0.68
|
RE3(Experiencing wastage of time in congestion)
|
<---
|
1.00
|
0.28
|
0.34
|
RE5(Negative road experiences in the beginning of the day impacting interactions during the day)
|
<---
|
2.02
|
0.40
|
0.72
|
RE6(Riding/Driving is influenced by the actions around in the street)
|
<---
|
1.55
|
0.33
|
0.58
|
Af3(Pressure)
|
<---
|
Negative affect
|
1.00
|
--
|
0.77
|
Af5(Anxiety)
|
<---
|
1.13
|
0.09
|
0.81
|
Af7(Negative Mood/Feelings)
|
<---
|
1.04
|
0.08
|
0.79
|
Af9(Tension)
|
<---
|
1.11
|
0.08
|
0.84
|
Af11(Resentfulness)
|
<---
|
0.96
|
0.09
|
0.72
|
Af15(Frustration)
|
<---
|
0.98
|
0.08
|
0.80
|
Af16(Emotional Distress)
|
<---
|
1.17
|
0.08
|
0.88
|
Af18(Feeling Helpless)
|
<---
|
1.04
|
0.09
|
0.72
|
Af19(Nervousness)
|
<---
|
0.94
|
0.08
|
0.76
|
Af20(Annoyance)
|
<---
|
0.88
|
0.08
|
0.71
|
Af23(Restlessness)
|
<---
|
0.97
|
0.08
|
0.74
|
Af24(Overwhelming Emotions)
|
<---
|
0.98
|
0.08
|
0.77
|
Af27(Feeling Edgy)
|
<---
|
0.91
|
0.08
|
0.75
|
ARO1(Aggressive Driving)
|
<---
|
Fight
|
1.00
|
--
|
0.74
|
ARO3(Mob-mentality)
|
<---
|
0.88
|
0.10
|
0.65
|
ARO4(Verbal Abuse)
|
<---
|
1.00
|
0.10
|
0.72
|
ARO5(Change in Driving/Riding Style)
|
<---
|
1.00
|
0.10
|
0.70
|
ARO6(Persistent and Recurring Honking)
|
<---
|
1.11
|
0.10
|
0.77
|
ARO11(Physical Altercations)
|
<---
|
Negative road occurrences
|
1.00
|
--
|
0.78
|
ARO14(Defiant Behaviour)
|
<---
|
1.31
|
0.10
|
0.85
|
ARO15(Entitled Behaviour)
|
<---
|
1.03
|
0.10
|
0.68
|
ARO16(Unethical Behaviour)
|
<---
|
1.11
|
0.09
|
0.84
|
ARO17(Racing Mentality)
|
<---
|
1.24
|
0.12
|
0.68
|
Note. Model fit indices: χ2 = 668.95, df = 347, p < .01; RMSEA = .066, LO90 = .058, HI90 = .073; CFI = .911; TLI = .904; IFI = .912; χ2/df = 1.928
RMSEA = root mean square error of approximation; LO = lower limit; HI = upper limit; CFI = Comparative Fit Index; TLI = Tucker Lewis index; IFI = Incremental Fit Index; χ2 = model chi square; df = degrees of freedom.
**p < .01
|
Results showed that travel adversity is a significant predictor of negative affect (β = .80, p < .01). Negative affect is a significant predictor of fight (β = .55, p < .01). Fight is a significant predictor of negative road occurrences (β = .78, p < .01). The factor loadings of the travel adversities ranged from .34 to .72. The factor loadings of negative affect ranged from .71 to .88. The factor loadings of fight ranged from .65 to .77. The factor loadings of negative road occurrences ranged from .68 to .85. The root mean square error of approximation (RMSEA = .066, LO90 = .058, HI90 = .073), Comparative fit index (CFI = .911), Tucker Lewis index (TLI = .904), Incremental Fit Index (IFI = .912), and Standardized Room Mean Square Residuals (SRMR = .0656) showed that the model is of adequate fit.
To complete the loop of the hypothetical vicious cycle, we examined whether negative road occurrences predicted travel adversities (Fig. 3C 6 & Table 4).
Table 4
Standardized and unstandardized coefficients in the SEM that tested if negative road occurrences is a predictor of travel adversities
Endogenous
|
|
Exogenous
|
B
|
S.E.
|
β
|
Travel adversities
|
<---
|
Negative road occurrences
|
0.32
|
0.08
|
0.40**
|
RE1(Congestion is a part of the daily commute)
|
<---
|
Travel adversities
|
1.00
|
--
|
0.46**
|
RE2(Being stuck in congestion feels chaotic)
|
<---
|
1.41
|
0.26
|
0.68**
|
RE3(Experiencing wastage of time in congestion)
|
<---
|
0.92
|
0.24
|
0.37**
|
RE5(Negative road experiences in the beginning of the day impacting interactions during the day)
|
<---
|
1.62
|
0.30
|
0.68**
|
RE6(Riding/Driving is influenced by the actions around in the street)
|
<---
|
1.31
|
0.26
|
0.58**
|
ARO1(Physical Altercations)
|
<---
|
Negative road occurrences
|
1.00
|
--
|
0.78**
|
ARO14(Defiant Behaviour)
|
<---
|
1.30
|
0.10
|
0.85**
|
ARO15(Entitled Behaviour)
|
<---
|
1.03
|
0.10
|
0.68**
|
ARO16(Unethical Behaviour)
|
<---
|
1.14
|
0.09
|
0.86**
|
ARO17(Racing Mentality)
|
<---
|
1.20
|
0.12
|
0.66**
|
Note. Model fit indices: χ2 = 89.30, df = 34, p < .01; RMSEA = .087, LO90 = .065, HI90 = .11; GFI = .914; CFI = .927; TLI = .904; IFI = .928; χ2/df = 2.627
RMSEA = root mean square error of approximation; LO = lower limit; HI = upper limit; GFI = Goodness of Fit index; CFI = Comparative Fit Index; TLI = Tucker Lewis index; IFI = Incremental Fit Index; χ2 = model chi square; df = degrees of freedom.
**p < .01
|
Results showed that negative road occurrences is a significant predictor of travel adversity (β = .40, p < .01). The factor loadings of the travel adversity ranged from .37 to .68. The factor loadings of negative road occurrences ranged from .66 to .85. The root mean square error of approximation (RMSEA = .087, LO90 = .058, HI90 = .073) and Standardized Room Mean Square Residuals (SRMR = .0861) were in the borderline of adequate fitness. Goodness of fit index (GFI = .914), Comparative fit index (CFI = .927), Tucker Lewis index (TLI = .904), and Incremental Fit Index (IFI = .928) showed that the model is of acceptable fit.
Based on our findings regarding the vicious cycle that intensifies travel adversity while commuting through traffic congestion, we developed the following mathematical model: NA(Negative Affect), FT (Fight), NRO (Negative Road Occurrences), and TA (Travel Adversities).
$$NA={f}_{1}\left(TA\right)+x$$
$$FT={f}_{2}\left(NA\right)+y$$
$$NRO={f}_{3}\left(FT\right)+z$$
where the functions \({f}_{1}, {f}_{2}, {f}_{3}\) incorporates the coefficients 0.80, 0.55, and 0.78 respectively. \(x, y, z\) are the Gaussian errors in both the model and the data, assuming that the data is normally distributed. This leads us back to how NRO affects TA, completing the feedback loop:
$$TA={f}_{4}\left(NRO\right)+w$$
\({f}_{4}\) involves the coefficient of 0.40, and \(w\), the Gaussian error.