Predictors of distractive activities to walking in Accra, Ghana

ABSTRACT

Walking is a fundamental mode of transport for many people globally, with immense health and environmental benefits. However, increased walking is associated with an increased risk of road traffic crashes and injuries, especially where traffic enforcement is poor, and pedestrians could easily be distracted. Despite this, a plethora of evidence exists on driver distraction; however, there is a dread of evidence of pedestrian distraction (i.e. distracted walking), particularly in Africa. This paper uses a quantitative methodology to examine the predictors of distractive activities to walking in Accra, Ghana. The study surveyed 400 pedestrians within Accra’s Central Business District (CBD). A questionnaire was deployed via Kobo Toolkit and Jamovi was used in analysing the data. The study observed that three of the top four distractive activities to walking were digital devices-related (e.g. the use of mobile phones). We found that listening to music, making, receiving phone calls, and conversing with other people while walking are the main distractive activities. The binary logistic regression model found sex, age, level of education, occupation, reasons for walking, weekly time for walking and time for common trips as significant predictors of distractive activities to walking.

KEYWORDS:

• Pedestrians
• distractive activities to walking
• central business district
• Accra
• Ghana

1 Introduction

Walking is a fundamental mode of transport for many people globally, with immense health and environmental benefits. As a form of physical activity, walking is vital for reducing cardiovascular and obesity-related diseases (Sam, 2022). Improved walking environments also contribute to urban renewal, social cohesion, improved air quality, and reduced traffic noise (Fonseca et al., 2022; Mikusova et al., 2021; Tobin et al., 2022). Accordingly, many countries have initiated policies to encourage walking as an essential mode of transport (Bassett et al., 2008; Rabl & de Nazelle, 2012; WHO, 2010).

Unfortunately, increased walking is associated with increased risk of road traffic crashes and injuries in some jurisdictions. This is typically where traffic enforcement is poor and pedestrians’ needs are neglected mainly in roadway design and land-use planning (Zegeer & Bushell, 2012; Job, 2012, cited in WHO, 2013). As a result, more than 270,000 (one-fifth) of persons killed (i.e. the 1.35 million people) on the roadways globally are neither travelling in a car nor on a motorcycle or bicycle but are pedestrians.

Current studies also show a disproportionate number of pedestrians and other road users in road traffic crashes and injuries (World Health Organization, 2017, 2018). Despite having fewer vehicles and transport networks than in Europe and U.S., road traffic crashes are the third leading cause of death and disability in Sub-Saharan Africa (Onywera & Blanchard, 2013). Africa has the worst rate of pedestrian crashes and fatalities. For instance, while only 22% and 27% of pedestrians are victims of road traffic crashes in the U.S. and Europe, Africa has 40% of its pedestrians involved in road traffic crashes (World Health Organization, 2018). In Ghana, pedestrian injury rates are almost similar to the aformentioned as 38.9% of pedestrians are annually involved in road traffic crashes (Building and Road Research Institute, 2020)

Pedestrians’ vulnerabilities are evident through a combination of key risk factors. The first key risk factor relates to the design of road infrastructure, which serves as a predictor of pedestrian crashes (Haghighatpour & Moayedfar, 2014). Amoako et al. (2014) have revealed (and intuitively so) that the absence of road facilities like sidewalks forces pedestrians into the road space, exposing them to motorised transport that heightens their risk of traffic crashes and injuries. For instance, Setorwofia et al. (2020) report that 8% of school children in the Cape Coast Metropolis are victims of traffic crashes due to the absence of sidewalks and crossing facilities (like zebra crossings) for their use. Additionally, Odame and Amoako-Sakyi (2019) also observed that the lack of sidewalks on 80% of high-volume routes at the University of Cape Coast exposes students to the danger of road crashes.

Useche et al. (2020) report that risky in-traffic pedestrian walking behaviour such as the consumption of alcohol, chatting with others, and the use of a mobile phone also exponentially heightens pedestrian’s risk of injuries in major cities like Accra, Lagos, Nairobi, and Banjul. In Li et al. (2018) ’s view, the impact of globalisation and the desire for technological devices to stay connected and informed via social media and other services account for the rising pedestrian inattentiveness on the roadways. In essence, pedestrians walk distracted. Kareem et al. (2021) reference pedestrians who use mobile phones while walking as aggressive road users who tend to engage in erratic activities like abrupt stopping, crossing the curb, and colliding with stationary objectives or other road users. In Ghana, Sam et al. (2019) revealed that indiscriminate pedestrian walking and crossing behaviour are major risk factors for pedestrian road crashes.

While the subject of distractive walking has not been extensively discussed in the Ghanaian context, walking dominates other travel modes as 65% of Ghanaians are estimated to reach various destinations through walking (Koinange, 2020). In Amegah’s (2022) view, walking is usually a preserve of low to middle-income people who do not have the means to acquire a private car or are limited by public transport. Despite this dominance, the literature on walking in Ghana has poorly described the level of walking and the conditions under which Ghanaians walk. For instance, Odame and Amoako-Sakyi (2019) indicate the absence of sidewalks on high-volume pedestrian streets in Cape Coast while exposing students’ daily struggle to reach lectures. In other instances where sidewalks are present,Damsere Derry et al. (2010) and Amoako et al. (2014) identify increased encroachment of pedestrian space by cars and hawkers as factors that inhibit pedestrian movement and further expose them to the risk of injuries. Additionally, indiscriminate car-following and road-crossing behaviours by street hawkers’ have also been identified by Sam et al. (2019) to increase pedestrian crashes.

Several studies, including Amin et al. (2022) and Damsere Derry et al. (2010), reveal that most pedestrian collisions occur when pedestrians are crossing or in the middle of the road and in Ghana, 68% of pedestrians are killed while crossing the roadway. Despite attributing these accidents to human error, a plethora of evidence exists on driver distraction (mainly talking and texting while driving a vehicle, i.e. distracted driving) (Ebel et al., 2023; Masello et al., 2023; Misokefalou et al., 2016). However, there is a dread of evidence of pedestrian distraction (i.e. distracted walking), particularly in Africa. As rightly opined by WHO (2013), distracted walking is more likely to be higher in countries with a greater mix of traffic or less controlled crossings as seen in Ghana’s national capital city of Accra. This paper uses a quantitative methodology to examine the predictors of distractive activities to walking in Accra, Ghana. This study offers essential insights into distractive walking in an urban environment in a developing country and policy-relevant interventions for targeting distracted walking. The remaining parts of the paper are as follows: methods, results, discussion, conclusion, and policy implications.

2 Study setting and methods

The study setting was Accra, Ghana’s capital city and, coincidentally, the country’s most populous city in terms of human and vehicular population. Accra also records the highest number of road traffic crashes yearly (i.e. top crash-prone city in Ghana) (Ghana Statistical Service, 2012, 2013), accounting for about 67% of Ghana’s pedestrian injuries. Undoubtedly, pedestrians in Accra are the most susceptible to traffic injuries in Ghana (Building and Road Research Institute, 2020).

This cross-sectional study surveyed pedestrians within Accra’s Central Business District (CBD). The CBD accounts for 57.6% of pedestrian crashes within the Accra Metropolitan Assembly (Debrah, 2012). The study surveyed respondents above 15 years (minimum age of consent in Ghana). Specifically, these include all persons who regularly walk within the CBD as walking is the predominant travel mode in the CBD. Using Miller and Brewer’s (2003) mathematical formula and a sampling frame of people aged 15 (i.e. 2,291,352) and above for Accra (Ghana Statistical Service, 2012), the study surveyed 400 pedestrians based on the computation below:

$=\frac{N}{1+N{\left(\alpha \right)}^2}$

where

‘N’ is the sample frame

‘n’ is the sample size

‘α’ is the error of margin, which in this case is 5%.

We chose a confidence level of 95% for this study.

By the formula, N = 2,291,352 and α= (0.05)²

The respondents were conveniently sampled at popular shopping malls around Rawlings Park in Accra’s CBD, including Melcom plaza, Despite stores, and SIC Mall. These malls were selected given their strategic location as commercial and leisure centres that records high foot traffic. Beyond this, engagement with respondents was easier at these selected venues since participants were inclined to stop and engage the researchers as compared to the sidewalk where commuters were focused on reaching an opportunity. The selected respondents answered an online questionnaire on Kobo Toolbox (an open-source suite of tools for data collection and analysis).

The questionnaire had three sections: respondents’ socio-demographic characteristics (e.g. sex, age, religion, occupation and martial status), ranking of distractive activities to walking (e.g. Which of these activities do you engage in whiles walking: make/receive phone calls, radio tuning, and reading sign post), and examining the extent to which respondents engaged in distractive activities to walking using a 5-point Likert scale (1-Never, 2-Rarely, 3-Sometimes, 4-Often, and 5-Very Often). The items measuring distractive activities to walking sought information on the extent to which respondents listen to music, make/receive phone calls, or browse the phone while walking. These items had a Cronbach alpha (α) value of 0.737, indicating the items’ internal consistency (Lättman et al., 2016). We transformed the outcome variable (i.e. the output of the Likert scale) into a composite value ranging from 1 to 5, where higher values denote increasing engagement in distractive activities to walking, as seen in other studies by Tamang et al. (2020) and Alrubaiee et al. (2020). All values below 2.5 were recoded as 0 for lower susceptibility to distractive walking (implying higher personal safety). Those above 2.5 were recoded as 1 for higher susceptibility to distractive walking (suggesting lower personal safety). This transformation created a dichotomous output variable forming the basis for the binary logistic regression. The study utilised eleven independent variables (sex, age, marital status, religion, level of education, occupation, residence, disability, reasons for walking, weekly time spent walking and walking time for common trips) in the model based on the literature review (Hasanat-E-Rabbi et al., 2021; Nagata et al., 2012; Useche et al., 2020).

Logistic regression was preferred to other multivariate techniques because of its robustness and sensitivity to outliers (Nwakeze, 2007). In addition, it has the advantage of estimating odds ratios for each variable to determine how much each is likely to explain the dependent variable (Huang et al., 2017). In a binary logistic model, let Pr represent the probability of engagement in disruptive walking activities, while the probability of not engaging in such activities is given as 1-Pr. Given that Y is a latent variable and that we do not actually observe Pr, but rather the outcome of Y = 1 if one engages in disruptive activity, and Y = 0 otherwise. The model is specified as:

(1) $\mathrm{P}\left(\mathrm{y}=\mathrm{j}|\mathrm{X}\right)=\frac{\mathrm{e}\mathrm{x}\mathrm{p}\left(\mathrm{X}{\beta }_{\mathrm{j}}\right)}{\left[1+{\sum }_{\mathrm{h}=1}^{\mathrm{j}}\mathrm{e}\mathrm{x}\mathrm{p}\left(\mathrm{X}{\beta }_{\mathrm{h}}\right)\right]},\hspace{0.17em};\mathrm{j}=0,1\hspace{0.17em},$(1)

where ${\beta }_{\mathrm{j}}$ is $\left(\mathrm{K}\times 1\right)$ vector, $\mathrm{j}=0,1$ and $\mathrm{K}=11$ (independent variables).

The probability of not engaging in a disruptive walking activity compared to the other option is specified in equation (2)(2) $\mathrm{P}\left(\mathrm{y}=0|\mathrm{X}\right)=\frac{1}{\left[1+{\sum }_{\mathrm{h}=1}^{\mathrm{j}}\mathrm{e}\mathrm{x}\mathrm{p}\left(\mathrm{X}{\beta }_{\mathrm{h}}\right)\right]},\hspace{0.17em};\mathrm{j}=0,1\hspace{0.17em}$(2)

(2) $\mathrm{P}\left(\mathrm{y}=0|\mathrm{X}\right)=\frac{1}{\left[1+{\sum }_{\mathrm{h}=1}^{\mathrm{j}}\mathrm{e}\mathrm{x}\mathrm{p}\left(\mathrm{X}{\beta }_{\mathrm{h}}\right)\right]},\hspace{0.17em};\mathrm{j}=0,1\hspace{0.17em}$(2)

where X is a vector of regressors and β is the vector of individual coefficients.

Following the review of the literature, the following binary logistic function was estimated to assess the likelihood of one’s engagement in disruptive walking activities.

(3) $P_{r}Ob\left(Y_i=\mathrm{j}\right)=\frac{\mathrm{e}\mathrm{x}\mathrm{p}\left(\mathrm{X}{\beta }_{\mathrm{j}}\right)}{{\sum }_{\mathrm{h}=0}^4\mathrm{e}\mathrm{x}\mathrm{p}\left(\mathrm{X}{\beta }_{\mathrm{h}}\right)};\mathrm{j}=0,1\hspace{0.17em}$(3)

From equation 3, j = 0,1 represents the likelihood of engaging in disruptive walking activities: 0 = lower susceptibility to distractive walking and 1 = higher susceptibility to distractive walking. The assumed uniform and independent distribution of the random disturbance terms form the foundation of the specified model (McFadden, 1974). The study employed the Statistical Package for Social Sciences (SPSS, version 24) and Jamovi (version 2.25) for data analyses to generate descriptive (frequencies and ranking) and inferential (binary logistic regression) statistics, respectively. The output was presented as Odds Ratio (OR) using a 95% Confidence Interval (CI). The study observed all ethical issues (including anonymity and confidentiality) during and after the data collection exercise.

3 Results

3.1 Socio-demographic characteristics

The study respondents were composed of 50.5% males and 49.5% females. More than 50% of the respondents were below 38 years. The single/unmarried were in the majority (52.5%), followed by the married (35.5%) (Table 1).

Table 1. Socio-demographic characteristics of the respondents.

Variable	Measure	Frequency	Percentages
Sex	Male	198	50.5
	Female	202	49.5
	Total	400	100
Age	16–26 years	124	31
	27–37 years	92	23
	38–48 years	141	35.3
	49–59 years	38	9.5
	60 or more years	5	1.3
	Total	400	100
Marital status	Married	210	52.5
	Single	142	35.5
	Divorced	48	12
	Total	400	100
Religion	Christian	245	61.3
	Muslim	101	25.3
	Traditionalist	53	13.3
	Deism (Others)	1	.3
	Total	400	100
Education	No formal education	46	11.5
	Primary	71	17.8
	Junior High School	34	8.5
	Senior High/Vocational/Technical	133	33.3
	Tertiary	116	29.0
	Total	400	100
Occupation	Private informal employment	211	52.8
	Private formal employment	118	29.5
	Public formal employment	56	14.0
	Public informal employment	15	3.8
	Total	400	100
Residence	Rural area	90	22.5
	Urban area	310	77.5
	Total	400	100
Observed Disability	No	385	96.3
	Yes	15	3.8
	Total	400	100
Reasons for walking	Daily commuting	193	48.3
	Making a short trip to a specific point in the city	34	8.5
	As a part of their job	173	43.3
	Total	400	100
Hours spent walking per week	1–5 hrs	94	23.5
	6–10 hrs	167	41.8
	11–15 hrs	139	34.8
	Total	400	100
Length of most common walking trip	0–15 minutes	83	20.8
	16–30 minutes	185	46.3
	31–45 minutes	106	26.5
	46–60 minutes	26	6.5
	Total	400	100

Table 1 also reveals that nearly 30% of the respondents had attained tertiary education during the survey. Furthermore, more than 80% were in private-sector employment (both formal and informal) compared to the public sector. On reasons why respondents walk, nearly half (48.3%) of respondents only engaged in walking as a mode of transport for their daily commute. These mostly include trips to transport terminals, the nearest facility or the last mile. Indeed, this was popular among respondents engaged in formal sector employment. On the other, nearly 44% identified walking as an active part of their job as most informal workers relied on walking to reach customers or sell their ware. On weekly duration spent walking, nearly 76% of respondents walk at least 6 hours while 46% of all respondents claimed to walk between 16 and 30 minutes for common trips.

3.2 Ranking of distractive activities to walking

The study sought to rank distractive activities to walking from the respondents’ perspective using the percentage of cases from the multiple-response questions presented to respondents. Figure 1 reveals that the respondents’ top four distractive activities to walking concerned using digital devices like mobile phones. Listening to music on a mobile phone emerged as the respondents’ (79%) major (first ranked) distractive activity to walking. Following were making or receiving phone calls and conversing with other people while walking (i.e. 2nd and 3rd). On the other hand, browsing the internet on mobile phones ranked 4th, which was widespread among Gen Z and millennials respondents. The least ranked distractive activities were reading signposts, eating/drinking while walking, tuning the radio, and reading a book.

Figure 1. Ranking of distractive activities to walking.

Source: Fieldwork, 2021

3.3. Binary logistic regression analysis of the influence of socio-demographic characteristics on distractive activities to walking

The study further investigated the influence of other variables on distractive activities to walking based on past studies. The model contained 11 independent variables (sex, age, marital status, religion, level of education, occupation, residence, disability, reasons for walking, weekly time spent walking and walking time for common trips). The model was statistically significant, χ2 (26, N = 400) = 218, p < .001, indicating that the model could predict the various factors that influence pedestrians’ engagement in distractive activities to walking. The model explained between 42% (Cox and Snell R square) and 57.7% (Nagelkerke R squared) of the variance in distractive activities to walking and correctly classified 64.5% of the cases. The model was also found to be fit (Hosmer and Lemeshow, chi square = 14.947, df = 8, p = 0.060).

As shown in Table 2, only seven independent variables (sex, age, level of education, occupation, reasons for walking, weekly time spent walking and walking time for common trips) significantly contributed to the model. The strongest predictor of distractive activities to walking was ‘reasons for walking’ and in this case, people who consider walking as a part of their job routines were about 14 times more likely to engage in distractive activities to walking than the referent (OR = 13.54, 95% CI = 2.93342–62.4720, p=< .001) using those who consider walking as their daily means of commuting as the reference category. Concerning sex, the study revealed that male pedestrians were more than twice as likely to engage in distractive activities to walking (OR = 2.20, 95% CI = 1.16135–4.1818, p = 0.016). Further, respondents with Secondary High School education (SHS) (reference category: no formal education) were five times (OR = 4.6, 95% CI = 1.04373–19.9062, p = 0.044) more likely to engage in distractive activities to walking. Regarding employment status, respondents working with the informal public sectors like errand workers and messengers showed a higher likelihood being about 13 times (OR = 13.29, 95% CI = 16.59522–1064.3168, p=< .001) more likely to engage in distractive walking activities while those in the formal public sector revealed a much lower chance of engaging in such activities (OR = 0.003, 95% CI = 1.87e-4–0.0664, p=< .001). With reference to weekly time spent on walking, respondents who claim to walk an average of 6–10 hours (relative to 1–5 hours) recorded lower chances (OR = 0.008, 95% CI = 0.00153–0.0506, p=< .001) of engaging in distractive activities while all variables under walking time for common trips (16–30 minutes, 31–45 minutes and 46–60 minutes) recorded higher likelihood to engage in such activities, relative to the reference category, 0–15 minutes.

Table 2. Binary logistic regression analysis on pedestrians’ engagement in distractive activities to walking.

Variable	p	Odds ratio	Confidence Interval
			Lower limit	Upper limit
Gender:
Female	Ref
Male	0.016 *	2.20	1.16135	4.1818
Age category
16–26 years	Ref
27–37 years	0.390	0.66	0.25836	1.6955
38–48 years	0.582	1.29	0.52099	3.1933
49–59 years	<.001 **	12.30	3.09036	48.9638
60 or more years	0.604	0.39	0.01104	13.7212
Religion
Christian	Ref
Muslim	0.957	0.98	0.47873	2.0077
Traditionalist	0.412	1.53	0.55544	4.2002
Others	0.989	0.00	0.00000	Inf
Marital status
Married	Ref
Single	0.275	1.56	0.70293	3.4525
Divorced	0.168	2.11	0.73132	6.0608
Education
No formal education	Ref
Primary	0.248	0.41	0.08949	1.8649
Junior High School	0.201	2.65	0.59425	11.8465
Senior High School	0.044*	4.56	1.04373	19.9062
Tertiary	0.700	1.28	0.37090	4.3830
Occupation
Private Informal	Ref
Private formal	0.590	0.62	0.11037	3.5025
Public informal	<.001**	13.29	16.59522	1064.3168
Public formal	<.001**	0.00	1.87e-4	0.0664
Residence
Rural area	Ref
Urban area	0.900	1.05	0.52132	2.0982
Observed Disability
No	Ref
Yes	0.708	1.43	0.22103	9.2263
Reasons for walking
Daily commuting	Ref
Short trip to a specific point in the city	0.966	0.97	0.24919	3.7801
As a part of job routine	<0.001**	13.54	2.93342	62.4720
Time spent walking (Weekly)
1–5 hours	Ref
6–10 hours	<.001**	0.01	0.00153	0.0506
11–15 hours	0.087	0.22	0.03924	1.2445
Walking time for common trips
0–15 minutes	Ref
16–30 minutes	<.001**	7.04	17.16630	288.4098
31–45 minutes	<.001**	4.22	68.05334	3139.0748
46–60 minutes	<.001**	3.05	944.17851	469464.8719
Pseudo R^2	0.419
Cox and Snell R	0.420
Nagelkerke R^2	0.577

p < 0.05*, p < 0.01**.

Ref = Reference category.

4 Discussions

The present study explored the predictors of distractive activities to walking in Accra, Ghana. At the outset, the study assessed respondents’ ranking of the distractive activities to walking. It came to light that listening to music (on a mobile phone) was the first-ranked (i.e. the most) distractive activity to walking. Earlier studies, including Schwebel et al. (2012) and Mikusova et al. (2021), have confirmed the distracting effect of music on pedestrians’ attention. These studies cite the distractive impact of listening to music on pedestrians’ auditory signals as a critical risk factor for evaluating the safety of a walking environment. Indeed, many busy streets in Accra are characterised by loud noise, making it difficult to detect the movement of other pedestrians or non-motorised transport like bicycles that approach from behind. Similarly, making or receiving phone calls and conversing with other people while walking is deemed equally distracting and risky (Haolan et al., 2021; Zhou et al., 2019)

In addition to the above, the 4th, 5th, 6th, and 7th ranked items are keenly related to one’s vision, including browsing the internet, buying items, using smartphones for directions, and reading signage while walking, respectively. The low rating accorded to visual distraction is inconsistent with findings from previous studies such as Kareem et al. (2021) and Adams et al. (2020). These studies identify visual distraction as millennials’ most popular activity in developing countries. Unlike mega cities like New York, Accra’s CBD has limited interactive signage or billboards than audio-powered vehicles, which may account for respondents’ proportionate lower engagement with reading signage while walking and their subsequent exposure to any risk of injury while walking.

This study also indicates that sex, age, level of education, occupation, reasons for walking and weekly time spent walking are significant predictors of distractive activities to walking. The data revealed that males have a significant association with distractive activities to walking, consistent with the views of Lewis and Duch (2021), who opine that male tolerance for risk and desire for adventure are primary factors informing their dominance in risky behaviours. Consequently, the World Bank (2021) statistics on susceptibility to injuries also identified male pedestrians to be three times as likely to sustain injuries than their female counterparts owing to a variety of risky walking behaviours. These findings further confirm Atkinson (1957) theory of risk-taking, which associates the male gender’s belief in success in a risky venture with their engagement in distractive activities.

On the age score, the data revealed a significant association between persons between 49 and 59 years and the pursuit of distractive walking behaviours despite accounting for less than 10% of the study population. This finding contradicts the findings of previous studies, such as Schwebel et al. (2012). Schwebel et al. (2012) identify millennials or young people as culprits for distractive walking, especially in today’s technological world. However, a growing body of literature has also identified the aged as engaging in distracted walking. The aged are less likely to accurately estimate their walking environment, given the relationship between ageing and body functionality. Therefore, in this light, the slightest engagement in distractive activities to walking like looking at signage/objects of interest, buying items, or conversing with other pedestrians may increase the risk of injury.

Regarding one educational attainment, respondents with Senior High School Education were also significantly associated with distractive activities to walking. Senior High School Education in this context refers to persons who have only graduated from the compulsory nine years of basic education, like grade Nine in America. Incidentally, the Ghana Statistical Service (2021) ranked this educational level to be the commonest terminal point for Ghanaians and known to dominate Ghana’s private informal economy (Ghana Statistical Service, 2014), including hawkers alongside busy walkways or other business entities that encroach sidewalks to reach prospective customers for their trade. This situation increases respondents’ likelihood of engaging in disruptive activities since informal economic agents in the CDB compete with pedestrians for space or run after moving cars to sell their wares.

Relative to respondents who only walk for short trips, respondents who identified walking as part of their job routines recorded a significant likelihood of engaging in distractive walking. These respondents may qualify as captive walkers since most part of their day is spent walking. From the data, this position was found to be influenced by the nature of respondents’ economic activity since sales/marketing agents, courier service providers, hawkers and other informal activities required active walking in a congested location like Accra. To account for captive walkers’ engagement in distractive walking practices, Anciaes et al. (2017) attribute this to poor planning and low accessibility in urbanised areas, fewer provision of pedestrian space, poor street layout and low enforcement on the removal of obstacles. Indeed, these insights may require some public policies to make walking safer for captive walkers as pedestrians in Ghana are exposed to the risk of traffic injuries more than other road users (Amoako-Sakyi, 2017)

Finally, time allocated to working also revealed a significant association with distractive walking. Whether weekly or daily, respondents who dedicate more time to walking in the CBD recorded higher likelihood of engaging in distractive walking activities. From the literature, no direct relationship between walk time and pedestrian engagement in distractive activities was observed but Russo et al. (2018) reveal the role of walk time in exaggerating the influence of other activities like making phone calls or browsing the internet on pedestrian safety. Specifically, pedestrians’ reaction time reduces by 0.25 seconds per minute and more walk time implies an increasing level of inattentiveness, further exposing pedestrians to the risk of injuries and crashes in major cities like Accra (United Nations Environment Programme, 2022).

5 Conclusion and policy implications

The study has shown that sex, age, level of education, occupation, reasons for walking and weekly time spent walking influence pedestrians’ engagement in distractive walking activities. Audio and visual distractive activities characterised the respondents’ walk in Accra. In sum, respondents who identified as males, aged, with lower educational attainment, informal workers and those with more walking time were more likely to engage in distractive activities to walking.

From the findings, this study suggests inclusion of a policy statement on pedestrian walking behaviour in the existing ‘Pedestrian Safety Action Plan in Accra’ document by the Accra Metropolitan Assembly (AMA). Even though this action plan seeks to promote pedestrians’ safety on sidewalks, crossing points and inner-city trips, no mention of pedestrian walking behaviour is seen in this action despite its heavy focus on the built environment. Therefore, the Accra Metropolitan Assembly may enact various laws that restrict pedestrians from listening to music with headphones, making phone calls while crossing or engaging in unwarranted conversations. Additionally, the city authority (AMA) can institute environmental planning schemes to minimise distractions to pedestrians. This may range from restricting the activities of hawkers who compete with pedestrians for space to the activities of outdoor advertisers, local preachers or other advertising agents.

6 Limitations of the study

The study was limited to specific reference points in Accra and may not represent all demographic groups like children, persons with disability and the aged. This implies that the research findings may not be generalised to the entire population in Accra. Beyond this, the choice of a quantitative inquiry may not offer a more profound and contextual perspective on why people engage in distractive activities to walking. We suggest a follow-up qualitative study to explore the contextual perspectives behind distractive walking in Accra.

Ethical approval

The research protocol was approved by the Institutional Review Board (IRB) of the Department of Geography Education, University of Education. All other activities involving human participants were done in accordance with the ethical standards of the University of Education which conforms to the 1964 Helsinki Declaration and other comparable ethical standards. Data analysis was executed using completely anonymised data and no animals were used as subjects in this study.

Informed consent

Informed consent was obtained from all individual participants included in the study.

Acknowledgments

We appreciate the commitment and efforts of all research participants, proof-readers and anonymous reviewers.

Disclosure statement

No potential conflict of interest was reported by the authors.

Data availability statement

The data is available via https://figshare.com/s/642a4b8da1e1efa29d17

Additional information

Funding

The author did not receive support from any organisation for the submitted work.