Factors Associated with Travel Patterns Among Mixed-Use Development Residents in Klang Valley, Malaysia, Before and During COVID-19: Mixed-Method Analysis

Abstract

Mixed-use development (MXD) is crucial for urban planning and travel. The COVID-19 outbreak had a significant impact on travel behaviour and MXD projects worldwide, particularly in high-income countries. However, limited studies have explored the predictors of MXD usage and travel patterns in low- and middle-income countries, including Malaysia, and how these events were affected by COVID-19. Using the Rowley and extended Hopenbrouwer and Louw models, this study investigates the travel patterns within MXD premises, their associated factors, and residents’ perspectives of internal and external trips before and during COVID-19 in Klang Valley, Malaysia. A mixed-method study was conducted by using a validated survey and performing a structured interview with MXD residents. A total of 134 and 52 respondents participated in the survey and qualitative interviews, respectively. Data were analysed using descriptive statistics, logistic regression models, and thematic analysis. A significantly higher proportion of MXD respondents engaged in external travel compared to internal travel before and during COVID-19. Before COVID-19, external travel was significantly higher among younger residents, government servants, higher-income earners, and those who owned a car and had recently moved to MXD areas. The odds of internal travel were significantly higher among private-sector employees, students, and low-income earners. During the pandemic, external travel frequency was significantly higher among male residents, older residents, government servants, high-income earners, and those with multiple vehicles. Residents with more parking lots tended to travel less internally compared to those with fewer parking lots allocated. Qualitative analyses revealed that cost-saving, convenience and comfort, social lifestyle, health and well-being, and green environment were the factors that shaped MXD residents’ perceived benefits of trip internalisation. Meanwhile, the barriers to internal trips included the lack of infrastructure, poor management, lifestyle activities/individual factors, and environmental factors. The recommended strategies to reduce external trips were to ensure diversified services and accessibility, inclusiveness in planning activities, promoting social interaction, and work-from-home policies. These findings reflect the strategies that can be incorporated to reduce external trips generated by MXD and enhance effective traffic management.

1. Introduction

With the advent of the information age, the urban functional structure in the industrial age is also facing substantial changes, leading to a new way of living and working [1]. These changes were further exacerbated following the COVID-19 pandemic, triggering a new lifestyle that has blurred the life–work boundary and increased the demand for land use mix (LUM), also known as mixed-use development (MXD) [1]. Mixed-use developments (MXDs) are now considered a core principle of contemporary planning strategy, together with the concepts of “New Urbanism” and “smart growth” [2,3].

MXD is a multifaceted concept that focuses mainly on the pattern of spatial development [1]. According to Tian et al. [4], MXDs are a single development project that integrates and interacts with different land uses, such as office, retail, restaurants, entertainment, hotels, and/or residential. Given its internal connectivity, users of MXD share the internal streets, walkways, driveways, and parking spaces. MXDs are more advantageous relative to urban expansion, particularly by enabling effective land use, less car dependence, better accessibility and fostering greener environments [5]. Research from several countries revealed that MXD facilitates active travel [6] and social interaction [7], while maintaining and promoting the performance and economic vitality of transit-oriented development (TOD) [8].

In recent years, many cities worldwide have experienced a decline in urban and economic vitality. The coronavirus disease (COVID-19) pandemic has further exacerbated the economic downturn, accompanied by economic shocks and complicating pre-existing urban crisis, including economic recession and long-term social inequities [9]. Modelling research forecasts that the negative effects of COVID-19 on MXD will prevail even after the public health threat is over [8,10]. Stakeholder engagement, relationship, attributes, influence, interest, needs, satisfaction, expectation and behaviour were identified as key stakeholder impact factors affecting the success of MXD projects during the COVID-19 pandemic [10].

The impacts of COVID-19 on travel behaviour and activity patterns have been explored in various countries worldwide [11,12,13]. Most studies focused on travel pattern changes and alterations in working and shopping behaviours. Different methods have been utilised, including online surveys [12,13] and objective data quantification via Travel Diary App [11] and GPS logger [12]. Accumulated data from a German study indicated that trip rates decreased by 39% during the strictest lockdown period between late March and early April 2020 [13], and 25% of the working population worked in home offices. These data reflect the changes in daily activities and travel-related behaviour due to the pandemic.

Further studies emphasise the implications of the global crisis for the relation between well-being and transport, either in the context of the pandemic or for the future agenda for policy and practice. For instance, Carteni et al. [14] explored the role of transport accessibility during the spread of COVID-19 and proposed tailored policy strategies for managing the public health threat depending on the area’s accessibility level. The concept of Responsible Transport—a transport policy approach, was proposed by Budd and Ison [15], which considers both environmental, public health and well-being aspects. These studies stress the significance of elucidating the impacts of the pandemic on travel behaviour in order to develop comprehensive and sustainable transport policy, measures and practice strategies.

In Malaysia, ever since the first confirmed cases of COVID-19 on 25 January 2020 [6], several waves of COVID-19 outbreaks have been reported [16]. During the pandemic, all types of travel were strictly regulated, including the enactment of non-medical interventions such as “stay-at-home” orders and travel restrictions [16]. Enacted policies led to the closure of many public services, which significantly affected economic activities, and in turn, altered people’s travel behaviour [13]. However, in Malaysia’s context, little is known about how MXD projects and infrastructures were affected by COVID-19, including travel behaviour and the potential implications for travel patterns after the pandemic. Apart from analysing potential mid and long-term effects, understanding how travel behaviours change during a global crisis and the underlying reasons for this is pertinent for developing appropriate strategies for similar future events. To our knowledge, no published research in Malaysia has looked into how different phases of the pandemic and national policies affected MXD projects. Prior studies focused on factors influencing travel intention during the post-pandemic or recovery phase [16] and changes in individual travel behaviour after the series of MCOs [17], without focusing on MXD. The analysis of MXD residents’ travel patterns offers the opportunity to elucidate the socio-demographic, socioeconomic, and land use data, which are vital for estimating trip rates.

Against this background, this study focuses on determining the travel patterns before, during and after the pandemic, and the associated factors among MXD residents in Klang Valley, Malaysia. The following research questions were addressed: (1) What were the changes in travel patterns (internal and external trips) during the strictest COVID-19-related lockdown period relative to the period before the global crisis? (2) What are the factors that shaped the changes in travel patterns between these periods? (3) What are the potential long-term impacts of these changes, particularly in the post-pandemic era?

The subsequent part, Section 2, presents the literature review, theoretical background and conceptual model to guide the research process. Section 3 describes the material and methods utilised for the quantitative and qualitative phases, while Section 4 presents the corresponding results. The findings are discussed in Section 5, before highlighting the study limitations, conclusions and recommendations for future research in Section 6.

2. Literature Review

The reviewed literature is presented under three broad sub-topics: (1) the relationship between land use, travel measures and trip internalisation, (2) factors influencing land use mix (LUM)/mixed-use developments (MXDs), (3) travel behaviour during COVID-19/global crises, and (4) MXD in Malaysia. Each sub-section is discussed below.

2.1. Relationship Between Land Use, Travel Measures and Trip Internalisation

Several studies have shown that a relationship exists between land use and travel measures [18], but the magnitude of such associations is debatable [19]. The main question is whether travel variations are sufficient to justify land use interventions. In some studies, the magnitude of a land use variable is considered small, but the effect becomes substantial when combined with other land use variables [20]. Other researchers found a weak [21] or insignificant relationship [22] between land use and travel measures. In addition, these authors posit that individual characteristics (i.e., based on RSS) exert significant effects on travel patterns rather than land use.

Given the inconsistent results regarding the effect of land use on travel measures such as travel distance/time and trip frequency, researchers have hypothesised that compact land use is designed to internalise travel within a locality rather than reducing the number or trip frequencies [21]. Thus, compact land use encourages commuters to walk and use bikes instead of automobiles. For instance, Ewing et al. [21] analysed samples from 20 communities in South Florida in the USA and found that land use mix and regional accessibility accounted for a significant portion of community internal trips.

The concept of job–housing balance has also been analysed using the trip internalisation approach [23], which is measured based on commuting distance/time and commuting vehicle miles travelled. Zhang et al. [24] found that MXD districts in Austin, Texas, recorded a 40% higher internal trip rate compared to the traffic analysis zones (TAZs). In Iran, trip internalisation was differentiated at the TAZ level by using five principal components of mixed land use and socio-demographic variables [25]. In Canada, local accessibility had a stronger effect on trip internalisation (measured via a local trip index (a combination of total travel distance and activity space internalisation)) compared to regional accessibility [26]. All the aforementioned studies investigated trip internalisation by measuring internal trips at a group level, whereas only limited studies have inferentially explored trip internalisation among individual travellers by measuring both external and internal trips.

2.2. Factors Influencing Land Use Mix

Various intensities and patterns of land use are usually predicted by complex and influential factors. This also applies to land use mix—a multifactorial concept that is influenced by socioeconomic, physical and policy factors. Socioeconomic factors are among the core perspectives for investigating urban spatial and land use structure. These factors encompass demographic characteristics, income level, industrial structure, regional culture, land price and other characteristic factors. As evidenced in a study by Talen [27], demographic characteristics such as age, race, income level, housing tenure and housing type were identified as factors influencing land use mix. The researchers further showed that LUM is a natural urban development state that stems from the drive of residents’ daily needs. In Taiwan, specific social and economic factors, particularly family-related (economic structure, surplus labour force, and organisational structure) and the balance between life and production needs, were the significant factors influencing land use mix [1]. In residential areas, LUM was strongly predicted by the size of commercial areas and population size [28].

As for physical factors, this dimension of land use mix refers to the influence of the physical environmental conditions and resource endowment, such as shape, plot size, terrain, location, traffic conditions, and geographical conditions [1,2]. For instance, the development of commercial and residential LUM was influenced by the building forms, frontage road widths, street corners and adjacency of the CBD in the Wanhua districts of Taipei. Similar findings were obtained in another study conducted by Chih-Hung Hsu [29] using regression models and canonical correlation analysis. In terms of land use mix, researchers were able to use regression analysis to prove that the LUM pattern is influenced by the spatial accessibility characteristics of the overall road network structure [30,31].

Recent findings have equally depicted the pivotal impacts of policies and regulations on land use change and MXD [1,31]. A strong and well-informed policy is pertinent for land use mix [32,33], as evidenced by the failure to effectively implement mixed-use projects when effective policy guidance is lacking [1]. An example is the success of LUM in Hong Kong’s statutory plan system, which provides dominant use and enables other related uses to enable land use mixing and compatibility [34].

2.3. Travel Behaviour During Disease Outbreaks/COVID-19

A health crisis ultimately affects travel patterns and the movement of people. This is primarily due to the intrinsic motivation to protect oneself from infection and the impacts of measures taken to prevent the spread of the disease, such as travel restrictions [35]. Apart from influencing travel intention to a specific destination and reductions in trip frequency, the decision not to travel at all is also affected [8,36]. This event was demonstrated in various studies conducted in the United States [37] and Sierra Leone [38] during the Ebola outbreak in 2014. On the contrary, a survey conducted by Leggat et al. [39] in Australia during the H1N1 outbreak in 2009 showed that most people were unwilling to cancel their travel, despite being aware and concerned about the virus. Similar results were reported by Lee et al. [40] in South Korea, wherein travel intention was not affected by the perception towards H1N1. These findings reflect the diverse changes in travel behaviour in response to a global health crisis, which appears to be shaped by several factors ranging from socio-demographic to perceptions towards the crisis and external determinants such as policies [41].

With the most recent COVID-19 pandemic, numerous studies on travel behaviour have been performed. De Vos [42] emphasised how the pandemic has affected travel patterns and suggested walking and cycling as sustainable alternatives while facilitating well-being. In the USA, large reductions in travel using public transit were observed in main cities such as Chicago and Washington [43]. Although several empirical studies have explored behavioural travel changes during and after the pandemic, most were conducted in high-income countries, with limited research in low- and middle-income nations [36,44]. Moreover, none of the available studies looked into how specific transportation systems, such as MXD, were impacted by the pandemic and the aftermath in the post-pandemic era.

2.4. Mixed Land Use in Malaysia and the Impacts of COVID-19

While socioeconomic, physical and policy factors are important predictors of mixed land use and development, there is a scarcity of data to understand these events in the Malaysian context. MXDs in Malaysia are driven by a non-planning-oriented mixing in which mixed land use is not comprehensively guided by government-driven policies [45]. Accumulated evidence from the literature reflects that MXDs in Malaysia still lag relative to high-income countries. One of the main issues is the development of a standardised and reliable manual for assessing land use mix.

Malaysia’s traffic engineers and planners usually use the Malaysian Trip Generation Manual (MTGM) published by the Highway Planning Unit (HPU), Ministry of Works Malaysia (MoWM). However, the manual is limited to forecasting single land use and is not suitable to be used for estimating the vehicular traffic at MXDs; the actual traffic may be overestimated [46]. The current MTGM is limited in estimating trips for new developments in urbanised areas (downtown) that support multimodal transportation options such as infill location, MXD and TOD. Moreover, MTGM omits transit, non-motorised transportation facilities (pedestrian infrastructure, bicycle facilities) and mixed-land use when deriving the trip generation models. A robust trip generation model needs to be established to estimate the future vehicular trips for MXD and TOD with consideration of multimodal transport. Analysis of (MXD) residents’ travel patterns offers the opportunity to elucidate socio-demographic, socioeconomic, and land use data, which are vital for estimating trip rates.

Research from high-income countries has shown the huge negative impacts of COVID-19 on MXD, with substantial changes in travel behaviours and patterns, extending into the post-pandemic era in some countries [8]. The first confirmed case of COVID-19 in Malaysia was reported on 25 January 2020, followed by multiple waves of outbreaks [36]. Between March 2020 and June 2021, Malaysia went through various phases of movement control orders (MCOs), ranging from partial to complete lockdown and movement restrictions in order to curb the spread of the virus [47]. Non-medical interventions such as “stay-at-home” orders and travel restrictions were enacted by Malaysia’s MOH to bring the outbreak under control [47]. Local studies revealed that these policies and control measures had a significant toll on travel behaviour and economic activities [16,17]. Nevertheless, the impacts on travel behaviour among users of MXD remain poorly understood. Previous studies in Malaysia focused on factors influencing travel intention during and after the pandemic [16] and alterations in travel behaviour during the different MCO periods [17]. Meanwhile, no study has looked into how COVID-19-related policies affected MXD projects and infrastructures, including travel behaviour and the implications for travel patterns in the post-pandemic era.

2.5. Conceptual Model

A conceptual model is pertinent to elucidate the concept of MXD. Rowley proposed a conceptual model by describing MXD as a multi-faceted concept comprising three main elements: socioeconomic setting, location and spatial texture [1,48]. Subsequently, Hopenbrouwer and Louw [49] extended Rowley’s model by including dimension, texture, scale and location, which has become an essential foundation for several MXD-related research.

The initial model developed by Rowley emphasised that MXD occurs in a specific dimension, whereby the mixing ensues between various land uses in a horizontal trajectory [49]. On the other hand, an extension of Rowley’s model is the multi-dimensional model synthesised by Hopenbrouwer and Louw [49], characterised by shared premise dimensions accounting for horizontal, vertical and time dimensions. The first scenario is the mixed use of several land functions/uses in a specific location, occurring within the same space unit. An example is the “family workshop” [50] in which production and living functions are integrated into the same unit to facilitate shared mixing. Other scenes are mixed-use between different land functions/uses, either on a two-dimensional plane or vertical dimension, such as the urban commercial complex [1].

As for scale, Rowley’s model was extended by adding four types of scales: building, block, district and city [49]. While some MXD researchers selected the scale system based on the local characteristics of specific countries, such as the studies by Bordoloi et al. [51] in India and Kong et al. [52] in Beijing, China, other scholars used the local scales accepted by the public [1]. Examples of the latter scales include the analysis of the LUM status of many Canadian cities based on the perspectives of suburb and downtown and the local scales available in Eric and Erik’s model, such as city centres, brownfield sites, suburban locations, and greenfield locations [1].

In terms of settlement texture, Rowley emphasised that MXD is a critical aspect of the internal texture of urban settlements, emanating from three areas: grain, density and permeability. In this context, grain refers to the size and division of city blocks [18]. Density, on the other hand, is defined as the number of residents per unit area and constitutes an important urban vitality indicator [49]. Urban vitality can be effectively achieved by ensuring 100–200 residents per unit area. Permeability stems from the city’s road network pattern and depicts the pedestrian movement of a given urban space. Hopenbrouwer and Louw [49] extended this component of Rowley’s model by maintaining two features of Alan’s model: density and grain, while introducing a new element, “interweaving”, to reflect the distribution characteristics of various land functions/uses [53].

2.6. Model Framework

This study uses the model global structure to explicitly introduce the frequency of simple and complex travel within a framework that accounts for long-term decisions relating to commuting distance, location patterns, car and motorcycle ownership, and shorter-term decisions such as daily distances travelled [54]. As shown in Figure 1, all variables are endogenous to the model, excluding the socioeconomic variables. Thus, the model incorporates self-selection effects, for example, household composition, income, and presence of children [55,56], that may correlate with socioeconomic variables. The effects progress from long-term decisions to short-term decisions, but also consider the feedback effects, whereby location patterns are equally shaped by car ownership [57]. Some studies investigated several relationships, including car ownership [55,58] and the link between mode choice and travel complexity [59]; these relationships are understudied in the Malaysian context.

3. Materials and Methods

3.1. Study Area and Study Design

The study area for this study was Klang Valley in Malaysia as shown in Figure 2. This study employed a cross-sectional design and mixed methodology, comprising quantitative and qualitative approaches. For the quantitative phase, a validated questionnaire developed by experts in traffic assessment research was distributed electronically to eligible participants. Due to the restriction of movement control order (MCO) in Malaysia between March 2019 and March 2022, the traffic survey was affected, because no traffic count could be conducted during this period. The initial plan was to conduct a face-to-face interview as one of the initiatives to estimate the internal capture trips. Unfortunately, due to the constraints of COVID-19 standard operating procedures (SOPs), the questionnaire study was performed online to avoid physical contact. Meanwhile, the qualitative survey entailed a structured online interview with MXD residents, which was conducted immediately after the relaxation of the MCO.

3.2. Study Duration and Study Population

The online questionnaire study was conducted over 6 months, starting from January 2021 and lasting until July 2021, targeting MXD residents. Meanwhile, the semi-structured interview was performed from September 2021 to January 2022. The respondents were selected based on convenience sampling since a random sampling technique could not be employed due to COVID-19 conditions. The study location was divided into 4 regions, and proportional sampling was performed to select suitable respondents in each study site based on the estimated MXD user population.

Four sites (A–D) located in Subang Jaya, Kuala Lumpur, Petaling Jaya, and Setapak were selected for the survey based on the concentration of different MXD designs in these areas and the appropriateness for traffic and travel behaviour surveys [60]. The first MXD has two types of vertically integrated land use. The upper part is a four-block residential condominium, while the bottom part is a four-storey shopping mall. The second site is a vertical MXD with retail shops beneath a residential apartment, whereas the third site is an MXD of horizontally integrated retail shops and luxury condominiums. The fourth MXD site is a vertically integrated condominium with a shopping mall.

The eligibility of the respondents was ascertained by filtering whether they were MXD residents or not. A few questions in the online questionnaire were purposely designed to validate the eligibility of respondents, such as specific residential location and the type of residential property. As for the qualitative study, respondents to the online survey were contacted by the researcher to ascertain whether they would be willing to participate in the interview. Those who provided a positive response were contacted on a later date to schedule the online interview.

3.3. Questionnaire Design

The questionnaire consisted of three main sections and was administered to the MXD residents of Klang Valley in Malaysia over 6 months, starting from April 2020 and lasting to September 2020 (Refer to Supplementary File S1). Travel behaviour and frequency before COVID-19 refer to practices before the announcement of the Movement Control Order (MCO) on 18 March 2020 and the government’s declaration of measures to address the pandemic in Malaysia. The time span for this period is from October 2019 to January 2020, which is approximately 6 to 2 months before the MCO was implemented. This time frame was considered in order to enable respondents to recall their most recent travel behaviours and trip frequencies, since a longer duration is more likely to increase recall bias. Meanwhile, travel behaviour during COVID-19 entailed practices following the MCO and the present time when the participants responded to the online survey (from March 2020 to September 2020). This information was provided in the preamble pages of the questionnaire. For the MXD travelling pattern, an external trip was classified as a vehicular trip that requires a mode of transport to travel; meanwhile, an internal trip was classified as a person or pedestrian trip where no mode of travel is required, and the destination is reachable by walking.

The main contents of the questionnaire began with 10 questions (Questions number 1 to 9 and 14) on residents’ socio-demographic characteristics, such as gender, age, marital status, employment status, monthly income, household size, private transport ownership, allocation or subscription of personal parking lot, residence property ownership and housing occupancy duration. The second part was designed to validate whether the respondents are staying on the MXD premises. A total of 11 questions were used to filter and validate the questionnaire (Questions 10 to 13 and 15 to 21).

The second and third sections of the survey focused on the MXD residents’ trip purposes and travel behaviour before and during the COVID-19 pandemic. For the second section, eight questions (Questions 22 to 29) were designed to gather data on the travel frequency (vehicular trips), reason(s) to stay at home (not to travel), purpose(s) of external trips, purpose of external vehicular trip and timing, use of public transport, purpose of internal person trips and frequency, preference to stay at MXD premise and choice of rapid transit. Similar questions (Questions 30 to 37) were asked to assess residents’ travel behaviour during the pandemic. Questions 22, 27, 30 and 35 were designed and selected as dependent variables for transportation modelling in the study.

As for the structured interview, two sections (i.e., A and B) were designed to explore MXD residents’ perspectives on travel patterns and behaviours. Section A focused on demographic characteristics. Section B entailed travel patterns and perceptions, including the perceived effect of COVID-19, strategies to reduce external trip frequency, and effective methods to encourage internal trips. Detailed information on the interview contents is provided in the Supplementary File S1.

3.4. Pilot Study

A pilot survey was conducted to assess the reliability and validity of the survey questionnaire and interview guide. For the quantitative survey, 40 respondents were selected randomly from nearby cities around the Universiti Putra Malaysia Serdang Campus to complete the instrument. This pilot was performed from November to December 2020. Meanwhile, 10 respondents were recruited for the pilot testing of the qualitative interview guide from mid-April 2021 until early March 2021. Resultantly, the questionnaire demonstrated an acceptable level of reliability with an overall Cronbach’s alpha of 0.85, with at least 0.79 and above for each section of the instrument. Feedback from two experts in qualitative research was used in modifying the interview questions. Overall, both research instruments were considered suitable to be used for the actual sampled population. None of the respondents for the pilot study was recruited for the final survey.

3.5. Ethical Approval

The instrument and procedures to be applied during the survey have been approved internally by Universiti Malaya Research Ethics Committee for the use of human subjects in Research (JKEUPM). The questionnaire form is attached as a Supplementary File S1.

3.6. Administration of the Questionnaire and Data Collection

To investigate the MXD residents’ travel behaviour in Klang Valley, an online questionnaire study was arranged and conducted at the MXD residences, via social media (e.g., Facebook, WhatsApp, Telegram, etc.) and professional networks (e.g., LinkedIn) using “Google Forms”. Compensation was given to encourage more reliable responses from MXD residents by providing a stipend of RM5 via a “Touch’ n Go” eWallet reload pin for each respondent.

As for the interviews (i.e., qualitative study), a total of 68 respondents from the online survey indicated their willingness to participate in the interview. Hence, respondents were selected via a purposive sampling strategy. The potential participants also noted down their preferred methods of conducting the interview, either via Google Meet or Zoom Meeting. This information was obtained following direct phone and email contacts with the respondents. All the interviews were organised by the first author of this manuscript. It took approximately 20 min to complete the interview with each respondent. Audio recordings of the interviews were transcribed by the researcher, and relevant data were documented in Microsoft Excel and prepared for analysis. Data saturation during the interview session was achieved when no new information could be gleaned from participants’ responses.

3.7. Data Management and Analysis

In the online questionnaire, nine different frequencies of travel were offered to MXD residents for selection. To simplify the test of significance and classification model, the travel frequencies were re-classified into three general classes as described by Mayo et al. [61]. The study recommended that travel frequency in urban cities during the pandemic can be grouped into three classes, as shown in

Table 1. Re-classification of Travel Frequency. 

Old Class	New Classes
Class 1	Seldom/Never
	Once monthly
Class 2	2 trips/month
	More than 2 trips/month
	Once weekly
Class 3	2 trips weekly
	>2 trips weekly
	Once daily
	2 trips daily
	>than 2 trips daily

.

Data analysis was performed using SPSS, version 23 (IBM Statistics). Data normality tests were performed by checking for the Kolmogorov–Smirnov, Shapiro–Wilk and levels of Kurtosis and skewness. Continuous variables were reported either as means and standard deviations or median and interquartile range, depending on the data distribution. For the non-normal distributed data, the mean rank was computed from the Mann–Whitney U test to compare the dependent variable according to the various demographic and socioeconomic characteristics. Upon completing the bivariate analysis, multinomial regression models were built to account for the multiple dependent variables and their association with residents’ demographic and socioeconomic characteristics.

Logistic regression models were built separately to identify the factors associated with MXD residents’ external and internal travel frequencies before COVID-19 and at present. The frequency of travel was re-classified into three classes, and a multinomial logistic model was built in the form of two distinct linear models. In this model, the coefficients represent the likelihood of an individual with a particular trait to fall in a particular class in comparison to the reference class (Class 2). The exponent of the coefficients represents the ratio of the probability of individuals being in a particular class given a particular trait. For example, participants who are a year older are 1.05 times more likely to fall in Class 1 compared to Class 2 and are 0.99 times more likely to fall in Class 3 compared to Class 2. As unordered categorical data, such as employment status and transport ownership, were present in the model, a category was chosen from the possible outcomes as the reference category. For employment status, the reference category was participants who were employed as government servants. For transport ownership, the reference category was participants who had no private transport. Hence, the exponent of coefficients is only interpretable in comparison to the reference category. A p-value of 0.05 was set for the analysis. Model fit was determined based on the Hosmer–Lemeshow and goodness-of-fit test, as well as the Akaike criterion.

Qualitative data were analysed thematically using NVIVO as described by Braun & Clarke [14] by identifying group patterns across the dataset, coding text segments and organising them into broader themes that represent shared experiences among participants. The analysis was performed until thematic saturation was reached, meaning no new themes or insights emerged. Themes were supported with relevant quotes from the participants.

4. Results

4.1. Survey Findings

A total of 254 responses were received from the online survey. Upon checking for missing data, 53 responses were incomplete, while 201 were eligible for further analysis. A total of 67 responses were from non-MXD residents, leading to 134 validated responses for final analysis.

4.1.1. MXD Residents’ Demographics

Table 2. MXD respondent profiles.

No.	Variable	Characteristic/Category	Frequency (n)	Percentage (%)
1	Gender	Male	80	59.70
		Female	54	40.30
		15–24 years (early working age)	38	28.36
		25–54 years (prime working age)	93	69.40
		55–64 years (mature working age)	3	2.24
3	Marital status	Single	80	59.70
		Engaged	21	15.67
		Married	31	23.14
		Divorced	2	1.49
4	Employment status	Student	35	26.12
		Employed as government servants	11	8.21
		Employed by the private sector	58	43.28
		Employed by a non-governmental organisation	4	2.98
		Self-employed/Own business/freelancer	14	10.45
		Unemployed (retired, housewife, graduate)	11	8.21
		Unable to work	1	0.75
5	Monthly income	>RM 10000 per month	11	8.21
		>RM 5000 < RM 10,000 per month	32	23.88
		<RM 5000 per month	54	40.30
		No source of income	37	27.61
6	Household size	Single staying	45	33.58
		Small family with 2 to 4 members	59	44.03
		Medium-sized family with 5 to 10 members	30	22.39
7	Private transport ownership	1 car	65	48.51
		1 motorcycle	15	11.19
		1 car and 1 motorcycle	1	0.75
		More than 1 unit of cars and motorcycles	4	2.98
		No private transport	49	36.57
8	Allocation/subscription of personal parking lot	1 unit of parking lot per household	58	43.28
		2 units of parking lot per household	40	29.85
		3 units of parking lot per household	16	11.94
		No/Not applicable	20	14.93
8	Residence property ownership	Sole Ownership	54	40.30
		Tenant	56	41.79
		Sub-tenant	24	17.91
10	Housing occupancy duration	Less than 1 year	24	17.91
		Between 1 and 2 years	25	18.66
		More than 2 years	85	63.43

presents the demographics of the MXD respondents for subsequent analysis. From the online questionnaire study, 59.70% were male respondents (80) and 40.3% were female respondents (54). A huge majority of MXD respondents were between 25 and 54 years old (69.4%), while the remaining ranged from 15 to 24 years old (28.4%) or 55 to 64 years old (2.2%). More than half of the respondents were single (59.7%), 15.6% were engaged, and 23.14% were divorced.

Most MXD respondents were employed by the private sector (43.3%), students (26.1%), self-employed (10.5%), government servants (8.2%), NGO (2.9%), unemployed (8.2%), or unable to work (0.75%). Based on occupations, most residents earned less than RM 5000 monthly (40.3%) compared to 8.2% earning more than RM 10,000 each month. In terms of household structure, 33.3% of respondents live alone, 44.0% live with a family of 2–4 members, and 33.3% live alone. For private transportation, almost half of the respondents (48.5%) own a car, 11.19% own a motorcycle, and 36.5% do not have private transportation. Regarding property ownership, respondents were either sole owners (40.3%), tenants (41.8%), or sub-tenants (17.9%). The majority of respondents have stayed in their current housing for more than 2 years (63.4%), while the rest have stayed for less than 1 year (17.9%) or between 1 and 2 years (18.6%).

4.1.2. Descriptive Findings Based on Classes of Travel Frequency

Table 3. Comparisons of MXD respondents’ external and internal travels according to classes of travel frequency. 

Before COVID-19 (October 2019 to January 2020)				During COVID-19 (March 2020 to September 2020)
	External travels	Internal travels		External travels	Internal travels
Categories	N	N	p-value	N	N	p-value
Class 1	6	4	0.025	1	21	0.001
Class 2	20	42		7	22
Class 3	85	88		47	17
Total	111	134		55	57

presents the number of MXD respondents who engaged in external and internal travel before and during COVID-19 in the study area. Although no significant difference was observed between the frequency of internal and external travel before and during COVID-19, significant differences (p < 0.05) were detected when comparing the classes of travel frequency. In the Class 2 category, a significantly higher number of MXD respondents engaged in internal travel compared to external travel before and during COVID-19. Meanwhile, in the Class 3 category, a significantly higher number of MXD respondents engaged in external travel compared to internal travel.

4.1.3. Factors Associated with External Travel Frequency Before and During COVID-19

The final predictors of external travel frequency before and during COVID-19 are shown in

Table 4. Multinomial logistic model for factors associated with external travel frequency before COVID-19.

Variable (Reference Category)		Covariates	Class 1		Class 2	Class 3
			Coefficients	Exp (Coef)	Coefficients	Coefficients	Exp (Coef)	p-Value
Intercept			−0.022		0	0.11
Age		$X_2$	0.054	1.05	0	−0.005	0.99	0.032
Employment Status (Government Servant)	NGO	$X_{4b}$	−0.056	0.94	0	−0.004	0.99	0.010
	Private Sector	$X_{4c}$	−0.019	0.98	0	−0.027	0.97
	Self-Employed	$X_{4d}$	−0.19	0.98	0	−0.028	0.97
	Student	$X_{4e}$	0.016	1.01	0	−0.075	0.92
	Unable to work	$X_{4f}$	−0.028	0.97	0	−0.056	0.94
	Unemployed	$X_{4g}$	−0.089	0.91	0	−0.11	0.88
Income		$X_5$	0.091	1.09	0	0.073	1.07	0.046
Transport Ownership (No private transport)	1 motorcycle	$X_{7b}$	−0.009	0.99	0	−0.008	0.99	0.021
	1 car	$X_{7c}$	0.069	1.07	0	0.10	1.10
	A car and motorcycle	$X_{7d}$	0.013	1.01	0	0.06	1.06
	Multiple cars and motorcycles	$X_{7e}$	0.014	1.01	0	0.028	1.02
Occupancy Duration		$X_{10}$	0.013	1.01	0	−0.016	0.98	0.047

and

Table 5. Coefficients of Multinomial Logistic Model for Factors associated with External Travel Frequency During COVID-19.

Variable (Reference Category)		Covariates	Class 1		Class 2	Class 3
			Coefficients	Exp (Coef)	Coefficients	Coefficients	Exp (Coef)	p-Value
Intercept			−0.05952		0	0.058
Gender		$X_1$	0.006	1.00	0	0.01	1.013	0.020
Age		$X_2$	−0.01	0.98	0	0.02	1.02	0.001
Employment Status (Government Servant)	NGO	$X_{4b}$	0.049	1.05	0	0.02		0.039
	Private Sector	$X_{4c}$	−0.01	0.98	0	0.008
	Self-employed	$X_{4d}$	−0.03	0.96	0	0.066
	Student	$X_{4e}$	−0.01	0.98	0	0.021
	Unable to work	$X_{4f}$	−0.00	0.99	0	0.065
	Unemployed	$X_{4g}$	6.78 0.00	1.00	0	0.021	0.046
Income		$X_5$	−0.02	0.97	0	−0.010	0.98	0.034
Transport Ownership (No private transport)	1 motorcycle	$X_{7b}$	0.16	1.17	0	−0.09	0.91	0.002
	1 car	$X_{7c}$	0.04	1.04	0	−0.083	0.92
	Both a car and motorcycle	$X_{7d}$	0.07	1.07	0	−0.10	0.89
	Multiple cars and motorcycles	$X_{7e}$	0.10	1.10	0	0.029	1.03

. Before the COVID-19 pandemic, younger participants (p = 0.032) tended to travel externally more frequently compared to their older counterparts. In terms of employment status, students travelled externally the least amount (p = 0.010) compared to government servants, who travelled externally most frequently. Residents with higher incomes exhibited significantly higher variation (p = 0.046) in external travel frequency relative to those with lower incomes. Residents who own a car exhibited higher variation in their external travel frequency before the COVID-19 pandemic compared to those with other statuses of transport ownership (p = 0.021). Meanwhile, external travel frequency before COVID-19 decreased significantly as respondents resided in their current residence for a longer period (p = 0.047).

During COVID-19, male residents displayed a significantly higher variation (p = 0.020) in external travel frequency compared to female residents. Older residents also reflected a higher likelihood (p = 0.001) of travelling externally relative to younger residents. Among all employment statuses, MXD respondents who were unemployed and those employed as government servants travelled externally the least amount, relative to self-employed residents (p = 0.039). Residents with a higher income were more likely to fall in Class 2 (p = 0.034), implying that these respondents have a higher variation in their travel frequency. Meanwhile, residents with a motorcycle travelled externally the least compared to those with multiple cars and motorcycles (p = 0.002).

4.1.4. Factors Associated with Internal Travel Frequency Before and During COVID-19

MXD residents who were employed in the private sector were most likely to travel internally less (p = 0.038), while students were the most likely to travel internally more before the COVID-19 pandemic (p = 0.002) (

Table 6. Coefficients of multinomial logistic model for internal travel frequency before and during the COVID-19 Pandemic.

			Before COVID-19
Variable (Reference Category)		Covariates	Class 1		Class 2	Class 3
			Coefficients	Exp (Coef)	Coefficients	Coefficients	Exp (Coef)	p-Value
Intercept			−0.04079		0	0.11
Employment Status (Government Servant)	NGO	$X_{4b}$	0.17	1.19	0	0.17	1.19	0.038
	Private Sector	$X_{4c}$	0.33	1.40	0	0.06	1.06
	Self-employed	$X_{4d}$	0.01	1.01	0	0.07	1.07
	Student	$X_{4e}$	0.29	1.33	0	0.19	1.21
	Unable to work	$X_{4f}$	0.20	1.23	0	0.07	1.07
	Unemployed	$X_{4g}$	0.05	1.05	0	0.01	1.01
Income		$X_5$	−0.05	0.94	0	0.05	1.06	0.002
During COVID-19
Intercept			−0.004		0	−0.01
Parking lot		X8	0.01	1.01	0	−0.00	0.99	0.021

). Meanwhile, residents with a higher income tended (p = 0.002) to travel internally more frequently compared to those with a lower income. During COVID-19, residents with more parking lots allocated for their household tend to travel less internally compared to those with fewer parking lots allocated (p = 0.021).

4.2. Qualitative Phase Findings

4.2.1. Descriptive Results

A total of 52 MXD residents were interviewed, comprising 28 females and 24 males. Their residential locations were mainly Kuala Lumpur and Petaling Jaya districts. Most residents lived in condominiums (n = 35) compared to service apartments (n = 15) and terrace buildings (n = 2). In terms of the type of trips frequently practised, 36 participants engaged more in external trips relative to 16 who preferred internal trips. Most respondents have been MXD residents for more than 1 year, with a mean duration of 2.2 (0.34) years in the studied population.

4.2.2. Thematic Analysis Findings

The main findings from the thematic analyses are presented under 3 main categories: perspectives on the benefits of internal trips, barriers to internal trips and strategies to reduce external trips.

Perspectives on the benefits of internal trips

Cost-saving and time-saving

A predominant finding from the qualitative survey regarding the benefits of internal trips was the cost-effectiveness and time savings. This important perspective was based on the easy access to grocery stores and restaurants. The relevant comments are presented below:

“Save time to travel and jam just to buy some groceries or necessities, and even dining”

(MXD resident 5)

I believe internalisation of trips is very important, especially in today’s fast-paced lifestyle. Having daily necessities like groceries, food, and beverage options within the residential area makes life much more convenient”

(MXD Resident 25)

Convenience and comfort

A few participants opined that internal trips simplify access to basic amenities, thereby helping to reduce the cost associated with external trips. However, the MXD residents also acknowledged the discomfort experienced when relying solely on internal trips. Examples include the possibility of overcrowding when most residents use the available MXD services such as grocery stores, eateries and parking lots. This theme is reflected in the following responses:

“It will always come with pros & cons. In my opinion, condos full of shopping & dining facilities will be convenient sometimes, but it will affect the comfort of staying in the above properties”

(MXD 18)

Health reasons

The positive impacts of internal trips on health were conveyed by several MXD participants. Given the availability of walkways, gyms and recreational centres in MXD areas, the facilities were perceived to offer opportunities to improve health and well-being. The effects of using such infrastructures in relieving stress and psychological health were equally highlighted by MXD residents.

“Internal trips mainly for health since there are swimming pools and gyms. Also able to jog around the condominium gated compound.”

(MXD 13)

“I prefer using the facilities around here because they help to release stress, which is definitely good for my health”

(MXD 28)

“Reduce travelling time and minimise the stress associated with traffic congestion

(MXD 32)

Social interaction

Participants had different views regarding how internal trips affect social lifestyle. While some respondents felt internal trips are crucial for strengthening family bonds and spending quality time with family members, others believed that they have a negative impact on social life. The following comments reflect this dichotomous finding:

“It’s not my preference as I still prefer heading outside of my residential area to maintain a healthy social lifestyle.”

(MXD 6)

“Maximise quality time with family members as more time could be spent on them”

(MXD 50)

“Internal trips are important for strengthening team bonding, enhancing communication, and fostering a sense of belonging”

(MXD 21)

“For social interactions, saving money by buying groceries, a breath of fresh air”

(MXD 39)

Green environment

Internal trips were perceived to facilitate the establishment of a green environment, particularly by lowering the risk of carbon footprint and emissions following a reduction in travel, traffic congestion and wastage. By fostering a green environment, some respondents posited that such events promote a sustainable lifestyle and long-term positive impact on the economy.

“Reduce carbon footprint (petrol, jams), supports the immediate local economy, reduces wastage (resources, time, accidents, etc.)”

(MXD 10)

“It promotes a more sustainable lifestyle by minimising the need for transportation, which can reduce traffic congestion and environmental impact.”

(MXD 22)

“I think internalisation of trips is essential, especially in condominiums, as it is able to reduce carbon emissions by residents to travel from their house to shopping malls or supermarkets for shopping.”

(MXD 48)

Barriers to Internal Trips

Four main themes were identified as barriers to trip internalisation, namely, (l) Lack of infrastructure, (2) Poor Management, (3) Lifestyle activities/individual factors, and (3) Environmental factors. Each of the theme are presented below.

Lack of infrastructure for active transport

Most MXD residents highlighted that the lack of infrastructure and facilities for active transport was the predominant barrier to internal trips. Despite being an MXD resident, several participants posited that the facilities were inadequate and insufficient to reduce the frequency of external trips. Typical examples include the absence of restaurants/cafes and healthcare centres. This finding is shown in the following responses.

“A residential area that has no development and bad management. Then there’s no reason for the occurrence of internal trips as no options are available internally”

(MXD 8)

Lack of options. No cafes/restaurants, only marts. Marts have limited options as well

(MXD 18)

“No such option available at my current residential area”

(MXD 32)

“The establishments in the MXD do not constitute to functional ecosystem”

(MXD 45)

Poor Management

Poor management of existing facilities and infrastructure at MXD areas was conveyed as a strong barrier to internal trips. These respondents affirmed that infrastructures to encourage internal trips are present in their respective areas, and poor management constitutes the main reason why they are sparingly utilised.

“I think the MXD areas are already designed and well-developed to enhance internal trips, but they are not managed properly”

(MXD 20)

“In order to improve internal trips, it goes beyond building condominiums and creating MXD areas; the management of these facilities is key for sustainable usage”

(MXD 41)

“There is a gym in my residential area, but I don’t use it because the place is always crowded and ill-managed”

(MXD 51)

Lifestyle and Individual Factors

Several MXD residents opined that personal lifestyle and individual factors were primarily responsible for the increasing frequency of external trips. Diverse events were used to convey this theme, ranging from time constraints to lack of interest, laziness, and poor communication. While some participants felt these factors played a significant role in increasing the rate of external trips, a few MXD residents perceived they would be bored with using the same facilities or performing repetitive activities.

“I feel we are just being lazy in using MXD services available at our disposal. Most people prefer driving all the time, even to the nearest mall or grocery store”

(MXD 4)

“People don’t just have the interest to reduce external trips; maybe they are unaware of its effects on well-being. However, sometimes it boils down to poor communication”.

(MXD 27)

“For me, I go outside because I would be sick of the repetition of activities”

(MXD 31)

Environmental factors

Some MXD residents posited that environmental factors such as overcrowding, poor design of shared spaces and noise shaped their decisions to engage in external trips rather than internal trips. These respondents affirmed that despite living in an MXD area, they prefer to travel externally to a more serene environment. Examples of the corresponding comments are as follows:

“I don’t like using the facility in my residential MXD area because it’s too noisy and poorly designed and currently used by so many people”

(MXD 21)

“Noise, overcrowding, or poor design of shared spaces can also reduce the appeal of internal trips”

(MXD43)

Strategies to Reduce External Trips

As for strategies to reduce external trips, four main sub-themes were synthesised from the thematic analysis: diversified services and accessibility, inclusiveness in planning activities, promoting social interaction, and work-from-home policies.

Diversified services and accessibility

Most residents highlighted that more diverse services need to be provided within the MXD environment to encourage trip internalisation. Examples of the services include essential spots such as markets, clinics, pharmacies, grocery and convenience stores, as well as mobile shopping and delivery services. Apart from expanding these services, residents also advocated for such services to be easily accessible.

“Good management, good development, more options. More essential spots, especially markets, clinics, barbershop surrounding the area”

(MXD 8)

“More online shopping & efficient delivery service”

(MXD 15)

“Developers can include more essential facilities within residential areas, such as grocery stores, convenience stores (like 7-Eleven), local eateries, pharmacies, and laundromats”

(MXD 36)

“Encouraging mobile services like food delivery, parcel lockers, and online grocery platforms (e.g., GrabMart, HappyFresh) is another effective way”

(MXD 42)

Organised layout and proper city planning

Another strategy recommended by MXD residents is to improve the layout in MXD areas, ranging from transportation options to last-mile connectivity and efforts to reduce reliance on personal vehicles.

“Strengthening public transport and last-mile connectivity options, such as feeder buses or e-scooters, can help reduce reliance on personal vehicles when external travel is unavoidable”

(MXD 6)

“The layout of MXD areas needs to be improved. Population density in these areas surpasses the available structures and facilities, so we need to balance the density. Completeness of amenities is also crucial”.

(MXD 44)

Inclusiveness in planning activities

Some respondents opined that MXD residents need to be taken along in the management, planning and designing of MXD areas. These respondents perceived that developers and government agencies usually neglect residents’ views and specific needs when making decisions relating to MXD management. In moving forward and reducing external trip frequency, primary users’ perspectives and needs were described as fundamental information to guide the identification and implementation of effective measures.

“The developers plan the MXD environment and facilities to their taste rather than considering the residents’ needs”

(MXD 24)

“We are not carried along in the planning and implementation of various policies relating to MXD development, no way to reduce external trips if you don’t understand the main problem faced by residents”

(MXD 37)

Promoting social interaction

A few respondents recommended hosting engaging events to promote social interaction and motivate residents to prioritise internal trips over travelling externally. These respondents perceived social engagement as a key reason why people travel externally, as such events are lacking in their immediate MXD environment.

“Reducing external trips can be done by hosting engaging events and forming interest groups”

(MXD 10)

“Organising community events such as weekend markets, workshops, or wellness activities can promote social interaction and engagement within the development itself”

(MXD 18)

Work-from-home policies

Given the availability of digital technologies and their potential in facilitating remote working modalities, as seen during COVID-19, some respondents perceived that work-from-home policies should constitute one measure of reducing external trips among MXD residents.

“Encourage flexible work-from-home policies”

(MXD 21)

“With the advent of digital technologies, businesses can implement or expand remote work opportunities to reduce the need for employees to commute daily. This could significantly lower the number of trips taken, particularly during peak hours”

(MXD 39)

5. Discussion

This mixed-method study was conducted to identify and understand key reasons that might lead an MXD resident to either travel internally or externally, which is essential in reducing vehicular traffic caused by these dual travel modes. We also looked into how key policy changes during various phases of COVID-19 might have shaped travel behaviours and usage of MXD infrastructures in the studied population. For instance, policies ranging from partial to strict lockdown had a significant impact on travel frequency and purpose in countries such as Japan and China [2,8], highlighting the need for a multi-dimensional analysis of travel patterns rather than a single-dimensional approach.

The situation in Malaysia offers an opportunity to understand how policy changes impact travel behaviours among residents of MXD properties since the MCO transitioned from partial lockdown to full lockdown before partial and complete reopening. By identifying the purpose of travel, residents of MXD properties can be better accommodated, thereby facilitating a further reduction in traffic by decreasing the frequency of external travel.

In line with studies on the impact of COVID-19 on travel behaviours and mixed land use [31,62], the present findings revealed a reduction in the travelling demand and frequency, either going out from the MXD premise (external trips with motor vehicles) or walking activities within the MXD premise (internal trips) before and during the COVID-19 pandemic. This finding might be valid at the beginning of the MCO relaxation period and may change as the pandemic situation improves. Nevertheless, the current situation might be influenced by the trend of working from home (WFH) and flexible working hours, with the advancement of online facilities. During the MCOs in Malaysia, strict SOPs were implemented to minimise the travelling frequency, and people began to rely on online meetings, e-commerce, e-hailing, and delivery services, which explains the reduction in travel demand.

Regarding the main purposes for travelling, leisure and recreation were MXD residents’ primary reasons, followed by working and visiting friends and relatives. These travel purposes were similar before the COVID-19 pandemic and the current situation. The main factor for MXD residents to stay in their residences before COVID-19 for their daily activities was the availability of stores where they could obtain their groceries within walking distance, without travelling by motor vehicle, and the presence of entertainment and recreation at MXD premises. Some of the qualitative study results also supported this finding, as the availability of service stores was considered cost-saving and time-saving. Moreover, interviewed participants also acknowledged the importance of diversified and accessible services, whereby the lack of such facilities in MXD areas led to a higher frequency of external trips. These findings correlate with reports from studies conducted in China during the post-COVID-19 era, whereby the location of socioeconomic resources and opportunities influenced the spatial distribution of travel demand [31,62].

Qualitative findings also indicated MXD residents’ perspectives regarding the benefits of internal trips, which encompass the impact on health and well-being, promoting a green environment and social interaction/lifestyle. The perceived health benefits of internal trips could stem from the positive relationship between physical activity and mental well-being, and how this association might be shaped by residents’ green travel awareness [63,64]. Such perceived benefits differed during the various stages of COVID-19, particularly the protective role of leisure-time physical activity against worsened mental health [65]. Pandemic-induced changes in travel had both negative and positive effects regarding health and well-being [66]. A similar finding was observed in the present study, as qualitative results reflected diverse perceptions about the effect of internal trips on MXD residents’ physical and psychological health. Increased knowledge of green travel awareness may contribute to residents’ perceived benefits of less external travel on the green environment [67].

In the present day, the main reasons for staying at home are consistent with the scenario before COVID-19. The most noticeable difference is the massive increase in MXD residents who prefer WFH. This result was also highlighted in the qualitative study, as residents recommended WFH policies as a strategy to encourage internal travel. A similar finding was reported by Huang et al. [68] and Rafiq et al. [69] as WFH groups/residents recorded significant reductions in trip distance, travel time and travel frequency following the COVID-19 pandemic, thereby mitigating negative transport externalities. Identifying the reasons and purposes of external trips is an important step in reducing traffic through the development of MXD and its facilities [62]. In addition, by identifying the timeframe where individuals perform external trips, the time of a potential surge in traffic due to increased travel can be effectively forecasted. As a result, the facilities that would accommodate the accomplishment of trip purposes with internal trips instead of external trips can be developed.

The logistic regression analysis reflected that before COVID-19, younger MXD residents tended to travel externally more frequently. Meanwhile, in the present situation, younger MXD residents tend to travel less while older MXD residents travel the most. According to the employment status, students were the least likely to travel externally, while government staff were the most frequent. The higher frequencies of travel among younger MXD residents may stem from the increased share of car-driver commuters among the young population that has risen substantially in several low- and middle-income countries, including Malaysia [70,71]. However, during the pandemic, apart from students accounting for the majority of the young population, policies requiring studying from home and WFH may have reduced the volume of external trips.

MXD residents with higher incomes exhibited higher variation in external travel frequency compared to MXD residents with lower incomes. MXD residents with lower income are likely to travel more, which may stem from the nature of their occupation, which is skill-based and necessitates reporting on duty at the workplace physically [31]. Likewise, MXD residents with more transport ownership tend to travel more regardless of the COVID-19 pandemic.

These effects of socioeconomic variables on external trips are consistent with prior literature findings [1,31,62], with income demonstrating a significant effect on private vehicle miles. MXD residents with higher incomes and those who are wealthier and more educated, alongside children belonging to younger households, have higher odds of engaging in more complex tours and external trips [72,73]. On the other hand, MXD residents who stay at their current residence for longer residential periods tend to decrease their external travelling with motor vehicles. In general, longer residential periods with proper entertainment and recreational facilities were the factors associated with less external travel before the COVID-19 pandemic. These results can be explained by the lower demand and need for external trips given the availability of socioeconomic resources in the MXD facilities, similar to the reports in studies conducted in China during the pandemic [31,62].

Among all employment statuses, MXD residents who are unemployed and government staff travel less, compared to MXD residents who are self-employed. As for government staff, a reduction in the amount of travel can be linked to the proximity to their place of work and a higher tendency to work in more urban and accessible environments. Moreover, many government offices implemented the WFH policies, and the travel demand has reduced significantly [74]. These events are either lacking or not readily applicable to self-employed MXD residents, thereby increasing the odds of more external trips either before or during COVID-19.

As for pre-pandemic internal travels, MXD residents who were employed in the private sector and those with lower incomes travelled less internally compared to students and those with higher incomes. These outcomes might be shaped by the level of concentration of jobs in the service sector in central and accessible areas, which are more aligned with skilled, educated and higher-income earners [75]. These socioeconomic characteristics reduce commuting distances while encouraging trip internalisation. For the current trend of internal trips, MXD residents with more parking lots allocated for their household tend to travel less internally compared to those with fewer parking lots allocated. Reducing parking lot ownership might be one of the methods to promote internal trips, as evidenced in prior studies [73,76,77].

This study has important limitations that need to be highlighted. The biggest challenge of this research was the interruption caused by the COVID-19 pandemic. Given the movement restrictions implemented by the MOH, a robust data approach, such as traffic surveys and face-to-face interviews, was not feasible. These limitations eventually affected the final sample size and response rates. Thus, the findings are not generalisable to the population of MXD residents in Klang Valley, as respondents were only recruited from 4 primary sites in the region. Nevertheless, the use of a mixed-method design provided a broader perspective of the survey results, and reduced the limitation associated with a relatively small sample size since the qualitative phase focused on data saturation. Based on the cross-sectional approach, the results are restricted to the research time span, as no follow-up observations were performed. Although the data provided in this study may not predict future changes in travel behaviour among MXD residents, it provides important insights regarding the impacts of global crises like COVID-19 on travel patterns and implications for planning transportation strategies.

Theoretical and Policy Implications

This study is the first attempt to use the model global structure in analysing mixed land use and factors influencing internal and external trips among MXD residents in Malaysia. The research findings have important theoretical implications. First, we have demonstrated that the model global structure can be used to develop a predictive model for understanding how MXD residents make informed decisions to engage in internal and external trips during an unexpected disaster such as COVID-19. Both trip patterns were successfully incorporated into the framework while accounting for factors influencing decisions relating to commuting distance, location patterns, and vehicle ownership, and daily distances travelled [20]. In addition, by recruiting MXD residents, this study reflects the suitability of the model in understanding how socioeconomic factors influence individual commuters’ external and internal trips.

A snapshot of MXD residents’ travel patterns in the Klang Valley area of Malaysia before and during COVID-19 was presented in this study. The survey results indicated how MXD residents altered their travel patterns, and possibly, the pandemic’s influence on their travelling behaviour. We found that MXD travel frequency reduced after the relaxation of the MCO in Malaysia. Private vehicles became popular during the pandemic, similar to the rebound in the modal share of cars in other countries [46]. Essential travel purposes appear to contribute to travel demand, whereas non-essential travel purposes declined significantly. MXD residents prioritised visiting resources and facilities that provided the most vital services, which were readily available at their premises. This led to a significant decline in external trips while encouraging trip internalisation. Policymakers and researchers need to pay attention to this directional change among MXD residents and the impacts on external and internal trips.

The qualitative findings, particularly the recommended strategies to enhance internal travel and mitigate negative transport externalities, have important implications for policymakers. For instance, policymakers may look into ensuring the availability and accessibility of diversified services, regular feedback from residents and their inclusiveness in planning activities, efforts to promote social interaction, and fostering WFH policies when necessary.

The pandemic might also have altered long-term behaviours among MXD residents, affecting the perceived importance of proximity to transit when selecting their residential locations [53]. Proximity to the workplace may no longer be considered since remote working has now become the norm. Although we did not explore the trend of suburbanization due to the pandemic, the significant socioeconomic factors identified in this study suggest behavioural changes among MXD residents. Policymakers ought to take advantage of work-from-home provisions established during COVID-19 and demographic factors that shaped MXD residents’ trip internalisation.

6. Conclusions

This mixed-method study provides evidence-based data on the predictors of MXD usage and travel patterns in a core metropolitan area in Klang Valley, Malaysia, and how these events were shaped by the COVID-19 pandemic. The rate of external trips decreased (i.e., more internal trips) significantly during the mid-phase of COVID-19 compared to the pre-pandemic period, reflecting changes in travel patterns among MXD residents at various phases of the global crisis. These alterations in travel patterns were influenced by several demographic factors, such as age, employment status, income levels, vehicle ownership, number of parking lots, and duration spent living in MXD areas. More insight into these events was gleaned from the qualitative interviews conducted during the pandemic, whereby cost-saving, convenience and comfort, social lifestyle, health and well-being, and green environment constituted MXD residents’ perceived benefits of trip internalisation. Potential strategies to reduce external trips were to ensure diversified services and accessibility, inclusiveness in planning activities, promote social interaction, and work-from-home policies. These findings reflect the strategies that could be incorporated to reduce external trips and enhance effective traffic management. Future studies may consider gathering longitudinal data during and after a global crisis for a comprehensive analysis of the impact on travel behaviour and MXD projects.