Undertaking a Household Travel Survey (HTS) in a pandemic: A Smart Approach

1. Introduction

As part of its broader work looking at the improvement of mobility, the Global Future Cities Programme (GFCP) in Malaysia has undertaken a month-long Household Travel Survey (HTS) Campaign in March 2021 to collect daily household travel data within Iskandar Malaysia and Melaka. Iskandar Malaysia is located at the southern tip of Peninsular Malaysia and on the border with Singapore, while Melaka is the state that is centred on the UNESCO World Heritage City and historic trading hub of the same name.

These data will enable the development of transport models that will be used to analyse travel demands in both these regions. The data gathered will include individual trips, preferred modes of transportation, trip lengths and other mobility information. They will inform improvements to mobility by supporting the development of GFCP interventions - Iskandar Malaysia’s Smart Integrated Mobility Management System (SIMMS) and Melaka’s Green Transport Masterplan.

2. About the survey

The survey was conducted using the PerjalananKu^[1] Mobile Application specifically developed for the project in partnership with UK’s behaviour change experts, BetterPoints^[2]. The participants who sign up for the survey download the app from the Google or Apple stores and then register their details. The app makes use of the participants’ own smartphone technology to monitor their journeys and prompts them to record both the purpose and mode used.

The app was developed to enable some of the challenges of social distancing and remote working to be overcome by reducing personal contact between interview teams and participants. Ideally this would be a person-less interaction with social media channels being used to sign up the participants. However, this proved more difficult and is discussed under the lessons learnt. Nevertheless, we consider that this use of technology and the overall approach can be used for other activities and sectors where traditionally, great dependence has been made on face to face methods. Furthermore, the use of smart technology enables the app to communicate with participants at any time by sending in-app prompts and enables automatic information capture so that participants can utilise the application in a hassle-free manner.

Figure 1 – User interface of PerjalananKu

Figure 2 – Some of the steps, including registration, location settings, and trip confirmation

3. Planning the survey

Proper planning of a survey is fundamental to delivering successful outcomes. Since this survey would require introducing innovative techniques, a focus group was formed to test the survey. A group of eight people from a range of backgrounds in both cities was established. Using interactive sessions, the focus group identified that some of the introductory app messaging was too long and may put off participants especially those from key stakeholder groups, such as the B40^[3] and the elderly who may not be technology proficient.

After incorporating the focus group’s comments, a pilot version of PerjalananKu was made available. This was intended to target 100 households in each city however slow sign up rates due to restricted access meant that only 64 individuals responded.

The pilot was used to trial and test the app with typical local participants to identify problems while using it, to ensure that it would record their journeys and could communicate with them. This trial was carried out over a period of a week and included participants from the GFCP in-country team, city partners as well as their friends and family. The feedback enabled the app to be modified to accommodate local requirements and terminology. Furthermore, it has enabled real data to be shared with stakeholders and potential participants to show how data are collected and how they would support the GFCP in developing the overall projects.

4. Data privacy

The concept of data privacy lies at the heart of the survey and data are collected strictly under the auspices of the Global Data Protection Regulations (GDPR), as BetterPoints is a UK-based company, and the Personal Data Protection Act (PDPA) in Malaysia.

The process for keeping participants data secure is based on creating a series of basic steps in the storing of data. During data collection all personal data are stored separately from the recorded travel data. Then, the data are further anonymised by collating the outputs into cells, these are areas to be used for transport modelling such as housing estates, hospitals, shopping areas and so on. This latter process breaks the connection with exact start and end points of journeys further anonymising the data.

5. Creating and incentivising participation

A major challenge was to sign up as many households as possible. Using our city partners and their channels provided one means and this created initial engagement and provided reference material as the survey appeared in the local media and on the Malaysia National News Agency’s website. However, after an initial flourish of several hundred participants in each city the need for greater on the ground activity was identified. Figure 3 shows the completion of the initial survey sign up survey reflecting when people downloaded and interacted with the app. Each of the major peaks occurred when the GFCP was able to visit the city, showing the continuing need for direct engagement.

Figure 3 - Dates when participants first used the app

1. • Interactive engagement using technology

The app features an engaging user interface through its use of a cool colour palette. It also aims to have a simple user experience to allow easy navigation for people of all ages.

Crucial for engagement, the app provides a simple summary of the environmental and health impacts of their trips. For example, if a participant chooses to walk instead of drive then the app generates information on their resultant carbon footprint as well as the number of calories burnt during their trip, as shown in Figure 4. This provides thought-provoking information to enable participants to make decisions and changes to their travel patterns to live a healthier lifestyle and benefit the environment.

Figure 4 – App feedback on environmental and health impact of each trip

1. • Providing an incentive

Since doing surveys are intrusive and requires both the participants’ goodwill and continued effort, it was viewed that a gesture should be made to thank them for the participation. Furthermore, it raised the opportunity to create an incentive to encourage greater participation and therefore, a headline prize of RM1,000 for each city was chosen to increase sign up rates. Additionally, in each city smaller prizes of RM100 were drawn for 25 lucky winners and for the first 1,000 participants in each city (early birds) a reward of RM20 vouchers was given.

Read more about the main prize winners in Appendix A - Our lucky winners which talks about their daily travel habits and how they interacted with the app.

1. • Support of GFCP partners and other stakeholders

The vast majority of the community groups approached were very open and welcomed the initiative, as it would benefit the people of Iskandar Malaysia and Melaka.

Clearly, the GFCP’s day-to-day working relationship with City Partners, Iskandar Regional Development Authority (IRDA), Unit Perancang Ekonomi (UPEN) Melaka and Perbadanan Teknologi Hijau Melaka (PTHM) was important in increasing both the publicity of the survey and increasing the number of participants.

Figure 5 – Press Conference held in partnership with the Melaka State Government for launch of PerjalananKu

The City Partners also provided access through the Gender Equality and Social Inclusion (GESI) groups they had helped set up as part of the programme as well as to other government and local agencies working with these groups.

1. • Leaving No One Behind

Running a survey requires that data are collected from as wide a range of stakeholders as possible and most importantly from the marginalised groups and communities. Accordingly, the GFCP team actively engaged members of the Gender Equality and Social Inclusion (GESI) Consultative Group (set up earlier to support the programme) and government departments to identify areas that could be targeted to identify participants. This work required numerous site visits, and these were conducted and used to engage these communities. Acknowledging that smartphone use may be lower in these groups, surveys were also undertaken to capture travel perceptions of these groups.

In Iskandar Malaysia, outreach activities (with strict COVID-19 procedures in place) reached out to over twenty Program Perumahan Rakyat (PPR), low-cost housing areas for the B40 group, Orang Asli (indigenous people) villages, the Cerebral Palsy Association of Johor and a homeless community. The locations were tracked in the field using a survey directly uploading to the GIS platform to record the details of each engagement. The locations visited were spread widely throughout the region with the sites in the city centre focussed on B40 and migrant communities, while those outside the city centre included a remote peri urban B40 community in Ulu Tiram and an industrial zone in Senai.

Figure 6 – Map of community engagement locations covered in Iskandar Malaysia

In Melaka, the community outreach took place at three PPR locations, ethnic minority villages such as the Portuguese and Chetti Villages, as well as Pusat Pemulihan Orang Kurang Upaya (PPOKU) Melaka, a centre supporting youth with disabilities. Remote industrial zones where B40 staff were likely to be employed in Jasin were covered as well, along with locations far from the city centre such as Sungai Putat, Paya Rumput and Durian Tunggal. The drive to ensure maximum inclusion in the survey by low-income communities was included reflecting their reliance on public transport for necessities, such as health care and education.

Figure 7 - Map of community engagement locations covered in Melaka

6. Initial Survey Findings

To obtain a statistically meaningful outcome from such a survey requires considerable input. The initial target was to sign up at least 2,000 full and complete households (1,000 in each city) for the survey to enable sufficient data collection for the development of an effective transport model. However, the survey concluded with 1,998 individual responses noting that this is below the targeted 1,000 complete households per city but still providing some base data for modelling.

1. • Using a dashboard during the main survey

To assist the team in increasing participation rates, a dashboard was included in the BetterPoints platform to monitor various aspects of the data being received by the app. These included daily sign-up rates by participants, basic gender and age demographics of participants, completion rates of the sign-up survey, journeys by mode of transport, and heatmaps of the modes and routes taken. These dashboards provided feedback to the team on potential modes, demographics and geographic areas that needed to be targeted.

Figure 8 - Participant totals by day

1. • Building a Demand Matrix using PerjalananKu Data

The survey data were analysed and key statistics such as trip length averages were derived. A more detailed explanation of how the origin and destination models were created will be shared in our future article in this series on how the HTS has been used in transport modelling. These data were combined with other data such as a trip generation manual to determine the total volume of trips per day.

Based on the behavioural patterns observed upon data analysis of PerjalananKu results and zonal-planning data, a gravity-distribution was developed to distribute trips or journeys to attraction zones by trip purpose and trip mode. HTS-derived factors were used to convert into time-period specific matrices for each trip mode.

With the demand matrix built, a transport model is being developed that will enable visualisation of data such as traffic flow, queue length forecasts and congestion hotspots. This working model will be clear enough to allow for continuous updates as and when new data becomes available.

Additionally, the data can also be included into the GIS^[4] to review mobility trends, speeds along links and overlays them with other data such as land use, building locations and demographic data.

1. • Travel insights

By its very nature the Household Travel Survey is based on direct input from its participants. This enables the team to gather greater knowledge on the participant’s journeys, confirming the mode, the distance travelled as well as understanding basic details of the participant’s demographics. These data can be extracted and plotted to show broader travelling habits exhibited in the city and potentially how they vary for different groups. We have created some initial insights for this report, noting that these will be expanded upon in our future article in this series on a detailed analysis of insights and statistics of travel habits from the survey.

Figure 9 shows how the mode of transport varies as the length of trip increases. Not surprisingly, this shows that for very short journeys of less than 500m, walking is dominant. However above 500m, these journeys are increasingly made by motorised vehicle and these represent over 90% of journeys when the distance reaches 3km.

Figure 9 - Changing modes of transport as distance increases

To encourage more active mobility to reach public transport, or for more general shorter journeys, then understanding how far people typically walk is a useful piece of evidence. Figure 10 has plotted out from the HTS the number of walking trips by distance. As expected, this is dominated by short trips of less than 400m but does have a significant number of journeys of up to 1.5km.

Figure 10 - Number of walking trips by distance

To further understand the distribution of how far people would walk we wanted to find out if it was multiple journeys by the same participant or if these journey numbers were generated by multiple participants. Figure 11  was developed to show both the length of trip and the number of participants. It reveals that there around a third of the participants who will walk up to 800m.

Figure 11 - Number of walking trips and participants by distance

These basic insights have started to inform us on how people travel around the cities. They provide insights to enable us to look at how to overcome some of the barriers to mobility that may inhibit the uptake of public transport.

7. Lessons learnt from running a Household Travel Survey

The lessons learnt focus on the key learnings from undertaking the survey in a pandemic and the impact that had on creating sufficient engagement and alleviating participants concerns. The technical lessons learnt will be shared in our future articles in this series.

1. • City authorities and stakeholder engagement

The city authorities will already have significant networks with the wider community and is, therefore, the fundamental channel for engagement. However, within any democracy there will always be those who object to state-led proposals or just do not have an interest in the state’s usual communication channels, especially the socially excluded. The views of these groups need to be sought by using wider methods of engagement such as targeted site visits, interest groups on social media and other more direct methods.

1. • Sign up and skepticism

One of the major hurdles to overcome from the use of an app-based product in Malaysia was found to be participant skepticism. Scarcely a day goes by without news of some online scam and whilst the younger generation were eager participants those from older generations proved much more cynical. Many expressed their concerns about data security and, although fully registered under the Personal Data Protection Act (PDPA), were hard to convince. However much of this was overcome by using credible forums such as public launches and formal gatherings to publicise the app. Inherently these tend to be populated by relatively narrow groups from similar backgrounds, hence the need to create and use as wide a stakeholder base as possible to access multiple communities although in a pandemic this is still an immense challenge.

1. • Language interface

The interface of PerjalananKu was planned and conceived in Bahasa Malaysia and English, however, it was found that many of the potential participants used vernacular languages, particularly Chinese dialects. This may have been a genuine excuse or a convenient means of avoiding the survey resulting from scepticism, see above. However, to mitigate this risk, it is recommended that as wide a range of known written language in use in the survey areas are built into such apps.

1. • Technology barrier

As a smart-technology tool, such apps will only run on a smartphone with their need for data. This is a barrier to participation in the survey for poorer communities. There still exists a perception that everyone has a smartphone, and this is not the case. Statista.com^[5] has estimated that there would be around 20 million smartphones in Malaysia in 2020, this is just over 60% of the population. Whilst we can discount very young children, this does leave many groups without such access to technology and leaves them unable to join these surveys.

As reflected previously, this leaves more traditional survey forms best suited to obtain household travel surveys from these groups. They are highly challenging since it means the deployment of personnel to undertake these interviews while maintaining strict adherence to physical distancing and other standard operating procedures.

8. What’s Next?

Through this survey, data was gathered on individual trips, preferred modes of transportation, trip lengths and purposes, and other mobility as well as demographic information, which will be collated to provide robust data for better-informed and data-driven transport models.

The data gathered will help support long-term improvements to transport and mobility networks in Iskandar Malaysia and Melaka through incorporation into traffic modelling and understanding how people in these cities travel at present. The opportunity exists to demonstrate the survey approach to encourage wider adoption and to support the development of transport models. These are identified in the 11th Malaysia Plan, and the HTS will be able to provide evidence to achieve the modal changes required by Malaysia to meet its climate obligations. PerjalananKu provides a basis for its development and use in other cities to support their transport master planning as they plan for inevitable growth in the population and reduced dependence on fossil fuels. PerjalananKu would act as a fast, efficient, cost-effective and safe data collection tool for a vast array of other industries and cities.

As the first-of-its-kind Household Travel Survey conducted in Malaysia, this data collection tool can be utilised for future surveys with minor changes to fit the survey requirements, reducing planning, design, and development costs.

The summary and findings of the survey will be published in a series of informative articles on the GFCP Knowledge Platform^[6], including the experiences engaging with the PerjalananKu participants, transport modelling analysis and derived statistics, all of which will inform a clearer understanding on improving the capacity for Urban and Transport Planning. These articles will act as important references on mobility patterns and their corresponding environmental impacts and can then be utilised by cities and organisations to make data-driven decisions and positively affect how people travel.

9. Appendix - Our lucky winners

   • Iskandar Malaysia

Fatimah Binti Othman, a retiree from the private sector, was drawn as the lucky Grand Prize winner for Iskandar Malaysia. She was tempted to download PerjalananKu after her daughter saw the poster on the Iskandar Malaysia Facebook page. She found the application interactive and easy to understand with straightforward questions.

She hopes to see improvements in the transport and mobility sector, noting that it would be good if public buses could display routes and locations in a clearer manner while ensuring real-time updates.

“Although I am usually a vehicle passenger and often accompany my daughter for deliveries and errands, being in the car a lot, I see first-hand the current shortcomings on roads. I was excited to hear about the electric bus that was recently launched in Iskandar Puteri. Hopefully the service will be extended to other areas as well for the benefit of the entire Johor”, said Fatimah. She went on to say that with the vouchers she has won, she will purchase some electrical appliances to replace her over 20-year old kitchen equipment.

Iskandar Malaysia PerjalananKu Grand Prize Winner, Fatimah Binti Othman

1. • Melaka

The Grand Prize winner from Melaka was 23-year old Dental Clinic Assistant Noraini Alias. She downloaded PerjalananKu after reading about it from the news portal Melaka Hari Ini. She is grateful that she was given an opportunity to contribute towards a better and more sustainable transport system. She found the participation hassle-free via the mobile application.

As someone who commutes at least twice daily, she looks forward to seeing improvements on the existing public infrastructure. She hopes the GFCP interventions will lead to smoother and shorter rides on public transport. She also wishes traffic congestion can be reduced on the streets of Melaka.

On winning vouchers worth RM1,000, she said that she had never expected to win but merely wanted to offer her contribution as a responsible citizen. “The win was a nice bonus, though. I will probably use the vouchers to purchase items for minor home upgrades”, said Noraini.

Melaka PerjalananKu Grand Prize Winner, Noraini Alias

*Disclaimer: All individuals photographed have given consent to having their information and photograph(s) taken and shared via GFCP Knowledge Platform, social media and other marketing tools