Issues, Challenges and Strategies in Obtaining Reliable and Quality Livelihood and Wealth Data Across B40 Community in Malaysia

Accurate information is vital in decision making. In poverty study, inaccurate, dishonest, incomplete, and misleading information especially related to income and wealth, disrupted the effectiveness of the analysis and the purpose the research. As a result, programs, strategies, and solutions implemented using the data to resolve problems and eradicate poverty will not be effective because it might not accurately reach authentic poor and extremely poor targeted groups. Therefore, researchers and enumerators must ensure the data is of high quality, well accurate and effective to prevent wrong decision making, including unethical behaviour and improper distribution of government grants and incentives. This study aims to investigate the issues and challenges in obtaining accurate, reliable, and quality livelihood and wealth data across B40 household in one district in Kedah, Malaysia. This data is important for the government to strategize the best way in helping them to improve their livelihood and reduce inequality of wealth. This study employed a qualitative research design and multi-method data collection including survey, physical observation, interview, and documents review. The respondents were heads and members of the household of B40 groups in this district. The data was analysed using a thematic analysis technique to transform the data into useful knowledge for strategies development. The main findings show that individual or household tend to provide incomplete, untrue, and enable government to develop appropriate poverty eradication policy as well as proper distribution of national wealth ecosystem in realizing Shared Prosperity Vision 2030, sustainability livelihood agenda; reach a high-income nation; and improve quality of live; consistent with SDG1 (No poverty).


Introduction
To measure is to manage. The same goes to the poverty and inequality contexts. Previously, Poverty Line Income (PLI) was used as a threshold to determine poverty. The total expenses of the basic needs were calculated and compared with the income earned. This measurement however received a lot of criticism as it was argued that poverty is not only should be measured monetarily, but also other qualitative elements such as health, wealth, education and entertainment. This measurement is known as Multidimensional Poverty Index (MPI). MPI was introduced officially by Malaysian government in 2015 to complement PLI (Kenneth et al., 2021).
In this study of tracking and profiling B40 and below households in one district in Kedah, Malaysia; the MPI is used and income and wealth data are among the main key dimensions. Data of income collected not just to determine whether they fall under the B40 group or below; but the data include their source of income to see whether their source is sustainable and can be further improved. As for wealth, data tracking includes all the wealth that under their position such as house, land, vehicles, saving, investment and so on. The reason for tracking this data is to develop the appropriate program to elevate their standard of living based on the resources that they have.
Matching programs and assistance with resources and wealth is not possible if the data is not correctly, completely and comprehensively gathered. When the researchers first came into the community, they had no idea of the challenges that they would face in this district as not much studies were done previously regarding poverty; despite of its status of one of the poorest place in Malaysia. Further, the existing study discuss more on the measurement index issues rather that issues and challenges in data collection. Researchers' experience shows that getting accurate, comprehensive, complete and reliable data is so challenging. They are many reasons for that. One is the methodological challenge in term of study design and practical limitations (Sweeney, et.al., 2016). From researchers' experience, the survey questions used in interview need to be modified many times to suite the respondents and to make it more practical in data collection process. With the sudden COVID-19 pandemic, things are getting worse as the interview process need to be delayed and SOP need to be followed.
Regardless of the challenges that the researchers face during data collection, what important is the urge to get accurate, complete and reliable data from respondents. The data is crucial for government, NGOs, agencies and philanthropists to strategies and plan for aid and assistance to be given to this group (B40 and lower). There should not be the case where aid and assistance are given to wrong person, and those who are really in need is left out. In addition, the programs and assistance planned should suite their needs and resources that they have, so that the sustainability of the program, as well as the outcome of the program become more impactful. Therefore, the objective of this paper is to highlight the issues and challenges pertinent to income and wealth data collection that are rarely discussed before by previous literatures. It is highly significant as it gives insight to the future researchers on what to expect when collecting poverty data in rural areas, as well as its impact to the data collected. In addition, this paper also highlight the strategies used by researchers in overcoming those issues and challenges; so the data collected are accurate, reliable and complete. The authors are hopeful that the findings of the study could contribute to the body of knowledge and literature on issues and challenges of qualitative collection of poverty data involving B40 community, and could spark future research on strategies to overcome them.

Conceptualising Poverty
According to the Britannica Dictionary, poverty is 'the state of one who lacks a usual or socially acceptable amount of money or material possessions'. Money or material possessions are understandably acceptable in this context if they are insufficient to cover fundamental requirements such as food, clothing, and shelter (Alcock, 1997). According to the World Bank, 10 percent of the global population, or 734 million people, lived on less than $1.90 a day in 2015. In addition, the World Bank projects that between 40 and 60 million people fell into extreme poverty (less than $1.90 per day) in 2020.
In Malaysia, the standard of life is categorised into three main categories benchmarked against the level of income range, namely T20, M40 and B40. This categorisation actually refers to the percentage of the community earning household income within a certain range, for example top 20% of the community earning the most income, middle 40% of the community earning moderate amount of income, and bottom 40% of the community earning limited or low amount of household income. The income range is shown in the following Table  1. Within the context of the present study, the scope of discussion is B40 community, i.e. the households earning RM4,849 or less per month.

Challenges of Poverty Data Collection involving B40 Community
Past literature has reported a number of challenges of data collection involving people with poverty, including B40 community. Listed below are among the key challenges, namely complexity of background of the respondents, research data collection is not core to family priorities, economic and IT restrictions, lack of quality assurance, instrumentation limitations, and lack of training among study researchers.
Among the major challenges of qualitative research purportedly to be conducted involving poor or underprivileged participants is the sampling and choice of interview participants. Based on certain inclusion criteria set by the researchers, the potential participants could be narrowed down. Nevertheless, when it involves B40 community, the background of the participants such as age, life experiences, working experiences and employment status, could be so complex so as to fit into the dimensions intended in the study (Adato et al., 2006). Certain groups of participants might have diverse complex backgrounds and experiences, to which the researcher should well anticipate and prepare to handle accordingly.
Another challenge is that research data collection is not core to family priorities. Banting (1979) highlighted that family priorities among poor and underprivileged community are financial resources, basic necessities and healthcare. Understandably, participating in research studies conducted by persons external to their lives is not a priority to them. Hence, data collection carried out involving this group might entail certain drawbacks to the study itself, such as dishonest responses, misleading information, and the feeling of not trusting the researchers (Zhou & Liu, 2019;Lister, et.al., 2019). What makes it even more problematic is when the data purportedly to be collected from them involves sensitive data, such as income and expense, wealth information and property ownerships. Therefore, the researchers should be smart to handle this problem, particularly to gain the trust of the intended participants, so that they can establish good rapport with the poor community before taking any steps to actually carry out the data collection via personal interviews, observation or any other tools for data collection.
Economic and IT restrictions also serve as a great challenge in data collection involving B40 community. With the advancement of economic and information technology (IT) opportunities, as technology, connectivity and devices become more affordable, economic activities take many shapes and forms, there still exist economic divides and IT restrictions among the people in rural areas, particularly poor and underprivileged community (Dawood et al., 2019). Money and financial resources remain one the greatest challenges among the community, as well as access to affordable IT opportunities (Umar, 2021). This could very well serve as one of the major challenges in collecting poverty data involving poor and underprivileged community. On this note, the researchers need to properly plan the strategies for their data collection of this community so as to get rich and useful, reliable data for their studies.
Apart from that, lack of quality assurance also serves as major challenges in data collection involving B40 households. Quality assurance is one strategy that is used to assure the validity of study results and maintain the integrity of data while it is being collected. In the process of carrying out the research study, quality assurance plays a crucial function by helping to ensure that the findings and conclusions are authentic and appropriate (Baigent et. al., 2008).
Particularly for data collection involving poor and underprivileged community, despite the internal administration and strategies for quality assurance set out by researchers, some variables are proven to be out of the control of the researchers (Meehan-Andrews et al., 2006). This is rightly so particularly if the subject matter of the data collection is sensitive personal data on wealth and income of the community. Henceforth, quality of the data and the data collection process could not be assured completely.
Literature also suggested that instrumentation limitations occasionally hamper the process of data collection when it comes to poor and underprivileged community (Islam, 2003). At times, researchers who are academically trained coupled with enumerators who are taught by the said academics would be familiar with the formal style of asking and probing questions (Jason, 2007;Jann et al., 2019). Accordingly, the instruments used for data collection might also be rigid, formal and lengthy. In turn, such instruments could become an obstacle for proper data collection, especially if it involves poor community, who might or might not be comfortable in the rigid and formal instruments and the way data collectors would approach them.
Another challenge highlighted by literature in the context of data collection is the suitability of the approach used by the researchers in collecting the relevant and useful data (Bentley & Meek, 2018). It is well understood that researchers should be well prepared to establish rapport and to be familiar with the participants, prior or during the actual meeting (Canivez, 2019;Oquendo et.al., 2018). Failure to approach the participants appropriately would entail inadequate or inappropriate data to be collected. Accordingly, the challenge facing the researchers in collecting poverty data among B40 community is the lack of training among the study researchers. Accordingly, the researchers should be able to anticipate the profiling and nature of their study participants, i.e. B40 community, and go through appropriate trainings and preparations prior to attempting the data collection process.

How to Overcome Data Collection Challenges
Based on the various challenges discussed above, literature has pointed about several strategies that could be adopted to overcome the challenges of data collection involving poor community.
The primary strategy for collecting data is to create a proper guide for research implementation (Fox & Bayat, 2008). This would include setting out step by step guidelines for the data collectors involved in the study, particularly the do's and don'ts during the study. All the researchers involved would abide by the guide, so that any issues and challenges that could be identified as early as possible, and eventually could be addressed and handled appropriately among the researchers (Moon & Blackman, 2014). Essentially, the research guide would cover at least three stages of the study, i.e. preparation before the data collection, during the data collection, and after the data collection (Merriam & Simpson, 1995). In the present study, what is highlighted in the guide would be the communication style with the study participants who are within the category of extremely poor and poor households, the type of data generated during the data collection, and the handling of the collected data.
An important aspect highlighted by literature is the training of research personnel particularly those who would be involved directly with the study participants (Heslop, Burns & Lobo, 2018). Even in cases involving standard community with diverse backgrounds, researchers would need to undergo training for implementing data collection (Tait & Williamson, 2019). The situation involving poor and underprivileged community would warrant even more specific and focused training of the researchers, mainly to train the skills of handling empathy in research, aptitude for listening to the sharing and the ability to establish rapport by the study participants, and many other research skills purportedly to be possessed by researchers (Strijker et al., 2020). With good handling and communication skills by the researchers, it is estimated that better experiences would take place during the data collection process involving B40 community.
Another strategy mentioned in the literature for the purpose of assisting data collection involving poor community is by using established data related indicators (Greville et.al., 2019). This essentially means that benchmarking could be done against numerous data that has already been reported in the existing literature on the required study needs, in this situation, the livelihood and income of poor and underprivileged households (Strijker et al., 2020). Established data such as property ownership, spending patterns, income range and health status could be used as indicators to probe information from the study participants. Nevertheless, it is important that the researchers do not lose their objectivity by totally relying on the established data as the only option during the data collection. The researchers should always be mindful of the inductive-deductive reasonings of qualitative data collection and analysis, and that new data might emerge from the data collection process with the study participants.
Finally, to address the instrumentation limitations as discussed earlier, the researchers could use pre-defined responses, and this includes adding in options such as, 'Prefer not to say', 'No comment' and 'Unable to ask'. Additionally, the options could also be in the form on ranking any given options, and thereafter by soliciting open-ended responses from the study participants (Ensor, et.al, 2019). The purpose of using such pre-defined responses would reduce the problem of lengthy, complex and complicated questions of the instruments adopted in the data collection process.
What could be gathered from the review of literature on the key concepts engaged in the study is that there are various numerous issues and challenges affecting qualitative data collection process involving poor and underprivileged households. Accordingly, literature has also pointed out a number of strategies for addressing and reducing such issues and challenges.

Methodology
This study adopted interpretivism research paradigm using a qualitative research design and multi-method technique, which using more than one data collection technique, but all are qualitative. The technique frames the research around interviews, surveys and observations. The textual data from respondent's source, which is among B40 households in one district in Kedah, Malaysia, relating to income and wealth data have been collected, skimmed, analysed, assessed and exploited by carrying out an in-depth analysis and drawing conclusions on issues, challenges and strategies that have been developed. Additionally, the researchers' experiences and observations are also taken into account.
These textual data and themes created from thematic analysis have permitted the researchers to use a contextual approach to construct rich and deep data. It also enables the researchers to extend their contribution in the research process to realise how issues, challenges and strategies were aligned together in obtaining reliable and quality livelihood and wealth data across B40 households in one district in Kedah, Malaysia. This approach helps the researchers in tracking a forceful and rich data in response to the research questions and objectives, containing the past and current critical issues and challenges that account for the practical strategies and solutions.

Issues and Challenges of Data Collection Involving B40 Community
Having carried out data collection involving B40 community at a certain district in Kedah, the researchers encountered numerous issues and challenges such as data disruption, gaining trust and confidence of the participants, sensitivity of data, evidence and validation, and isolated and disintegrated database. Each of the issues and challenges are as provided in the following Figure 1 and deliberated thereafter.

Data Disruption
COVID-19 has badly affected not only poverty and extreme poverty groups (B40 group) household income, but also M40 and T20 groups across this particular district in Kedah, Malaysia (as well as throughout Malaysia and worldwide). Indeed, it has caused many M40 and T20 groups fell under B40 category. Moreover, the poverty data that is normally drawn from household surveys has turned invalid and obsolete during COVID-19. Hence, it is almost impossible to carry out appropriate surveys due to movement control order and lockdown that triggered economic and social disaster such as loss of job, loss of livelihood and life, bankruptcy, financial loss, closure of business, business collapse, disruption of business operations and health crisis. In fact, all the data that the researchers have collected prior to COVID-19, become no longer relevant. Furthermore, the profile of income and wealth have changed significantly. Economic and poverty parameter seem need to be reset, including data to be recollected and updated to take into account the new norm, risks, opportunities and business model.

Difficulty to Gain Trust and Confidence
It is not easy for researchers and enumerators to build rapport to gain trust and confidence of respondents to provide and share information about their income and wealth. The enumerators and researchers are outsiders to the respondents and it is normal not to disclose data to the people that they do know well. Consequently, it takes quite some time for the data collector to explain the purpose of the visit and convinced them the data will be treated as confidential data. Further, the research will at the up benefit them and therefore it is important for them to disclose accurate data.

Sensitivity of Data
People are turning out to be more reluctant and increasingly worried to share their personal data, especially data related to income and wealth. This is in consequence of data privacy, confidentiality, sensitivity and security. In addition, disclosure of income and wealth data might exposure respondents to unethical activity including scammer, fraud and other crimes. Besides exhausted with so many surveys conducted on them, the respondent were also reluctant to share information because they do not always even understand why and how data is collected, shared, stored, and used. Sometimes the unwilling to share the data are driven by respondent's negative perception and anxiety, which the disclosure may cause trouble and backfire them such as they possibly will loss of all the aids and supports that currently enjoyed or obtain from various government agencies and other parties. Therefore, respondents are getting more curious, less transparent and also exhibit tendency to hide the complete and accurate data for the sake of their own good.

Evidence and Validation of Data
In ensuring the reliability, accuracy, completeness, relevance and timeliness of the income and wealth data, tracking and obtaining evidence for the collected data are very challenging. Normally people will refuse to share their wealth and income data with other, unless it is mandatory or required by law and authority. Without legitimate power and authority, who are researchers and enumerators to ask respondents among household in this district to prove their data and validate the ownership of assets by providing evidence such as documents, pay slip, income statement, bank statements and grants for houses, land, vehicles and other assets. Besides some of the document requested are not available, especially the pay slip because of most of the respondents are self-employed, small traders, farmers, working in stalls or shops and doing various other village work.

Isolated and Disintegrated Database
There are many poor and hardcore poor households in this district. Therefore, they are eligible for various financial and non -financial assistance from various parties and government agencies (via eKasih, Zakat, NGOs and lots more). In general, the financial assistance obtained on a monthly basis including contributions from children (that working outside the district) is accounted for as passive income to them. There are cases where some of these poor households received financial aids from more than one agency separately and overlap due to the lack of a centralized and integrated database. Hence, income data reported by this group is inaccurate because they do not report all income received (missing income data on the census files), definitely because fear of losing the overlapping income due to possibility of being removed from the list of aid recipients.

Unmatched Data
There are cases the income data provided by the respondents in this district unmatched and not consistence with the observation made by the researchers and enumerators. This can be seen when the income earned by the respondents are less than expenses made (income VS expenses). Moreover, the income declare also does not reflect what researchers and enumerators observe (observation VS interview) such as respondents proclaim low income yet living in a luxury house with comfortable lifestyle, lavishness home furnishing, expensive vehicle and also owns lots of assets. Apart from not fully declaring income and property, this unmatched data happens attributable to lots of reasons including respondents have inherited property, insurance money, have huge past savings (example, EPF, Tabung Haji, ASB) and get extra support from immediate family and extended family members or others. In this context, in-depth interview and observation from qualitative research strategy is necessary to prevail this issues, which is missing if data just collected through survey questionnaire (quantitative).

Complexity and Comprehensiveness of Survey Instruments
In order to obtain accurate and complete data, indicators, instruments and number of questions tend to be too long and questionnaire form seem unclear, very confusing and complicated -require more time and effort/be more tactical. For instance, the questionnaire form for the household income section (which does not yet include wealth and other data) has become complicated and lengthy because it seeks to investigate income from various sources and angles, with the unit of analysis including salary, wages, income from business, income from services rendered, revenue, rent, lease, interest, dividend, royalty, financial aid, zakat, grants, endowment, profit, fee, allowance, and cash inflow. Tracking the forms and types of household income, whether daily, weekly, monthly, yearly, indefinite, fixed, or seasonal, allows for the collection of more data regarding household income. It is essential to obtain information on the amount of revenue earned by the family's head, spouse, and other adults in order to determine the household's total income. In addition, the authors recorded the details of passive income such as monthly financial assistance, annuity, gifts (monetary and non-monetary), and school (children) aid that a family or its members received from the government, non-government organisations, corporate entities, Lembaga Zakat Negeri Kedah, relatives, and other parties to be included in the household income.
In order to fully comprehend household income, it is essential to track information about household source of income and type of occupation, such as self-employed, rubber tappers, farmers, fruit sellers, chef, home builders, meat butcher, fishmonger, foreman, gardeners, motor/car mechanic, sole proprietorship, small and medium entrepreneurs, farmers, contractor, government sector employee, private sector employee, fishermen, food store worker, or unemployed. The tracking should also take into account the number of adults in a home who create revenue, the number of jobs held by the head of the household and other adults, and whether the employment is part-time, full-time, temporary, or permanent. Last but not least, it is vital to know the size of a household or the number of household members because many households consist of a single individual and this information is used to calculate the median household income. Comparing poverty, wealth, and standard of living between different regions, cities, states, or countries is facilitated by household income. To ensure smooth, accurate, reliable, comprehensive, effective and efficient data collection, the questionnaire form and design need to be simplified and tested in real setting for resolving confusing part and further improve before it can be used or otherwise it will remain too long and complicated for respondent to understand.

Other Issues and Challenges
Other issues and challenges discovered during the data collection in this district contain respondents' unwillingness to participate including by shutting and do not want to open their house door. Respondents also found to be very difficult to participate and give full commitment start from the second round onward of the data collocation phase. It is a result of the long period of time consumed to complete the interview and survey questionnaire, which have disrupting respondents working hours and daily activities that implicating loss of their daily income. Furthermore, there are respondents that really struggle to obtain a sustainable income, which the demand, patent and nature of their work (or service) are fluctuate and inconsistent across daily, weekly, monthly and sessional timeframe.

Strategies to Overcome Issues and Challenges of Poverty Data Collection Involving B40 Community
The following are few strategies used by the researchers to overcome the issues and challenges faced in collecting poverty and wealth data involving B40 community, including analytic, cross check and validation, appropriate data collection management and administration, as well as prior rapport building.

Analytic, Cross Check and Validation
Data gathered from researchers and enumerators will go through few validations stage. Among other thing, the data collector will observe respondents' lifestyle and activities during the interview and later match with the information given. If they think that there is confusion between data and observation, data collector will go for second visit and clarify the issue. During the interview, data of inflow (income) will be collected together with the outflow (expenses). At the end of the interview, these two data (total revenue and total expenses) will be match. If total expenses are more than revenue, respondent need to justify or some adjustments are made with them on their income or/and revenue data. Data on wealth such as vehicles, livestock, furniture and fixtures can be validated through data collectors' observation during interview. Certain income data related to government assistance and incentives are verified through head of village, and matched with the data maintained by the agencies and authorities.

Appropriate Data Collection Management and Administration
At first, the survey form used by the researchers was quite lengthy and comprehensive. Consequently, the data collection process took too long, complicated to understand and has high potential to bore respondents. Through the pilot test which was carried out with real respondents, some improvements were done, where some questions were combined, adjusted and simplified. This eventually made the data collection easier and not too complicated especially for the data collectors.
Other than that, based on the pilot test, the authors strategised the approach in terms of the sequence and chronology of the questions to be asked, as to increase convenience and flow of the data collection. Simulation was also carried out with enumerators with real respondents in order to guide enumerators of the right way for approaching respondents and asking questions. Continuous communication is carried out between enumerators and researches for any arising issues in data collection.
In addition to the above, data collectors also bring some token to the respondents in term of daily necessities such as sugar, coffee and cooking oil. The intention of the researchers is to reward and to ease the burden of the study participants for committing to participate in the study.

Prior Rapport Building
Normally, no one would reveal any income and wealth information to other people, especially if the authors do know that person closely. Therefore, it is important for data collector to gain trust of the respondents first. Researchers encourage enumerators to build rapport with the villagers before they do to the field for data collection. Among the strategies the authors use to be accepted and recognizable by the villagers is by doing social responsibility activities such as providing assistance to the villagers, especially involving school children, small entrepreneurs, farmers, the sick, single mothers and the disabled. In addition, the authors also held meetings with village heads, imams and mosque committees, local leaders and village development committees to provide information and awareness on the study conducted, as well as to get help, cooperation and support from them to facilitate the data collection process. It is crucial to make us visible and acknowledged.

Conclusion
In this article the authors have highlighted important considerations in collecting accurate, comprehensive and reliable data in order to obtain, assess and analysing quality income and wealth data for decision making purposes. Quality of income and wealth data use in determining and categorizing poverty groups are crucial to correctly and effectively classify, manage and eradicate poverty in this district or anywhere else worldwide. Going forward in these contexts, the government and other relevant bodies tasked with developing programmes, initiatives, and policies for alleviating poverty must improve and ensure the accuracy, completeness, and dependability of data as critical requirements in the study design phase and advocate for the collection of high-quality data as a crucial component of the poverty evaluation. Even though data can be validated but if the data are incomplete and not reported by the respondents, it remains inaccurate, incomplete and unreliable, which will lead to wrong identification and recognizing poverty groups as well as incorrect strategies to mitigate or eradicate the poverty.
For the effectiveness of the analysis, findings, and conclusions, which are becoming increasingly crucial to policymakers and poverty programmes, the data must be of high quality. Unavoidably, a degree of variation in data accuracy and availability may have severe consequences for poverty management. How and what type of income and wealth data were obtained and measured in poverty studies is crucial, since it will affect how the authors interpret the meaning of poverty and how policy is formulated. As a consequence, accurate, complete and reliable data of income and wealth can help researchers to understand and interpret findings more effectively.
Finally, it may be possible to minimize the inaccurate data and less impact poverty initiatives, through scrutinizing and ensuring data accuracy and performing in-depth interview, tracking and observation of data collection methods. For quantifying the impact of erroneous, incomplete, and unreliable data, researchers should be encouraged to address underlying data shortages. The concern of this paper is to improve the accuracy and quality of income and wealth data so that the authors can effectively combat poverty and extreme poverty. Eventually the authors are in the right track to effectively manage and get rid of poverty as well as to promote Shared Prosperity Vision 2030 and SDG 1 goals via generating sustainable income for B40 groups in one district in Kedah, Malaysia and globally.