Page 150 - AJWEP-22-4

Jusoh and Alwi

humanitarian operations during previous national The minimum sample size necessary for multivariate
disasters identified. The verification of NADMA analysis was determined using G*Power 3.0 software.
14
contacts as legitimate flood responders began with The established parameters comprised two predictors,
the extraction of a list of 3,732 individuals from the a power level of 95%, an effect size of 15%, and a
NADMA’s operational database. This list comprised statistical significance threshold of 5%, collectively
personnel registered as having participated in indicating a minimum required sample size of 129
humanitarian operations during past national disasters. respondents. Data were collected from government
To ensure the accuracy of the sampling frame, the list agencies, specifically the APM, and NGO offices
was cross-referenced with official deployment records, nationwide to ensure comprehensive national
including duty rosters, incident reports, and operational coverage. A cross-sectional design was employed to
logs from specific flood events. This step was critical to gather responses within a single timeframe, ensuring
confirm each individual’s active involvement in flood consistency in data conditions. The data collection
13
response rather than in other types of disasters. Such process lasted 4 months, from February to May 2019.
15
triangulation methods have been widely recommended Two thousand printed questionnaires were distributed
in disaster management research for ensuring data to identified respondents at the headquarters of the
validity and reliability. 16 participating humanitarian organizations. A reminder
Further verification was conducted through direct was sent 1 month after the initial distribution to
consultation with regional NADMA offices. Regional enhance the response rate. A total of 593 completed
coordinators played a key role in validating the roles questionnaires were considered valid for analysis,
and responsibilities of personnel during specific flood surpassing the minimum sample size required by
events. This included confirming names, locations, G*Power 3.0, thereby providing robust statistical power
duration of deployment, and specific tasks performed. for further analyses. A total of 86.2% of the respondents
In cases where documentation was incomplete, verbal were associated to government agencies, while 12.8%
confirmation and supplementary materials from regional were connected to NGOs. The structural dominance of
records were used. Moreover, to strengthen the validity the public sector in Malaysia’s disaster management
of the list, external verification was sought through system helps to mostly explain the disproportionate
collaborating agencies, such as the APM and the Fire presence of government entities (86.2%) compared
and Rescue Department (BOMBA), which often operate to NGOs (12.8%) in the respondent pool. Under a
alongside NADMA during flood crises. These multi- centralized disaster governance system, Malaysia’s
17
agency consultations helped to eliminate individuals NADMA arranges official channels, including the
whose disaster involvement did not pertain to floods. Royal Malaysian Police, BOMBA, Armed Forces, and
The final stage involved cleaning the data to exclude APM, to coordinate the flood disaster response. These
18
non-flood responders and ensure that only individuals government players are constitutionally obligated and
with direct flood-related experience remained in the methodically mobilized during national catastrophes,
sampling frame. This process was aligned with best particularly floods, which often impact low-lying and
practices in disaster research, emphasizing specificity coastal areas. On official planning and implementation,
in respondent selection to ensure accurate analysis and NGOs active and vital during humanitarian crises are
policy recommendations. The resulting list, having sometimes viewed as supplemental players rather than
18
undergone thorough cross-verification and validation, fundamental stakeholders. NGOs, especially smaller
17
was then approved for use in subsequent research or community-based ones, often lack consistent inclusion
operational planning. By employing a robust verification in such databases unless they are involved in formal
approach supported by practices documented in the memoranda of understanding or registered under
literature, the study ensured a credible and focused collaborative frameworks. In addition, the educational
20
representation of flood responders. background indicates that the majority of individuals
In addition, a non-probabilistic sampling strategy possess the Sijil Pelajaran Malaysia certificate (42.2%),
was utilized to facilitate accessibility and participation, followed by diploma holders (39%), bachelor’s degree
enabling researchers to efficiently target respondents holders (16.7%), master’s degree holders (1.2%), and
from diverse organizations. This method aligns with PhD holders (1%). Volunteers represented the largest
previous studies indicating that access and logistical segment of occupational roles at 44.7%, followed by
challenges restrict the application of probabilistic administrative staff at 22.3%, clerical workers at 17%,
sampling in field-based humanitarian research. 10,19 supervisors/executives at 8.1%, managers at 4.1%,

Volume 22 Issue 4 (2025) 142 doi: 10.36922/AJWEP025220182