Assessing the vulnerability and climate change impact on the marginalized communities in Dhaka division: a rasch score and mixed methods approach

Israt Rayhan; Nayeem Sultana

doi:10.15406/bbij.2026.15.00447

eISSN: 2378-315X

Biometrics & Biostatistics International Journal

Review Article Volume 15 Issue 1

Assessing the vulnerability and climate change impact on the marginalized communities in Dhaka division: a rasch score and mixed methods approach

Israt Rayhan, Nayeem Sultana

Received: April 15, 2026 | Published: May 6, 2026

Citation: Rayhan I, Sultana N. Assessing the vulnerability and climate change impact on the marginalized communities in Dhaka division: a rasch score and mixed methods approac. Biom Biostat Int J. 2026;15(1):45-49. DOI: 10.15406/bbij.2026.15.00447

Download PDF

Abstract

Bangladesh remains one of the most climate-vulnerable nations in the world, yet rigorous, community-disaggregated vulnerability measurement for its marginalized populations remains scarce. This study assesses climate and service-access vulnerability across four marginalized community types in Dhaka Division, namely Ethnic Minorities (Adivasi), Dalits, Disabled or Begging populations, and Urban Slum dwellers, covering three critical sectors: Health, Education, and Social Safety Net. A cross-sectional survey of N = 295 respondents was conducted using a multi-stage stratified random sampling design. The Rasch Item Response Theory (IRT) model was applied to dichotomized response matrices, while Principal Component Analysis (PCA) was performed on the polytomous Likert items to validate the latent dimensional structure. The Rasch model produced highly reliable vulnerability scales (KR-20 = 0.826 for Health, 0.919 for Education, 0.821 for Social Safety Net). PCA confirmed unidimensional pattern, with the first principal component explaining 34.2%, 53.3%, and 46.6% of total variance in the three sectors respectively. Convergent validity between Rasch and PCA scores was very high (r = 0.943 to 0.978, p < 0.001). Oneway ANOVA revealed statistically significant community-level vulnerability differentials in all sectors, most strongly in Education (F = 26.48, p < 0.001). Ethnic Minorities emerged as the most vulnerable overall (mean = −0.618 logits), followed by Dalits (−0.196). Sexbased analysis showed females as significantly more vulnerable than males in Education (F = 21.66, p < 0.001), while Third Gender respondents exhibited a distinctive mixed profile. Item-level analysis identified mental health services, specialist care, VGD/VGF registration, and housing assistance as the most inaccessible entitlements, pointing to priority areas for intersectional, community-sensitive climate adaptation and social protection policy.

Keywords: climate vulnerability, marginalized communities, rasch model, principal component analysis, social safety net, Bangladesh