Skip to main content
. 2026 Jul 8;26:2742. doi: 10.1186/s12889-026-28395-w

Table 7.

Sensitivity analysis comparing Firth penalized logistic regression with survey-weighted and unweighted models

Variable OR (95%CI) % Change from M1 Verdict Verdict
M1 M2 M3 M1→M2 M1→M3 M1→M2 M1→M3
Gender
 Male (rf)
 Female 0.64(0.10, 4.21) 0.00 (0.00, 0.00) 0.00 100 100 Unstable Unstable
Age
 15–24 (rf)
 25–34 0.85(0.47,1.53) 0.66 (0.28, 1.57) 0.86 (0.47, 1.56) 22.36 1.18 Unstable Stable
 35–44 1.00(0.56,1.80) 0.67 (0.28, 1.59) 1.02 (0.56, 1.86) 33 2.00 Unstable Stable
 45–54 0.96(0.53,1.76) 1.07 (0.47, 2.42) 0.97 (0.52, 1.80) 11.46 1.05 Unstable Stable
 ≥55 1.20(0.64,2.23) 1.00 (0.42, 2.38) 1.25 (0.66, 2.36) 16.67 4.17 Unstable Stable
Residence
 Urban (rf)
 Rural 0.90(0.67,1.21) 1.18 (0.75, 1.86) 0.90 (0.67, 1.22) 31.12 0.00 Unstable Stable
Division
 Barisal (rf)
 Chattogram 0.30(0.17,0.53) 0.15 (0.07, 0.34) 0.27 (0.15, 0.49) 50 10 Unstable Unstable
 Dhaka 1.12(0.67,1.89) 0.72 (0.38, 1.35) 1.13 (0.66, 1.93) 35.72 0.90 Unstable Stable
 Khulna 0.80(0.46,1.42) 0.53 (0.23, 1.21) 0.79 (0.44, 1.40) 33.75 1.25 Unstable Stable
 Mymensingh 0.60(0.35,1.05) 0.40 (0.18, 0.86) 0.58 (0.33, 1.03) 33.34 3.34 Unstable Stable
 Rajshahi 0.81(0.48,1.39) 0.42 (0.20, 0.91) 0.80 (0.46, 1.38) 48.15 1.24 Unstable Stable
 Rangpur 0.82(0.50,1.35) 0.63 (0.31, 1.30) 0.80 (0.48, 1.34) 23.18 2.44 Unstable Stable
 Sylhet 0.50(0.29,0.87) 0.35 (0.15, 0.79) 0.49 (0.28, 0.86) 30 2.00 Unstable Stable
Educational Level
 No Formal School (rf)
 Less Than Primary 1.33(0.91,1.95) 1.33 (0.81, 2.18) 1.36 (0.92, 2.01) 0.00 2.26 Stable Stable
 Primary School Completed 0.98(0.61,1.58) 1.15 (0.59, 2.27) 0.98 (0.61, 1.60) 17.35 0.00 Unstable Stable
 Less Than Secondary 1.10(0.72,1.67) 1.44 (0.79, 2.64) 1.10 (0.72, 1.69) 30.91 0.00 Unstable Stable
 School Completed 0.98(0.53,1.79) 1.01 (0.38, 2.72) 0.98 (0.52, 1.82) 3.07 0.00 Stable Stable
 High School Completed 1.62(0.76,3.48) 4.16 (1.36, 12.70) 1.65 (0.75, 3.63) 156.8 1.86 Unstable Stable
 Graduation Completed 2.23(1.02,4.88) 1.09 (0.21, 5.58) 2.31 (1.03, 5.17) 51.13 3.59 Unstable Stable
 Postgraduate Degree 3.13(1.13,8.70) 3.59 (0.67, 19.18) 3.30 (1.14, 9.53) 14.7 5.44 Unstable Stable
Occupation
 Government Employee (rf)
 Non-Government Employee 1.32(0.44,3.96) 4.90 (1.27, 18.92) 1.40 (0.44, 4.45) 271.22 6.07 Unstable Stable
 Business (Small) 2.03(0.69,5.91) 3.67 (0.80, 16.83) 2.22 (0.72, 6.84) 80.79 9.36 Unstable Stable
 Business (Large) 1.86(0.55,6.31) 2.86 (0.50, 16.41) 1.99 (0.55, 7.17) 53.77 6.99 Unstable Stable
 Farming (Land Owner) 3.03(1.00,9.17) 7.32 (1.53, 35.10) 3.35 (1.05, 10.70) 141.59 10.57 Unstable Unstable
 Agricultural Worker 2.49(0.76,8.17) 7.03 (1.53, 32.24) 2.75 (0.80, 9.50) 182.33 10.45 Unstable Unstable
 Industrial Worker 1.26(0.25,6.35) 4.96 (0.57, 42.88) 1.24 (0.22, 6.94) 293.66 1.59 Unstable Stable
 Daily Laborer 1.49(0.49,4.51) 3.58 (0.75, 17.21) 1.63 (0.51, 5.21) 140.27 9.40 Unstable Stable
 Other Self-Employed 2.05(0.66,6.39) 3.98 (0.81, 19.46) 2.23 (0.68, 7.33) 94.15 8.79 Unstable Stable
 Student 2.87(0.61,13.43) 2.03 (0.18, 22.63) 3.18 (0.63, 16.11) 29.27 10.81 Unstable Unstable
 Retired 3.39(0.87,13.24) 19.85 (3.11, 126.60) 3.69 (0.89, 15.29) 485.55 8.85 Unstable Stable
 Unemployed, Able to Work 1.41(0.36, 0.56) 3.40 (0.51, 22.86) 1.55 (0.37, 6.57) 141.14 9.93 Unstable Stable
 Unemployed, Unable to Work 1.97(0.50,7.80) 5.28 (0.63, 44.64) 1.88 (0.43, 8.10) 168.03 4.57 Unstable Stable
 Other (Specify) 2.88(0.86,9.63) 10.05 (1.74, 58.18) 3.20 (0.91, 11.28) 248.96 11.12 Unstable Unstable
Wealth Index
 Poorest (rf)
 Poorer 0.58 (0.29, 1.18) 0.56 (0.28, 1.11) 0.83 (0.51, 1.34) 3.45 43.11 Stable Unstable
 Poor 0.63 (0.31, 1.28) 0.58 (0.29, 1.16) 0.71 (0.44, 1.15) 7.94 12.7 Stable Unstable
 Rich 1.15 (0.65, 2.02) 1.10 (0.62, 1.93) 1.10 (0.70, 1.72) 4.35 4.35 Stable Stable
 Richest 0.82 (0.48, 1.4) 0.79 (0.46, 1.35) 0.96 (0.62, 1.48) 3.66 17.08 Stable Unstable
Constant 0.27 (0.04, 1.73) 0.28 (0.04, 1.84) 0.32(0.08, 1.27) 3.71 18.52 Stable Unstable

M1 refers to the Firth penalized logistic regression model, M2 denotes the survey-weighted logistic regression model, and M3 represents the unweighted logistic regression model. OR indicates odds ratio, and 95% CI refers to the 95% confidence interval. Percentage change has been calculated as |OR_M1 − OR_Mx| / |OR_M1| × 100, where OR_Mx corresponds to either OR_M2 or OR_M3. Results have been deemed stable if the percentage change has been < 10% and unstable if it has been ≥ 10%