Skip to main content

Table 3 School attendance outcomes from school attendance tracking instrument among girls menstruating at baseline and interviewed at endline

From: Effects of sanitary pad distribution and reproductive health education on upper primary school attendance and reproductive health knowledge and attitudes in Kenya: a cluster randomized controlled trial

  Arm 1
Control
Arm 2
Pads only
Arm 3
RH only
Arm 4
Pads and RH
Respondents in analytical sample (N) 627 632 629 656
Attendance was taken for all 60 days (% (N))a 84.7 (531) 87.3 (552) 86.5 (544) 85.7 (562)
Intra-cluster correlation coefficient 0.000
Coefficient (95% CI) Reference 0.022 (− 0.047, 0.092) 0.015 (-0.049, 0.079) 0.007 (− 0.069, 0.083)
P-value   0.524 0.431 0.862
Mean # of days attended (mean (SD))a 55.6 (6.5) 56.0 (6.3) 55.8 (6.8) 56.2 (6.1)
Intra-cluster correlation coefficient 0.059
Coefficient (95% CI) Reference 0.37 (− 0.73, 1.46) 0.14 (− 0.99, 1.26) 0.58 (− 0.37, 1.52)
P-value   0.507 0.812 0.230
Observed attendanceb (% (N)): 91.4 (619) 91.8 (625) 91.6 (622) 92.2 (647)
Intra-cluster correlation coefficient 0.056
Coefficient (95% CI) Reference 0.42 (− 1.94, 2.78) 0.28 (− 2.00, 2.57) 0.75 (− 0.97, 2.48)
P-value   0.725 0.806 0.389
  1. The table reports post-intervention means for the control arm, and the estimated effect of the intent-to-treat for each study arm relative to the control arm. Difference-in-differences were estimated from regressions with girl-level fixed effects and robust standard errors accounting for clustering at the school level
  2. Higher scores equate to higher knowledge and more positive/equitable norms and attitudes
  3. aAttendance data only includes those who remained in the same school throughout the 60 days
  4. bDifferences at endline were estimated using ANCOVA. Regressions controlled for the following covariates measured at baseline: cognitive, math and literacy test scores, socio-economic quintile, age, parental living status, subcounty, and were estimated with robust standard errors accounting for clustering at the school level