Skip to content

Improving Women's Mental Health during a Pandemic

icpsr_167601

Michael Vlassopoulos, Abu Siddique, Tabassum Rahman, Debayan Pakrashi, Asad Islam, Firoz Ahmed (2024). American Economic Journal: Applied Economics 2024, 16(2): 422–455

DOI: 10.1257/app.20210655

field value
venue American Economic Journal: Applied Economics 2024, 16(2): 422–455
paper 10.1257/app.20210655
replication data openICPSR 167601
data license CC BY 4.0 (openICPSR project page 167601 (maintainer-read 2026-07-20)) — “This work is licensed under a Creative Commons Attribution 4.0 International (CC BY 4.0) License.”
regression datasets (in this corpus) 4
graduated / eligible / exported 4 / 38 / 411 (policy claims_manual_regression_cap)
routed / gap / residue (at source) 38 / 0 / 0

Claims & tasks

Every claim the paper makes in its main body about a parameter, joined to the benchmark task (dataset · coefficient) that captures it. The effect, s.e. and p are the captured regression's own reported values (the paper's Stata figures, carrying the paper's SE method); the row shows the paper's verbatim quote (page · exhibit), so you can confirm the regression matches what the paper says. The prominence column is the within-study inclusion signal (⭐ headline → primary → secondary). The sensitivity column is the privacy level of the most-sensitive variable in the regression (🔴 high / 🟠 medium / ⚪ low).

# prominence sensitivity claim effect s.e. p task instance (dataset · coef)
C1 ⭐ primary 🔴 high At the one-month endline, treated women experienced a 0.70 SD reduction in perceived stress ( p < 0.01) and a 0.65 SD reduction in depression severity ( p < 0.01) relative to untreated women (column 2, panel A, Table 2).
p. 18 · Table 2, col 2 (1-month endline With covar. (2))		 -0.696 0.059 2.9e-27 Linear Regression reg_103__1 · treat
n=2220 · d=12
C2 ⭐ primary 🔴 high “At the ten-month endline, the respective effects are reductions of 0.55 SD in perceived stress and 0.51 SD in depression severity ( p < 0.01for both), suggesting that the intervention had a lasting effect on the mental health of treated women ten months after the end of the intervention while the pandemic was still raging and a second lockdown was underway.” — p. 18
“These effects persisted ten months after the intervention ended when we find that stress levels and depression severity in the treatment group were a respective 0.55 SD and 0.51 SD lower than in the control group. These impacts translate into a reduction of 19.5 percentage points in the prevalence of moderate or severe stress and 19.1 percentage points in the prevalence of depression compared to the control group in which 95.7 percent of participants were moderately or severely stressed and 58.3 percent were depressed.” — p. 3
Table 2, col 5 (10-month endline With covar. (5))		 -0.551 0.075 1.3e-12 Linear Regression reg_107__1 · treat
n=2254 · d=12
C3 ⭐ primary 🔴 high At the one-month endline, treated women experienced a 0.70 SD reduction in perceived stress ( p < 0.01) and a 0.65 SD reduction in depression severity ( p < 0.01) relative to untreated women (column 2, panel A, Table 2).
p. 18 · Table 2, col 2 (1-month endline With covar. (2))		 -0.652 0.050 4.8e-32 Linear Regression reg_119__1 · treat
n=2220 · d=11
C4 ⭐ primary 🔴 high “At the ten-month endline, the respective effects are reductions of 0.55 SD in perceived stress and 0.51 SD in depression severity ( p < 0.01for both), suggesting that the intervention had a lasting effect on the mental health of treated women ten months after the end of the intervention while the pandemic was still raging and a second lockdown was underway.” — p. 18
“These effects persisted ten months after the intervention ended when we find that stress levels and depression severity in the treatment group were a respective 0.55 SD and 0.51 SD lower than in the control group. These impacts translate into a reduction of 19.5 percentage points in the prevalence of moderate or severe stress and 19.1 percentage points in the prevalence of depression compared to the control group in which 95.7 percent of participants were moderately or severely stressed and 58.3 percent were depressed.” — p. 3
Table 2, col 5 (10-month endline With covar. (5))		 -0.513 0.063 6.3e-15 Linear Regression reg_123__1 · treat
n=2254 · d=11

Regression datasets

The 4 regression dataset(s) graduated into the benchmark corpus from this study's replication package — each backs one or more claims above. Expand each for its features, response, (sound, data-independent) public bounds, and reproduction grade.

icpsr_167601_reg_103__1 — end_pss_score_sdidx ~ treat + pss_score_sdidx + age_respondent + Edu_respondent + occ_res…

Sensitivity proposal: 🔴 high maximum across columns (advisory; reviewed manually). Reproduction grade: ✅ match (benchmark-eligible).

Features

name description type sensitivity bound lo bound hi estimate s.e.
_cons intercept continuous — 1 1 +0.388 0.285
treat Treatment arm assignment indicator continuous ⚪ low 0.0 1.0 -0.6962 0.059
pss_score_sdidx Baseline standardized PSS stress score continuous 🔴 high -6.0 6.0 +0.06193 0.0235
age_respondent Respondent's age in years continuous 🟠 medium 0.0 120.0 -0.006142 0.0027
Edu_respondent Respondent's level/years of education continuous 🟠 medium 0.0 25.0 -0.02205 0.00894
occ_respondent Respondent's occupation category categorical 🟠 medium — — — —
income_loss Household pandemic-related income loss indicator categorical 🔴 high — — — —
nmember Number of household members continuous 🟠 medium — — -0.04417 0.0184
nchild Number of children in household continuous 🟠 medium — — +0.009317 0.0341
household_head Indicator respondent is household head continuous 🟠 medium 0.0 1.0 -0.005141 0.198
occ_husband Husband's occupation category categorical 🟠 medium — — — —
chores Respondent's household chore burden/time continuous 🟠 medium — — -0.07751 0.0535
UNION_CODE Administrative union (local geographic) code categorical 🟠 medium — — — —

Response

name description type sensitivity bound lo bound hi estimate s.e.
end_pss_score_sdidx Endline standardized PSS stress score continuous 🔴 high -6.0 6.0 — —

Public bounds (data-independent) sourced from: binary indicator; binary treatment indicator; plausible ceiling for years of completed schooling; plausible human age range; standardized (sdidx) index.

n = 2,220 samples.

icpsr_167601_reg_107__1 — end2_pss_score_sdidx ~ treat + pss_score_sdidx + age_respondent + Edu_respondent + occ_re…

Sensitivity proposal: 🔴 high maximum across columns (advisory; reviewed manually). Reproduction grade: ✅ match (benchmark-eligible).

Features

name description type sensitivity bound lo bound hi estimate s.e.
_cons intercept continuous — 1 1 -0.593 0.351
treat Treatment arm assignment indicator continuous ⚪ low 0.0 1.0 -0.5507 0.0748
pss_score_sdidx Baseline standardized PSS stress score continuous 🔴 high -6.0 6.0 -0.003764 0.0267
age_respondent Respondent's age in years continuous 🟠 medium 0.0 120.0 +0.009077 0.00308
Edu_respondent Respondent's level/years of education continuous 🟠 medium 0.0 25.0 +0.01246 0.00894
occ_respondent Respondent's occupation category categorical 🟠 medium — — — —
income_loss Household pandemic-related income loss indicator categorical 🔴 high — — — —
nmember Number of household members continuous 🟠 medium — — -0.03244 0.0181
nchild Number of children in household continuous 🟠 medium — — -0.06419 0.0366
household_head Indicator respondent is household head continuous 🟠 medium 0.0 1.0 +0.03376 0.209
occ_husband Husband's occupation category categorical 🟠 medium — — — —
chores Respondent's household chore burden/time continuous 🟠 medium — — +0.0278 0.0619
UNION_CODE Administrative union (local geographic) code categorical 🟠 medium — — — —

Response

name description type sensitivity bound lo bound hi estimate s.e.
end2_pss_score_sdidx Endline2 standardized PSS stress score continuous 🔴 high -6.0 6.0 — —

Public bounds (data-independent) sourced from: binary indicator; binary treatment indicator; plausible ceiling for years of completed schooling; plausible human age range; standardized (sdidx) index.

n = 2,254 samples.

icpsr_167601_reg_119__1 — end_ces_score_sdidx ~ treat + age_respondent + Edu_respondent + occ_respondent + income_l…

Sensitivity proposal: 🔴 high maximum across columns (advisory; reviewed manually). Reproduction grade: ✅ match (benchmark-eligible).

Features

name description type sensitivity bound lo bound hi estimate s.e.
_cons intercept continuous — 1 1 -0.3029 0.252
treat Treatment arm assignment indicator continuous ⚪ low 0.0 1.0 -0.6524 0.05
age_respondent Respondent's age in years continuous 🟠 medium 0.0 120.0 +0.00578 0.00225
Edu_respondent Respondent's level/years of education continuous 🟠 medium 0.0 25.0 -0.0003016 0.00716
occ_respondent Respondent's occupation category categorical 🟠 medium — — — —
income_loss Household pandemic-related income loss indicator categorical 🔴 high — — — —
nmember Number of household members continuous 🟠 medium — — -0.02963 0.014
nchild Number of children in household continuous 🟠 medium — — +0.0494 0.0254
household_head Indicator respondent is household head continuous 🟠 medium 0.0 1.0 +0.09704 0.176
occ_husband Husband's occupation category categorical 🟠 medium — — — —
chores Respondent's household chore burden/time continuous 🟠 medium — — -0.01512 0.0416
UNION_CODE Administrative union (local geographic) code categorical 🟠 medium — — — —

Response

name description type sensitivity bound lo bound hi estimate s.e.
end_ces_score_sdidx Endline standardized CES-D depression score continuous 🔴 high -6.0 6.0 — —

Public bounds (data-independent) sourced from: binary indicator; binary treatment indicator; plausible ceiling for years of completed schooling; plausible human age range; standardized (sdidx) index.

n = 2,220 samples.

icpsr_167601_reg_123__1 — end2_ces_score_sdidx ~ treat + age_respondent + Edu_respondent + occ_respondent + income_…

Sensitivity proposal: 🔴 high maximum across columns (advisory; reviewed manually). Reproduction grade: ✅ match (benchmark-eligible).

Features

name description type sensitivity bound lo bound hi estimate s.e.
_cons intercept continuous — 1 1 -0.5997 0.256
treat Treatment arm assignment indicator continuous ⚪ low 0.0 1.0 -0.5132 0.0629
age_respondent Respondent's age in years continuous 🟠 medium 0.0 120.0 +0.01248 0.00265
Edu_respondent Respondent's level/years of education continuous 🟠 medium 0.0 25.0 +0.01708 0.00678
occ_respondent Respondent's occupation category categorical 🟠 medium — — — —
income_loss Household pandemic-related income loss indicator categorical 🔴 high — — — —
nmember Number of household members continuous 🟠 medium — — +0.005921 0.0162
nchild Number of children in household continuous 🟠 medium — — -0.0494 0.0297
household_head Indicator respondent is household head continuous 🟠 medium 0.0 1.0 +0.05857 0.131
occ_husband Husband's occupation category categorical 🟠 medium — — — —
chores Respondent's household chore burden/time continuous 🟠 medium — — +0.05576 0.0531
UNION_CODE Administrative union (local geographic) code categorical 🟠 medium — — — —

Response

name description type sensitivity bound lo bound hi estimate s.e.
end2_ces_score_sdidx Endline2 standardized CES-D depression score continuous 🔴 high -6.0 6.0 — —

Public bounds (data-independent) sourced from: binary indicator; binary treatment indicator; plausible ceiling for years of completed schooling; plausible human age range; standardized (sdidx) index.

n = 2,254 samples.