Spillover Effects of Handwashing Behavior in Children

Last registered on April 20, 2022

Pre-Trial

Trial Information

General Information

Title
Spillover Effects of Handwashing Behavior in Children
RCT ID
AEARCTR-0004746
Initial registration date
November 26, 2019

Initial registration date is when the trial was registered.

It corresponds to when the registration was submitted to the Registry to be reviewed for publication.

First published
November 26, 2019, 10:39 AM EST

First published corresponds to when the trial was first made public on the Registry after being reviewed.

Last updated
April 20, 2022, 4:32 PM EDT

Last updated is the most recent time when changes to the trial's registration were published.

Locations

Region

Primary Investigator

Affiliation
Harvard University

Other Primary Investigator(s)

PI Affiliation
Harvard University

Additional Trial Information

Status
Completed
Start date
2019-10-14
End date
2020-03-31
Secondary IDs
Prior work
This trial does not extend or rely on any prior RCTs.
Abstract
This study examines (1) the impact of introducing a handwashing campaign and handsoap dispensers into primary schools in rural Bangladesh and (2) the precise process of information transfer of such practices from the school to the home and vice versa. Using time-stamped sensor technology in handsoap dispensers distributed both at the school level and at the household level to a random subset of primary school attendees, paired with a hygiene curriculum randomized at the school level, we can track the process of spillovers along two key dimensions: home-to-school and school-to-home spillovers.
External Link(s)

Registration Citation

Citation
Hussam, Reshmaan and Da Yea Oh. 2022. "Spillover Effects of Handwashing Behavior in Children." AEA RCT Registry. April 20. https://doi.org/10.1257/rct.4746-1.1
Sponsors & Partners

There is information in this trial unavailable to the public. Use the button below to request access.

Request Information
Experimental Details

Interventions

Intervention(s)
Intervention Start Date
2019-10-14
Intervention End Date
2020-03-31

Primary Outcomes

Primary Outcomes (end points)
The primary outcome variables are the daily handwashing rates at each classroom and each household and the child health measures collected at the households.

1) Daily handwashing rates are precisely measured by and extracted from the time stamped sensor embedded in each of the liquid soap dispensers. At each press of the dispenser, a sensor records the exact time when the button is pressed. The identity of the user is unknown, but we proxy for distinct users by collapsing presses that occur two or fewer seconds apart into a single press. Furthermore, since the exact time the dispenser was pressed is known, meal-time specific handwashing rates (the time most advocated by the public health campaigns) can be calculated, by asking the schools and households their reported lunch time or dinner time.

2) Child health measures on the selected subset of students and their siblings who receive liquid soap dispensers at the households will be collected. Data is collected in the form of self-reported monthly incidence of child diarrhea and respiratory illness and anthropometric measures of height, weight, and mid-arm circumference.
Primary Outcomes (explanation)

Secondary Outcomes

Secondary Outcomes (end points)
Secondary Outcomes (explanation)

Experimental Design

Experimental Design
This study employs a randomized field experiment in 26 primary schools and 858 households in rural Bangladesh over a five month period. The experimental design involves two objectives: (1) Trace the institutional spillovers of handwashing behavior from the school to the home; (2) Trace the spillovers of handwashing behavior from the home to the school. We first generate exogenous variation in handwashing behavior at the school level by randomizing the administration of a simple media information campaign across classrooms. The campaign involves a series of children’s cartoons revolving around hand hygiene, the entirety of which takes approximately fifteen minutes to air (aired via a projector in the classroom) and is done twice per week for the duration of the experiment. At the household level, the exogenous variation is generated by distributing the soap dispensers at homes to a randomized subset of students from each classroom. We vary the number of students who receive the dispensers at home by again randomly assigning each class to be either a high saturation group or a low saturation group and then randomly selecting at least 25% of the students to receive the dispenser at their homes for the high saturation classes and randomly selecting two students to receive the dispenser at their homes for the low saturation group classes.

If the households of students from treated classrooms wash more at home than the households of students from control classrooms, then cross-institutional information transfer from school to home is established. And by comparing the handwashing rates of the low-saturation classes to those of the high-saturation classes, whether students who cultivate a behavior at home are more likely to perform it at school can be tracked.
Experimental Design Details
Randomization Method
There are three randomizations in the study. The first randomization is done for the educational campaign exposure treatment and is randomized at the class level. We first randomly select one half of the classes at each school to be treatment classes and assign the rest half of classes to be control classes. The randomization is done at the partner organization's local office in Bangladesh using a statistical software on a computer. It should be noted that it is very common for two more more classes to share one classroom in rural Bangladesh. If classes do share a classroom, we check if the sharing of the class happens simultaneously at any given time period. If they do share the classroom simultaneously, and if their treatment assignments are different, then we change the treatment assignment of the control class such that both classes will be assigned as treatment classes. If they do share the classroom simultaneously, and if their treatment assignments are the same (both are treated or both are controls), we leave their treatment assignment as is.

The second randomization is done for the saturation treatment and is also randomized at the classroom level within each school. We randomly assign classes to be a high-saturation class or a low-saturation class by using a statistical software on a computer at the local office in Bangladesh. The high saturation classes will select 25% of its students to receive the dispenser at their homes and the low saturation classes will select 2 of its students to receive the dispenser at their homes.

The third randomization is done for selecting household participants and is randomized at the individual student level within each class. We randomly select students from each class who will be receiving the soap dispenser at their homes. This randomization is also done at the local office using a statistical software on a computer. We randomly sort the list of class roll number for all the students in the class, and start from the top of the list and work down until we select the right number of students from each class. If the student has missed more than 8 days of school in a given month, we moved on to the next student in the randomly sorted list of the students.
Randomization Unit
School level, information treatment: The unit of randomization is at the class level
School level, saturation treatment: The unit of randomization is at the class level
Household level: The unit of randomization is at the individual student level
Was the treatment clustered?
Yes

Experiment Characteristics

Sample size: planned number of clusters
School level: 26 schools and 156 classes
Household level: 784 households.
Sample size: planned number of observations
School level:156 classes Household level: 784 households
Sample size (or number of clusters) by treatment arms
School level I: 82 treatment classes, 74 control classes
School level II: 66 high saturation classes, 90 low saturation classes
Household level: 604 high saturation households, 180 low saturation households
Minimum detectable effect size for main outcomes (accounting for sample design and clustering)
IRB

Institutional Review Boards (IRBs)

IRB Name
Harvard University-Area Committee on the Use of Human Subjects
IRB Approval Date
2019-10-23
IRB Approval Number
IRB19-1355
Analysis Plan

Analysis Plan Documents

PAP for Spillover effects

MD5: 80b9baadab062813dd7a4576fb5a6de3

SHA1: 241a2331e7e77ae9842022d133fb2bb652585fad

Uploaded At: November 26, 2019

Post-Trial

Post Trial Information

Study Withdrawal

There is information in this trial unavailable to the public. Use the button below to request access.

Request Information

Intervention

Is the intervention completed?
No
Data Collection Complete
Data Publication

Data Publication

Is public data available?
No

Program Files

Program Files
Reports, Papers & Other Materials

Relevant Paper(s)

Reports & Other Materials