x

We are happy to announce that all trial registrations will now be issued DOIs (digital object identifiers). For more information, see here.
Using behavioral interventions to fight teacher sorting
Last registered on September 24, 2019

Pre-Trial

Trial Information
General Information
Title
Using behavioral interventions to fight teacher sorting
RCT ID
AEARCTR-0004676
Initial registration date
September 09, 2019
Last updated
September 24, 2019 7:55 PM EDT
Location(s)
Primary Investigator
Affiliation
Sao Paulo School of Economics - FGV
Other Primary Investigator(s)
PI Affiliation
Inter-American Development Bank
PI Affiliation
Inter-American Development Bank
PI Affiliation
Inter-American Development Bank
PI Affiliation
Inter-American Development Bank
PI Affiliation
Inter-American Development Bank
Additional Trial Information
Status
On going
Start date
2019-07-01
End date
2019-10-31
Secondary IDs
Abstract
Do intrinsic motivations matter more than extrinsic rewards for teacher's school choice? In this paper, we answer this question by presenting the results of a novel low-cost, large-scale intervention put in place in Peru to reduce teacher sorting (e.g., good teachers choosing the best schools). Following the experimental and theoretical literature on internal versus external rewards in decision-making, we designed an implemented a package of interventions to prime (a) intrinsic motivations (e.g., satisfaction for having high social impact in a disadvantaged school) or (b) extrinsic motivations (e.g., receiving a monetary reward for choosing a disadvantaged school) on prospective teachers, just before they have to choose the set of schools where they would prefer to be allocated. We implemented our intervention through two channels: the online platform they use to apply for schools and text messages. We analyze the effect on their likelihood of choosing disadvantaged schools.
External Link(s)
Registration Citation
Citation
Ajzenman, Nicolas et al. 2019. "Using behavioral interventions to fight teacher sorting ." AEA RCT Registry. September 24. https://doi.org/10.1257/rct.4676-2.0.
Sponsors & Partners

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

Request Information
Experimental Details
Interventions
Intervention(s)
The main objective of the intervention is to reduce teacher sorting in terms of quality (e.g., teachers primarily applying to the best schools). In order to do this, we test two different treatment arms in one region and one treatment arm in another region.

Region 1) The first arm was designed to prime extrinsic motivations (i.e., monetary or career-related rewards that receive teachers that choose "disadvantaged" schools). The second one was designed to prime intrinsic motivations (i.e., the joy of helping students of lower-income families to thrive). The idea is to test the effect of each of these arms versus a control/placebo group and also between them, so as to understand if intrinsic and/or extrinsic motivations are important to shape teachers' preferences related to school choice. One important thing to note is that, although in one arm extrinsic 'rewards' were emphasized in the messages while in the other one the emphasis was put on intrinsic 'rewards', in both (as well as in the control group) there are monetary incentives to teachers that choose 'disadvantaged schools'.

The three arms (controls and two treatments) were designed as "packages" of interventions, mainly implemented through two channels: text messages (SMS) and the online platforms prospective teachers use to apply for schools to work. The intervention started in early August, a week before teachers typically enter the platform to select schools, which happens after prospective teachers passed the national exam. The intervention covered the entire country at all levels (basic, primary and high school teachers).

Treatment 1 ("extrinsic motivations"):

1) 6 text messages (2, 4, 6, 7, 13, 15 and 20 of August): messages suggested teachers take a look in the platform to the schools that include monetary incentives AND where they could have a faster progression in their careers (note: in "disadvantaged" schools career progression is faster as an incentive for teachers to choose them). Some of them refer explicitly to the potential amount they could make in Peruvian Soles. Others were more generic, but always referring to the extrinsic/career-related incentives.

2) In the platform: an online "exercise" was implemented. Teachers have to do it before choosing the schools they want to apply. The instructions show the following text (in Spanish):

"Thanks for being part of this process. In which way do you think the monetary incentives for disadvantaged schools promote teachers' wellbeing? We would like you to take a few minutes to analyze this question and share your thoughts with us. Your opinion is very valuable for us and will only be used for informative purposes of the Ministry. The answer will not affect your allocation. Thank you!"

The text was clearly related to the extrinsic motivations to choose some schools (those with monetary incentives) and to reflect on that.

Treatment 2 ("intrinsic motivations"):

1) 6 text messages (2, 4, 6, 7, 13, 15 and 20 of August): messages suggested teachers to take a look in the platform to the schools in which they could have a more significant impact on the learning of the students. They were written in a way in which the role of a teacher as a "life changer" and the joy of helping students to thrive was emphasized.

2) In the platform: an online "exercise" was implemented. Teachers have to do it before choosing the schools they want to apply. The instructions show the following text (in Spanish):

"Thanks for choosing to be a teacher and helping to generate an improvement in students' learning! We would like you to share with us the main reasons that motivated you to become a teacher. We would like you to take a few minutes to think and share your thoughts with us, about what motivated you to choose this path. Your opinion is very valuable for us and will only be used for informative purposes of the Ministry. The answer will not affect your allocation. Thank you!"

The message was written in a way that referred clearly to the intrinsic motivations of a teacher. Moreover, "teacher" was referred as an identity (e.g., the individual chose "to be" a teacher).

Control ("Placebo):

1) 6 text messages (2, 4, 6, 7, 13, 15 and 20 of August): messages were neutral, just reminding teachers about the deadlines to select a school.

2) In the platform: an online "exercise" was implemented. Teachers have to do it before choosing the schools they want to apply. The instructions show the following text (in Spanish):

"What is your opinion about the application process to choose a school? Your opinion is very valuable for us and will only be used for informative purposes of the Ministry. The answer will not affect your allocation. Thank you!"

The message was completely neutral.

In the three arms (control and the two treatments) there are monetary incentives in place for those that choose 'disadvantaged schools'.

Region 2 (Lima+Callao):

Only T2 was implemented. This is because in those regions there were no schools with monetary incentives. Therefore, the interpretation in this case will be different than in the other regions: none of the arms have monetary incentives for teachers choosing disadvantaged schools and therefore in this region we have a more "pure" measurement of the "intrinsic motivations" intervention.
Intervention Start Date
2019-08-02
Intervention End Date
2019-09-09
Primary Outcomes
Primary Outcomes (end points)
School Choice:

1) Dummy = 1 if the teacher chose at least one "disadvantaged school" among his/her options, 0 otherwise.
2) The proportion of "disadvantaged schools" chosen by the teacher.
3) The rank in which the first "disadvantaged school" is included in the list of schools chosen by the teacher.
4) Other measures related to teachers' preferences (e.g., if she/he chose a disadvantaged school in the 1st two or three choices of the list).
5) Allocation efficiency outcomes: if the teacher was finally assigned to any school at all.

We are also planning to test heterogeneity (mainly between men and women but also in terms of pre-treatment quality measures such as teachers' grade in the national exam and regional characteristics)
Primary Outcomes (explanation)
We divided our experiment in two regions:

(a) All country except Lima and Callao:

"Disadvantaged schools": the Ministry defines specific schools that are "hard to staff", typically in the poorer areas or in rural regions. Consequently, teachers that pick these schools receive a monetary incentive.

(b) Lima and Callao:

"Disadvantaged schools": schools that are in the lowest quintile of performance in math in the 2018 standardized test (ECE 2018) in those regions.

We decided to do this because in Lima and Callao there were not "disadvantaged schools" as defined in (a). This also allows us have a "pure" measure of the effect of the "intrinsic" arm, as there are no monetary incentives in this region.
Secondary Outcomes
Secondary Outcomes (end points)

We are also planning to implement a post-experiment survey to analyze potential channels but it has not yet been designed. Probably, we are going to implement a choice experiment with two characteristics (commuting time and social impact) and analyze an outcome related to teachers' "willingness to travel" to increase the social impact of their work.
Secondary Outcomes (explanation)
Experimental Design
Experimental Design
We implemented a 3-arm design (two treatments, one control) at the individual level (teachers). The randomization was stratified only by region (26 regions). Each treatment included a combination of SMS and messages in a platform teachers use to apply for jobs (schools).
Experimental Design Details
Randomization Method
Randomization method: within each of 26 regions in Peru, we randomized the individuals through a random number. Randomization was done in office by a computer (Government).
Randomization Unit
The level of randomization is the individual (teacher), with no clusters.
Was the treatment clustered?
No
Experiment Characteristics
Sample size: planned number of clusters
There are no clusters.
Sample size: planned number of observations
15,000 individuals
Sample size (or number of clusters) by treatment arms
15,000 individuals
Minimum detectable effect size for main outcomes (accounting for sample design and clustering)
With this N we are able to detect an effect as small as of 0.04sd (no clusters).
IRB
INSTITUTIONAL REVIEW BOARDS (IRBs)
IRB Name
IRB Approval Date
IRB Approval Number
Post-Trial
Post Trial Information
Study Withdrawal
Intervention
Is the intervention completed?
No
Is data collection complete?
Data Publication
Data Publication
Is public data available?
No
Program Files
Program Files
Reports and Papers
Preliminary Reports
Relevant Papers