The power of giving advice

How to cite

Ajzenman, Nicolas. 2023. "The power of giving advice." AEA RCT Registry. AEARCTR-0012373. https://doi.org/10.1257/rct.12373-1.1

Download: BibTeX · RIS — for reference managers without a browser translator. Zotero and most others pick up this page's citation metadata directly.

Nicolas Ajzenman*

* Primary investigator

Registry ID
AEARCTR-0012373
DOI
https://doi.org/10.1257/rct.12373-1.1
First registered
Last updated
Status
Completed
Trial dates
2023-09-01 – 2023-12-31
Countries
Peru (Nationwide)
Registry record
http://www.socialscienceregistry.org/trials/12373

Abstract

Recent evidence has shown that giving advice can benefit the advisor (Eskreis-Winkler et al., 2018; 2019). There are several potential
reasons underlying this result. For instance, when people give advice, they assess whether they are doing what they claim should be
done and thus adjust their behavior accordingly (e.g., advising to eat healthy food to lose weight would be inconsistent with not eating
healthy food if the goal is to lose weight). Also, being asked to provide advice may raise confidence (Eskreis-Winkler et al., 2018;
Schaerer et a., 2018). The literature on mentoring has also acknowledged that mentors can substantially benefit from mentoring.
Career mentoring has been associated with career success and improved job performance and job satisfaction, among others (Eby,
Durley, Evans, and Ragins, 2006; Ghosh and Reio, 2013).

While most of the papers on giving advice have been implemented in the lab or very controlled environments, the objective of this
paper is to experiment advice-giving in a real-world decision-making context: teachers' school application decisions. From a policy
point of view, the study aims at improving teacher understaffing, a long-standing concern of the government of many governments
worldwide, as the unequal access to good quality teachers, perpetrates the student socioeconomic achievement gap and also affects
teachers' well-being, as systems become congested and many good candidates end up with no job (Ajzenman et al., 2021; Sass et al.
2012; Thiemann 2018).

Teachers in Peru apply to schools using a centralized system designed and implemented by the federal government. A short
description of the platform's work (regardless of the proposed intervention) follows. After passing a qualifying exam, teacher candidates
must log in to an online platform and select and rank their preferred vacancies, choosing as many available posts as they like.
Applications are made through an online platform containing basic school information. The government characterizes some institutions
as disadvantaged institutions (they typically serve disadvantaged populations). On the platform, these schools are labeled with an icon
that signals that the school is considered disadvantaged by the government and includes a pop-up message that says that teachers
working there can have a significant social impact on students. These schools typically include a monetary (or non-monetary) incentive
for teachers working there. If that is the case, the system also shows an icon making that attribute explicit.

The experiment consists of a motivational exercise that uses the application platform to place teachers as a dvice givers. The
government of Peru will randomize the candidates into two groups: a treatment group, in which the candidates are asked to advise
future candidates on what strategies they would recommend to teacher candidates who want to have a significant impact on
disadvantaged students' learning, and a control group, in which the candidates are asked to answer neutral questions related to the
application processes. Although the effect of advice-giving has not yet been explored in the literature in this specific context, we
hypothesize that the channels by which it worked in a different context could also apply to the context of this proposal. More
specifically, our hypothesis is that, by making explicit ways in which a teacher could have a social impact (typically, teaching
disadvantaged students), prosocial teachers - those who care about having a social effect - would prefer to have consistent behavior
and apply to schools that will maximize their impact.

We will estimate the treatment's effect on the candidates' preferences—in the probability of candidates listing high social impact
schools in their rankings -- and in the final allocation of the candidates. It is important to note that all candidates will have access to the
same information on the platform, regardless of their experimental arm. Our experiment does not affect teachers' access to information
(or vacancies). Also, our experiment is completely embedded in the official government's platform, as teachers will go through the
business-as-usual processes designed and implemented by the government almost every year. The wording of the questions is of
everyday use in Peru, not sensitive, and it was designed/approved by the ministry.
The outcomes are also business-as-usual: successful teachers will receive an offer from a school within the pool of the schools that
recruit teachers during this cycle. If any, our interventions should increase the efficiency of the process (that is, the probability of a
teacher receiving a job offer) because disadvantaged schools are typically the less-demanded schools in Peru (Ajzenman et al., 2020;
Bobba et al., 2021). In other words, the intervention does not imply any additional risk to teachers, as it is a context in which teachers
usually interact. If any, the intervention will help (if it works) to reduce congestion and thus increase the probability of teachers securing
a job.

Ajzenman, N., Bertoni, E., Elacqua, G., Marotta, L., & Vargas, C. M. (2020). Altruism or money? Reducing teacher sorting using
behavioral strategies in Peru. Journal of Labor Economics (forthcoming).
Bobba, M., Ederer, T., Leon-Ciliotta, G., Neilson, C., & Nieddu, M. G. (2021). Teacher compensation and structural inequality: Evidence
from centralized teacher school choice in Perú (No. w29068). National Bureau of Economic Research
Eby, L. T., Durley, J. R., Evans, S. C., & Ragins, B. R. (2006). The relationship between short-term mentoring benefits and long-term
mentor outcomes. Journal of Vocational Behavior, 69(3), 424-444.
Eskreis-Winkler, L., Fishbach, A., & Duckworth, A. L. (2018). Dear Abby: Should I give advice or receive it?. Psychological Science,
29(11), 1797-1806.
Eskreis-Winkler, L., Milkman, K. L., Gromet, D. M., & Duckworth, A. L. (2019). A large-scale field experiment shows giving advice
improves academic outcomes for the advisor. Proceedings of the national academy of sciences, 116(30), 14808-14810.
Ghosh, R., & Reio Jr, T. G. (2013). Career benefits associated with mentoring for mentors: A meta-analysis. Journal of Vocational
Behavior, 83(1), 106-116.
Sass, T. R., Hannaway, J., Xu, Z., Figlio, D. N., & Feng, L.

Intervention

Teachers in Peru apply to schools using a centralized system designed and implemented by the federal government. A short
description of the platform's work (regardless of the proposed intervention) follows. After passing a qualifying exam, teacher candidates
must log in to an online platform and select and rank their preferred vacancies, choosing as many available posts as they like.
Applications are made through an online platform containing basic school information. The government characterizes some institutions
as disadvantaged institutions (they typically serve disadvantaged populations). On the platform, these schools are labeled with an icon
that signals that the school is considered disadvantaged by the government and includes a pop-up message that says that teachers
working there can have a significant social impact on students. These schools typically include a monetary (or non-monetary) incentive
for teachers working there. If that is the case, the system also shows an icon making that attribute explicit.

The experiment consists of a motivational exercise that uses the application platform to place teachers as advice givers. The
government of Peru will randomize the candidates into two groups: a treatment group, in which the candidates are asked to advise
future candidates on what strategies they would recommend to teacher candidates who want to have a significant impact on
disadvantaged students' learning, and a control group, in which the candidates are asked to answer neutral questions related to the
application processes. Although the effect of advice-giving has not yet been explored in the literature in this specific context, we
hypothesize that the channels by which it worked in a different context could also apply to the context of this proposal. More
specifically, our hypothesis is that, by making explicit ways in which a teacher could have a social impact (typically, teaching
disadvantaged students), prosocial teachers - those who care about having a social effect - would prefer to have consistent behavior
and apply to schools that will maximize their impact.

We will estimate the treatment's effect on the candidates' preferences—in the probability of candidates listing high social impact
schools in their rankings -- and in the final allocation of the candidates. It is important to note that all candidates will have access to the
same information on the platform, regardless of their experimental arm. Our experiment does not affect teachers' access to information
(or vacancies). Also, our experiment is completely embedded in the official government's platform, as teachers will go through the
business-as-usual processes designed and implemented by the government.

Important: we will present the results for the full sample and also for a restricted sample following Ajzenman et al. 2023 procedure in Peru. The reason is that there are are certain districts in which teachers do not have any hard-to-staff school to apply (or all the schools they can apply are hard-to-staff schools). We don't know exactly which districts will these be in advance, but we anticipate that we will need to exclude them.

Experimental design

Teachers in Peru apply to schools using a centralized system designed and implemented by the federal government. A short
description of the platform's work (regardless of the proposed intervention) follows. After passing a qualifying exam, teacher candidates
must log in to an online platform and select and rank their preferred vacancies, choosing as many available posts as they like.
Applications are made through an online platform containing basic school information. The government characterizes some institutions
as disadvantaged institutions (they typically serve disadvantaged populations). On the platform, these schools are labeled with an icon
that signals that the school is considered disadvantaged by the government and includes a pop-up message that says that teachers
working there can have a significant social impact on students. These schools typically include a monetary (or non-monetary) incentive
for teachers working there. If that is the case, the system also shows an icon making that attribute explicit.

The experiment consists of a motivational exercise that uses the application platform to place teachers as advice givers. The
government of Peru will randomize the candidates into two groups: a treatment group, in which the candidates are asked to advise
future candidates on what strategies they would recommend to teacher candidates who want to have a significant impact on
disadvantaged students' learning, and a control group, in which the candidates are asked to answer neutral questions related to the
application processes. Although the effect of advice-giving has not yet been explored in the literature in this specific context, we
hypothesize that the channels by which it worked in a different context could also apply to the context of this proposal. More
specifically, our hypothesis is that, by making explicit ways in which a teacher could have a social impact (typically, teaching
disadvantaged students), prosocial teachers - those who care about having a social effect - would prefer to have consistent behavior
and apply to schools that will maximize their impact.

We will estimate the treatment's effect on the candidates' preferences—in the probability of candidates listing high social impact
schools in their rankings -- and in the final allocation of the candidates. It is important to note that all candidates will have access to the
same information on the platform, regardless of their experimental arm. Our experiment does not affect teachers' access to information
(or vacancies). Also, our experiment is completely embedded in the official government's platform, as teachers will go through the
business-as-usual processes designed and implemented by the government.

Important: we will present the results for the full sample and also for a restricted sample following Ajzenman et al. 2023 procedure in Peru. The reason is that there are are certain districts in which teachers do not have any hard-to-staff school to apply (or all the schools they can apply are hard-to-staff schools). We don't know exactly which districts will these be in advance, but we anticipate that we will need to exclude them.

Design details

Randomization method
By a computer
Randomization unit
Teachers
Clusters
0
Observations
10000 approximately
Arms
10000 approximately, 5000 per arm
Primary outcomes
Proportion of hard to staff schools (as defined by the government) in teachers' choice set Ranking a hard-to-staff school in the first place Ranking a hard-to-staff school in the first place or second place Ranking a hard-to-staff school in the first place or second place or third place We will analyze heterogeneity by gender. This is based on the findings of Ajzenman et al. (Journal of Labor Economics, forthcoming), in which the authors show in the same context (teachers allocation in Peru) that the effect of a similar intervention in the rural areas of Peru was successful only among male teachers (the interpretation being that female candidates are less flexible to travel longer distances). In this experiment a similar constraint could operate in Peru's countryside (although less so in Lima and Callao, which are urban areas with better infrastructure and not so longer distances).
Secondary outcomes
Being offered a position in a hard-to-staff school. This outcome is secondary because it's partially endogenous.

Cited by

No citation data available for this registration yet.

Keywords