|
Field
Abstract
|
Before
This study plans to determine to what extent learning through games is an effective tool for financial education in schools. In collaboration with a foundation, teachers are trained to play a board game introducing personal finance knowledge with their students. I randomly assign classes for each teacher, so that they play with one class. Then, there are three different control groups: i) other class of the same teacher, other class from other teacher in the same school and other class in a different school (and different teacher). Using a survey, I measure students' financial behaviour before and after being exposed to play a board game aimed to introduce personal finance topics. Thus, I am able to observe the causal effect of playing the board game in financial literacy. Also, I measure peer effects for the classes that did were not exposed to the game. Finally, I measure the persistence of these effects one year after the intervention.
|
After
This study aims to evaluate the effectiveness of learning through games as a tool for financial education in schools. In collaboration with the Foundation Entrepreneur, teachers are trained to use a board game called Financity to introduce personal finance concepts to their students. Each participating teacher is randomly assigned to play the game with one of their classes, while another class taught by the same teacher serves as a control group. Students' financial knowledge and behaviors are measured through surveys conducted before and after the intervention. This design allows me to estimate the causal effect of playing the board game on financial literacy.
|
|
Field
Trial Start Date
|
Before
May 14, 2025
|
After
July 20, 2025
|
|
Field
Trial End Date
|
Before
October 31, 2025
|
After
April 30, 2026
|
|
Field
Last Published
|
Before
May 14, 2025 10:43 AM
|
After
July 18, 2025 08:32 AM
|
|
Field
Intervention (Public)
|
Before
I study the effect of a board game called Financity as an instrument to increase financial education in schools. The game provides knowledge on: Balancing money with well-being, responsible use of money, preparation of budget, risks of over-indebtedness, responsible purchase, relationship with financial entities, importance of saving. It can be played by 3 to 6 players, from 12 years old, with a duration of approximately 60 minutes.
We train teachers in-person how to play the game and teach it to their students. Also, continuous support is provided during the intervention. The teachers then play with their assigned class for approximately 4 months. The other classes do not get to play and keep having the normal curriculum, which includes financial education in theory within certain subjects, but not in practice.
|
After
I study the effect of a board game called Financity as an instrument to increase financial education in schools. The game provides knowledge on: Balancing money with well-being, responsible use of money, preparation of budget, risks of over-indebtedness, responsible purchase, relationship with financial entities, importance of saving. It can be played by 3 to 6 players, from 10 years old, with a duration of approximately 60 minutes.
The Foundation trains teachers in-person how to play the game and teach it to their students. Also, continuous support is provided during the intervention. I randomly assign one treated class (to play) and one control class (not to play) to each teacher. The teachers will then play with their assigned class for approximately 4 months. The other classes do not get to play and keep having the normal curriculum, which includes financial education in theory within certain subjects, but not in practice.
|
|
Field
Intervention Start Date
|
Before
May 14, 2025
|
After
July 20, 2025
|
|
Field
Primary Outcomes (End Points)
|
Before
Debt and saving levels
Intention to save index
Purchasing behavior
Financial autonomy index
Financial savviness
Financial literacy
|
After
Money for happiness
Money vs. well-being
Willingness to pay for well-being
Saving index
Financial Behavior index
Financial Knowledge index
Financial Attitude index
|
|
Field
Primary Outcomes (Explanation)
|
Before
Intention to save index comes from 21 questions with answers ranging from extremely negative to extremely positive (7point scale) on attitudes toward behavior, subjective norms, perceived control over one's own behavior, attitudes about the possible effects of behavior, motivation for following subjective norms and perceived benefits of oneäs own behavior.
Financial autonomy index comes from 15 questions with answers agree or disagree on reflexive autonomy, emotional autonomy and functional autonomy.
Financial savviness is built from four binary outcomes: keeping a budget, saves before buying something that cannot be afforded, compares prices and bargains before shopping.
Financial literacy comes from 3 questions from Lusardi and Mitchell (2014) on interest rate, inflation and risk diversification.
|
After
All the main outcomes come from the survey that I developed with respect to the main goals of the game.
Money matters for happiness: This variable is constructed from the question “Think of someone close to you who is very happy. How much do you think money matters in their happiness?". Subjects are offered five response categories (A lot; Quite a bit; Some; A little; Not at all). Responses are recoded on a 0–4 scale, where 0 corresponds to Not at all and 4 to A lot and then standardized as mentioned above. An interesting measure that might be studied later is the binary outcome, as a robustness check.
Money vs. well-being: This variable builds from the question “Imagine you had to pick just one of these. Which one would you choose?". The question has five possible answers (Having a ton of money, but feeling not happy at all; Having a lot of money, but feeling only a little happy; Having some money, and feeling somewhat happy; Having a little money, but feeling very happy; Having no money, but feeling super happy). Given there is an unclear a priori variation, I choose to give a numerical value to each answer: 1 being money over happiness and 5 being happiness over money. Thus, I will be able to measure variation from different initial scenarios. Still, it remains interesting to create a dummy variable for each preference (money for 1 and 2, balance for 3 and happiness for 4 and 5), which might be done as a robustness check post hoc. Also, the variance (or distance to the balanced answer (3)) of the variable might be studied to observe if students tend to approach to balance between money and well-being.
Willingness to pay for well-being: This variable comes from the question “Imagine you are having fun doing your favorite thing. If someone asked you to stop doing that activity to do something else you don’t like, how much money would they have to give you?". The possible answers are (in parentheses how I code them): I would stop for any amount of money; I would stop for \$10,000; I would stop for \$50,000; I would stop for \$100,000; I wouldn’t stop my favorite activity, even for a lot of money. This variable will be coded to use a logarithmic scale. Therefore, respectively the values are: 2,000; 10,000; 50,000; 100,000; 500,000. This way, each step of answer represent a 5-time increase. Again, I standardize with respect to the control group.
Saving index: This variable is composed by 6 questions from the survey that will be added to build this index. First, if their main use of their work money is “save it", the variable takes the value 1, otherwise 0. Second, similarly, if their main use of money from their parents is “save it", the variable takes the value 1, otherwise 0. Third, if they answer “yes" to “Do you set saving goals to buy something in the future?" the variable is equal to one, otherwise 0. Fourth, if they answer “yes" to “Do you currently have any savings?" the variable is equal to one, otherwise 0. Fifth, if they answer “Very safe" or “Safe" to “How do you feel when you get to save some money?" the variable takes the value 1, otherwise 0. Last, if they answer “Somewhat easy" or “Easy" to the question “How difficult is it for you to save money regularly?" the variable is equal to 1, and 0 otherwise. Thus, the saving index takes values from 0 to 6.
Financial Behavior index: This variable is composed by 6 questions from the survey, and will be added to build the index. The first variable takes the value 1 if the student answers “yes" to the question “Do you plan your spending before using your money?". The second is equal to 1 if the subject answers “yes" to the question\lq \lq Do you keep track (on paper, app, memory) of your income and expenses?". The third is equal to one when replying “yes" to \lq \lq Do you feel prepared to face an unforeseen event?". The fourth correspond to the value 1 when answering “yes" to “Do you compare prices (or look at different brands/models) before buying something?". The fifth takes the value 1 for students replying “yes" to \lq \lq Do you negotiate the price?". Finally, those repaying on time what they borrowed take the value 1, and 0 otherwise. Then, the behavior index takes values from 0 to 6.
Financial Knowledge index: This is constructed by 3 questions from the survey, added to build the index. First, a variable that takes the value 1 for those answering “yes" to the question “Do you know what “debt” or “loan” means?". Second, using the question \lq \lq Which of the following statements best describes what insurance does?" the variable takes the value 1 for those students answering correctly. Finally, an indicator equal to 1 for students replying correctly to “Which of the following statements best describes what an investment does?", and 0 otherwise.
Financial Attitude index: This index builds on 3 questions from the survey. All of them will take the value 1 for those students replying “agree" or “strongly agree" to the following statements: "Planning my budget helps me make better decisions", "It is important to pay debts on time" and "It is important to learn about finances from a young age". This index takes values from 0 to 3.
|
|
Field
Experimental Design (Public)
|
Before
The universe of interest is restricted to students of signed-up teachers from three regions: Maule, Biobío and Ñuble. The final universe included 20-30 schools. The restricted universe is stratified class size and cohort.
|
After
The universe of interest is restricted to students of signed-up teachers from three regions: Maule, Biobío and Ñuble. They will be approximately 60 teachers coming from approximately 20-30 schools. In average, each class has 25 students, therefore there will be approximately 3.000 students (1,500 students in the control group and 1,500 in the treated group).
|
|
Field
Randomization Method
|
Before
Randomization done in office by computer
|
After
Randomization done in office by computer.
|
|
Field
Was the treatment clustered?
|
Before
Yes
|
After
No
|
|
Field
Planned Number of Clusters
|
Before
2 classes per teacher, 120 classes
(2 classes per teacher + 1 class of other teacher + 1 class from different school, 240 classes.)
|
After
60 teachers
|
|
Field
Planned Number of Observations
|
Before
Each class has an average of 25 students. Each teacher has two classes, the total sample could reach 3,000 students (60 teachers × 25 students × 2 classes), with 1,500 in the treatment group and 1,500 in the control group. Additionally, teachers may be asked to invite a colleague to complete the survey and administer it to another class, creating a secondary control group in which neither the teacher nor the students receive the intervention (1,500 students). Additionally, a third control group with classes from a different school in which neither the teacher nor the students nor anyone at the school receive the intervention (1,500 students).
|
After
Each class has an average of 25 students. Each teacher will have two classes (one treated and one control), the total sample could reach 3,000 students (60 teachers × 25 students × 2 classes), with 1,500 in the treatment group and 1,500 in the control group.
|
|
Field
Sample size (or number of clusters) by treatment arms
|
Before
60 treatment classes (1,500 students) and 60 control classes (1,500 students) + possible 60 second control classes (1,500 students) + possible 60 third control classes (1,500 students)
|
After
60 treatment classes (1,500 students) and 60 control classes (1,500 students)
|
|
Field
Power calculation: Minimum Detectable Effect Size for Main Outcomes
|
Before
Power calculations were performed with the following parameters: significance level of 0.05, statistical power of 0.8, minimum detectable effect of 0.1 SDs, R2 of the outcome equation of 0.1, intra-cluster correlation of 0.05 and a sample size of 25 students per class. Under these assumptions, 52 classes were required, 26 in each treatment arm. (For the case with the second control group, 69 classes were required, 23 in each group). (For the case with the second and third control group, 80 classes were required, 20 in each group).
|
After
Power calculations were performed with the following parameters: significance level of 0.05, statistical power of 0.8, minimum detectable effect of 0.1 SDs, R2 of the outcome equation of 0.1, intra-cluster correlation of 0.05 and a sample size of 25 students per class. Under these assumptions, 52 classes were required, 26 in each treatment arm.
|
|
Field
Intervention (Hidden)
|
Before
The foundation will conduct in-person training sessions for the enrolled teachers (one day per region). During this session, teachers will complete the baseline survey to ensure a higher response rate. After the training, they will continue with an online course that provides further details about the game, along with virtual support from foundation staff. Teachers will also receive the game so they can begin implementing it in their classrooms. The school will provide the remaining necessary set of games to have a full class playing (because one set is for maximum 6 players).
Some teachers will come from the same school, meaning approximately 20-30 schools will be participating. The students involved will be from 7th to 12th grade (ages 12 to 18: 12 is the minimum age to play). Each teacher may instruct a certain topic (math, history, geography) to multiple classes/cohorts, so they will be asked to specify which classes/cohorts they teach and how many students they have. Then, each teacher's class will be randomly assigned to either the treatment group (where students play the game) or the control group (where students do not play), ensuring balanced groups and preventing teachers from favouring specific classes.
Before playing the game for the first time, teachers will make their students complete the baseline survey. The implementation phase will last approximately 4 months, with a tentative completion date around October. This timeline allows students ample time to play the game, avoids survey administration during winter (when absenteeism is higher), and takes place before the tournament scheduled for October (and before the end of the academic year in December). The end of the implementation phase will be marked by the endline survey, which both teachers and students will complete simultaneously. The survey will be identical to the baseline, except for the demographic module that will not be included.
|
After
The foundation will conduct in-person training sessions for the enrolled teachers (one day per region). During this session, teachers will complete the baseline survey to ensure a higher response rate. After the training, they will continue with an online course that provides further details about the game, along with virtual support from foundation staff. Teachers will also receive the game so they can begin implementing it in their classrooms. The school will provide the remaining necessary set of games to have a full class playing (because one set is for maximum 6 players). Teachers from 3 regions (Maule, Biobío and Ñuble) signed-up to participate in the training. I only use teachers that have classes from 4th grade and higher.
Some teachers will come from the same school, meaning approximately 20-30 schools will be participating. The students involved will be from 4th to 12th grade (ages 10 to 18: 10 is the minimum age to play). Each teacher may instruct a certain topic (math, history, geography) to multiple classes/cohorts, so they will be asked to specify which classes/cohorts they teach and how many students they have. Then, each teacher's class will be randomly assigned to either the treatment group (where students play the game) or the control group (where students do not play), preventing teachers from favouring specific classes.
In a setting with a larger sample, all teachers would be randomly assigned to one control and one treated class. However, some teachers in the sample only teach a single class. Due to requirements set by the Foundation, each participant teacher must play with at least one class in order to complete the training and certification. As a result, all one-class teachers are assigned to the treatment group.
For teachers with at least two classes, one class is randomly assigned to the treatment group and another to the control group. Since each school has at least two participating teachers, some classes are taught by more than one teacher (i.e. both math and history teachers participate and share a class). In cases where a class is assigned to both treatment and control (due to different teachers), I resolve the conflict by reassigning one teacher to another class. For instance, if teacher 1 is assigned class A as treated and class B as control, and teacher 2 is assigned class B as treated and class C as control, I reassign teacher 2 to class D as treated while keeping class C as control.
After this correction process, each participating class is uniquely assigned to either the treatment or control group. Every teacher is assigned one treated class, and those with more than one class are also assigned one control class. Ideally, one-class teachers would also have been randomly assigned to treatment or control (rather than all to treatment), but this was not feasible due to the constraints that were mentioned above. Consequently, some classes are treated by two teachers.
Before playing the game for the first time, teachers will make their students complete the baseline survey. The implementation phase will last approximately 4 months, with a tentative completion date around October. This timeline allows students ample time to play the game, avoids survey administration during winter (when absenteeism is higher), and takes place before the tournament scheduled for October (and before the end of the academic year in December). The end of the implementation phase will be marked by the endline survey, which both teachers and students will complete simultaneously. The survey will be identical to the baseline, except for the demographic module that will not be included.
|
|
Field
Did you obtain IRB approval for this study?
|
Before
No
|
After
Yes
|
|
Field
Secondary Outcomes (End Points)
|
Before
Aspirations and expectations of going to university
Spillover effects (primary outcomes for control group)
School performance
School attendance
Gender differences
|
After
Money for happiness alternative
Money vs. well-being alternatives
Saving amounts
Borrowing money
Purchasing behavior
Impatience
Aspirations and expectations of going to university
School performance
School attendance
Gender and age heterogeneity
|
|
Field
Secondary Outcomes (Explanation)
|
Before
|
After
Money for happiness alternative: The binary outcome related the question on how much money matters for happiness, providing a extensive margin to whether money matters or not for happiness.
Money vs. well-being alternatives: indicator variables that were previously mentioned from question regarding money vs well-being. These will be 3 indicator variables, one for those choosing money, one for choosing balance and other for choosing well-being.
Saving amounts: there is a question about the amount of savings for those students that answer “Yes" to having savings. This intensity factor might be interesting to observe.
Borrowing money: question about having borrowed money in the last 3 months (yes or no).
Purchasing behavior: there is a question regarding the main use of money, giving different options such as clothes/shoes, food/drinks, video games/entertainment (movies, concerts), transportation, gifts or others. In that same line, a following question asks about the main motivation when deciding to spend money and about having regretted a purchase.
Impatience: “I prefer to spend now rather than think about the future" (agree or disagree) is not the main interest of the study, but could also be included.
Aspirations and expectations of going to university (only for 9th-12th grade): "How much do you want to continue your studies in higher education (technical/professional institute or university) after finishing school?" and "How likely is it that you will continue your studies in higher education (technical/professional institute or university) after finishing school?"
School performance: I will be able to see school grades from the participating classes and observe changes.
School attendance: I will be able to see school attendance from the participating classes and observe changes.
Gender and age heterogeneity: I will know age, cohort and gender of the students to observe whether there exist some differences of the main outcomes with respect to this variables.
|