Parental Motivated Reasoning

Lab-in-the-field experiment run online and at Norwegian secondary schools (7th grade).

2024-10-09

2025-05-31

PARENTS:
I) Prior beliefs (mathematics task)
II) Signals (mathematics task)
III) Posterior beliefs (mathematics task)
IV) Prior beliefs (sharing task)
V) Signals (sharing task)
VI) Posterior beliefs (sharing task)

CHILDREN:
i) Performance in the mathematics task
ii) Choices in the sharing task

PARENTS
I) Prior beliefs (mathematics task): perceived probability of the child’s performance in the mathematics task being in the top half (0-100%)
II) Signals (mathematics task): Each signal can be positive (“Your child is in the top half.”) or negative (“Your child is not in the top half.”) and is accurate with p=2/3. Signals are independent and drawn with replacement.
III) Posterior beliefs (mathematics task): perceived probability of the child’s performance in the mathematics task being in the top half (0-100%)
IV) Prior beliefs (sharing task): perceived probability of the child’s choices in the sharing task being in the top half (0-100%)
V) Signals (sharing task): Signal can be positive (“Your child is in the top half.”) or negative (“Your child is not in the top half.”) and is accurate with p=2/3. Signals are independent and drawn with replacement.
VI) Posterior beliefs (sharing task): perceived probability of the child’s choices in the sharing task being in the top half (0-100%)

CHILDREN:
i) Performance in math quiz: this variable measures the number of correct questions on the mathematics quiz.
ii) Choices in sharing task: this variables measures the aggregate share of tokens given to the other child across the four dictator games.

For parents, I am interested in examining belief updating after receiving positive or negative signals in each domain.

PARENTS:
I) Reasoning signals and guesses explanation (mathematics task)
II) Reasoning signals and guesses explanation (sharing task)
III) Ex-post rationalization by parent (effort, importance for child's outcomes, reflective of parent) (mathematics task)
IV) Ex-post rationalization by parent (effort, importance for child's outcomes, reflective of parent) (sharing task)

The variables listed above are secondary variables in the sense that I will use them to learn more about the variation in my primary outcomes of interest.

PARENTS:
I: Reasoning signals and guesses explanation (mathematics task): this is an open-ended question, asking parents to write a short explanation how they would explain to their child what signals they received and how they made their guesses, i.e., primary outcome III.
II: Reasoning signals and guesses explanation (sharing task): this is an open-ended question, asking parents to write a short explanation how they would explain to their child what signals they received and how they made their guesses, i.e., primary outcome VI.

CHILDREN:
i: Performance task (mathematics): this variable measures the number of correct questions on the mathematics task, taking values from 0-15.
ii: Sharing task: this variable measures the number of tokens allocated to the other child across four dictator games that vary 1) in the initial endowment of tokens that are allocated to the child (dictator) and the other child (receiver), as well as 2) in the possibility whether the dictator can take away tokens from the initial token endowment of the receiver. The aggregate number of tokens across the four dictator games that can be allocated to the other child ranges from minus fifteen to forty tokens where a negative number of tokens indicates that the child has taken away tokens from the initial token endowment of the other child.

ADMINISTRATIVE DATA OUTCOMES:
I will be able to link the experimental data with administrative outcomes (past and
future) from the children and parents. These variables will not be outcomes in the sense that
I expect my treatment variation to influence these variables, but I do plan to examine the relationship between parental belief updating and child outcomes in the administrative data. These variables are discussed more in the attached document describing the experimental design.

My experimental design is described in detail in the attached PDF.

Not available

The randomization will be implemented by the computer program oTree with an equal probability (p=1/6) of being assigned to each of the six treatment conditions.

Individual: Randomization is implemented at the level of the parent.

No

The study plans to collect a sample size of approximately 750 children-parent pairs. Since parents and children complete the survey and experiment separately, this implies that I will collect data from some children who are not part of a matched pair. Therefore, I expect that a fraction of parents will not participate (perhaps 25%). The child who has already participated will still be in the sample, but this data from children whose parents do not participate will not form part of my main dataset. Therefore, I will continue to collect data until 750 parent-child pairs participate.

Since I expect this non-participation to take place before exposure to any treatment, I expect that it will be orthogonal to my treatment conditions.
My anticipated effective sample size will therefore be 750 parent-child pairs, with 1/6 randomized into each of the treatment conditions.

1500

N=375 in mathematics task:
- N=125 in NoInfo
- N=125 in Info – Transmission
- N=125 in Info – Outcomes

N=375 in sharing task:
- N=125 in NoInfo
- N=125 in Info – Transmission
- N=125 in Info – Outcomes