Addressing Dropout Behavior Through Behavioral Science

Research | Behavioral Science | Intervention Design

Industry

Education

Year

2022

Role

Applied Behavioral Scientist

The Challenge

Education is a privilege and opportunity highly valued in the US. In some circumstances, the costs of attending school, such as financial, social, or psychological, can outweigh the perceived immediate benefits for individuals. Schools not only impart valuable knowledge and skills, but are also institutions that shape students’ identity, values, and potential. Thus, education is key in exploring opportunities to manifest positive future social, economic, and health outcomes for individuals.

In the United States, dropping out contributes negatively to society and brings no return on investment for students and interferes with the development of the country. These consequences mark the demand for more effective dropout prevention interventions.


My Role

Methodology

 

Discovery

Behavioral Barriers of Students

At-risk students encounter several psychological and contextual barriers that either create friction against the decision to attend school or provide no perceived benefits to them. In terms of barriers that create friction, students may not attend school due to having a scarcity mindset, identity threat, social pressure, or low academic performance, which leads to attentional and anchoring bias. Students may also choose not to attend school due to lack of perceived benefits, which may include having low engagement and motivation, feeling like immediate benefits outweigh the long-term, or not relating to the school and its values.

★ Insights 

Behavioral Barriers of Teachers

Teachers play a critical role in the students’ engagement with the school, which can affect their academic performance, attendance, and graduation completion. However, teachers may have their own set of biases when dealing with students who are at risk. Teachers may have confirmation bias, lack of information, ostrich effect, or low self-efficacy. While teachers’ behavioral barriers are examined, the intervention will address students’ barriers. This will allow random assignment since students in different treatments may have the same teacher. By only focusing on students’ behavioral drivers, we can manipulate the conditions without interfering with the teachers’ behaviors.

★ Insights 

Predictors of Dropout Behavior

While there are many behavioral barriers involved that hinder students from attending school, the literature proposes student motivation and growth mindset are observed to be contributing factors of dropping out of school. Given the short duration of the intervention, that it targets students in their first year of high school, and graduation rate will not be the main outcome variable for the intervention, it is essential to determine the predictor variables for dropout behavior using existing evidence.

 

Type of Behavior Change

Based on Fogg’s Behavior Grid, as shown, the type of behavior that attending school can be categorized as is a path behavior since the behavior change is intended to be long-lasting and not for a short duration. Since students will have been attending school already before high school, the behavior would be a familiar one where the change aims to strengthen the intensity of the behavior through increasing the school attendance rate.

Type of Behavioral Constraint 

Fogg states that path behaviors are the most difficult type of behavior to change since they require shifting the target individual’s identity (social opportunity) or lifestyle (environment). This aligns with the analysis, with a total feasibility score of 1.5 out of 3.8. However, it suggests there is potential for a technical opportunity in addressing school attendance.

Ideation

Here, I conducted a prioritization analysis of the student’s behavioral barriers along with associated design implications and ideation.

Selecting the treatments

After doing an analysis, three possible behavioral science treatments have been selected to implement for this prevention program based on the priority scores.


All three behavioral science treatments will consist of elements that address the student’s identity, since identities are what motivate people to act in accordance with their goals. The intervention will include a treatment inspired by neuroscience since brain architecture can impact the learning process, and students’ ability to learn can affect their motivation and growth mindset.

Type of Motivation

Research recommends cultivating a range of motivations in at-risk students. An overview of the different motivation types based on Organismic Integration Theory and how it relates to the treatment.

Theory-Driven Elements

Student motivation and growth mindset are predictors of school dropout behavior. These two factors have parallels with elements in behavior change models, such as the Theory of Planned Behavior, Fogg Behavior Model, and the COM-B Model.


Considering these theories, the behavioral science treatments are likely to influence students’ beliefs and motivation, while the executive function training is likely to affect actual control, ability, or capacity (growth mindset). While growth mindset is the belief in ability, improving actual physical control through executive function training can reinforce this belief since improving brain architecture leads to tangible results.

Implement

Hypotheses

Given the 2 x 4 factorial design, the intervention has six hypotheses to encompass the multiple treatment conditions.

Procedure

Intervention Design

An overview of the intervention design, with the stages: Participant Recruitment, Pre-Intervention, Intervention, and Post-Intervention.

Feasibility

An overview of the estimated costs of the intervention based on a sample size of 66 per treatment condition.

Treatment Designs

Treatment #1 : Teacher’s Expectations

Treatment #2 : Student’s Precommitment

Treatment #3 : Loss Framing

Treatment 4: Executive Function Training

Feasibility

An overview of the estimated costs of the intervention based on a sample size of 66 per treatment condition.

Evaluate

Measurable Outcomes

The intervention mainly targets the behavioral drivers: low motivation, present bias, and commitment bias. Motivation will be assessed using the same Student Motivation questionnaire. Students who report motivations as “identified/introjected” or “external” will be classified as at-risk. Present bias will be measured using the Hypothetical Monetary Present Bias Elicitation. Lastly, goal commitment will be measured using an adapted version from another study, which can reveal the level of commitment students feel towards their goal. All groups will be assessed for all outcome variables, but some are more relevant for assessing efficacy.

Analysis

Given the 2x4 factorial design, two-way ANOVA will be used to see if there was a statistically significant interaction between the effects of Executive Function Training and the Behavioral Science treatments on the outcome variables (school attendance, student motivation, growth mindset, present bias, commitment bias, and grades). We will use regression analysis to see how the Executive Function training interacts with the other treatments and how student attendance changes with the different conditions. Regression analysis can also reveal the socio-economic characteristics of the students that are relevant for the different treatments.

Advantages

Given the support of theories, behavioral elements, comprehensive factorial design, cost-effectiveness, and feasibility, this intervention has the potential to increase school attendance rates for students.

Limitations

Next Steps

Future Considerations

Conclusion

By targeting the two variables examined to be predictors of dropout behavior: student motivation and growth mindset - the intervention design ensures it can address dropout behavior while maintaining a feasible timeline that does not require four years. These two variables align with the Fogg Behavior Model, which suggests targeting both motivation and ability influences behavior. It overlaps with the Theory of Planned Behavior, which proposes addressing beliefs (behavioral, normative, control) and actual control to target intention and behavior.

The constructs of student motivation and growth mindset also align with the core components of the COM-B model: motivation and capability. The treatment conditions have elements that address various motivation types and the students’ identity. The potential impact of the intervention is strengthened due to targeting both individual psychology and brain architecture.

Given the support of well-grounded theories, behavioral elements, a comprehensive factorial design, cost-effectiveness, and feasibility, this intervention has the capacity to effectively increase school attendance rates for high school students and ultimately lower dropout rates.