Social learning abilities are crucial for our greatest accomplishments as a species. Computers, spaceships, and scientific theories stem from a collective process wherein each generation adds innovations to existing cultural traits, resulting in complex solutions beyond any individual's capacity. However, social learning biases can hinder progress by promoting the dissemination of current solutions while potentially restricting exploration and innovation. Indeed, in many cultures, people adopt practices and theories that are apparently ineffective. In this talk, I will give an overview of social learning biases and showcase a study in which we investigate one of these biases, the subjective appeal bias, suggesting that people tend to adopt appealing solutions and mistakenly evaluate them as effective. We employ a costly production task in which participants aim to maximize earnings by interacting with the task, which has an unknown and noisy underlying production function. Using a transmission chain design, where a group of participants play a task and pass on information to the next player in the chain, the experiment manipulates the solution space to control for subjective appeal.