Latent variables are variables that are not directly observed but are inferred from other variables that are measured. They represent underlying factors that can influence observable outcomes. For example, in psychology, a person's intelligence may be considered a latent variable, as it cannot be measured directly but can be assessed through tests and behaviors. In statistical modeling, latent variables help explain relationships between observed data. They are often used in techniques like factor analysis and structural equation modeling to identify patterns and reduce complexity in data. By understanding latent variables, researchers can gain insights into hidden influences on measurable phenomena.