Identifying MAR involves statistical techniques and a deep understanding of the dataset. Researchers often use logistic regression models to examine the relationship between the probability of missingness and the observed data. If the missingness can be explained by the observed variables, then the data may be considered MAR.
