Using Hierarchical Linear Modeling Methods

2178 WordsApr 27, 20169 Pages
As research across all fields and topics evolve, techniques on how to improve research endeavors, such as statistical modeling techniques, become important to utilize. Hierarchical linear modeling, similarly known as multilevel modeling, has been a statistical approach that has gained attention and improved the analysis and interpretation of research data (Osborne, 2000). Hierarchical linear modeling is a regression-based statistical analysis that considers the hierarchical (i.e., multiple levels; nested) nature of variables within a dataset. Hierarchical data is data that has an organizational structure consisting of units of groups (i.e., levels) that are clustered under a larger group. Therefore, the variables within a dataset may have mediating effects and such effects can only be explored when taking into account the clustering of groups in the data. Prior to the wide spread use of hierarchical linear analyses, data that was structured hierarchically in nature was not analyzed appropriately due to the neglect of considering the correlated, underlying relationships among independent and dependent variables. Considering the correlated parameters of a proposed relationship within a dataset is of importance because the relating parameters would indicate that the variables are not independent. Thus, violating the assumption of independence and increasing the likelihood to make a Type 1 error, in addition to resulting in biased standard errors and hypothesis tests.
Open Document