The Case for Multilevel Analysis

Bray and Thomas (1995) appraised and criticized the predominance of comparative educational studies targeted at countries and world regions. According to the pair of researchers, this focus on geography has several shortcomings that impair the full understanding of the phenomena being studied. In order to address some of the shortcomings inherent in the geographic focus, the authors developed a model for analysis that includes three dimensions they labeled as Geographic/Locational, Aspects of Education and of Society, and Nonlocational Demographic Groups. These dimensions and their sublevels allow for a more complete comparison and form what is known as multilevel analysis. The authors described each of the seven levels of the Geographic/Locational dimension, providing examples of their usage, limitations and potential insights from the interaction with other dimensions.

The first level is World Regions/Continents. This level presents a special challenge to the researcher, as world regions may not represent true lines of dissimilarity of culture, language, history or other factors. Integration with other dimensions (e.g. Nonlocational Demographic Groups) may allow researchers to understand a region that does not follow the artificial borders found on a map, but contains the same ethnic or age group.

The second level is Countries. Studies focused on the Countries level do not address the many-faceted social, economic and political reality of smaller territories or subgroups within the larger unit being studied. Many ethnic, religious, age or gender groups may constitute subsets that could provide special insight into the phenomenon being studied.

The third and fourth levels are States/Provinces and Districts. Both levels provide additional points of reference for a fuller analysis and comparison of a region, while taking into consideration factors that are more particular to specific areas or sectors of that region. This type of analysis allows researchers to better understand a smaller territory, and to recommend policies that would fit what is actually happening in that territory, avoiding the aggregation of data that could obscure the identification of localized problems or issues.

The fifth level is Schools. This level allows researchers to investigate the culture of the institutions and helps them determine what factors play a definite role in the success or failure of that specific school in that specific district. The authors pointed out that this type of research may result in “personalized portraits” (Bray & Thomas, 1995, Schools, para. 4) and that statistical analysis at this level is manageable due to the ability of researchers to conduct random sampling.

The sixth and seventh levels are Classrooms and Individuals. These two levels deal with contrasts and similarities, and allow researchers to delve into the fields of Psychology and Sociology in order to better understand how teaching and learning take place. This, in turn, enables them to compare those findings with other classrooms or individuals in the same region or in a different area.

The authors proposed that multilevel analysis – where more than one of the dimensions and levels discussed above is utilized – can provide researchers a fuller picture of the phenomenon under study. This better, more complete picture may allow researchers to gain a greater understanding of the object of study and compare data more effectively and comprehensively. Therefore, multilevel analysis is not only advisable, but “required if appropriate answers are to be obtained” (Bray & Thomas, 1995, Multilevel Analysis, para. 3).

This article did not advance knowledge directly, but it has profound implications for how knowledge is obtained. If the authors are correct, comparative educational research that does not incorporate multilevel analysis is at least incomplete, and at most flawed. By discussing only one of the dimensions, Bray and Thomas demonstrated how limited researchers are in terms of knowledge that can be derived from the type of research that has been conducted without the application of multilevel analysis principles. When all three dimensions are used along with several levels, the possibilities for research and discovery are multiplied, and can result in a better understanding and global perspective of educational factors or issues being studied.



Bray, M. & Thomas, R. (1995). Levels of comparison in educational studies: Different insights from different literatures and the value of multilevel analyses. Harvard Educational Review, 65(3), 472-490.


Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s