Identity and Citizenship: Methodological Considerations



Download PDF Article of Pyatt


This essay describes some conceptual and methodological challenges for future comparative scholarship on national identity and citizenship. Conceptually, linking individual attributes with emotional/ psychological awareness in civic, political, or institutional realms is a challenge and distinguishing between ethnic and civic components is important. Methodologically, we need to capitalize on analytical techniques that allow us to frame our research questions in harmony with our selected analytical technique to maximize variability that exists among individuals, groups, and nations, and that further takes into account the nested structure of these relations. Here, I briefly discuss two techniques that are well-placed for future research on national identity and citizenship: structural equation modeling with latent variables and multilevel or hierarchical linear modeling.

Keywords: national identity, hierarchical linear modeling, confirmatory factor analysis

Full Text of Article

In an increasingly globalized world, national identity continues to strongly connect individuals, nations, and supranational entities. Variation in these multiple, complex identities is present both within countries and between countries, the study of which is a useful direction for future research. This essay describes some conceptual and methodological challenges for future comparative scholarship on national identity and citizenship.

Although some predicted diminishing importance (Huntington 1990), research on national sentiment has broadened in scope and encountered new challenges. Attitudes expressing love of country, pride, sense of belonging, loyalty, and views toward political authorities, for instance, are no longer restricted to geopolitical boundaries. Linking attributes such as place of birth, cultural markers, religious faith, and citizenship with emotional/ psychological awareness in civic, political, or institutional realms is a challenge. Although nationalism, national identity, and patriotism are often considered as interchangeable concepts, modern scholars articulate its many facets (Davidov, 2009; Kunovich, 2009).

In addition, delineations between ethnic and civic dimensions is important: Assuming that national attachment means blind acceptance of authorities does not adequately address cultural traditions, religious views, and perceptions of institutions. Specifying the intersection between national identity and place is also of interest. In this regard, comparative and cross-national research offers significant potential for gains. Cross-national research on national identity suggests regional differences, such as Western and Eastern European differences (Haller and Ressler, 2006; Shulman, 2002). Others argue that variation exists not only across geographic divides, but also within specific national contexts, articulating the importance of group-based differences within geographic boundaries (Zubrzycki, 2001). Stated succinctly, accounting for variation both across and within countries is essential for advancing the field. An emerging line of inquiry investigates how national identity is taking shape in the post-socialist region among mass publics to shed light on how multiple collective identities can emerge simultaneously among different social groups. Both comparative and case study research offer important insights on national identity and citizenship. Comparative studies that build on investigations of individual cases are also fruitful for thinking about how processes of identity formation are similar or different across countries within a geographic region or region sharing an ideological legacy.

Posing a challenge for research is the question of how to reconcile complexities related to identity coupled with a comparative perspective that fundamentally rests on traditional modes of data collection and analytical techniques for data analysis. In short, our point of entry from a data and methods perspective remains the geopolitical boundary of the state. Given this, we need to capitalize on analytical techniques that allow us to frame our research questions in harmony with our selected analytical technique to maximize variability that exists among individuals, groups, and nations, and that further takes into account the nested structure of these relations. Here, I briefly discuss two techniques that are well-placed for future research on national identity and citizenship: structural equation modeling with latent variables and multilevel or hierarchical linear modeling.

Structural equation modeling with latent variables (SEM) is an analytical technique with many advantageous properties for research on national identity and citizenship including its ability to incorporate direct and indirect effects, reciprocal relations, feedback loops, measurement error, and observed and latent variables. In a full SEM with latent variables model, numerous latent exogenous and endogenous variables are constructed and evaluated using confirmatory  factor analysis (CFA). Specifying CFA shifts the investigation to latent levels, which is essential for research on national identity and citizenship as these are abstract, multidimensional constructs that are likely imperfectly measured via survey instruments. CFA is theoretically driven, as the model and the relations between individual measures indicators and the latent construct(s) are detailed in advance, providing a range of fit statistics that enables comprehensive assessment of model fit including values assessing the quality of survey items. CFA thus offers a number of methodological advantages for future research to consider (Bollen, 1989; Kaplan, 2009; Kline 2011). For national identity and citizenship research, SEM provides the tools for creating and assessing multiple, complex measures of identity and citizenship. It also affords the opportunity for including national distinctiveness in a comparative framework through the specification of the error structure and introducing nation or group-specific correlated errors of measurement.

Multilevel or hierarchical linear modeling (HLM) is appropriate when there is nesting of observational units at one level within another level of aggregation (Raudenbush and Bryk, 2002; Snijders and Bosker, 2012). These models account for micro-macro linkages through a number of model specifications articulating separate within-group and between-group regressions to empirically link parameter estimates with their respective clusters. Across various multilevel specifications, terminology of within and between groups is essential, as multilevel models allow for estimation of models including fixed and random effects associated with the levels of aggregation articulated as groups. Two are described briefly here: the typical macro-to-micro proposition and a cross-classified model. Briefly, the former proposes that individuals are nested within countries, and human sentiments are embedded within a series of processes at increasingly aggregate scales. The latter proposes complex relations among micro-level variables and variables at other levels of aggregation, which are referred to as imperfect hierarchies where the nesting structure takes into account multiple proposed nesting hierarchies including individuals, groups, communities, nations, and supranational entities. For national identity and citizenship scholarship, HLM provides an appropriate analytical technique that enables researcher to examine variation both within and between nations.

Prospects for future research on national identity and citizenship remain bright, given their continued salience in the world. Future scholars are tasked with describing existing variability in ways that show how to apply analytical techniques like SEM and HLM to pressing social issues. This scholarship remains critical for comparative research in the social sciences.


Bollen, K. (1989). Structural Equations with Latent Variables. New York: Wiley.

Davidov, E. (2009). Measurement Equivalence of Nationalism and Constructive Patriotism in the ISSP: 34 Countries in Comparative Perspective. Political Analysis, 17(1):64-82.

Haller, M. and Ressler, R. (2006). National and European Identity. Revue francais de Sociologie 47(4):817-850.

Kaplan, D. (2009). Structural Equation Modeling: Foundations and Extensions. Thousand Oaks, CA: SAGE Publications.

Kline, R. (2010). Principles and Practice of Structural Equation Modeling. New York: Guilford Press.

Kunovich, R.M. (2009). The Sources and Consequences of National Identification. American Sociological Review, 74:573-593.

Raudenbush, S. and Bryk, A. (2002). Hierarchical Linear Models, Second Edition, Thousand Oaks, CA: SAGE Publications.

Shulman, S. (2002). Challenging the Civic/Ethnic and West/East Dichotomies in the Study of Nationalism. Comparative Political Studies, 35:554–85.

Snijders, T.A.B. and Bosker, R. (2012). Multilevel Analysis: An Introduction to Basic and Advanced Multilevel Modeling, second edition. Thousand Oaks, CA: SAGE Publications.

Zubrzycki, G. (2001). ‘We, the Polish Nation’: Ethnic and Civic Visions of Nationhood in Post-Communist Constitutional Debates. Theory and Society, 30:629–68.