ECPR Winter School
University of Bamberg, Bamberg
22 February - 1 March 2019




WD219 - Age Period Cohort Analysis

Instructor Details

Instructor Photo

Roula Nezi

Institution:
GESIS, Leibniz

Instructor Bio

Roula Nezi is a Senior Researcher at the GESIS-Leibniz Institute for the Social Sciences. Previously, she was a postdoctoral fellow at the University of Konstanz.

Her research interests lie in public opinion and political attitudes, political representation, and political accountability. She is also interested in political methodology, including comparative analysis and survey design.

In recent years, Roula has conducted research in a wide range of comparative European projects on political representation, political behaviour, and political elites. She is currently leading a project on the support of authoritarian beliefs and the vote for rightwing populist parties among young cohorts in Europe.


Course Dates and Times

Monday 25 February - Friday 1 March 2019
14:00 to 17:30 (finishing slightly earlier on Friday)
15 hours over 5 days

 

Prerequisite Knowledge

Note from the Academic Convenors to prospective participants: by registering to this course, you certify that you possess the prerequisite knowledge that is requested to be able to follow this course. The instructor will not teach again these prerequisite items. If you doubt whether you possess that knowledge to a sufficient extent, we suggest you contact the instructor before you proceed to your registration.

Participants are expected to have a good understanding of multivariate linear regression and of multilevel (hierarchical) models before they take this class. Good knowledge includes fixed and random effects and understanding of interaction terms. If you don't have already a solid grasp of multilevel modelling I recommend you first to take a class in multilevel (hierarchical) models. This class (Age-Period-Cohort models) will not teach you multilevel modelling, but rather how to run a Hierarchical (HAPC) model for estimating Age-Period-Cohort effects. This class will also provide you the tools and the syntax needed to apply those models. I personally use R and I will strongly encourage the use of R in this class, though, some examples will be given in STATA as well. Please make sure that you bring your own laptop in class with an up to date version of R.

Short Outline

This course is designed to give participants a deep knowledge of Age-Period-Cohort (APC) Models. Those are models that help social scientists to examine whether time varying changes in social and political phenomena are due to the biological and social process of aging (age effects), social or economic factors, for example an economic crisis (Period effects) or due to the unique experience of a group, for example children socialised under an authoritarian regime (cohort effects). What makes the analysis of those effects interesting and methodologically challenging is the so called “identification problem”; The age-period-cohort conundrum refers to the problem of separating the effects of age, periods and cohorts. Upon completion of the course participants will have an understanding of the identification problem in the study of aging and cohort analysis and an understanding of the different methods developed to overcome the identification problem. Participants will also be familiar with the limitations of these methods and will also be able to know which method to use depending on their own research question.

In this class special attention is given to the Hierarchical Age-Period-Cohort-Model (HAPC). In addition to the methods behind the study of APC models you will also learn how to conduct your own APC models in R. The overall goal of the course is to provide participants with a general overview of the APC models, familiarise participants to the different methods of studying age-period-cohort effects, and to teach participants how to produce and interpret their own findings.

Tasks for ECTS credits

  • Participants attending the course: 2 credits (pass/fail grade). The workload for the calculation of ECTS credits is based on the assumption that students attends the course and fully participates in in-class activities. Carries out the necessary reading and/or other work prior to, and after, classes.
  • 3 ECTS Credits - Graded, take home exam distributed on Friday, due Monday (after class)
  • 4 ECTS Credits - same as above and a research paper due 2 weeks after the class using your own data and APC method.
Long Course Outline

The aim of this course is to introduce students to the major methodological tools related to the study of ageing and cohort analysis. In times of major political events studies of Age-Period-Cohort are of significant importance; APC models help us understand time varying social and political changes, and identify the trends and the dynamics of those trends among the population. This course aims at providing students with the guidelines on how to conduct such analysis. It will first start by giving an overview of the importance of studying aging, cohorts and time and the importance of the identification problem. This will be followed by an introduction to the key methods and models in the study of APC models and by the newest developments in the study of Age-Period-Cohort models. By the end of this course students will be able to understand the key concepts and theories behind APC models, to apply different statistical methods related to the study of Age-Period-Cohorts and to produce their own findings. Though we will not cover those topics extensively, participants will be introduced to the concept of priors and Bayesian statistical inference for APC models.

The course is structured in the following way:

The first class will introduce participants to the study of Age-Period-Cohorts and on why APC models are important for the study of social change. We will continue by discussing the identification problem in the study of Age-Period-Cohort and on how different statistical methods can help us overcome this problem. In our first session in the lab you will learn how to get your dataset ready for APC analyses. This is a very important first step in the study of Age-Period-Cohort models.

At the second class we will examine the first generation of models in the study of APC models. Those include the use of age-by-time period contingency tables, the use of proxy variables in the study of APC models, and the coefficient constraint approach. All these methods will be practised in the lab session.

The third class will introduce participants to the Intrinsic Estimator (IE). The Intrinsic Estimator is a method to estimate and capture the effects of age, period, and cohorts. During the lecture we will talk about the advantages of this method the statistical model and the criticism to the Intrinsic Estimator. At the lab session we will practice on Intrinsic Estimator models.

The fourth class will introduce students to the Hierarchical Age-Period-Cohort Model. Hierarchical modelling is considered the most up-to-date method in the study of the APC models. Those models take into consideration the multilevel structure of the data; individuals are nested in a cross-classification of periods and birth cohorts. In the lab we will learn how to run a Hierarchical Age-Period-Cohort-Model.

On our last day we will talk about the use of priors (Bayesian statistical inference) in estimating Age-Period-Cohort effects and about the Non-parametric Generalised Additive models (GAM). Generalised Additive Models allow researchers to plot nonparametric smoothed curves for the effect of year or of the cohort. This approach help us observe differences among the estimated effects and differences of those effects among different countries.

Day-to-Day Schedule

Day 
Topic 
Details 
Day 1Why Age-period-cohort (APC) models?

90 min lecture: Why it is important to study APC models and the importance of the identification problem                       

90 min lab: First illustration and how to prepare your data

Day 2The first generations of APC models and the Constrained Generalized Linear Model

90 min lecture: Descriptive APC analysis, the proxy variable approach, Constrained Generalized Linear Model (CGLM)                             

90 min lab: Empirical illustration in the lab

Day 3Intrinsic Estimator

90 min lecture: The theory behind the Intrinsic Estimator                     

90 min lab: Empirical Illustration of the Intrinsic Estimator (IE) in the lab

Day 4Hierarchical Age-Period-Cohort models (HAPC)

90 min lecture: The basics and advanced HAPC analyses                 

90 min lab: Illustration of the HAPC model in the lab

Day 5HAPC Continued and Generalised Additive Models

90 min lecture: Criticism of the HAPC model and the inclusion of priors, Non parametric Generalised Additive models                                                                   

90 min lab: Empirical Illustration of GAM models

Day-to-Day Reading List

Day 
Readings 
Day 1

Ryder, N.B. (1985) The Cohort as a Concept in the Study of Social Change. In: Mason W.M., Fienberg S.E. (eds) Cohort Analysis in Social Research. Springer, New York, NY

Riley, M. (1987). On the Significance of Age in Sociology. American Sociological Review, 52(1), 1-14.

Mason, Karen , Oppenheim, William , M. Mason, H., H. Winsborough, W. Kenneth Poole.(1973) Some Methodological Issues in Cohort Analysis of Archival Data. American Sociological Review 38: 242-258

Day 2

Yang, Y., & Land, K. C. (2016). Age-period-cohort analysis: New models, methods, and empirical applications. Chapman and Hall/CRC. (Chapters 3 and 4)

Preston, S. H., & Wang, H. (2006). Sex mortality differences in the United States: The role of cohort smoking patterns. Demography, 43(4), 631-646. Chicago

O’Brien, R. M. (2011). The age–period–cohort conundrum as two fundamental problems. Quality & Quantity, 45(6), 1429-1444.

Day 3

Yang, Y., Fu, W. J., & Land, K. C. (2004). A methodological comparison of age‐period‐cohort models: the intrinsic estimator and conventional generalized linear models. Sociological methodology, 34 (1), 75-110.

Yang, Yang, Schulhofer-Wohl, Sam, Fu, Wenjiang, and Land, Kenneth. 2008. The Intrinsic Estimator for Age-Period-Cohort Analysis: What it is and How to Use it. American Journal of Sociology 113 (6):1697–1736.           

Day 4

Yang, Yang and Land, Kenneth. 2008. Age-Period-Cohort Analysis of Repeated Cross-Section Surveys: Fixed or Random Effects? Sociological Methods & Research 36 (3):297–326.

Frenk, S. M., Yang, Y. C., & Land, K. C. (2013). Assessing the Significance of Cohort and Period Effects in Hierarchical Age-Period-Cohort Models: Applications to Verbal Test Scores and Voter Turnout in U.S. Presidential Elections. Social Forces; a Scientific Medium of Social Study and Interpretation, 92(1), 221–248. http://doi.org/10.1093/sf/sot066

Gelman, A., & Hill, J. (2006). Data analysis using regression and multilevel/hierarchical models. Cambridge university press. (Chapters 11,12,13)

Neundorf, Anja and Niemi, Richard G. 2014. Beyond Political Socialization: New Approaches to Age, Period, Cohort Analysis. Electoral Studies 33:1–6.

Day 5

Grasso, M. T., Farrall, S., Gray, E., Hay, C., & Jennings, W. (2017). Thatcher’s children, Blair’s babies, political socialization and trickle-down value change: An age, period and cohort analysis. British Journal of Political Science, 1-20.

Grasso, M. T. (2014). Age, period and cohort analysis in a comparative context: Political generations and political participation repertoires in Western Europe. Electoral Studies, 33, 63-76.

Andersen, R. (2009). Nonparametric methods for modelling nonlinearity in regression analysis. Annual Review of Sociology, 35, 67-85.

Bell, A., & Jones, K. (2015). Bayesian informative priors with Yang and Land's hierarchical age-period-cohort model. Quality and Quantity, 49(1), 255-266.

Yang, Y. (2006). 2. Bayesian Inference for Hierarchical Age-Period-Cohort Models of Repeated Cross-Section Survey Data. Sociological Methodology, 36(1), 39-74.

Tilley, J. (2002). Political Generations and Partisanship in the UK, 1964–1997. Journal of the Royal Statistical Society: Series A (Statistics in Society), 165(1), 121-135.

Software Requirements

Latest version of R (3.5.1) but will also provide examples in Stata

Hardware Requirements

PC, MAC and Linux are appropriate.

Literature

Mason, K. O., Mason, W. M., Winsborough, H. H., & Poole, W. K. (1973). Some methodological issues in cohort analysis of archival data. American sociological review, 242-258.

Mason, W. M., & Fienberg, S. (Eds.). (2012). Cohort analysis in social research: Beyond the identification problem. Springer Science & Business Media.

Glenn, N. D. (1976). Cohort analysts' futile quest: Statistical attempts to separate age, period and cohort effects. American sociological review, 41(5), 900-904.

Grasso, Maria Teresa. 2016. Generations, Political Participation and Social Change in Western Europe. London: Routledge.

Neundorf, Anja. 2010. Democracy in Transition: A Micro Perspective on System Change in Post-Socialist Societies. Journal of Politics 72(4):1096–1108.

Dinas, E., & Stoker, L. (2014). Age-Period-Cohort analysis: A design-based approach. Electoral Studies, 33, 28-40.

The following other ECPR Methods School courses could be useful in combination with this one in a ‘training track .
Recommended Courses Before

Summer School

Applied Multilevel Regression Modelling

Winter School

Regression Refresher
Multilevel Regression Modelling

Recommended Courses After

Winter School

Introduction to Bayesian Inference

Additional Information

Disclaimer

This course description may be subject to subsequent adaptations (e.g. taking into account new developments in the field, participant demands, group size, etc). Registered participants will be informed in due time.

Note from the Academic Convenors

By registering for this course, you confirm that you possess the knowledge required to follow it. The instructor will not teach these prerequisite items. If in doubt, contact the instructor before registering.


Share this page
 

"Wherever you have an efficient government you have a dictatorship" - Harry S. Truman


Back to top