Methods Flashcards
Aisenbrey and Fasang 2010 – New Life for Old Ideas: The “Second Wave” of Sequence Analysis Bringing the “Course” Back into the Life Course
Life course topics typically focus on trajectories and events. EHA can’t interpret trajectories. people interested in trajectors have used optimal-matching, but these come with criticisms about: lack of theoretical underpinning (data-based transformation costs and reference sequences), inability to validate, handling of missing/incomplete data, and no focus on the timing and ordering in sequences. The authors walk through approaches that overcome these criticisms. Stovel uses a decay function to test an event relative to the occurrence of other events. Lesnard’s dynamic Hamming measure is similar to optimal matching but it doesn’t insert distances, it finds them inductively. Dijkstra and Taris (1995) use DT coefficients where sequences remove events that do not match and test the number of changes required to match in order. This gives a distance (although removing events is a bad move). Still, you can use this to find how common states or sequences are. Elzinga’s nonalignment metrics test the distance between events in a sequence. Elzinga also has a method to test sequence complexity (variability within individual sequences over time). They end by comparing the methods and use that as a call to action: sequence analysis may be useful, but the method needs to be tailored to the data and theory.
Alwin, Hofer, and McCammon 2006 – Modeling the Effects of Time: Integrating Demographic and Developmental Perspectives
Growth curve models are well-positioned for life course research because they measure intraindividual change across time. Structural equation models or hierarchical models achieve the same thing.
Cohler and Hostetler 2003 – Linking Life Course and Life Story
To understand how major historical events impacted people, they propose life histories. This aligns with “Verstehende Soziologie”.
They propose four ways that qualitative data can be analyzed: holistic content (grounded in close textual reading and focuses on salient themes in the life story as a whole), categorical content (content analysis, focuses on tagging text and assigning codes to specific content largely irrespective of the place in the account), holistic form (focuses on the structure rather than the content of an account or graphically representing the distance between events and/or relationships), and the categorical form (discourse analysis, focuses on the structure of narrative discourse as contained within the interview transcript).
They work from a Foucault framework, focusing on the narrative rather than the “truth”. They offer an empirical example, dissecting the life histories of 20 gay men from different cohorts. They find that their age during the Stonewall Riots influences how they perform their gay identities. Those who were older during the riots subdue the salience of their identities and eschew long-term relationships, while those who were younger follow the trend of centralizing their gay identity and are more susceptible to long-term relationships. This is a tool through which sociologists can capture the social structure that influences individual decision-making processes. They propose they use a holistic-form analysis (holistic content and categorical content approaches).
Fosse and Winship 2019 – Analyzing Age-Period-Cohort Data: A Review and Critique
They show the basic arithmetic relationship between age, period, and cohorts to present a theoretical technique that can explore the causal effects of these three principles. To begin, when all three are included in a regression model, causal claims cannot be made bc they’re interconnected (according to them, although the end of their piece just shows how to manipulate this connection to make causal claims…). The authors walk through techniques people can use to overcome this: drop-one technique and proxy variable. Proxy variables just replace a time-related variable for something else that should be causally related. Finally, they produce graphs with basic arithmetic concepts to show how effect sizes must impact an outcome.
George 2009 – Conceptualizing and Measuring Trajectories
Advocates for people studying trajectories, but can be retrospective. Important features of trajectories include timing and critical periods, duration, sequencing, and turning points. Trajectories can offer new angles for causality research. E.g. do people’s health change over their lifetime as their SES changes? Cross-fertilization with other research can be fruitful. E.g. studying stress and health over the life course can improve causal claims. The methods most commonly used in this line of research are qualitative, (dis)aggregated, hierarchical models, and latent class analysis.
MacMillan and Eliason 2003 – Characterizing the Life Course as Role Configurations and Pathways
Life course theory and methods often fail to converge. Two-stage latent class models could connect period and cohort effects. First, assign people to binary indicators reflecting: school, employment, married, and parents statuses. Use latent class models to predict groups. Then, predict life course patterns across age periods. They find 3: extended education, full time work w/o marriage/fam, quick exit from educ w/ marriage/fam using the NLSY79 w/ kids who were 15-16 at wave I. Multinomial logistic regression could be used to predict variables that correlate with pathways. Also, cohort differences can be seen.
O’Rand 1998 – The Craft of LIfe Course Studies
She highlights how life-course theory was predicated on a need for the data to resemble real-life experiences. Thomas (polish study), Elder (depression), and Neugarten (age norms) were really the pioneers pushing the field forward. Life course studies is interdisciplinary, varies theoretically and methodologically. The author doesn’t give any good descriptions of methodologicaac techniques, but does offer some interesting datasets.
Ryder 1965 – Cohort as a Concept in the Study of Social Change
Using a functionalist perspective, they argue that cohorts are the major engine of social change. He compares societies to animal species; individuals don’t change species, but generations do. Schools are central to social change because they shift socialization from families to society. Peer groups, sub-cohort shared within a school, become the focal point for youth’s attention and socialization. But peer groups expire. Once they enter the workforce, people live in an age-hierarchical system and must either defer to or defy conventional norms. Reaction influences opportunity. At times when cohorts see dramatic change at the top of their systems (oldest age groups), we will likely see the greatest opportunity for newcomers to thrive in that system. Ryder resists the idea that personality and ideas become permanent once people become adults. Instead, he proposes the type of training people receive when younger is either conducive to or repellent against new ideas. In coming economies, he predicts that people will be required to re-educate constantly, which will further promote social change. On the other hand, older age groups tend to maintain practices that limit their exposure to new ideas (maintaining particular age groups, not exposed to new media). For that reason, we likely see conservatives in the older age groups. “The feasibility of personal transformation is probably limited more by restricted membership than by physiological aging.”
Settersten and Mayer 1997 – The Measurement of Age, Age Structuring, and the Life Course
They define important terms: age, age structuring, and the life course. Age is the property of individuals which stands as a proxy for biological maturation, psychological development, or membership in a cohort. Age structuring refers to the social conditions tied to age (experiences, roles, statuses of individuals). Life-course refers to the way in which social institutions shape and institutionalize individual lives in the interconnected domains of education, family, and work. A cohort is reserved for an aggregate of individuals anchored together in historical time. Birren and Cunningham (1985) differentiate between biological age, social age (roles and habits), and psychological age (propensity for responsibility). They also discuss subjective age. This includes cognitive age (feel age, look age, do age, interest age), desired age, relative age identification, self-perceived ages, other perceived ages. Age norms can be statistical (averages), optimal (“best” or “ideal”), and prescriptive/proscriptive (collective, shared expectations). Life trajectory (Elder 1985) “a pathway defined by the aging process or by movement across the age structure”. An event is an abrupt change and a transition is gradual. Timing refers to the age when transitions occur, the sequence is the order, duration is the length of time spent in a state, and spacing refers to time between states.
