16.1 Introduction
Due in part to technological development the last few decades have witnessed an explosion in the availability of time-series data.
Common Sources of Time Series:
- ecological and immunological phenomena
- sensor-based physiological measurements (e.g. heart rate and skin conductance)
- health and movement data (calorie tracking, fitbit, GPS, etc.)
- daily diary and ecological momentary assessment data
- measures of emotional states
- measures of social interdependence (social network data)
16.1.0.1 Theories of Processes
Theories in the social, health and behavioral sciences often describe mechanisms or systems occurring within individuals.
A large portion of applied psychological work is focused on the analysis of variation within individuals, across time.
Discussion Question
In relation to your own work, can you come up with a theory that operates entirely at the between-person level? If so what is theory? If not, why is this difficult?
16.1.0.2 Interesting Directions for Time Series Analysis
Questions and Opportunities:
- Integration of multiple time-scales and individuals
- Complex interactions across levels of analysis
- Characterizing heterogeneity between and within individuals
16.1.0.3 Between-Person Heterogeneity
A number of promising methods exist for addressing between-person heterogeneity in time series.
Grouping Individuals with Similar Dynamics
- Mixture modeling of dynamic processes
- Clustering individuals by common dynamics
Multilevel Approaches
- Multilevel VAR modeling
Common and Unique Variation
- Joint and Individual Variance Explained
- Group Iterative Multiple Method Estimation
- multi-VAR and extensions