Sampling design
This is the landing page for everything related to how you select a subset of occasions and a subset of cases (e.g., persons) from all possible occasions and the population that you want to generalize your results to.
For instance, sampling faster and more frequently allows you to see more detail in the fluctuations of a process over time; sampling for a longer period, allows you to see whether the process is characterized by slow changes over time; and sampling multiple cases allows you to see whether the process is characterized by individual differences.
Below we have specified multiple articles where you can read more about important aspects to consider when designing your sampling scheme.
Think more about your sampling design
The timing of your measurement occasions is essential to what patterns you may capture from your process of interest.
- Occasions
- Temporal lens
- [Timing of measurements]
- Time span, process coverage, and granularity
- [Ecological momentary assessment versus ecological retrospective assessment]
- [Measuring rare events]
The characteristics of the participants determine to what extent your results are generalizable to the population of interest.
- Cases
- Population
- [Idiographic versus nomothetic]
Missing data may be introduced due to the nature of the sampling procedure or participant noncompliance.
- [Missing observations]
- [Night gaps in sampling designs]
Ideally, your sample distribution closely resembles the real-world distribution of your construct.
- Distributions
- [Discrete vs continuous constructs]
- Floor and ceiling effects