Specialized Outcome Types
Use this overview to decide whether your outcome belongs in count, bounded-response, compositional, or ordinal guidance, and to choose the right leaf page without overstating current product support.
Use this section when your outcome is not a standard unrestricted continuous variable and you need to decide whether count data, bounded responses, compositional structure, or ordinal categories better describe the problem.
This section is for orienting yourself when the response variable has a special structure that changes how it should be modeled or interpreted. It is not the place to assume that Licklider offers every specialized distribution or fully customizable model family; if your main question is about ordinary group differences or standard regression on a continuous outcome, start with Group Comparison or Regression and Modeling instead.
Some pages in this category are guidance-first or still planned. Read this category as a careful routing layer for special outcome structures and support boundaries, not as proof that every listed specialized method already has a confirmed end-to-end product workflow.
Is this the right section?
Use the table below to choose the leaf page that best matches the structure of your outcome.
| If your outcome looks like... | Start here |
|---|---|
| counts of events, items, or occurrences, often with many zeros or small integers | Count Data Models |
| proportions, fractions, percentages, or other values constrained by natural bounds such as 0 to 1 | Proportion and Bounded Response Data |
| parts of a whole whose components are interpreted relative to each other and often sum near a common total | Compositional Data Analysis |
| ordered categories such as scores, stages, ratings, or severity levels where rank matters but spacing is not truly continuous | Ordinal Outcome Analysis |
If your outcome is a standard continuous measurement and your main task is choosing a familiar t-test, ANOVA, or regression path, this is usually not your first stop. Use Group Comparison or Regression and Modeling before returning here.
What this category covers
This category covers four special outcome structures that often need different reasoning from ordinary continuous analyses:
- Count Data Models for event counts and other non-negative integer outcomes
- Proportion and Bounded Response Data for outcomes constrained by natural limits such as 0 and 1
- Compositional Data Analysis for parts-of-a-whole data with shared row-wise constraints
- Ordinal Outcome Analysis for ordered categories that should not be treated as simply binary or continuous
Support boundary
This category is about deciding whether a specialized outcome type changes the analysis path. It is not a general promise that any custom distribution, specialized link function, or domain-specific modeling workflow can be configured freely in the current product.
Child pages
- Count Data Models
- Proportion and Bounded Response Data
- Compositional Data Analysis
- Ordinal Outcome Analysis
Start here
- Start with Count Data Models if your response is a count of events or items.
- Start with Proportion and Bounded Response Data if your response is bounded, fraction-like, or proportion-like.
- Start with Compositional Data Analysis if your variables represent parts of a whole under a shared total constraint.
- Start with Ordinal Outcome Analysis if your response is ordered categorical rather than truly continuous.
Related
- Related links are registered in frontmatter and rendered below this stub.