Creating Analyses and Dashboards
Analyses versus dashboards
An analysis is a single query with criteria, filters, and views (table, chart, pivot, gauge). A dashboard is a page composed of multiple analyses, prompts, and other objects arranged in columns and sections. Analyses are saved in the catalog; dashboards reference them. Understanding this containment relationship resolves many questions: editing an analysis changes every dashboard that uses it, whereas dashboard-level prompts filter all embedded analyses that are mapped to them.
Prompts and filters
Prompts collect user input at runtime and can be dashboard prompts or analysis prompts. Filters restrict rows and can be column filters, measure filters, or saved named filters. Know the difference between a filter applied at the analysis level and one applied at the dashboard level. Selection steps and groups let users build ad hoc sets. Interaction features include drill-down, master-detail linking, and action links that navigate or invoke actions.
Traps to watch
A frequent trap is confusing a dashboard prompt with an analysis filter: prompts are interactive and reusable, filters are fixed in the analysis. Another is assuming a dashboard stores its own copy of data - it does not; it renders the referenced analyses live. Questions may test that column prompts can be constrained by other prompts, or that a dashboard page's layout uses columns and sections rather than free positioning in all cases. Watch for wording about "saving" a customized dashboard - personal customizations are saved separately from the shared original.
Rule of thumb
If the question mentions reusable runtime input, think prompt. If it mentions a fixed restriction baked into a query, think filter. If it mentions multiple views on one page, think dashboard.