Data Analysis¶
Most days, a glance at your metrics overview is enough. When you want to go deeper, Cohearence gives you the same tools a scientist would reach for — without needing to know how to use them.
Statistical tools we support¶
Across the Analysis page, the Signal page, and the MCP analysis tools, Cohearence supports:
- Trends — direction and magnitude of change over a window, with trend lines drawn directly on your charts.
- Correlations — Pearson or Spearman correlation between any two metrics, with confidence intervals and significance (p-values, and FDR-adjusted q-values across the whole network on the Signal page).
- Lag analysis — whether one metric's effect on another shows up a day (or several) later, rather than the same day.
- Regression — fitting a relationship between metrics rather than just describing it.
- Anomaly detection — flagging measurements that don't fit the recent pattern.
- Period comparison — stats for the same metric across two different date ranges, side by side.
Every one of these is exposed as a real statistical computation, not a canned insight — you (or your AI assistant, see Exposing Analysis Over MCP) control the date range, the method, and how missing data is handled.
Where to find them¶
- Analysis Page — build your own multi-metric charts and correlation views by hand.
- Signal Page — explore correlations across your entire dataset at once, ranked and navigable as a network.
- Exposing Analysis Over MCP — hand the same tools to your AI assistant for open-ended, conversational analysis.