Model health summary card
Model health summary card
Model health score
99.8
How is health score measured?

The Meridian model health score gives a unified view of model integrity by combining six individual health checks into one metric ranging from 0 to 100. While individual diagnostics provide granular insights into specific model components, the model health score provides an aggregate measure on a model's health for causal inference. See Model health score to learn more.

check_circle Passed: No major quality issues were identified.
Metrics check
Metric check Status Recommended action
Convergence Pass The model has likely converged, as all parameters have R-hat values < 1.2.
Baseline Pass The posterior probability that the baseline is negative is 0.00. We recommend visually inspecting the baseline time series in the Model Fit charts to confirm this.
Bayesian p-value Pass The Bayesian posterior predictive p-value is 1.00. The observed total outcome is consistent with the model's posterior predictive distribution.
Goodness of fit Pass R-squared = 0.7713, MAPE = 0.0269, and wMAPE = 0.0269. These goodness-of-fit metrics are intended for guidance and relative comparison.
Prior-posterior shift
Pass
5/5 channels passed
 The model has successfully learned from the data. This is a positive sign that your data was informative.
Calibration score
Pass
0/5 channels recommended for calibration
 The overall calibration score is 93.5/100. No channels require calibration.
Meridian GeoX calibration summary
Calibration score
93.5
How is calibration score measured?

The Meridian calibration score gives a view of the overall calibration health of your Meridian media and reach & frequency channels. The score combines the number of calibrated channels along with potential channel recommendations for the remaining channels. The calibration score is a factor in your overall model health score in the previous section. See Channel Calibration Recommendation and Score to learn more.

Category Overview Summary
Prior calibration 0 channel(s) No prior calibration has been completed using incrementality experiments.
Channel recommendation 0 channel(s) No channels require calibration.
Channel calibration recommendation
info No channels require calibration. We recommend reviewing the table and plots below to check for channels near the boundaries that may be good candidates for calibration via an incrementality experiment such as those run with Meridian GeoX.
Score High ROI Low ROI High Variance ROI Potential Bias
Channel0
94.9
Non-Driver Non-Driver Non-Driver Non-Driver
Channel1
92.0
Non-Driver Non-Driver Non-Driver Non-Driver
Channel2
95.1
Non-Driver Non-Driver Non-Driver Non-Driver
Channel3
89.4
Non-Driver Non-Driver Non-Driver Non-Driver
Channel4
96.2
Non-Driver Non-Driver Non-Driver Non-Driver
Spend vs. ROI (Implausible ROI Check)
info

This plot displays your media and reach and frequency channels according to their spend and ROI. Channels closer to the boundaries of the Implausible High ROI and Implausible Low ROI regions may be strong candidates for calibration. For readability, ROIs between 0.6 and 19 are clustered together on this plot. Please hover over points or use MeridianAnalyzer.roi to view the exact ROI for specific channels.
Spend vs. Relative Credible Interval (High Variance ROI Check)
info

This plot displays your media and reach and frequency channels according to their spend and Relative Credible Interval (RCI), a measure of their variance. Channels close to the boundary of the High Variance ROI region may be strong candidates for calibration.
Correlation with Controls (Potential Bias Check)
info

This plot displays your media and reach and frequency channels along with their correlation with your available controls.