Compare Serum IgG
The Compare Serum IgG utility allows researchers to evaluate how well the CMC model predicts serum IgG concentrations measured in their own calves.
Rather than comparing CMC with another published equation, this utility compares model predictions directly with observed laboratory measurements.
Preparing the data
The utility accepts a spreadsheet containing calf-level information, including:
- Birth weight
- Colostrum IgG concentration
- Feeding volumes
- Feeding ages
- Pasteurization status
- Heat stress status
- Measured serum IgG concentration
Each row represents one calf.
CMC uses the management information provided for each calf to calculate a predicted serum IgG concentration using the selected prediction equation.
Statistical comparison
After predictions are generated, the utility compares predicted and observed serum IgG concentrations for every calf.
Summary statistics include:
- Number of calves
- Mean observed serum IgG
- Mean predicted serum IgG
- Mean prediction error (bias)
- Standard deviation of prediction errors
- Median prediction error
- Mean absolute error (MAE)
- Root mean square error (RMSE)
- Pearson correlation coefficient
Optionally, the utility can export calf-level results so individual predictions and residuals may be examined further.
Why this utility is important
The strongest evaluation of any prediction model is comparison with independent data that were not used during model development.
Researchers can use this utility to determine how well the CMC model performs under different management conditions, breeds, climates, or experimental protocols. The resulting statistics help identify systematic bias, quantify prediction error, and evaluate overall model performance.
Because the comparison is performed using the investigator’s own data, this utility provides one of the most informative methods for validating the model in new populations.
