.png)
Excel Upload Guide
To ensure successful data analysis, please follow these requirements when preparing your Excel file:
- File must be in .xlsx or .xls format
- Data should be organized in columns with headers
- Reference method values should be in Column A (starting from A1)
- Test method values should be in Column B (starting from B1)
- Each row should contain paired measurements from both analyzers
- No missing values or non-numeric data in measurement columns
- Minimum of 30 paired measurements recommended for reliable analysis
- Maximum of 1000 paired measurements to ensure optimal performance
Example Excel Format:
| Reference Method | Test Method |
|---|---|
| 120.5 | 118.7 |
| 95.2 | 94.8 |
About LabValid
LabValid is a powerful tool designed for laboratory method comparison studies. It helps validate new analytical methods against established reference methods by providing comprehensive statistical analysis and visualization.
Passing-Bablok Regression
Non-parametric regression analysis that helps determine systematic and proportional differences between methods.
Bland-Altman Analysis
Evaluates agreement between two measurement methods by plotting differences against means.
CUSUM Test
Assesses the linearity of the relationship between methods.
Statistical Reports
Generate comprehensive reports with statistical parameters and interpretations.
How to Use
- Upload your Excel file containing paired measurements from both methods.
- Select the columns containing your reference and test method values.
- Click "Analyze" to generate comprehensive statistical analysis.
- Review the results, including regression parameters, correlation, and agreement measures.
- Export your results as images or a complete Word document report.
Interpreting Results
Key parameters to consider:
- Slope: Ideally close to 1, indicates proportional differences
- Intercept: Ideally close to 0, indicates systematic differences
- Correlation: Measures strength of linear relationship
- Bland-Altman Limits: Shows agreement between methods
Upload Data
Enter your data with column headers in the first row, followed by your measurements. Use commas to separate values.
Column Selection
Sample Count
Total Paired Samples
0
This represents the number of valid paired measurements used in the analysis
Data Range
Summary of Key Results to Report
| Statistic | What It Tells You | Ideal Value |
|---|---|---|
| Slope (β₁) | Proportional bias | Close to 1, CI includes 1 |
| Intercept (β₀) | Constant bias | Close to 0, CI includes 0 |
| Correlation (r) | Strength of association | r > 0.95 for strong agreement |
| P-value for slope/intercept | Significance of bias | p ≥ 0.05 (no significant bias) |
| CUSUM test | Linearity check | p ≥ 0.05 (linear relationship) |
| Bland-Altman (Mean Bias, LOA) | Agreement range | Mean bias close to 0 |
Current Analysis Results
Passing-Bablok Regression - Axis Configuration
X-Axis (Reference Method)
Smaller intervals may create crowded axis labels
Y-Axis (Test Method)
Smaller intervals may create crowded axis labels
Passing-Bablok Regression
CUSUM Test for Linearity
This plot shows the cumulative sum of residuals. A random pattern around zero indicates linearity. Systematic deviations suggest non-linearity.
Residual plot shows individual residuals vs. reference method values. Random scatter around zero indicates linearity. Any visible pattern suggests non-linearity.