Method Validation Edition Validate and verify your analytical and diagnostic methods to meet the demands of regulatory compliance.

Quantitative methods

• Supports singlicate, duplicate, and replicate measurements
• Compare commutability of samples with prediction intervals new in v4.90
• Reduce measuring interval to linear range, or partition into multiple intervals with different relationships (e.g. constant / relative differences) new in v4.00
• Ordinary and Weighted linear regression average bias with confidence intervals
• Deming and Weighted Deming regression average bias with Jacknife confidence intervals
• Passing Bablok regression average bias with Passing-Bablok or Bootstrap confidence
• Predict bias with confidence intervals at important decision levels
• Test equality (no difference) or equivalence (difference within an allowable difference) at decision levels
• Scatter plot, with average bias, average bias confidence bands, identity line, and allowable difference band
• Vary color of points by a factor
• Difference/relative difference/ratio plot against X or mean of methods with allowable difference band and histogram of differences
• Bland-Altman limits of agreement with mean, median, and linear fit bias new in v3.75
• Mountain plot with allowable difference band new in v3.71
• Residual plot, raw and standardized, with histogram of residuals
• CUSUM linearity plot and Kolmogorov-Smirnov linearity test
• Precision (SD or CV) and precision plots for each method
• Pearson r correlation coefficient
• Supports CLSI EP09-A3 Measurement Procedure Comparison and Bias Estimation Using Patient Samples and CLSI EP21 Estimation of Total Analytical Error for Clinical Laboratory Methods

Qualitative methods

• Proportion in positive agreement / negative agreement with Clopper-Pearson exact or Wilson score confidence intervals
• Kappa and Weighted Kappa for chance-corrected agreement with Wald Z confidence interval
• Kappa test for agreement

• Unified analysis to examine performance characteristics of a method (bias, precision, linearity, interferences, and detection limits)
• Flexible balanced and unbalanced experiment design: up to 3 random nested factors new in 4.60 (e.g. day, run, laboratory), and 1 fixed factor (e.g. level)
• Variance function fit: 4 parameter function with turning point new in v4.91
• Generalized ESD outlier identification (Grubb’s test for more than one outlier) new in v4.95
• Scatter plot and difference plots
• Variability of measurements plot

Bias / Trueness new in v4.00

• Bias with confidence interval
• Difference plot of bias against assigned values with allowable bias bands
• Test equality (no bias) or equivalence (bias within allowable bias)
• Supports CLSI EP15-A3 User Verification of Precision and Estimation of Bias and EP10-A3-AMD Preliminary Evaluation of Quantitative Clinical Laboratory Measurement Procedures

Linearity

• Fit models: linear, polynomial (up to 5th order), forward stepwise polynomial, and best (2nd or 3rd) polynomial regression
• Weighted models for non-constant SD across the measuring interval new in v5.00
• Adjust measuring interval to find linear range new in v5.00
• Difference between linear and nonlinear fit with Hsieh-Liu confidence intervals
• Test equality (no difference) or equivalence (difference within allowable nonlinearity)
• Difference plot of difference against assigned values with allowable nonlinearity band
• Supports Emancipator-Kroll linearity and CLSI EP06-A Evaluation of the Linearity of Quantitative Measurement Procedures and EP10-A3-AMD Preliminary Evaluation of Quantitative Clinical Laboratory Measurement Procedures

Precision

• Precision as variance, SD, or CV% with Exact, Satterthwaite, and MLS confidence
• Abbreviated reproducibility/repeatability, and detailed intermediate precision components
• ANOVA table
• Terminology depending on conditions of measurement: total / within, reproducibility / repeatability, or laboratory / repeatability
• X2 test against precision claim
• Precision profile of SD or CV
• Variance function fit: constant variance, constant CV, mixed constant / proportional variance, and Sadler 3 parameter power, and 3 parameter alternative power functions new in v4.00
• Supports CLSI EP05-A3 Evaluation of Precision of Quantitative Measurement Procedures , EP10-A3-AMD Preliminary Evaluation of Quantitative Clinical Laboratory Measurement Procedures and EP15-A3 User Verification of Precision and Estimation of Bias

Detection limits new in v4.00

• Limit of blank (LoB): of a blank material (parametric SD or non-parametric quantile), or using precision profile variance function
• Limit of detection (LoD): pooled SD of non-blank materials, or using precision profile variance function
• Limit of detection (LoD) using Probit regression new in v5.50
• Limit of quantitation (LoQ) using precision profile variance function
• Frequency density histogram of detection capability, LoB, and LoD
• Supports CLSI EP17-A2 Evaluation of Detection Capability for Clinical Laboratory Measurement Procedures

• Partition by factor(s) for separate analysis of groups new in 4.0
• Frequency distribution histogram with normal overlay and reference limits
• Tukey outlier box plot
• Transform reference values with reciprocal, log, square and cube root, Box-Cox new in 3.52, and Manly exponential & 2-stage exponential / modulus new in 4.0 functions
• Shapiro-Wilk or Anderson-Darling normality test of reference distribution
• Normal quantile plot
• Supports CLSI EP28-A3C (Formerly C28-A3C) Defining, Establishing, and Verifying Reference Intervals in the Clinical Laboratory

Establish reference limits

• Normal quantile
• Nonparametric quantile, Harrell-Davis quantile, and Bootstrap quantile
• Robust bi-weight quantile for symmetric and skewed small samples
• Various quantile computation methods (N+1)p, Np+1/2, and (N+1/3)p+1/3

Transfer / verify limits

• Transfer existing reference interval using method comparison regression function
• Binomial test for proportion inside reference interval

ROC

• Diagnostic performance for 1 test, up to 10 paired tests, or up to 10 independent tests/groups
• Number of true positives (TP), true negatives (TN), false positives (FP), and false negatives (FN)
• Sensitivity/Specificity
• Likelihood ratios
• Predictive values
• Odds ratio, and Youden's index
• Bi-histogram and dot-plot of positive/negative outcomes new in v4.0
• ROC curve plot: Sensitivity (TPF) vs 1-Specificity (FPF) with no discrimination line
• Wilcoxon-Mann-Whitney area under curve with DeLong-DeLong-Clarke-Pearson confidence interval
• Z test of area under curve is better than chance decision
• Compare DeLong-DeLong-Clarke-Pearson difference in area under curves and test for equality (no difference), equivalence (difference negligible), or non-inferiority (not unacceptably worse than a standard test)
• Decision plot of accuracy over all possible decision thresholds: Sensitivity vs Specificity, Likelihood ratios, Predictive values, or Cost
• Find optimal decision threshold based on cost of diagnosis/misdiagnosis
• Predict false positive fraction (FPF) at: Fixed sensitivity, Sensitivity at fixed FPF, or Sensitivity/FPF at fixed threshold.
• Supports CLSI EP24-A2 (Replaces GP10-A) Assessment of the Diagnostic Accuracy of Laboratory Tests Using Receiver Operating Characteristic Curves

Qualitative

• Diagnostic performance for 1 test, 2 paired tests, or 2 independent tests/groups
• Sensitivity/Specificity with Clopper-Pearson exact or Wilson score confidence interval
• Likelihood ratios with Miettinen-Nurminen score confidence interval
• Predictive values with Mercado-Wald logit confidence interval
• Odds ratio, and Youden's index
• Mosaic plot of outcomes
• Difference between sensitivity/specificity with Newcombe or Tango score confidence intervals
• McNemar-Mosteller exact, Fisher exact, and Score Z test for equality of sensitivity/specificity
• Supports CLSI EP12-A2 User Protocol for Evaluation of Qualitative Test Performance

Continuous

• Sum, Mean, Variance, SD, CV%, Skewness, Kurtosis
• Geometric Mean, Harmonic Mean new in v4.50
• Median, Minimum, Maximum, Range, IQR
• Quantiles
• Mode
• Transform variable new in v5.50
• Histogram with optional normal overlay
• Frequency polygon
• Dot plot – jittered, aligned, spread points and vary point symbol/color
• Skeletal box plot, Tukey outlier box plot, Quantile box plot
• Mean error bar plot, Mean confidence diamond plot
• CDF plot with optional Kolmogorov-Smirnov confidence band
• Normal Q-Q plot with optional Lilliefors confidence band
• Shapiro-Wilk, Anderson-Darling, and Kolmogorov-Smirnov tests for normality
• Z test for mean with known population SD
• Student’s one sample t-test for mean
• Wilcoxon test for mean/median
• Sign test for median
• X2 test for variance
• Mean estimate with t-based or Z-based confidence interval
• Median with Thompson-Savur confidence interval
• Hodges-Lehmann pseudo-median with Tukey confidence interval
• Variance with X2-based confidence interval

Discrete

• Frequency table – frequency, cumulative frequency, relative frequency, cumulative relative frequency
• Frequency bar plot with optional cumulative frequency line
• Frequency pie (whole-to-part) plot
• Binomial exact test for proportions
• Score Z test for binomial proportions
• Pearson X2 and Likelihood ratio G2 test for multinomial proportions
• Proportion with Clopper-Pearson exact or Wilson score confidence interval

• Descriptive statistics by group
• Side-by-side dot plots, mean plots, box plots by group
• Z test for difference in means with known population SDs
• Student’s independent samples t-test for difference between means
• Welch’s t-test for difference between means with unequal variances
• Wilcoxon-Mann-Whitney test for difference between means/medians
• 1-way between-subjects ANOVA for equality of means
• Welch’s ANOVA for equality of means with unequal variances
• Kruskal-Wallis test for equality of mean/medians
• Mean difference with t-based, Welch-Satterthwaite t-based, or Z-based confidence interval

• Cohen’s and Hedge’s g standardized mean difference with non-central t-based confidence interval
• Hodges-Lehmann location shift with Moses confidence interval
• Multiple comparisons procedures: Student’s t (individual comparisons), Tukey-Kramer (all pairs), Dunnett (against control), Hsu (with best), Scheffe (all contrasts), Steel (non-parametric against control), Dwass-Steel-Critchlow-Fligner (non-parametric all pairs), Wilcoxon (non-parametric individual comparisons)
• Mean-Mean scatter plot for multiple comparisons
• F-test for variance ratio
• Bartlett, Levene, and Brown-Forsythe tests for homogeneity of variances

• Descriptive statistics for each group
• Side-by-side dot plots, mean plots, box plots by group
• Difference plot with identity line and optional histogram of differences
• Z-test for difference in means with known population SDs
• Student’s paired samples t-test for difference between means
• Wilcoxon signed ranks test for difference between means/medians
• Sign test for difference between medians

• 1-way within-subject ANOVA for equality of means
• Friedman test for equality of medians
• Mean difference with t-based or Z-based confidence interval
• Cohen’s and Hedge’s g standardized mean difference with non-central t-based confidence interval
• Median difference with Thompson-Savur confidence interval
• Hodges-Lehmann location shift with Tukey confidence interval

• Contingency table
• Grouped frequency plot
• Stacked frequency plot
• Pearson X2 test for independence / equality of proportions
• Likelihood ratio G2 test for independence
• Mosaic plot for association with color by category or residual

2x2 related tables

• McNemar-Mosteller exact test for symmetry / marginal homogeneity
• Score Z test for difference between proportions
• Proportion difference (risk difference) with Newcombe score or Tango score confidence interval
• Odds ratio with Binomial exact or Wilson score confidence interval

2x2 tables

• Fisher exact test for independence
• Score Z test for difference between proportions
• Proportion difference (risk difference) with Miettinen-Nurminen score, Newcombe score, or Tango score confidence interval
• Proportion ratio (risk ratio) with Miettinen-Nurminen score or Newcombe confidence interval
• Odds ratio with Hypergeometric exact or Miettinen-Nurminen score confidence interval

Linear Fits

• Simple linear regression
• Polynomial regression (2nd to 6th order)
• Logarithmic regression
• Exponential regression
• Power regression
• Probit regression new in v5.50
• Multiple linear regression
• ANOVA new in v4.80
• ANCOVA new in v4.80
• Advanced models with simple, crossed, polynomial and factorial terms, with categorical explanatory variables coded as dummy variables
• Transform X and Y variable in simple regression new in v5.50

Other Fits

• Binary logistic regression

• Model equation
• Summary of fit – R2, AIC, BIC
• Parameter estimates – beta, confidence intervals, VIF, standardized beta
• Scatter plot with fit line and optional confidence bands
• F-test effect of model
• Predicted against actual plot
• F-test effect of each term in model
• Leverage plot for effect of each term
• Residual plot – raw, standardized
• Sequence and Lag-1 plots
• Outlier and Influence plot
• Cook's D influence
• Predict Y for X
• Effect means for categorical variables new in v4.80
• Main effect and interaction plots for categorical variables new in v4.80
• Multiple comparisons of effect means: Student’s t (individual comparisons), Tukey-Kramer (all pairs), Dunnett (against control), Hsu (with best), Scheffe (all contrasts) new in v4.80
• F-test for lack of fit test for simple regression models
• Save model variables back to the dataset: Fitted Y, Residuals, Standardized Residuals, Studentized Residuals, Leverage, Cook's Influence

• Correlation matrix with color map on coefficients
• Covariance matrix
• Scatter plot
• Scatter plot matrix
• Vary points by color of based on a factor

Correlation / Association

• Pearson r correlation coefficient with Fisher’s Z confidence interval
• Pearson test for linear association
• Spearman rs correlation coefficient with Fisher’s Z confidence interval
• Kendall tau correlation coefficient with Samara-Randles confidence interval
• Kendall test for monotonic association

Item reliability new in v4.80

• Cronbach’s alpha (standardized and unstandardized)
• Deleted Cronbach’s alpha for each item

PCA new in v3.80

• Eigenvalues / Eigenvectors
• Coefficient matrix with color map to reveal relationships
• Classic Gabriel bi-plot with variables as vectors and observations as points
• Gower-Hand bi-plot with variables and observations as points
• Predict new observations / variables
• Reflect, rotate, and scale bi-plot
• Correlation mono-plot to show relationship between variables
• Scree plot

Common Factor Analysis new in v3.90

• Maximum likelihood factor extraction
• Factor pattern / structure matrices with color map to reveal structure
• 12 factor orthogonal/oblique rotations including Varimax, Oblimin