ad Science, SARL : Software : Version Overview

Home Page: Software : Version Overview

StatEL Base Module

Utility for choice of the right statistic test
Normality test
- Shapiro-Wilks test
- Khi² test
- Kolmogorov-Smirnov test
Extreme values analysis by :
- Dixon test
- Rosner test
Descriptive statistics:
- for quantitative variables
- for qualitative variables
Calculation of subjects number:
- to evaluate one parameter in a population
- to compare one parameter between 2 populations
- to compare one parameter between K samples
Preparation of data (counting of a qualitative variable) :
- calculation of descriptive statistics by group
Fisher’s exact test
Comparison of an observed distribution with a theoretical distribution
Comparison of 2 distributions of a qualitative variable
Khi² test of association
Comparison of an observed mean with a theoretical mean
T-test for independent groups
T-test for dependent groups
One-way Anova for independent and dependent groups :
- Dunnett posthoc test
- Newman & Keuls posthoc test
Two-way Anova :
- without repeated measurements
- with repeated measurements :
  - on independent groups
  - on dependent groups
  - on both dependent and independent groups
Simple and multiple linear regression, Pearson correlation :
- complete model
- Stepwise selection of variables
- analysis of diagnosis (linearity, homoscedasticity, normality of residuals, analysis of studentized residuals and the levers, Durbin-Watson test for serial variable)
Ancova for adjusted means comparison
Ancova for simple linear regression models comparison
Nonparametric test of Mann et Whitney (for independent groups)
Nonparametric test of Wilcoxon (for dependent groups)
Nonparametric test of Kruskall et Wallis (for independent groups)
Nonparametric test of Friedman (for dependent groups)
Nonparametric test of Spearman correlation
Inter-rater Reliability test :
- Cohen’s Kappa for qualitative variables
- Kendall’s correlation for quantitative and ordinal variables

StatEL Med Module (General and Biomedical Statistic Tests)

In addition to the above tests from the base version, StatEL Med offers the following:

Calculation of Sensitivity / Specificity indexes
- From raw or counted data
- With a qualitative or quantitative “Medical Test” variable
ROC
- Elaboration of ROC curves (qualitative or quantitative diagnostic test)
- Comparison of 2 ROC curves by analysis of area under curve
- Calculation of N for comparison of 2 ROC curves
Survival curve :
- With actuarial method
- With Kaplan-Meier’s method
- Calculation of :
  - Survival indexes
  - 95% confidence intervals
  - Hazard rate
  - Cumulated hazard
Comparison of survival curves
- approach by:
  - actuarial method
  - Kaplan-Meier’s method
- tests:
Bland & Altman analysis for comparison of clinical measurement methods with:
- single measurements
- duplicate measurements

StatEL AD Module (General statistical tests and multifactorial analysis)

This version includes al the features of the StatEL base pluse the below tests:

Principal Components Analysis
- Weight of variables and observations
- Added or illustrative variables and observations
- Replacement of missing data
- Agglomerative Hierarchical Clustering after PCA
Correspondences Analysis
- Added or illustrative variables and observations
Agglomerative Hierarchical Clustering
- For a group of observations or des modalities of a qualitative variable
- Replacement of missing data
- Weight observations
- Agglomeration criteria (minimum, maximum, mean or Ward distance)
- Dendrogram construction
Discriminant Analysis
- Descriptive
- Decisional
- Added or illustrative observations
Logistic regression
- analysis of quantitative and qualitative covariates
- gestion of design variables
- log-likelihood, test of Wald and calculation of Odds Ratio
- tests of Pearson's Khi², Deviance, Hosmer & Lemeshow
- diagnostic analysis of the regression model