NCSS Upgrade Information

Procedures added in the NCSS 2020 Upgrade from NCSS 2019

Random Sampling

• Simple Random Sampling
• Stratified Random Sampling
• Simple Random Sampling with Group Assignment
• Stratified Random Sampling with Group Assignment

 

Stratified Random Sampling Input Example

 

Group-Sequential Tests (Futility Boundary Options, Sample Size Re-estimation)

For each of these group-sequential analysis and sample-size re-estimation procedures, there are corresponding group-sequential power and sample size procedures in PASS 2020.

• Group-Sequential Analysis for Two Hazard Rates
• Group-Sequential Non-Inferiority Analysis for Two Hazard Rates
• Group-Sequential Superiority by a Margin Analysis for Two Hazard Rates
• -
• Group-Sequential Non-Inferiority Analysis for Two Means with Known Variances
• Group-Sequential Superiority by a Margin Analysis for Two Means with Known Variances
• Group-Sequential Non-Inferiority T-Tests for Two Means
• Group-Sequential Superiority by a Margin T-Tests for Two Means
• -
• Group-Sequential Non-Inferiority Analysis for Two Proportions
• Group-Sequential Superiority by a Margin Analysis for Two Proportions
• -
• Group-Sequential Analysis for One Mean with Known Variance
• Group-Sequential T-Tests for One Mean

 

Procedures that were Updated and/or Improved in NCSS 2020

These procedures have been updated to improve functionality and flexibility.

At-Risk Tables in Survival/Reliability Procedures

For each of these procedures, at-risk tables were added to the survival-type plots. Additional input and output improvements were also made.

• Kaplan-Meier Curves (Logrank Tests)
• Life-Table Analysis
• Cumulative Incidence
• Gamma Distribution Fitting
• Beta Distribution Fitting
• Distribution (Weibull) Fitting

 

At-Risk Table Example

Between-Group Mean Survival Comparisons (RMST and RMTL) in the Kaplan-Meier Curves (Logrank Tests) Procedure

The Kaplan-Meier Curves (Logrank Tests) Procedure now computes Restricted Mean Survival Time (RMST) and Restricted Mean Time Lost (RMTL) statistics, along with pairwise difference and ratio comparisons among groups.  

Heatmaps in Factor Analysis and PCA Procedures

Visual interpretation was improved in these procedures by adding heatmaps to the output. Additional examples were added as well.

• Factor Analysis
• Principal Components Analysis

 

Heatmap Example

 

Contingency Tables (Crosstabs / Chi-Square Test)

In the Contingency Tables (Crosstabs / Chi-Square Test) procedure, the Tukey-Pairwise and Dunnett's Many-to-one (control) multiple comparisons results were added. These include the adjustments suggested by Agresti and Caffo (2000). Both p-values and confidence limits for Wald and Score tests are now included.

Randomization Lists

A variety of improvements were made to the Randomization Lists procedure, including block randomization, the option of specifying a random seed, group titles, sample size entry options, and strata specification.

Enhancements

Memory

A major memory performance enhancement was integrated into the spreadsheet, resulting in up to a 70% decrease in required memory.

Standardized Range Distribution

Improved algorithms for the Standardized Range probability distribution were integrated, to improve accuracy.

Additions and Improvements in the NCSS 2019 Upgrade from NCSS 12

Group-sequential analysis was introduced, along with Mediation analysis and many improvements.

Procedures Added in NCSS 2019

Group-Sequential Tests (Futility Boundary Options, Sample Size Re-estimation)

For each of these group-sequential analysis and sample-size re-estimation procedures, there are corresponding group-sequential power and sample size procedures in PASS 2019.

• Group-Sequential Analysis for Two Means with Known Variances
• Group-Sequential T-Tests for Two Means
• Group-Sequential Analysis for Two Proportions

Mediation Analysis

• Mediation Analysis

 

Mediation Analysis Output Example

 

Updated and/or Improved Procedures in NCSS 2019

These procedures were updated to improve functionality and flexibility.

• Frequency Tables
• Correlation Matrix

Frequency Tables Procedure

In the Frequency Tables procedure, one can now sort the frequency table by count. Additional bar charts and pie charts were also added.

 

Correlation Matrix Procedure

More correlation statistics can be reported in a more flexible way. Heat map and eigenvector plots have also been added.

 

Enhancements

Numeric Axis Value Labels

The ability to substitute numeric values with value labels for numeric axis reference labels is now available for all plots with numeric axes.

New Transformations

The following transformations have been added:

• DATEDT: Returns the date portion of a date-time value.
• TIMEDT: Returns the time portion of a date-time value.
• RANDOMSAMPLEROWS: Creates a binary indicator variable by randomly selecting k non-empty rows.
• RANDOMSAMPLEBIN: Creates a binary indicator variable by randomly selecting k values from a column.
• RANDOMSAMPLE: Returns a random sample of k values from a column.
• FIRSTDIFF: Returns the first-order differencing of a column (D1[i] = X[i] - X[i-1]).
• SECONDDIFF: Returns the second-order differencing of a column (D2[i] = X[i] - 2X[i-1] + X[i-2]).
• LOESS: Returns the smoothed Loess regression values based on Y and X.

Improved Chart Headings

For many procedures, particularly regression procedures, additional headings and navigation links were added for the plots.

Compatibility Improvements

The underlying framework was converted to .NET 4.6 for better compatibility with Windows 10 and general compatibility improvements. This framework significantly improved the responsiveness of the 64-bit version of NCSS 2019, as well.

Procedures and Enhancements added in the NCSS 12 Upgrade from NCSS 11

Many new procedures, plots, and enhancements were added in the upgrade from NCSS 11 to NCSS 12.

Procedures Added in NCSS 12

• Paired T-Test for Superiority by a Margin
• One-Sample T-Test for Non-Inferiority
• One-Sample T-Test for Superiority by a Margin
• One-Sample T-Test for Equivalence
• Two-Sample T-Test for Superiority by a Margin
• -
• Analysis of 2x2 Cross-Over Designs using T-Tests for Non-Inferiority
• Analysis of 2x2 Cross-Over Designs using T-Tests for Superiority by a Margin
• Analysis of 2x2 Cross-Over Designs using T-Tests for Equivalence
• -
• One Proportion - Non-Inferiority Tests
• One Proportion - Superiority by a Margin Tests
• One Proportion - Equivalence Tests
• -
• Two-Sample Non-Inferiority Tests for Survival Data using Cox Regression
• Two-Sample Superiority by a Margin Tests for Survival Data using Cox Regression
• Two-Sample Equivalence Tests for Survival Data using Cox Regression
• -
• Cluster Randomization - Create Cluster Means Dataset
• Cluster Randomization - Create Cluster Proportions Dataset
• Cluster Randomization - Create Cluster Rates Dataset
• -
• General Linear Models (GLM) for Fixed Factors
• -
• One-Way Analysis of Covariance (ANCOVA)
• Analysis of Covariance (ANCOVA) with Two Groups
• -
• Clustered Heat Maps (Double Dendrograms)

Examples of new plots:

• Clustered Heat Map (Double Dendrogram)

• Analysis of Covariance Plots

• Multiple Comparisons Plots

Enhancements

Spreadsheet Summary Statistics Table

A new Summary Statistics table has been added to the bottom of the spreadsheet window. This spreadsheet extension contains column summary statistics, such as number of unique values, minimum, maximum, mean, median, standard deviation, number of cells with data, and many others.

Autorecovery

A data autorecovery function was implemented.

Importing Files

Imported Files are now added to the list of recent files. It is now easier to re-import files that have been imported before.

Procedure Window - Recent Templates

Each procedure window now has a menu item displaying the recently opened templates. This can be used to quickly open previously used (and saved) template settings.

Procedure Setting - Autosave

An option has been added to autosave the procedure settings with each run.

Procedure Input Settings Report

The option to give a report in the output showing all the procedure options used has been added.

Report Headers and Colors

New system options give the user the ability add (or remove) lines to the titles to improve separation of report sections. Title line color and text color for page headers, titles, and reports may also be specified individually.

Spreadsheet Find and Replace

The find and replace functionality of the spreadsheet data table has been extended to the Column Info table and the new Summary Statistics table.

Copy Table

A function giving the ability to copy the entire spreadsheet data table to the clipboard has been added.

Transformations

The deviation from the mean and absolute deviation from the mean options were added to the transformation system.

Paired T-Test Procedure

The Mann-Whitney U Test was improved and the Shapiro-Wilk Normality Test was added to the Paired T-Test procedure.

One-Sample T-Test Procedure

The Mann-Whitney U Test was improved and the Shapiro-Wilk Normality Test was added to the One-Sample T-Test procedure.

Two-Sample T-Test Procedure

The Mann-Whitney U Test was improved and the Shapiro-Wilk Normality Test was added to the Two-Sample T-Test procedure.

One- and Two-Factor Box Plot

Numeric axis log scaling and the option to reverse the order of the groups or subgroups were added to the one- and two-factor box plots.

Error-Bar Chart

Numeric axis log scaling and the option to reverse the order of the groups or subgroups were added to the error-bar charts.

Percentile Plot

Numeric axis log scaling and the option to reverse the order of the groups or subgroups were added to the percentile plots.

Comparative Histogram

Numeric axis log scaling and the option to reverse the order of the groups or subgroups were added to the comparative histograms.

Density Plot

Numeric axis log scaling and the option to reverse the order of the groups or subgroups were added to the density plots.

Dot Plot

Numeric axis log scaling and the option to reverse the order of the groups or subgroups were added to the dot plots.

Combo Chart

Numeric axis log scaling and the option to reverse the order of the groups or subgroups were added to the combo charts.

Bar Chart

The option to reverse the order of the groups or subgroups was added to the bar charts.

2x2 Cross-over Analysis using T-Tests Procedure

A numeric test for Normality and the Mann-Whitney U Non-parametric test were added to the 2x2 Cross-over Analysis using T-Tests procedure.

Fractional Polynomial Regression Procedure - Predicted Values and Residuals

The ability to store predicted values and residuals to the spreadsheet from a fractional polynomial regression run has been added.

Reference Interval Procedure - Predicted Values and Residuals

The ability to store predicted values and residuals to the spreadsheet from a Reference Interval - Age-Specific procedure run has been added.

Procedures and Enhancements added in the NCSS 11 Upgrade from NCSS 10

A wide range of new procedures, plots, and enhancements were added in the upgrade from NCSS 10 to NCSS 11.

Procedures Added in NCSS 11

• Conditional Logistic Regression
• Multiple Regression - Basic
• Negative Binomial Regression
• Zero-Inflated Negative Binomial Regression
• Zero-Inflated Poisson Regression
• Geometric Regression
• Fractional Polynomial Regression
• Passing-Bablok Regression for Method Comparison
• Robust Linear Regression (Passing-Bablok Median-Slope)
• Logistic Regression (including confidence intervals for AUC)
• Two-Stage Least Squares
• -
• Scatter Plots with Error Bars
• Scatter Plots with Error Bars from Summary Data
• Error-Bar Charts from Summary Data
• Error-Bar Charts from Summary Data (2 Factors)
• -
• Descriptive Statistics - Summary Tables
• Descriptive Statistics - Summary Lists
• -
• One ROC Curve and Cutoff Analysis
• Comparing Two ROC Curves - Independent Groups Design
• Comparing Two ROC Curves - Paired Design
• Single Sample Binary Diagnostic Test Analysis
• -
• Correlation
• Circular Data Correlation
• -
• Reference Intervals
• -
• Appraisal Ratio Studies
• Comparables Appraisal
• Hybrid Appraisal Models
• Multiple Regression for Appraisal
• -
• Acceptance Sampling for Attributes
• Operating Characteristic Curves for Acceptance Sampling for Attributes
• -
• Linear Programming with Bounds
• Mixed Integer Programming
• Quadratic Programming
• Transportation
• Assignment
• Minimum Spanning Tree
• Shortest Route
• Maximum Flow
• Minimum Cost Capacitated Flow
• Transshipment
• -
• Dwass-Steel-Critchlow-Fligner MC Test (in the One-Way Analysis of Variance Procedure)

Examples of new plots:

• Scatter Plot with Error Bars

• ROC Curves

• Summary Table Plots

• Operating Characteristic Curves for Acceptance Sampling for Attributes

• Appraisal Ratio Horizontal Equity Plot

• Appraisal Ratio Price-Related Bias Scatter Plot

Enhancements

Data Simulation Procedure

The Data Simulation procedure was enhanced to include a much larger selection of distributions.

One-Way ANOVA Residuals

The One-Way Analysis of Variance procedure now gives the ability to store residuals.

Contingency Tables Table Entry

The table entry in the Contingency Tables (Crosstabs / Chi-Square Test) procedure was improved.

One Proportion Data Entry

A Database Data Entry option is now available in the One Proportion procedure.

Export Tool

The Export tool now has the ability to select variables for export.

Output Titles

Software version titles are given in the output.

Error-Bar Plots

The error-bar plots procedures now give the options for Confidence Interval and Range.

Spreadsheet Controls

The Spreadsheet controls have been updated.

3D Charting

The 3D charting control has been updated for additional processing and display speed.

Column Selection Tools

The column selection tools have been dramatically enhanced.

Auto-Scaling of Ticks

The Auto-Scaling of the ticks of the numeric axis of plots has been improved.

Filters and Missing Values in Time Series and Forecasting

Filter and missing values options are now available in all Forecasting and Time Series procedures.

Procedure Menus

The procedure menus were enhanced to be more intuitive.

Features and Procedures added in the NCSS 10 Upgrade from NCSS 9

The upgrade from NCSS 9 to NCSS 10 contained a very large number of new procedures, plots, and features.

New Procedures

Addition of new procedures and tests:

• Box-Cox Transformation
• Box-Cox Transformation for Two or More Groups (T-Test and One-Way ANOVA)
• Box-Cox Transformation for Simple Linear Regression
• Point-Biserial and Biserial Correlations
• Cochran's Q Test with Multiple Comparisons
• Michaelis-Menten Equation
• Bland-Altman Plot and Analysis
• Deming Regression
• Harmonic Regression
• 3D Bar Charts
• 3D Bar Charts (2 Factors)
• 3D Line Charts
• 3D Line Charts (2 Factors)
• 3D Scatter Plots (Revised) - Regression Surfaces Added
• Grubbs' Outlier Test
• Paired T-Test for Equivalence
• Paired T-Test for Non-Inferiority
• Testing Non-Inferiority with Two Independent Samples
• Testing Equivalence with Two Independent Samples
• Stem-and-Leaf Plots
• Back-to-Back Stem-and-Leaf Plots
• X-bar and R Charts
• X-bar and s Charts
• X-bar Charts
• R Charts
• s Charts
• CUSUM Charts
• Moving Average Charts
• EWMA Charts
• Individuals and Moving Range (I-MR) Charts
• P Charts
• NP Charts
• C Charts
• U Charts
• Capability Analysis
• Normality Tests
• Two Proportions (Revised) - Data Input Enhanced
• Two Proportions - Non-Inferiority Tests
• Two Proportions - Superiority Tests
• Two Proportions - Two-Sided Tests vs. a Margin
• Two Proportions - Equivalence Tests
• Two Correlated Proportions (McNemar Test) (Revised) - Data Input Enhanced
• Two Correlated Proportions - Superiority Tests
• Two Correlated Proportions - Equivalence Tests
• Two Correlated Proportions - Non-Inferiority Tests
• Contingency Tables (Revised) - Data Input Enhanced with New Summary Table Input Option
• Cochran-Armitage Proportion Trend Test (in Contingency Tables)
• Gray's Test for Comparing Groups (in Cumulative Incidence)
• Pepe and Mori's Test for Comparing Groups (in Cumulative Incidence)
• Lag Plots
• Analysis of Runs including Wald-Wolfowitz and Other Runs Tests for Randomness
• Normal Probability Plots
• Weibull Probability Plots
• Uniform Probability Plots
• Exponential Probability Plots
• Gamma Probability Plots
• Chi-Square Probability Plots
• Half-Normal Probability Plots
• Log-Normal Probability Plots
• Probability Plot Comparison
• Reference Intervals with a Covariate
• One-Way ANOVA (Revised) - 3 New Normality Tests
• One-Way ANOVA (Revised) - Welch's Unequal Variance Test
• One-Way ANOVA (Revised) - Sample Group Standard Deviations

Examples of new plots:

• Michaelis-Menten Equation Plot

• Lag Plot

• 3D Bar Charts

• 3D Line Charts

• Regression Surfaces

• Box-Cox Transformation Plot

• Quality Control Chart

• Reference Interval Plot

• Stem-and-Leaf Plot

• Model from Harmonic Regression

 

New Features

64-bit option

With the 32-bit version of NCSS, analysis of datasets with greater than 25 million cells can result in memory errors. The 64-bit version of NCSS 10 increases this limit by many, many times, so that the analysis limits are only those imposed by the amount of memory on the computer. The purchase of NCSS 10 gives you access to both the 32-bit and 64-bit versions.

Enhanced Data Stacking and Unstacking Tools

NCSS 10 includes improved tools for stacking and unstacking data. New options in the stack data tool allow you to automatically rename output columns and include label columns in the stacking process so that information about each observation is maintained in the stacked data. The new data unstacking tool also allows you to include label variables and automatically renames output columns. It also allows you to unstack data using a grouping column, which provides greater flexibility in data processing.

New Contrast Variable Tool

NCSS 10 includes an all-new data tool to help you create a set of contrasts or binary indicator variables from categorical data columns. These new variables are stored right on the data table.

New Transformations

NCSS 10 includes new date-variable processing transformations that allow you to work directly with Date-Time values without having to first translate them to Julian dates before extracting Month, Day, Year, etc. from the date values. NCSS 10 also includes a new transformation to convert Julian dates to Date-Time values and Date-Time values to Julian dates.

Copy with Filter

NCSS 10 includes a new data copy function that only copies values from rows that pass the filter.