Вы находитесь на странице: 1из 54
Decisions IBM SPSS Product cAtALoGuE drive better outcomes through predictive analytics check out what’s new

Decisions

IBM SPSS Product cAtALoGuE

Decisions IBM SPSS Product cAtALoGuE drive better outcomes through predictive analytics check out what’s new IBM

drive better outcomes through predictive analytics

check out what’s new

IBM SPSS Statistics 19 IBM SPSS Advanced Statistics IBM SPSS Direct Marketing IBM SPSS Amos IBM SPSS Modeler 14.1

www.ibm.com/spss/uk

and

what’s cool

IBM SPSS Statistics 14-day free trial

The Fall of the Wall: Statistics and Modeler as one application

Make Better Decisions Through Predictive Intelligence

Business Analytics software

Dear valued customer: Welcome to our 201 1 catalogue, Decisions . This is an exciting

Dear valued customer:

Welcome to our 2011 catalogue, Decisions. This is an exciting time to be in analytics because analytics is increasingly seen as the key to a smarter planet. Smarter schools. Smarter businesses. Smarter healthcare. Smarter government.

In the following pages, you’ll find a wealth of information about the products that support effective, reliable analytics … and about how some of your peers have used these products to drive improved outcomes and achieve outstanding results.

For instance, check out how Avis cut the cost of email marketing as a percentage of revenue by almost half; and how Zorg en Zekerheid, an independent health insurer in the Netherlands, uses predictive analytics to detect and prevent fraud.

We continue to add to both the technical sophistication of our products and their ease of use.

We’ve made our core IBM SPSS Statistics products even easier to use, with syntax enhancements and Automatic Linear Models

IBM SPSS Statistics Server offers greater speed and performance

We’ve improved our highly popular IBM SPSS Direct Marketing product by adding new scoring capabilities

And IBM SPSS Modeler continues to lead the field in data mining, and now offers direct integration with Cognos Business Intelligence software

Perhaps, in the coming year, you’ll use our software to bring greater focus to your college’s recruitment efforts. Or perhaps you’ll manage the product assortment at multiple store locations more effectively. Or maybe uncover surprising insights in your customers’ blog posts that will lead to a product breakthrough.

Whatever your needs, we’re committed to continuing to offer you a full range of innovative software to help you reach your goals. And we look forward to building a lasting relationship that will help us all build a smarter planet.

Sincerely,

that will help us all build a smarter planet. Sincerely, Deepak Advani VP, Predictive Analytics table

Deepak Advani VP, Predictive Analytics

table of contents

Introduction to predictive analytics

The Data Analysis Path to Success ��������������������4 The SPSS end-to-end story

Get More Value with Every Release ���������������������6 See what you’ll gain when you upgrade from an earlier version of IBM SPSS Statistics

IBM ® SPSS ® Statistics 19�������������������������������������8 Analyse data with the world’s leading statistical software

Most popular products

IBM ® SPSS ® Statistics Base 19 �������������������������10 Your first choice for data analysis

IBM ® SPSS ® Data Preparation 19 ����������������������12 Improve data preparation for more accurate results

IBM ® SPSS ® Decision Trees 19 ��������������������������14 Create classification trees for better identification of groups and relationships

IBM ® SPSS ® Custom Tables 19 ������������������������16 Analyse data easily and communicate results effectively

IBM ® SPSS ® Advanced Statistics 19 �����������������18 Use powerful techniques to analyse complex data

IBM ® SPSS ® Regression 19��������������������������������20 Make better predictions using regression procedures

IBM ® SPSS ® Missing Values 19��������������������������22 Build better models when you fill in the blanks

IBM ® SPSS ® Forecasting 19�������������������������������24 Build expert time-series forecasts — in a flash

Special-purpose products

IBM ® SPSS ® Bootstrapping 19���������������������������26 Ensure the stability of your models

IBM ® SPSS ® Complex Samples 19 ��������������������28 Correctly and easily compute statistics for complex samples

IBM ® SPSS ® Statistics Server 19 �����������������������30 Analyse “big data” and data in dispersed organisations

IBM ® SPSS ® Direct Marketing 19�����������������������31 Easily identify the right contacts and improve campaigns

IBM ® SPSS ® Developer 19 ���������������������������������32 Complete flexibility to build customised functional- ity and procedures in R ® or Python ® your way

IBM ® SPSS ® Neural Networks 19�����������������������33 Discover complex relationships in your data more easily

IBM ® SPSS ® Visualization Designer 1����������������34 Easily create and share customised visualisations

IBM ® SPSS ® Exact Tests 19 �������������������������������35 Reach accurate conclusions with small samples or rare occurrences

IBM ® SPSS ® Categories 19 ��������������������������������36 Predict outcomes and reveal relationships through perceptual maps of categorical data

IBM ® SPSS ® Conjoint 19 ������������������������������������38 Discover what drives your customers’ purchase decisions

IBM ® SPSS ® Data Collection Data Entry 6 ��������40

A faster, more effective way to collect and manage

survey research data

Complementary products

IBM ® SPSS ® Text Analytics for Surveys 4 ���������42 Easily make your survey text responses usable in quantitative analysis

IBM ® SPSS ® SamplePower ® 3 ���������������������������44 Save time, effort and money by finding the appropriate sample size for your study

IBM ® SPSS ® Amos 19 ��������������������������������������46 Get your research noticed — take your analysis

to the next level

Data mining and modeling

IBM ® SPSS ® Modeler 14�1����������������������������������48 Solve your toughest challenges with data mining

IBM ® SPSS ® Modeler Premium 14�1������������������50 Improve model accuracy with unstructured data

IBM ® SPSS ® Modeler Professional 14�1 ������������52 Make better decisions through predictive intelligence

Who uses IBM SPSS analytics?

Businesses use analytics for…

Sales and marketing forecasting and budgeting

Database and direct marketing

Product attribute testing

New product development

Financial account balancing

Risk and credit management

Customer and employee satisfaction surveys

Planning for facility and staffing resources

Market basket analysis

Operational excellence

…and more

Colleges and universities use analytics for…

Teaching and student assessment

Administration

Enrollment management

Alumni development

Research

…and more

Schools and school districts use analytics for…

Student assessment

Program assessment

Community and staff surveys

Planning and budgeting

Facility maintenance scheduling

…and more

Government agencies use analytics for…

Human Capital Management

Program evaluation

Community and employee surveys

Fighting crime and protecting public safety

Promoting public health

Preventing fraud, waste and abuse

Environmental impact studies

…and more

Medical and healthcare organisations use analytics for…

Evidence-based medicine

Treatment outcome analysis

Behavioural and biomedical research

Outcome management

…and more

To order, phone 0845 345 0935

3

Follow the Path to Improved Outcomes through Predictive Analytics

CAPTURE information, PREDICT outcomes and ACT on insights

To get the answers you need for successful decision making, it’s important to follow all the steps in the data analysis process – and using the right data analysis tools along the way will help you arrive at those decisions faster and more accurately.

When your organisation follows the analytical path illustrated here, you will benefit in a number of ways. You’ll gain a better understanding of your current situation, be able to consider the most appropriate options, predict what is likely to happen next and take actions to

improve outcomes. Specific benefits to your business include the ability to acquire, grow and retain customers; reduce costs; minimise risk and fraud; and improve efficiency.

IBM SPSS predictive analytics products are delivered in an easy- to-integrate, open technology platform. So take a look at how you can turn your organisation’s data into a strategic asset that gives your organisation a competitive advantage.

Data Collection

IBM SPSS Data Collection Web Interviews

IBM IBM SPSS SPSS Data Data Collection C

Planning

IBM SPSS SamplePower

IBM SPSS Complex Samples

Data Access

IBM SPSS Statistics Base

Data Data Entry Entry
Data Data Entry Entry
Data Access IBM SPSS Statistics Base Data Data Entry Entry All All Modules Modules D a
All All Modules Modules D a IBM IBM IB Data Data Analysis
All All Modules Modules
D
a
IBM
IBM
IB
Data
Data Analysis

IBM SPSS PSS Conjoint Conjoint

Data Management

and an Preparation

IBM SPSS Statistics Base

IBM IB M SPSS Data Preparation

IBM SPSS Missing Values

Deployment Deployment
Deployment
Deployment

IBM SPSS Collaboration and Deployment Services

Reporting Reporting
Reporting
Reporting

IBM IBM SPSS Te xt Analytics for Surveys

IBM SPSS Modeler Professional IBM SPSS Modeler Premium

IBM SPSS Statistics Base IBM SPSS Amos IBM SPSS Advanced Statistics IBM SPSS Bootstrapping IBM SPSS Categories IBM SPSS Direct Marketing IBM SPSS Exact Tests IBM SPSS Forecasting IBM SPSS Neural Networks IBM SPSS Regression

IBM SPSS Statistics Base IBM SPSS Custom Tables IBM SPSS Visualization Designer IBM SPSS Data Collection

4

To order, phone 0845 345 0935

Rely on IBM SPSS products at every stage of the analytical process

1.

Planning

Beginning with the end result in mind, the

first steps are to set objectives, identify data sources and carefully craft the process. Products:

IBM SPSS SamplePower

IBM SPSS Conjoint

IBM SPSS Complex Samples

2. Data collection

Data is collected through surveys, online

activity, call centers and more. Products:

IBM SPSS Data Collection Web Interviews

IBM SPSS Data Collection Data Entry

3. Data access

Data is brought in from available sources,

using ODBC or direct file input. Products:

IBM SPSS Statistics Base

All modules

4. Data management and data

preparation

Data is reviewed for suspicious, invalid,

or missing cases, variables and data values. Products:

IBM SPSS Statistics Base

IBM SPSS Data Preparation

IBM SPSS Missing Values

IBM SPSS Text Analytics for Surveys

IBM SPSS Modeler Professional

IBM SPSS Modeler Premium

5. Data analysis

In this step, the data is examined, tested, explored and transformed. Patterns are identified, hypotheses are tested and information is extracted.

Products for understanding data:

IBM SPSS Statistics Base

IBM SPSS Data Preparation

IBM SPSS Complex Samples

IBM SPSS Modeler

Products for predicting numerical outcomes:

IBM SPSS Statistics Base

IBM SPSS Regression

IBM SPSS Advanced Statistics

IBM SPSS Complex Samples

IBM SPSS Neural Networks

IBM SPSS Modeler

IBM SPSS Amos

Products for identifying groups:

IBM SPSS Statistics Base

IBM SPSS Direct Marketing

IBM SPSS Regression

IBM SPSS Advanced Statistics

IBM SPSS Complex Samples

IBM SPSS Custom Tables

IBM SPSS Categories

IBM SPSS Decision Trees

IBM SPSS Exact Tests

IBM SPSS Neural Networks

IBM SPSS Modeler

Product for forecasting time-series data:

IBM SPSS Forecasting

6. Reporting

Data is summarised, put in tables and charts and ready for consumption. Products:

IBM SPSS Statistics Base

IBM SPSS Custom Tables

IBM SPSS Visualization Designer

IBM SPSS Data Collection

7. Deployment

Data, reports and procedures are distrib-

uted to end users globally, with interaction

and access managed centrally. Products:

IBM SPSS Collaboration and Deployment Services

8.

Success!

Take a moment to celebrate – you’ve done it! But now it’s back to planning how to maintain your competitive advantage –

once again using IBM SPSS products to

help you reach the next level of success.

To order, phone 0845 345 0935

5

Get More Value with Every Release of IBM SPSS Statistics

We continually update our IBM SPSS Statistics product family to deliver the power, versatility and ease of use you need to address your toughest analytical challenges. With each release, we add new procedures, features and platform support to help you work faster and more effectively.

If you’re using an earlier version of IBM SPSS Statistics, you’ll gain all of these time-saving features - and many more - when you upgrade to version 19.

Added in IBM SPSS Statistics 18

Prepare data in a single step using the new Automated Data Preparation feature

New Nonparametric tests in IBM SPSS Statistics Base

Post computed categories in IBM SPSS Custom Tables

IBM SPSS Direct Marketing Module

IBM SPSS Bootstrapping Module

Rule checking on Secondary SPC Charts

IBM SPSS Statistics Developer

Ability to view significance tests in the main results table in IBM SPSS Custom Tables

Interactive Model Viewer on Two- Step Cluster Analysis and Automated Data Preparation procedures

Improved display of large pivot tables

Improved performance on procedures within IBM SPSS Statistics Base Server for Frequencies, Descriptives, Crosstabs

Support for 64-bit hardware on desktop for Windows ® and Mac ®

Support for Snow Leopard ® on Mac OS X 10.6

Added in IBM SPSS Statistics 17

6

Syntax Editor with features to make it easier to create, test and deploy syntax jobs

Switch user interface language

Mac OS X and Linux ® platforms can connect clients to IBM SPSS Statistics Server

Updated plug-ins for Python, .NET ® and R

Support for graphic packages written in R

Create user-defined interfaces for existing procedures and user-defined procedures with Custom Dialog Builder

Call front-end Python scripts or scripting APIs explicitly from within back-end Python programs

Support for Predictive Enterprise View, a common data interface that can be defined once and used by all IBM SPSS analytic tools

Administrative enhancements in IBM SPSS Statistics Server, including opti- mised multithreading, virtualisation support and a “file in use” message to reduce errors in data created by more than one person writing to an IBM SPSS Statistics file at the same time

Read access to IBM SPSS Statistics data files as an ODBC/JDBC data source, allowing these files to be read using SQL

Codebook procedure to automatically describe the dataset

Spell-checking of long text strings

IBM SPSS EZ RFM module

Multiple imputation of missing data in IBM SPSS Missing Values module

Regularisation methods: Ridge regression, the Lasso, Elastic Net all in IBM SPSS Categories

Model selection methods: 632(+), bootstrap, cross validation (CV) all in IBM SPSS Categories

Nearest Neighbor analysis in IBM SPSS Statistics Base

Median transformations function in COMPUTE procedure

Option to use aggressive versus conservative rounding in COMPUTE procedure

Create new variables that contain the values of existing variables from preceding or subsequent cases

Graphboard integration, enabling users of IBM SPSS Statistics products to deploy new or customer graph templates created in the new IBM SPSS Visualization Designer stand-alone module

Wrapping and shrinking of wide tables in Word and PowerPoint ®

Smartreader feature to allow the viewing and pivoting of IBM SPSS Statistics output

Added in IBM SPSS Statistics version 16

Mac and Linux versions of IBM SPSS Statistics

Several multithreaded procedures for improved performance and scalability

In the Data Editor: ability to customise variable view, spell checking for value labels and variable labels, sort by variable name, type, format, etc.

Unicode support

Import/export Excel 2007 data

Syntax to change string length and basic data type of existing variables

Creation of value labels and missing values on strings of any length

Ability to set a permanent default working directory

IBM SPSS Neural Networks module

Complex Samples Cox Regression added to IBM SPSS Complex Samples

Latent Class Analysis in IBM SPSS Amos

Partial Least Squares regression

Support for R algorithms

Find and Replace feature in the Output Viewer

To order, phone 0845 345 0935

What’s new in version 19

IBM SPSS Statistics is now even faster and easier to use. The new release offers a number of new and improved procedures, as well as ease-of-use enhancements and support for additional platforms including:

Faster performing tables – Save time when creating reports that involve large tables or a large number of smaller tables. Creating tables in the output is up to 200 percent faster than before. In addition, tables take up less memory (an optional setting).

Automatic Linear Models – A new family of algorithms makes it possible for business analysts and analytic professionals to build powerful linear models in an easy and automated manner.

Generalized Linear Mixed Models (GLMM) – Create more accurate models for predicting non-linear outcomes based on hierarchical/ nested data or data which includes repeated measures. GLMM extends the linear model so that the target is linearly related to the factors and covariates through a specified link function, the target can have a non-normal distribution and the observations can be correlated.

IBM SPSS Direct Marketing – Enhancements make it easier to access pre-built models, score data and leverage customer information in Salesforce.com. The smart output feature enables you to easily understand your results.

Syntax Editor – We’ve implemented more than a dozen performance and ease-of-use enhancements for the syntax editor, including tip displays of the “name,” improved scrolling, improved indentation of lines and many more.

Statistics portal – Provide any internal or external user with interactive online access to the analysis they require in the server version of Statistics 19. The Statistics portal allows you to filter the data being analysed, and provides users with rich capabilities for interacting with the output – but administrators remain in complete control of what results various analysts or departments can access and edit.

Optimisation techniques – Use methods such as boosting and bagging to improve performance and stability when working with large datasets.

Easier modeling and scoring within UI - A single dialog box enables you to specify model and transformation files, identify discrepancies in variable attributes and change attributes as necessary before applying a model.

IBM System z Support – Run IBM SPSS Statistics Base Server on your IBM System z machines using Linux for powerful, enterprise- wide analysis.

To order, phone 0845 345 0935

7

IBM ® SPSS ® Statistics 19

Analyse data with the world’s leading statistical software

With IBM SPSS Statistics you can:

Easily access, manage and analyse any kind of dataset

Gain reliable results with a broad range of tests and procedures

Report results in easy-to-understand formats

Whether you are a beginner or an experienced analyst or statistician, IBM SPSS Statistics puts the power of advanced statistical analysis in your hands.

Solves business and research problems

IBM SPSS Statistics offers superior capabilities, flexibility and usability that are not avail- able in traditional statistical software. Organisations around the world rely on it to:

Identify which customers are likely to respond to specific promotional offers

Forecast future trends to better plan organisational strategies, logistics and manufacturing processes

Detect fraud and minimise business risk

Report test and program results to the government and other regulatory agencies

Identify groups, discover relationships between groups and predict future events

Improves the entire analytical process

IBM SPSS Statistics addresses the entire analytical process from planning and data

preparation to analysis, reporting and deployment. It enables you to get a quick look at your data, formulate hypotheses for additional testing and then carry out a number of statistical and analytic procedures to help clarify relationships between variables, create clusters, identify trends and make predictions.

Quickly access and analyse massive datasets

IBM SPSS Statistics makes it easy for you to quickly access, manage and analyse any kind of dataset, including survey data, corporate databases or data downloaded from the Web. It can also process Unicode data, enabling your organisation to view, analyse and share data written in multiple languages.

Prepare your data in a flash

With IBM SPSS Statistics, you can prepare data for analysis quickly and easily. Eliminate the time-consuming task of labeling all your data by creating your labels once and applying them to your entire dataset. Make sure your data is clear and organised properly for analysis by automatically identifying duplicate cases.

In addition, Visual Binning allows you to easily create bands (such as breaking income into bins of 10,000 or ages into demographic groups). A histogram then enables you to interactively create cutpoints and automatically create data value labels for them.

Powerful statistics for better analysis

IBM SPSS Statistics includes an extensive variety of procedures for descriptive analysis, numerical prediction, group identification and forecasting that help you quickly generate the most accurate results for specific data types. And through the IBM SPSS Statistics Programmability Extension, users can create procedures in third-party programming languages such as Python and Microsoft .NET, as well as access and use the many statistical procedures written in R.

Easy, flexible reporting options

IBM SPSS Statistics makes it easy to integrate your output into your reports by enabling you to automatically export results into Microsoft Word, PowerPoint and Excel (including Excel 2010), and as a PDF. In addition, through the Custom Dialog Builder, you can create procedures and make them available to others, making it easier to ensure that all analyses are performed correctly.

specifications

System requirements

For IBM SPSS Statistics 19 for Windows • Operating system: Microsoft Windows XP (Professional, 32-bit) or Vista ® (Home, Business, 32- or

64-bit), Windows 7 (32- or 64-bit) Hardware

– Intel or AMD x86 processor running at 1 GHz or higher

– Memory: 1 GB or more recommended

– Minimum free drive space:

800 MB*

– DVD drive

– Super VGA (800x600) or higher- resolution monitor

– For connecting with an IBM SPSS Statistics Server, a network adapter running the TCP/IP network protocol

Browser

– Internet Explorer ® 7 or 8

For IBM SPSS Statistics 19 for MAC OS X

Operating system

– Apple Mac OS 10�5 (Leopard ) or

10�6 (Snow Leopard) (32-bit or 64-bit versions)

Hardware

– Intel processor

– Memory: 1 GB or more

recommended

– Minimum free drive space: 800 MB*

– DVD drive

– Super VGA (800x600) or a

higher-resolution monitor

Browser

– Web browser: Mozilla ® Firefox ®

2�x and 3�x

IBM SPSS Statistics 19 for Linux

Operating system: Any Linux OS that meets the following require-

ments (32 bit only):

– Kernel 2�6�3

– glibc 2�8

– libstdc++5

Hardware

– Processor: Intel or AMD x86 processor running at 1 GHz or

XFree86-4�7

higher

– Memory: 1 GB or more

recommended*

– Minimum free drive space: 800 MB

– DVD drive

– Super VGA (800x600) or a higher- resolution monitor Browser

– Mozilla Firefox 2�X and 3�X

* Installing Help in all languages requires 1.1 GB free drive space

Available on the following platforms:

Windows n Mac n Linux

Driving to greater customer insight Avis cuts email marketing costs with predictive analytics

Avis Europe is a leading car rental company with a network of more than 2,800 locations. Approximately 86 percent of Avis Europe’s revenues in 2008 were generated in France, Germany, Italy, Spain and the United Kingdom.

Brand leadership, service differentiation and cost effectiveness are part of Avis Europe’s strategic focus and ‘We Try Harder’ philosophy. For the Avis customer, this translates into quick, professional services, a high quality vehicle at a reasonable price and targeted communications.

Avis Europe turned to IBM SPSS software to help create targeted and cost-effective email campaigns and build customer retention through timely and relevant contact.

The car rental group selected the IBM® SPSS® Modeler data mining workbench to develop customer profiles and segment its data more accurately.

IBM sPss training

As a result of using IBM SPSS predictive analytics software, the cost of email marketing as a percentage of revenue was cut almost by half in 2009 compared to 2008.

According to Chris Parker, direct analytics specialist at Avis Europe, “We are now better at sending the right emails to the right people at the right time. This new targeted approach helps Avis cut email marketing costs and, hence, maximise revenue.

“The Customer Segmentation project allows us to keep in touch with our large database, but with all the benefits of a one-to-one relationship,” he adds. “The ability to identify and stay ahead of customers’ ever-changing activities and needs is one of the biggest benefits provided by IBM SPSS predictive analytics software.”

Get the maximum return on your investment in IBM SPSS technology by taking advantage of our flexible and convenient training options.

We offer a variety of training options and courses customised to suit your needs and quickly get you up and running with IBM SPSS software and solutions.

Our instructors and consultants use an interactive, engaging hands-on approach in a small group environment. Such focused instruction can dramatically increase your performance Click here for more information.

1.800.253.2575 Order now!

9

IBM ® SPSS ® Statistics Base 19

Your first choice for data analysis

With IBM SPSS Statistics Base, you can:

Quickly access and manage data

Get a “first look” through descriptive statistics

Perform fundamental tests to classify data, uncover relationships and make predictions

You can take the analytical process from start to finish with IBM SPSS Statistics Base. In addition to the data preparation, data management, output management and charting features now available in all IBM SPSS Statistics modules, IBM SPSS Statistics Base offers the most frequently used procedures for statistical analysis – the fundamental toolset that every analyst should have.

8.59% 19.53% 16.02% 17.58% 38.28%
8.59%
19.53%
16.02%
17.58%
38.28%
8.59% 19.53% 16.02% 17.58% 38.28%

See and graph the characteristics of your dataset quickly and efficiently with IBM SPSS Statistics Base.

These procedures will enable you to get a quick look at your data, formulate hypotheses for additional testing and then carry out a number of procedures to help clarify rela- tionships between variables, create clusters, identify trends and make predictions.

Get a quick data “snapshot”

You can quickly understand the basic structure of your data through the foundational statistics in IBM SPSS Statistics Base. These include crosstabulations, frequencies, descriptives, descriptive ratio statistics, ANOVA and ANCOVA, correlation and the procedures for comparing means and exploring data.

Map relationships in your data

IBM SPSS Statistics Base also includes a

number of procedures to help you identify groups and predict outcomes. These proce- dures include:

Factor analysis

K-means Cluster Analysis

Hierarchical Cluster Analysis

TwoStep Cluster Analysis

Discriminant analysis

Linear regression

Analysis • Discriminant analysis • Linear regression Download datasheet and full specification • Ordinal

Download datasheet and full specification

Ordinal regression (also called PLUM)

Nearest Neighbor analysis

regression (also called PLUM) • Nearest Neighbor analysis The Data Editor makes it easy to manage

The Data Editor makes it easy to manage data from IBM SPSS Statistics and IBM SPSS Data Collection files, as well as text files and data from other applications and databases.

Even if you use one or more of the other modules in the IBM SPSS Statistics family for specific kinds of analysis, IBM SPSS Statistics Base will continue to form the basis of many deployments, since it contains statistical tests and procedures that are fundamental to many analyses.

specifications

Key features

Descriptive statistics

Crosstabulations

Frequencies; Descriptives; Explore; Descriptive ratio statistics

Bivariate statistics

Means; t tests; ANOVA; Correlation (Bivariate, Partial, Distances); and Non-parametric tests

Prediction for numerical out- comes and identifying groups

Factor Analysis

K-means Cluster Analysis

Hierarchical Cluster Analysis

10

TwoStep Cluster Analysis

Linear Regression

Ordinal regression — PLUM

Multithreaded algorithms: SORT, correlation, partial correlation, linear regression, factor analysis

Nearest Neighbor analysis, which can be used for prediction or for classification

Non-parametric tests provide multiple comparisons and perform

Discriminant

efficiently on large datasets

Data Editor

– Configure attributes, so that some

can be hidden

– Spell check value labels and

variable labels

– Sort by variable name, type, format, etc�

– Use Find and Replace functionality

Easily eliminate duplicate records with the Identify Duplicate Cases tool

Make sense and keep track of your data files by adding notes to them with the Data File Comments

command

Create read-only datasets

More accurately describe your data using longer variable names (up to

64 bytes)

Create value labels up to 120

characters

Clone or duplicate datasets

Apply an extended Variable

Properties command to customise properties for individual users

Longer text stings (up to 32,000

bytes)

• Define Variables Properties tool

Copy Data Properties tool

Data Restructure Wizard

Continued on PG. 11

To order, phone 0845 345 0935

“ Automatic Linear Models (ALM) is a great procedure for handling modeling of data - this is something we can add to our list of tools we can now offer our clients. While ALM is the feature I liked best, the faster table rendering will likely have the biggest impact on our business given the large amount of IBM SPSS Statistics output we produce, most of which uses IBM SPSS Custom Tables functionality. ”

– Brian Robertson, PhD Director of Research, Market Decisions

Aggregate data to external or to the active data file

Automatically convert string variables to numeric with Autorecode

– Spell-checking of long text strings

Date and Time Wizard:

– Easily work with data containing time and dates in IBM SPSS Statistics

– Create a time/date variable from a string containing a date variable

– Create a time/date variable from variables that include individual date units, such as month or year

– Calculate times and dates

– Separate a date unit from a time/ date variable

Apply splitters in the Data Editor for easier viewing of wide or long data files

Create your own dictionary informa- tion for variables by using Custom Attributes

Customise the viewing of extremely wide files with Variable Sets

Use syntax to change string length and basic data type

Set a permanent default working directory

Transformations

• Easily find and replace text strings in your data using the find/replace function

Recode string or numeric value

Recode values into consecutive integers

Create conditional transformations using DO IF, ELSE IF, ELSE and END IF statements

Use programming structures, such as do repeat-end repeat, loop-end

loop and vectors

Compute variables using arithmetic,

cross-case, date and time, logical,

missing-value, random-number,

statistical or string functions

Create variables that contain the values of existing variables from

preceding or subsequent cases

Count occurrences of values across

variables

Make transformations permanent or temporary

Execute transformations immedi-

ately, batched or on demand

Reporting

Reports

– OLAP cubes

– Case summaries

– Report summaries

Graphs

Categorical charts

– 3-D Bar: Simple, cluster and

stacked

– Bar: Simple, cluster, stacked,

dropped shadow and 3-D

– Line: Simple, multiple and drop-line

– Area: Simple and stacked

– Pie: Simple, exploding and 3-D effect

– High-low: High-low-close,

difference area and range bar

– Box plot: Simple and clustered

– Error bar: Simple and clustered

– Error bars: Add to bar, line and area charts; and confidence level, S�D or S�E�

– Dual-Y axes and overlay

Scatterplots

– Simple, grouped, scatterplot

matrix and 3-D

– Fit lines: Linear, quadratic, or cubic regression; Lowess smooth-

er; confidence interval control;

and for total or subgroups, display spikes to line

– Bin points by color or marker size

to prevent overlap

Density charts

– Population pyramids: Mirrored axis to compare distributions; with

or without normal curve

– Dot charts: Stacked dots show distribution; symmetric, stacked

and linear

– Histograms: With or without nor-

mal curve; custom binning options

Quality control charts

– Pareto, X-Bar, range, Sigma, indi-

vidual chart or moving range chart

– Rule-checking performed on

primary and secondary charts

– Automatic flagging of points that

violate Shewhart rules, the ability

to turn off rules and the ability to

suppress charts

Diagnostic and exploratory charts

– Caseplots and time-series plots

– Probability plots

– Autocorrelation and partial auto-

correlation function plots

– Cross-correlation function plots

– Receiver-Operating Characteristics

Multiple use charts

– 2-D line charts (with 2 scale axes)

– Charts for multiple response sets

Custom charts

– Charts for multiple response sets • Custom charts View UK prices and purchase – Graphics

View UK prices and purchase

– Graphics Production Language

(GPL), a custom chart creation

language, enables advanced users to attain a broader range of chart

and option possibilities than the interface supports to create mixed charts and more

Layout options

– Paneled charts: Create a table

of subcharts, one panel per level or condition; multiple row and

columns

– 3-D effects: Rotate, modify depth and display backplanes�

Chart templates

– Save selected characteristics of a chart and apply them to others automatically

– Apply the following attributes at creation or edit time: Layout, titles,

footnotes and annotations; chart

element styles; data element styles; axis scale range;

axis scale settings; fit and reference lines; and scatterplot point binning – Tree-view layout and finer control of template bundles

System requirements

Please see p� 10 for complete

system requirements

Available on the following platforms:

Windows n Mac n Linux

11

IBM ® SPSS ® Data Preparation 19

Improve data preparation for more accurate results

With IBM SPSS Data Preparation you can:

Streamline the data preparation process

Eliminate labor-intensive manual checks

Reach more accurate conclusions

IBM SPSS Data Preparation enables you to easily identify suspicious and invalid cases, variables and data values; view patterns of missing data; and summarise variable dis- tributions. You can streamline the data prep- aration process so that you can get ready for analysis faster and reach more accurate conclusions.

Prepare data in a single step — automatically

Manual data preparation is a complex and time-consuming process. When you need results quickly, the Automated Data Preparation (ADP) procedure helps you detect and correct quality errors and impute missing values in one efficient step. The ADP feature provides an easy-to-understand report with complete recommendations and visualisations to help you determine the right data to use in your analysis.

Additional options for data preparation

Perform automatic data checks Eliminate time-consuming, tedious, manual checks by using the Validate Data procedure. This procedure enables you to apply rules to perform data checks based on each variable’s measure level (whether categorical or con- tinuous). Then, determine data validity and remove or correct suspicious cases at your discretion prior to analysis.

Quickly find multivariate outliers Easily detect multivariate outliers so you can further examine them and determine if they should be included in your analyses. The Anomaly Detection procedure searches for unusual cases based upon deviation, enabling you to flag outliers by creating a new variable.

enabling you to flag outliers by creating a new variable. Play a 10-minute demonstration Download datasheet

Play a 10-minute demonstration

Download datasheet and full specification

Bin or set cutpoints for scale variables With the Optimal Binning procedure, you can more accurately use algorithms designed for nominal attributes (such as Naïve Bayes and logit models). Optimal Binning enables you to bin – or set cutpoints for – scale

variables. Select from three types of optimal binning for preprocessing data prior to model building:

Unsupervised: Create bins with equal counts

Supervised: Take the target variable into account to determine cutpoints. This method is more accurate than unsupervised; however, it is also more computationally intensive.

Hybrid approach: Combines the un- supervised and supervised approaches. This method is particularly useful if you have a large amount of distinct values.

specifications

Key features

Automated Data Preparation

Recommends steps to speed up model building and improve predictive power

Determine objective, prepare dates and times for modeling, exclude low-quality input fields, prepare fields to improve data quality, rescale fields, continuous input and target fields, transform fields, perform feature selection and construction, name fields and apply transformations to data

Validate data

Validate data in the working data file

Basic checks:

– Maximum percent of missing values, single category cases and cases with a count of 1 – Minimum coefficient of variation

12

– Minimum standard deviation

– Flag incomplete IDs, duplicate IDs

and empty cases

Standard rules: Describe the data,

view single variable rules and apply

them to analysis variables

– Description of data

Distribution: Shows a thumbnail- size bar chart for categorical variables or histogram for scale variables

Min�/max� data values shown

– Single-variable rules:

Apply rules to identify missing or invalid values

• User-defined

• Custom rules: Define cross-variable rule expressions in which respon- dents’ answers violate logic

Output: Reports for invalid data

– Casewise report, specify by case

Specify the minimum number of

violations needed for a case

Specify the maximum number of

cases in the report

– Standard validation rules reports

Summarise violations by analysis

variable and rule

Display descriptive statistics

Save: Save variables that record rule violations and use them to clean data and filter out bad cases

– Summary variables:

Empty case indicator

Duplicate ID indicator

Incomplete ID indicator

Validation rule violation

– Indicator variables that record all validation rule violations

Identify unusual cases

Anomaly Detection searches

for unusual cases, based upon deviations from peer group, and

reasons for deviations

VARIABLES subcommand: Specify categorical, continuous and ID

variables, and list variables that are excluded from the analysis

HANDLEMISSING subcommand:

Specify the methods of handling

missing values in this procedure

The CRITERIA subcommand

specifies the following settings:

– Number of peer groups

– Adjustment weight on the

measurement level

– Number of reasons in the anomaly list

Continued on PG. 13

To order, phone 0845 345 0935

Identify suspicious or invalid cases, variables and data values easily with IBM SPSS Data Preparation

and data values easily with IBM SPSS Data Preparation Variables tab: The Validate Data dialog is

Variables tab: The Validate Data dialog is used to validate your data. The Variables tab shows variables in your file. Start by selecting the variables you are interested in and moving them to the Analysis Variables list.

in and moving them to the Analysis Variables list. Define standard rules: The Validate Data dialog

Define standard rules: The Validate Data dialog lets you create your own rules or apply predefined rules.

lets you create your own rules or apply predefined rules. Basic checks: You can specify basic

Basic checks: You can specify basic checks to apply to variables and cases in your file. For example, you can obtain reports that identify variables with a high percentage of missing values or empty cases.

with a high percentage of missing values or empty cases. Define custom rules: Create cross-variable rules

Define custom rules: Create cross-variable rules in which respondents’ answers violate logic (such as “pregnant males”).

answers violate logic (such as “pregnant males”). Standard rules: Apply rules to individual variables that

Standard rules: Apply rules to individual variables that identify invalid values, such as values outside a valid range or missing values.

such as values outside a valid range or missing values. Recommendations: Automated Data Preparation delivers

Recommendations: Automated Data Preparation delivers recommendations and allows users to drill in and examine the recommendations.

– Percentage and number of cases considered as anomalies and included in the anomaly list

– Cutpoint of the anomaly index to determine whether a case is considered as an anomaly

Save additional variables to the working data file including:

– Anomaly index

– Peer group ID, size and size in percentage

– Variable, variable impact measure, variable value and norm value associated with a reason

OUTFILE subcommand: Write a model to a filename as XML

PRINT subcommand prints:

– Case-processing summary

– Anomaly index list, anomaly peer ID list and anomaly reason list

– The Continuous Variable Norms table, for continuous variable, and the Categorical Variable Norms,

for categorical variable

– Anomaly Index Summary

– Reason Summary Table

Optimal Binning

Preprocess data with Optimal

Binning� Categorises one or more

continuous variables by distributing the values of each into bins

Select from the following methods:

– Unsupervised binning via the

equal frequency algorithm� It uses

the equal frequency algorithm to

discretise the binning input vari-

ables� Guide variable not required�

– Supervised binning via the MDLP (Minimal Description Length Principle) algorithm� Discretises binning input variables using the

MDLP algorithm without any

preprocessing� Ideal for small datasets� Guide variable required�

– Hybrid MDLP binning� Involves preprocessing via the equal fre- quency algorithm, followed by the MDLP algorithm� Ideal for large datasets� Guide variable required�

Specify the following criteria:

– How to define the minimum and maximum cutpoint for each bin- ning input variable, and the lower limit of an interval

– Whether to force-merge sparsely populated bins

interval – Whether to force-merge sparsely populated bins View UK prices and purchase – Whether missing

View UK prices and purchase

– Whether missing values uses listwise or pairwise deletion

Save new variables with binned values and syntax to a IBM SPSS Statistics syntax file

PRINT subcommand prints:

– The binning input variables’

cutpoint sets

– Descriptive information for all binning input variables

– Model entropy for binned variables

System requirements

Requirements vary according to platform

Available on the following platforms:

Windows n Mac n Linux

13

IBM ® SPSS ® Decision Trees 19

Create classification trees for better identification of groups and relationships

With IBM SPSS Decision Trees you can:

Create classification trees using your choice of algorithms

Identify patterns, segments and groups in your data

Choose from four established tree-growing algorithms

IBM SPSS Decision Trees creates classifica- tion and decision trees to identify groups, discover relationships and predict future events. By creating visual trees, you are able to present results in an intuitive manner – so you can more clearly explain results to non-technical audiences.

Why IBM SPSS Decision Trees should be added to your desktop:

Identify groups, segments and patterns in a highly visual manner with classification trees

Choose from CHAID, Exhaustive CHAID, C&RT and QUEST to find the best fit for your data

CHAID, C&RT and QUEST to find the best fit for your data Play a 10-minute demonstration

Play a 10-minute demonstration

Download datasheet and full specification

Present results in an intuitive manner – perfect for non-technical audiences

Save information from trees as new variables in data (information such as terminal node number, predicted value and predicted probabilities)

node number, predicted value and predicted probabilities) Greater Manchester Police make contact Community-based

Greater Manchester Police make contact Community-based policing addresses ‘signal crime’

Formed in 1974, Greater Manchester Police (GMP) has a workforce of more than 7,000 police officers and 3,500 support staff. Serving some 2.5 million people covering an area of 500 square miles and 10 metropolitan boroughs, it is one of the largest police forces in the United Kingdom. Public consultation plans, which report on community concerns, safety, crime and disorder, are a statutory function of GMP.

As part of the National Reassurance Project, GMP sent a survey to 9,000 households to establish a community-based policing program. The concept focused on identification of “signal crime.”

Before implementing an IBM SPSS solution, GMP had been using a paper system and Microsoft® Excel®. This method proved unfruitful due to Excel’s limited analysis capabilities.

Keith Bentley, chief superintendent (retired), said, “We wanted a solution that would not only find answers to questions that basic packages would miss, but also make data entry faster and more reliable.”

GMP decided to adopt an IBM SPSS solution to gather public views on policing that could be incorporated into the National Reassurance Project. The questionnaires were seamlessly created within the system, and the responses were scanned in electronically.

The solution has enabled GMP to reduce the cost and time spent on survey research and analysis. “Not only did we benefit from a massive reduction in time for this project, saving approximately €20,000 in two weeks, but these results are now being referenced by other Greater Manchester Police divisions,” said Bentley.

14

To order, phone 0845 345 0935

Uncover patterns in your data with powerful tree-growing algorithms

The four algorithms in this module differ from more traditional statistics, such as logistic regression, because these algorithms produce trees that enable you to explore your results and visually determine how your model flows. IBM SPSS Decision Trees makes it easier to identify specific subgroups and relationships in your data than if more traditional statistics were used. It breaks your data into branches and nodes, so you can easily see where a group splits and terminates.

IBM SPSS Decision Trees includes four established tree-growing algorithms. Find the best fit for your data by trying different algorithms or let IBM SPSS Decision Trees suggest the most appropriate algorithm.

CHAID: A statistical multi-way tree algorithm that explores data quickly and builds segments and profiles with respect to the desired outcome

Exhaustive CHAID: A modification of CHAID that examines all possible splits for each predictor (independent) variable

• Classification & Regression Trees (C&RT): A complete binary tree algorithm, which partitions data and produces accurate homogeneous subsets

• QUEST: A statistical algorithm that selects variables without bias and builds accurate binary trees quickly and efficiently

Once you have produced your classification tree(s), you can dig deeper into your data and gain more insight by identifying a particular subset of the data via the tree, and then run further analysis on this group.

Segment and group cases directly within the data As you are creating classification trees, you can use the results to segment and group cases directly within your data. Additionally, you can generate selection or classification/prediction rules in the form of IBM SPSS Statistics syntax, SQL statements or simple text. Display these rules in the Viewer and save them to an external file for later use to make predictions about individual and new cases.

specifications

Key features

• Create tree-based classification models for:

– Segmentation

– Stratification

– Prediction

– Data reduction and variable screening

– Interaction identification

– Category merging and discretising

continuous variables

Classify cases into groups or predict values of a dependent (target) variable based on values of inde- pendent (predictor) variables

Validation tools for exploratory and confirmatory classification analysis

View nodes using one of several ways: Show bar charts of your tar- get variables, tables or both in each node

Collapse and expand branches without deleting the model

Generate syntax automatically from the UI

Re-run tree building using syntax in production mode

Score data based on results or use results in further analysis using other IBM SPSS Statistics procedures

Algorithms

Four powerful tree-modeling algorithms:

– CHAID by Kass (1980)

– Exhaustive CHAID by Biggs, de

Ville and Suen (1991) – Classification & Regression Trees

(C&RT) by Breiman, Friedman,

Olshen and Stone (1984)

– QUEST by Loh and Shih (1997)

Evaluation

Evaluation graphs enable visual rep-

resentation of gains summary tables

• Misclassification functionality

Gains chart: Identify segments by

highest (and lowest) contribution

Deployment

Export output objects to any of IBM

SPSS Statistics’ available output formats

• Generate rules that define selected segments in SQL to score databases or IBM SPSS Statistics syntax to score IBM SPSS Statistics files

Export XML models to score new cases

System requirements

Requirements vary according to

platform

Available on the following platforms:

Windows n Mac n Linux

on the following platforms: Windows n Mac n Linux Use the highly visual trees to discover

Use the highly visual trees to discover relationships that are currently hidden in your data. IBM SPSS Decision Trees’ diagrams, tables and graphs are easy to interpret.

Trees’ diagrams, tables and graphs are easy to interpret. Create tree models in IBM SPSS Decision

Create tree models in IBM SPSS Decision Trees using CHAID, Exhaustive CHAID, C&RT or QUEST.

Trees using CHAID, Exhaustive CHAID, C&RT or QUEST. Use tree model results to score cases directly

Use tree model results to score cases directly in IBM SPSS Statistics Base.

results to score cases directly in IBM SPSS Statistics Base. Directly select cases or assign predictions

Directly select cases or assign predictions in IBM SPSS Statistics Base from the model results, or export rules for later use.

in IBM SPSS Statistics Base from the model results, or export rules for later use. View

View UK prices and purchase

15

IBM ® SPSS ® Custom Tables 19

Analyse data easily and communicate results effectively

With IBM SPSS Custom Tables you can:

Quickly create tables with a drag-and-drop interface

Preview tables as you create them to get it right the first time

Customise tables to make them easier for your audience to understand

IBM SPSS Custom Tables enables you to display your analyses as presentation-quality, production-ready tables. It gives you all the tools you need to easily create and work with tabular reports.

One of the most popular modules in the IBM SPSS Statistics family, IBM SPSS Custom Tables delivers many valuable capabilities:

Create new fields directly in output tables to perform calculations (e.g., sum, difference, percentage difference) on output categories

See the results of significance tests directly in output tables

Preview tables as you build them:

Take out the guesswork by previewing your table as you select variables and table options with a simple drag-and- drop interface

Control your table output: Choose from a variety of formats to represent multi-way information in a two-way table and generate the view you want

Customise your table structure:

Exclude specific categories, display missing value cells and add subtotals to your table

Use in-depth analyses: Run Chi- square, column proportions and column means tests, and add more insight to your tables by identifying differences, changes or trends in your data

by identifying differences, changes or trends in your data Play a 10-minute demonstration Download datasheet and

Play a 10-minute demonstration

Download datasheet and full specification

Automate frequent reports: Run large production jobs and complex table structures with ease to automatically build similar tables with new data

Build tables interactively

Generate the table you envision using the Graphical User Interface. You can easily work with output and present survey results by using nesting, stacking and multiple response categories, and handle missing values and change labels and formats – you can even include missing values in your results.

specifications

Key features

Graphical user interface

Simple drag-and-drop table builder interface allows you to preview tables as you select variables and options

• Single, unified table builder instead of multiple menu choices and dialog boxes for different table types Control contents

Create tables with up to three display dimensions: Rows (stub), columns (banner) and layers

Nest variables to any level in all dimensions

Crosstabulate multiple independent variables in the same table

16

Display frequencies for multiple variables side-by-side with tables of

frequencies

Display all categories when multiple variables are included in a table even if a variable has a category

without responses

Display multiple statistics in rows, columns or layers

Place totals in any row, column or

layer

Create subtotals for subsets of categories of a categorical variable

Perform calculations (e�g�, sum, difference, percentage difference) on output categories, and display

the results in new fields created directly in Custom Tables output – No limit on the number of calculated fields

• Show significance test results directly in Custom Tables output

instead of in a separate table – No need to combine findings in a

Word document – Complies with APA guidelines

Control category display order and the ability to selectively show or hide categories

To order, phone 0845 345 0935

Statistics

Select from over 40 summary statistics

Calculate statistics for each cell, subgroup or table

Calculate percentages at any or all levels for nested variables

Calculate counts and percentages for multiple-response variables based on the number of responses or the number of cases

Select percentage bases for missing

values to include or exclude missing

responses Formatting controls

Sort categories by any summary statistic in the table

Continued on PG. 17

Create presentation-quality tables from IBM SPSS Statistics data — in a snap

Work with IBM SPSS Statistics Base seamlessly and easily

Drag and drop your variables on the table preview builder and see what your table
Drag and drop your
variables on the table
preview builder and see
what your table looks
like as you create it.

Customise your table to show the information you need. Add summary statistics, inferential stats and subtotals to make it easier to understand results.

Create your table and easily export your results

results. Create your table and easily export your results Add additional variables by dragging and placing

Add additional variables by dragging and placing them where you want.

variables by dragging and placing them where you want. Once all your variables are in place,

Once all your variables are in place, push the “OK” button to create your final table. Apply the optional TableLooks for a more polished appearance.

Hide the categories that make up subtotals – remove a category from the table without removing it from the subtotal calculation

Directly edit any table element, including formatting and labels

Sort tables by cell contents in ascending or descending order

Automatically display labels instead of coded values

Specify minimum and maximum width of table columns (overrides TableLooks)

Show a name, label, or both for each table variable

Display missing data as blank, zero, “.” or any other user-defined term

Add titles and captions

Output as IBM SPSS Statistics pivot tables

Specify corner labels

Customise labels for statistics

Display the entire label for variables,

values and statistics

Choose from a variety of numerical

formats

Apply pre-formatted TableLooks

• Define the set of variables that are related to multiple response data and save it with your data definition

for subsequent analysis

Use both long- and short-string

elementary variables

• Define an unlimited number of sets,

or variables that can exist in a set

• Tests of significance:

– Column means

– Column proportions

Exclude categories from

Chi-square

significance tests

• Significance tests for multiple

response variables

Syntax and printing formats

Simpler, easy-to-understand syntax

Syntax converter (for upgrade users)

Specify page layout: Top, bottom,

left and right margins and page length

layout: Top, bottom, left and right margins and page length View UK prices and purchase •

View UK prices and purchase

Use the global break command to produce a table for each value of a

variable when the variable is used in

a series of tables

System requirements

Requirements vary according to

platform

Available on the following platforms:

Windows n Mac n Linux

17

IBM ® SPSS ® Advanced Statistics 19

Use powerful techniques to analyse complex data

With IBM SPSS Advanced Statistics you can:

Move beyond basic analysis

Build flexible models using a wealth of model-building options

Achieve more accurate predictive models using a wide range of modeling techniques

predictive models using a wide range of modeling techniques Play a 10-minute demonstration Download datasheet and

Play a 10-minute demonstration

Download datasheet and full specification

Created to provide you with more statistical power, IBM SPSS Advanced Statistics enables you to reach more accurate conclusions. Consider what it would be like to harness sophisticated univariate and multivariate analytical techniques and unleash them on your data.

You can reach more dependable conclusions with procedures designed to fit the inherent characteristics of data describing complex

relationships. IBM SPSS Advanced Statistics provides a powerful set of sophisticated

techniques designed to help you solve real- world problems, such as:

Medical research: Analyse patient survival rates

Manufacturing: Assess production processes

Pharmaceutical: Report test results to the FDA

Market research: Determine product interest levels

It’s time to step up your analysis when you require multiple outcomes, want to measure outcomes over time, analyse data with hiearchical structure or estimate length of time until an event.

Break through the barrier between general analysis and advanced modeling, and begin reaping the rewards today.

specifications

Key features

Generalized Linear Mixed Models (GLMM)

Extends the linear model so that:

1) The target is linearly related to the factors and covariates through a specified link function, 2) The target can have a non-normal distribution, 3) The observations can be correlated� Generalized linear mixed models cover a wide variety of models, from simple linear regression to complex multilevel models for non-normal longitudinal data�

Specify the subject structure for repeated measurements and how the errors of the repeated measure- ments are correlated

Choose among the 8 covariance types

Specify the target, optional offset and optional analysis (regression) weight

Choose among the following probability distributions: binomial, gamma, inverse Gaussian, multi- nomial, negative binomial, normal, Poisson

Choose among the following link functions: identity, complementary log-log� log-link, log complement, logit, negative log-log, power, probit

GENLIN and GEE

GENLIN and GEE procedures provide

a unifying framework for a wide variety of model types� Together, they enable you to predict more types of out- comes than ever, including:

Ordinal outcomes such as customer satisfaction

Outcomes that are a combination of discrete and continuous outcomes, such as claim amount, with a Tweedie distribution

Provide a common framework for the following outcomes: continuous

outcomes, count data, event/trial data, claim data, ordinal outcomes, combination of discrete and con- tinuous outcomes, and correlated

responses within subjects

MIXED

Expands the general linear model

used in the GLM procedure so that data can exhibit correlation and non-

constant variability

Fit the following types of models:

– Fixed effects ANOVA model, randomised complete blocks design, split-plot design, purely

random effects model, random coefficient model, multilevel analysis, unconditional linear growth model, linear growth model with person-level covariate,

repeated measures analysis, and repeated measures analysis with time-dependent covariate

Use one of six covariance structures offered

Select from 11 non-spatial

covariance types

GLM

Describe the relationship between a dependent variable and a set of independent variables

n Select univariate and multivariate

lack-of-fit tests

Regression model

Fixed effect ANOVA, ANCOVA, MANOVA and MANCOVA

Random or mixed ANOVA and ANCOVA

Repeated measures: Univariate or multivariate

Doubly multivariate design

VARCOMP

Variance component estimation

Estimation methods: ANOVA

MINQUE, maximum likelihood and restricted maximum likelihood

Type I and Type III sums of squares for the ANOVA method

Choices of zero-weight or uniform- weight methods

Choices of ML and REML calculation methods

Save variance components estimates and covariance matrices

LOGLINEAR and HILOGLINEAR

(For a full description, see www�spss�

com/statistics/advanced_statistics)

GENLOG

Fit loglinear and logit models to count data by means of a generalized linear model approach

• Model fit, using ML estimation

under Poisson loglinear model and

multinomial loglinear models

Accommodate structural zeros

Generalized log-odds ratio facility

tests the specific generalized log-

odds ratios are equal to zero, and

can print confidence intervals

Diagnostic plots: Scatterplots and

normal probability plots of residuals

System requirements

Requirements vary according to platform

Available on the following platforms:

Windows n Mac n Linux

18

To order, phone 0845 345 0935

Comprehensive tools for today’s analyst

IBM SPSS Advanced Statistics meets the wide range of statistical needs of today’s analyst. With its wealth of features and capabilities, you will never be limited to basic analytical techniques again.

General linear models (GLM) multivariate: Gain more flexibility to describe the relationship between a dependent variable and a set of independent variables.

Linear mixed models (Mixed): Use the Mixed procedure to model means, variances, and covariances when working with nested-structure data or repeated measures data, including when there are different numbers of repeated measurements or different intervals for different cases, or both.

Survival analysis: Analyse event history and duration data to better understand events. IBM SPSS Advanced Statistics includes state-of-the-art survival procedures such as Kaplan- Meier and Cox Regression.

Variance component estimation (VARCOMP): Choose from a number of methods to estimate the variance component for each random effect in a mixed model.

Loglinear analysis: Fit loglinear and logit models to count data so you can easily model and predict your outcomes.

New in IBM SPSS Advanced Statistics 19

Generalized Linear Mixed Models (GLMM): Allows more accurate models when predicting nonlinear outcomes (for example, what product a customer is likely to buy) by taking into account hierarchical data structures (customer nested with an organisation)

• Interactive and improved visualisations enable a more intuitive explanation of model predictors and outcomes

more intuitive explanation of model predictors and outcomes Structure your data A marketing group tests three

Structure your data

of model predictors and outcomes Structure your data A marketing group tests three campaigns to determine

A marketing group tests three campaigns to determine which promotion has the greatest effect on sales. First the marketing group specifies the structure of the data with MarketID as the “subjects”.

Achieve more accurate modelsstructure of the data with MarketID as the “subjects”. Next, they specify units sold as the

MarketID as the “subjects”. Achieve more accurate models Next, they specify units sold as the target

Next, they specify units sold as the target variable.

Identify random effectsNext, they specify units sold as the target variable. Market size, age of store location and

units sold as the target variable. Identify random effects Market size, age of store location and

Market size, age of store location and promotion are selected as the fixed effects. MarketID is chosen as the random effect.

the fixed effects. MarketID is chosen as the random effect. Analyse results using various methods Output

Analyse results using various methods

as the random effect. Analyse results using various methods Output shows that promotion is signifi -
as the random effect. Analyse results using various methods Output shows that promotion is signifi -

Output shows that promotion is signifi- cantly related to units sold. Specifically, pro- motion 1 has a higher amount of units sold than promotion 3 and promotion 2 has a lower number of units sold than promo- tion 3.

than promotion 3 and promotion 2 has a lower number of units sold than promo- tion

View UK prices and purchase

19

IBM ® SPSS ® Regression 19

Make better predictions using regression procedures

With IBM SPSS Regression you can:

Predict categorical outcomes with more than two categories

Easily classify your data into two groups

Gain more control over your model

IBM SPSS Regression gives you an even wider range of statistics so you can get the most accurate response for specific data types. Do you build predictive models but find ordinary least squares regression too limiting? If so, IBM SPSS Regression can make your life easier.

Use IBM SPSS Regression for:

Market research: Study consumer buying habits

Medical research: Study response to dosages

Loan assessment: Analyse good and bad credit risks

Institutional research: Measure academic achievement tests

Find the best predictor from dozens of possibilities

Apply more sophisticated models with IBM

SPSS Regression’s wide range of nonlinear modeling procedures.

Multinomial logistic regression (MLR): Predict categorical outcomes with more than two categories. Free

of constraints such as yes/no answers, MLR allows you to model which factor predicts if the customer buys product A, B or C.

Choose from four methods for choosing predictors: forward entry, backward elimination, forward stepwise and backward stepwise

– Stepwise function in MLR: Save time and easily find the best predictors for your data

– Use Score and Wald methods for a faster and more accurate conclusion for variable selection

– Apply a highly scalable, high- performance algorithm to handle big datasets

– Save time by specifying the ref- erence category in your outcome variable in the user interface. You no longer need to recode the dependent variable set up in the desired reference category.

– Use Akaike Information Criterion (AIC) and Bayesian Information Criterion (BIC) to better assess model fit

Information Criterion (BIC) to better assess model fit Play a 10-minute demonstration Download datasheet and full

Play a 10-minute demonstration

Download datasheet and full specification

Binary logistic regression (BLR):

Predict dichotomous variables such as buy or not buy, vote or not vote. This procedure offers many stepwise methods to select the main and interaction effects that best predict your response variable.

Nonlinear regression (NLR) and constrained nonlinear regression (CNLR): Get control over your model and your model expression. These procedures give you two methods for estimating parameters of non-linear models.

Weighted least square regression (WLS): Give more weight to measurements within a series

Probit analysis (PROBIT): Analyse potency of responses to stimuli, such as medicine doses, prices or incentives. Probit evaluates the value of the stimuli using a logit or probit transformation of the proportion responding.

specifications

Key features

Multinomial logistic regression

Control the values of the algorithm- tuning parameters

Include interaction terms

Customise hypotheses by directly specifying null hypotheses as linear combinations of parameters

Specify a dispersion scaling value

Build equations with or without a constant

• Use a confidence interval for odds ratios

Save the following statistics:

predicted probability, predicted

response category, probability of the predicted response category and

probability of the actual response

category

Find the best predictor from dozens of possible predictors using step-

wise functionality

Use Score and Wald methods to quickly reach results with a large

number of predictors

• Assess model fit

• Diagnostics for the classification table

Binary logistic regression (BLR)

Forward/backward stepwise and

forced entry modeling

Transform categorical variables by

using deviation contrasts, simple

comparison, difference (reverse Helmert) contrasts, Helmert contrasts, polynomial contrasts, comparison of adjacent categories, user-defined contrasts or indicator variables

Criteria for model building: probability

of score statistic for entry, probability of Wald, or likelihood ratio statistic for removal

Save the following statistics:

Predicted probability and group, residuals, deviance values, logit,

Studentised and standardised residuals, leverage value, analog of Cook’s influence statistic and difference in Beta

Export the model using XML

Continued on PG. 21

20

To order, phone 0845 345 0935

University of Ulster Teaching data mining to students

The University of Ulster is the largest in Ireland, and its Faculty of Computing and Engineering is one of the most substantial providers of undergraduate Computing Science courses in the UK. The Faculty also has a large Data Mining Research Group, and there is considerable collaboration with blue-chip companies, other universities, government agencies and hospitals.

However, there was a substantial shortfall in students’ knowledge and skill levels in data mining. Ulster University decided it would be beneficial to devote a full module to the subject. The overriding aims were that students should be able to assess the suitability of a data mining solution; that they should have a good understanding of a range of commonly used techniques; and that they should be able to interpret and evaluate the results.

In addition, it was felt to be important that they should use of industry standard software. IBM® SPSS® Modeler seemed a natural choice.

The module began in the autumn of 1998. “It was an immediate success,” says Ray Hickey, Lecturer in Computing Science. Students found it fascinating that a computer could discover knowledge that seemed impossibly hidden in data.

The CRISP-DM (CRoss Industry Standard Practice for Data Mining) methodology was introduced as a feature of the module. Mr Hickey concludes: “This use of CRISP-DM has been very successful. Yet CRISP-DM does not stifle creativity – as some methodologies do. Many students reported that they found this coursework very stimulating and rewarding.”

Powerful regression procedures help you find the best predictors

regression procedures help you find the best predictors B u i l d p r e

Build predictive models

i l d p r e d i c t i v e m o d

Predict the presence or absence of a characteristic/ outcome based on values of a set of predictor vari- ables. In this example, a wireless telephone service provider is interested in identifying dissatisfied customers so they can intervene before they defect and switch to a competitor.

intervene before they defect and switch to a competitor. Predict categorical outcomes Use the binary logistic

Predict categorical outcomes

and switch to a competitor. Predict categorical outcomes Use the binary logistic regression procedure to test

Use the binary logistic regression procedure to test the impact of service offers (new or expanded) on customer satisfaction. The classification table indicates that this model can correctly predict satisfaction with 90 percent accuracy.

Constrained nonlinear regression

Save predicted values, residuals and derivatives

• Choose numerical or user-specified derivatives

Nonlinear regression (NLR)

Specify loss function options

Use bootstrap estimates of standard errors

Weighted least squares (WLS)

Calculate weights based on source variable and Delta values or apply from an existing series

Output for calculated weights:

Log-likelihood functions for each

value of Delta; R, R2, adjusted R2, standard errors, analysis of variance

and t tests of individual coefficient for Delta value with maximised log-likelihood function

Display output in pivot tables

Two-stage least squares (2SLS)

Structural equations and instrumen- tal variables

Control for correlations between predictor variables and error terms

Display output in pivot tables

Probit

Transform predictors: Base 10, natural or user-specified base

Natural response rate estimates or specified

Algorithm control parameters:

Convergence, iteration limit and

heterogeneity criterion probability

• Statistics: Frequencies, fiducial

confidence intervals, relative median potency, test of parallelism, plots of

observed probits or logits

Display output in pivot tables

System requirements

Requirements vary according to platform

Available on the following platforms:

Windows n Mac n Linux

according to platform Available on the following platforms: Windows n Mac n Linux View UK prices

View UK prices and purchase

21

IBM ® SPSS ® Missing Values 19

Build better models when you fill in the blanks

With IBM SPSS Missing Values you can:

Overcome missing data issues

Use multiple imputation to replace missing data

Build models taking missing data into account

When you ignore or exclude missing data, you risk obtaining biased or insignificant results. Use IBM SPSS Missing Values to impute your missing data and draw more valid conclusions.

IBM SPSS Missing Values is a critical tool for anyone concerned about data validity. You can easily examine your data to uncover missing data patterns. Then, estimate summary statistics and impute missing values through statistical algorithms.

For example, improve survey questions that you’ve identified as possibly confusing based on observed missing data patterns. You can even determine if missing values for one variable are related to missing values of another with the percent mismatch of patterns table.

You might find that respondents who skip a question on income might also bypass a question about education level. Use this information to enhance the quality of your surveys in the future.

to enhance the quality of your surveys in the future. Play a 10-minute demonstration Download datasheet

Play a 10-minute demonstration

Download datasheet and full specification

Reach more valid conclusions

Replace missing values with estimates and increase the chance of receiving statistically significant results. Remove hidden bias from your data by replacing missing values with estimates to include all groups in your analysis – even those with poor responsiveness.

specifications

Key features

Analyse patterns

Display missing data and extreme

cases for all cases and all variables using the data patterns table

– Display system-missing and three types of user-defined missing values

– Sort in ascending or descending order

– Display actual values for specified Variables

Display patterns of missing values

for all cases that have at least one missing value using the missing patterns table

– Group similar missing value patterns together

– Sort by missing patterns and variables

– Display actual values for specified variables

Determine differences between missing and non-missing groups for

a related variable with the separate variance t test table

t test, degrees of freedom, mean, p value and count

22

Show differences between present and missing data for categorical

variables using the distribution of categorical variables table

– Produce crosstabs showing

product and missing data for

each category of one variable

by the other variables

Assess how much missing data for

one variable relates to the missing

data of another variable using the

percent mismatch of patterns table

– Sort matrices by missing value patterns or variables

Identify all unique patterns with

the tabulated patterns table, which

summarises each missing data pat-

tern and displays the count for each

pattern plus means and frequencies

for each variable

– Display count and averages for

each missing value pattern using the summary of missing value patterns table

Multiple Imputation

Specify which variables to impute and specify constraints on the imputed values, such as minimum

and maximum values� You can also specify which variables are used as predictors when imputing missing values of other variables�

Impute values for categorical and continuous variables� Logistic regression is used for categorical variables and linear regression for

continuous variables� Predictive mean matching is an option for

continuous outcomes; this ensures that the imputed values are reason- able (within the range of the original data)�

Missing data pattern detection helps determine which imputation

method to use

Three imputation methods are offered:

– Monotone: an efficient method for data that have a monotone pattern of missingness

To order, phone 0845 345 0935

– Fully conditional specification

(FCS): an iterative Markov Chain Monte Carlo (MCMC) method that is appropriate when the data have an arbitrary (monotone or no monotone) missing pattern

– Automatic: scans the data to determine the best imputation

method (monotone or FCS) Specify:

– The number of imputations

– The range of imputed values

– Whether or not interaction effects are used when imputing

– Optionally, turn off imputation for variables that have a higher percentage of missing values

– Tolerance levels, to check for

singularity You can also specify a variable containing analysis (regression)

weights� The procedure incorporates analysis weights in regression and classification models used to impute missing values� Analysis weights are also used in summaries

Continued on PG. 23

Länsförsäkringar Swedish insurance company increases the number of direct marketing campaigns

Länsförsäkringar comprises 24 independent, regional insurance companies collaborating through the jointly owned Länsförsäkringar AB and its subsidiaries. Länsförsäkringar offers non-life insurance, accident and medical insurance, life assurance, pension-saving plans, fund savings and various banking services.

Länsförsäkringar’s campaign selection was based on the experience of the sales force. The work carried out to obtain information from different data sources was inefficient and administratively cumbersome.

There was a great need for a more effective solution whereby information could be structured and analysed more systematically to optimise campaign selections and the products offered to customers.

Länsförsäkringar implemented IBM® SPSS® Modeler to analyse large quantities of data from various data sources and to identify patterns of customer behaviour. By building and deploying models, Länsförsäkringar can now analyse customer behaviour and needs, and leverage this information to customise its marketing campaigns.

According to Ola Gustafsson, Customer Data Analyst at Länsförsäkringar, “The efficiency of the analyst team has been enhanced significantly – work which used to take a week can now be done in a day and a half.”

“We have also seen a qualitative improvement, with results that are up to four times more reliable and accurate,” he added. “In addition, it has been possible to standardise the way the work is done.”

Today, Länsförsäkringar has a very straightforward way to predict the likelihood of a customer having an interest in a specific insurance policy. As a result, Länsförsäkringar haS moved from executing two nationwide campaigns per year to running 13 unique, customer-specific campaigns a day.

of imputed values (e�g�, mean, standard deviation and standard error)�

Display an overall summary of missingness in your data as well as an imputation summary and the imputation model for each variable whose values are imputed� You can obtain analysis of missing values by variable as well as tabulated patterns of missing values� Optionally, you can obtain descriptive statistics for imputed values�

Graphically summarise missingness for cases, variables and individual data (cell) values

Request an IBM SPSS Statistics data file containing imputed values and/or an FCS iteration history

Multiple imputation datasets can be analysed using supported analysis procedures to obtain final (combined) parameter estimates that take into account the inherent uncertainty in the various sets of imputed values

Analysis

Supported analysis procedures for Multiple Imputation (note: you must

have purchased the proper module in which the procedure is located)

Descriptive procedures: frequencies, descriptives, crosstabs, correlations, non-parametric correlation, partial correlation

Comparison of means: means, t test, non-parametric tests, one-way ANOVA, univariate ANOVA

Models: General Linear Models,

Generalized Linear Models, linear regression, multinomial logistic regression, binary logistic regres-

sion, discriminant analysis, ordinal regression, linear mixed models

Survival analysis techniques: Cox

regression

Pooling

Pooling of output: output is pooled using one of two levels of pooling produces pooled parameters

Pooling Diagnostics

– Relative Increase in Variance:

measure of relative variability in parameter estimate across imputations

– Fraction of Missing Information:

relative increase in variance scaled as a proportion� A measure

of uncertainty due to nonre- sponse� – Relative Efficiency: efficiency of estimate for M imputations rela- tive to that for an infinite number of imputations

Obtain Model PMML for pooled parameter estimates: Linear regression, Generalized Linear Models, multinomial logistic regression, binary logistic regression, discriminant analysis, Cox Regression

logistic regression, discriminant analysis, Cox Regression View UK prices and purchase System requirements •

View UK prices and purchase

System requirements

Requirements vary according to platform

Available on the following platforms:

Windows n Mac n Linux

23

IBM ® SPSS ® Forecasting 19

Build expert time-series forecasts – in a flash

With IBM SPSS Forecasting you can:

Support time-series analysis

Find the best model for your data using the Expert Modeler

Apply saved models to “what-if” scenarios to optimise your decisions

models to “what-if” scenarios to optimise your decisions Play a 10-minute demonstration Download datasheet and full

Play a 10-minute demonstration

Download datasheet and full specification

Time-series analysis is the most powerful procedure you can use to analyse historical information, build models, predict trends, and forecast future events. IBM SPSS Forecasting is the best way to quickly create powerful forecasts with confidence. With better forecasts, long-term goals can be set – with insight on how to achieve them – based on your organisation’s past performance and knowledge of your industry.

Unlike spreadsheet programs, IBM SPSS Forecasting has the advanced statistical techniques you need in order to work with time-series data. But you don’t need to be an expert statistician to use it. Regardless of your level of experience, you can analyse historical data and predict trends faster, and deliver information in ways that your organisation’s decision makers can under- stand and use.

IBM SPSS Forecasting will help you find answers to tough questions:

If I increase my advertising budget, how will it affect sales by product or region?

How will increasing assembly line capacity affect production?

Will a change in fees affect the number of new customers we gain?

How will tuition increases affect enrollment?

If you’re new to building models from time-series data, IBM SPSS Forecasting helps you by:

Generating reliable models, even if you’re not sure how to choose exponential smoothing parameters or ARIMA orders, or how to achieve stationarity

Automatically testing your data for seasonality, intermittency and missing values, and selecting appropriate models

Detecting outliers and preventing them from influencing parameter estimates

Generating graphs showing confidence intervals and the model’s goodness of fit

If you’re experienced at forecasting, IBM SPSS Forecasting allows you to:

Control every parameter when build- ing your data model

Or use IBM SPSS Forecasting’s Expert Modeler recommendations as a starting point or to check your work

Use IBM SPSS Forecasting to:

Develop reliable forecasts quickly, regardless of the size of the dataset or number of variables

Update and manage forecasting models efficiently

Reduce forecasting errors by automating appropriate model selection and parameters

Gain more control over choices affecting models, parameters and output

Deliver high-resolution graphs and communicate results effectively

specifications

Key features

TSMODEL

Model a set of time-series variables by using the Expert Modeler or speci- fying the structure of an ARIMA or exponential smoothing model

Allow Expert Modeler to select the best- fitting predictor variables and models – Limit search space to only ARIMA or only exponential smoothing models

– Treat independent variables as events Specify custom ARIMA models

– Produces maximum likelihood estimates for seasonal and non-seasonal univariate models

– General or constrained models specified by autoregressive or moving average order, order of differencing, seasonal autoregres- sive, or moving average order, and seasonal differencing

– Two dependent variable transfor- mations: Square root and natural log

– Automatically detect or specify outliers: Additive, level shift, innovational, transient, seasonal

additive, local trend and additive

patch

– Specify seasonal and nonsea-

sonal numerator, denominator

and difference transfer function orders and transformations for each independent variable

Specify custom exponential smoothing models

– Four non-seasonal model types:

Simple, Holt’s linear trend, Brown’s linear trend and damped trend

– Three seasonal model types:

Simple seasonal, Winters’ additive and Winters’ multiplicative

– Two dependent variable transfor-

mations: Square root and natural log

• Display forecasts, fit measures, Ljung-Box statistic, parameter estimates and outliers by model

Generate tables and plots to compare statistics across all models

• Eight goodness of fit measures available: stationary R 2 , R 2 , root mean square error, mean absolute

percentage error, mean absolute error, maximum absolute percentage

error, maximum absolute error and normalised BIC

Tables and plots of residual autocor-

relation function (ACF) and partial autocorrelation function (PACF)

Plot observed values, forecasts,

fit value, confidence intervals for forecasts and confidence intervals for fit values for each series

• Filter output to fixed number or percentage of best or worst fitting models

Save predicted values, lower confidence limits, upper confidence

limits and noise residuals for each series back to the dataset

Continued on PG. 25

24

To order, phone 0845 345 0935

Yamaha Motor Europe

IBM SPSS software gives insight into customers’ design wishes

Founded in 1968, Yamaha Motor Europe N.V. (YME) is the regional headquarters of the Yamaha Motor Company. YME is responsible for coordinating the marketing and sales activities of numerous Yamaha Motor products in the European marketplace.

Desiring more information into customers’ design wishes, YME turned to the web as an economical and efficient means of collecting comparable consumer insight. The result was the Yamaha Design Café. The site has links directing readers to surveys, which encourage visitors to express their opinions and thoughts on current designs.

A critical component of this online user portal is IBM SPSS software. It has provided the technology that helps YME optimise its new product evaluation process while increasing the competitiveness of its existing products.

Hennes Fischer, YME’s product planning consultant, used IBM® SPSS® Data Collection Web Interviews – a component of the IBM SPSS Data Collection survey research platform – to develop and field the Design Café surveys. IBM® SPSS® Translation Utility, along with Web Interviews, coordinates the translation of responses gathered across Europe.

Now, any rider of a Yamaha motorcycle who visits the Yamaha Design Café online portal is helping to determine the Japanese manufacturer’s future production. The portal helps the company tailor its new product designs closely to customer tastes.

The company has achieved more customer-oriented product development thanks to information from these surveys. YME has also experienced minimal product development failures, and has improved their ability to identify purchasing trends.

has improved their ability to identify purchasing trends. This screenshot of the time-series modeler shows how

This screenshot of the time-series modeler shows how it provides you with the ability to model multiple series simultaneously. Because the module presents results in an organised fashion, you can concentrate on the mod- els that need closer examination.

Specify forecast period, treatment of user-missing values, and confidence intervals

• Export models to an XML file for later use by TSAPPLY

TSAPPLY

Apply saved models to new or updated data

Simultaneously apply models from multiple XML files created with TSMODEL

Re-estimate model parameters and goodness of fit measures from the data or load from the saved model file

Selectively choose saved models to apply

Override the periodicity (seasonality) of the active dataset

• Same output, fit measure, statistics

and options as TSMODEL

Export re-estimated models to an XML file

SEASON

Estimates multiplicative or additive

seasonal factors for periodic time series

Multiplicative or additive model

Moving averages, ratios, sea- sonal and seasonal adjustment factors, seasonally adjusted series, smoothed trend-cycle components

and irregular components

smoothed trend-cycle components and irregular components This screenshot displaying a fore- cast for women’s

This screenshot displaying a fore- cast for women’s apparel shows how you can automatically determine which model best fits your time-series and independent variables.

SPECTRA

Decomposes a time series into its

harmonic components, a set of regular periodic functions at different

wavelengths or periods

Produces/plots univariate or bivariate periodogram and spectral density estimate

Bivariate spectral analysis

Smooth periodogram values with

weighted moving averages

Spectral data windows available for smoothing: Tukey-Hamming, Tukey, Parzen, Bartlett, equal weight, no smoothing and user-specified weights

High-resolution charts available:

Periodogram, spectral and cospectral density estimate, squared coherency, quadrature spectrum estimate, phase spectrum, cross amplitude

and gain

System requirements

Requirements vary according to

platform

Available on the following platforms:

Windows n Mac n Linux

according to platform Available on the following platforms: Windows n Mac n Linux View UK prices

View UK prices and purchase

25

IBM ® SPSS ® Bootstrapping 19

Ensure the stability of your models

With IBM SPSS Bootstrapping you can:

Quickly estimate the sampling distribution of an estimator

Eliminate outliers and anomalies that degrade accuracy

Bootstrap many analytical procedures

degrade accuracy • Bootstrap many analytical procedures Play a 10-minute demonstration The models your organisation

Play a 10-minute demonstration

The models your organisation creates drive important decisions. They may be used to shape public policy, to prevent the spread of disease or to determine a multi-million dollar investment. It’s important that your models are stable, so that they will produce accurate, reliable results. Bootstrapping is a useful technique for testing model stability, and IBM SPSS Bootstrapping makes it easy to do.

This module of IBM SPSS Statistics provides an efficient way to ensure your models are stable and reliable. It estimates the sampling distribution of an estimator by re-sampling with replacement from the original sample. With IBM SPSS Bootstrapping, you can reliably estimate the standard errors and confidence intervals of a population param- eter like a mean, median, proportion, odds ratio, correlation coefficient, regression coefficient and numerous others.

Reliable models for critical projects

When you require the most reliable model be created to predict an outcome or map a

sample to a population, simply running the model once on the sample data on hand may not be the best approach because results are dependent on your sample data. Resampling with replacement will provide you with more accurate estimates of the reliability of your data.

In order to identify precisely how suitable your model is, you will want to bootstrap the model to assess its stability.

A complete view of your data

Computing a statistic on a large number of alternate datasets helps you determine the variability of that statistic. Through re-sampling, IBM SPSS Bootstrapping can create thousands of alternate versions of your dataset, providing a more accurate view of what is likely to exist in the population. (Its default setting is 1,000 samples but this setting can be modified upward or down- ward.) IBM SPSS Bootstrapping also helps you eliminate the outliers and anomalies that can degrade the accuracy or applicability of your analysis. As a result, you have a clearer view of your data for creating the model you are working with.

of your data for creating the model you are working with. Through the IBM SPSS Bootstrapping

Through the IBM SPSS Bootstrapping dialog box, you can easily control the numbers of bootstrap samples, set a random number seed, specify confidence intervals and indicate if a simple or stratified method is appropriate.

IBM SPSS Bootstrapping is available for installation as client-only software but, for greater performance and scalability, a server-based version is also available.

specifications

Key features

IBM SPSS Bootstrapping provides the ability to bootstrap a number of analytical procedures found throughout the IBM SPSS Statistics product family, including:

Descriptive Procedures

IBM SPSS Statistics Base:

Examine

Means

Crosstabs

Descriptives

Frequencies

t tests

Correlations/nonparametric correlations

Partial correlations

Modeling Procedures

IBM SPSS Statistics Base:

One-way

UniAnova

PLUM

Discriminant

IBM SPSS Advanced Statistics:

GLMM

GENLIN

Linear Mixed Models

Cox Regression

IBM SPSS Regression:

Regression

Nominal Regression

Logistic Regression

Available on the following platforms:

Windows n Mac n Linux

26

• Logistic Regression Available on the following platforms: Windows n Mac n Linux 26 View UK

View UK prices and purchase

what if…

you

could easily identify the best customers for a marketing campaign?

Today’s marketers understand the value of leveraging their data to gain insight into customer behaviour and preferences. But that value is only delivered when the insight (and in the case of predictive analysis, the foresight based on customer propensities and predicted behaviour) is turned into action at the point of customer contact.

IBM SPSS Predictive Analytics lets you embed improved decision-making and automation into CRM and other customer-facing systems, based on data insights.

Find out how IBM SPSS Predictive Analytics can turn data into actions.

Business Analytics software

IBM, the IBM logo, ibm.com are trademarks or registered trademarks of International Business Machines Corporation registered in many jurisdictions worldwide. Other product and service names may be trademarks of IBM or other companies. A current list of IBM trademarks is available on the Web at www.ibm.com/legal/copytrade.shtml © 2011 IBM Corporation. All rights reserved.

27

trademarks is available on the Web at www.ibm.com/legal/copytrade.shtml © 2011 IBM Corporation. All rights reserved. 27

IBM ® SPSS ® Complex Samples 19

Correctly and easily compute statistics for complex samples

With IBM SPSS Complex Samples you can:

Get a more accurate picture of your data when working with large-scale surveys

Achieve more statistically valid inferences for populations

Reach correct point estimates for statistics such as totals, means and ratios, and obtain standard errors of these statistics

Predict numerical and categorical outcomes from non-simple random samples

Take up to three stages into account when analysing data from a multistage design

into account when analysing data from a multistage design Play a 10-minute demonstration If you’re working

Play a 10-minute demonstration

If you’re working with complex sample designs, such as stratified, clustered or multistage sampling, you need specialised statistical techniques to account for the sample design and its associated standard error.

IBM SPSS Complex Samples has everything you need to correctly and easily compute statistics and their standard errors from complex sample designs.

You can apply it to:

Survey research – Obtain descriptive and inferential statistics for survey data

Market research – Analyse customer satisfaction data

Health research – Analyse large public-use datasets on public health topics such as health and nutrition or alcohol use and traffic fatalities

Social science – Conduct secondary research on public survey datasets

Public opinion research – Characterise attitudes on policy issues

In addition to giving you the ability to assess your design’s impact, this module also produces a more accurate picture of your data because subpopulation assessments take other subpopulations into account.

assessments take other subpopulations into account. IBM SPSS Complex Samples helps agencies drive better

IBM SPSS Complex Samples helps agencies drive better outcomes for public policy

Analyzing data for government agencies poses special challenges. For some projects, enough data exists so that obtaining a random sample for analysis is relatively straightforward. For other projects, however, if an analyst drew a simple random sample, there would probably be too few cases in some of the categories, because these conditions are relatively rare. Therefore, an analyst must oversample certain groups in which there are relatively few cases.

For example, if an analyst for a city public health department wanted to evaluate the effectiveness of several different substance abuse programs, there might not be enough cases in each age group, in each geographic area, for each treatment option being studied – resulting in some sub-groups being over-sampled. The procedures in IBM SPSS Complex Samples enable analysts to apply scientific sample designs in the sample selection process in order to reduce the risk of a distorted view of the population.

In addition, analysts can document how a sample was selected, so that subsequent evaluations can replicate the sample design – and obtain results that can be used to reliably identify trends and make predictions about future developments. This not only helps agencies evaluate the effectiveness of their programs but also anticipate and plan for future needs.

28

To order, phone 0845 345 0935

Incorporate complex sample designs into your data analysis

Only IBM SPSS Complex Samples makes understanding and working with your complex sample survey results easy. It is one of the most comprehensive software programs available.

IBM SPSS Complex Samples provides you with everything you need to produce more accurate results …

Logistic regression: Predict categorical outcomes (such as: Who is most likely to buy my product?) while taking the sample design into account to accurately identify groups

Ordinal regression: Predict ordinal outcomes such as customer satisfaction (low, medium or high)

General linear models: Predict numerical outcomes while taking the sample design into account

Cox regression: Predict time to an event for samples drawn by complex sampling methods

Intuitive Sampling Wizard: Guides you step-by-step through the process of designing and drawing a sample

Easy-to-use Analysis Preparation Wizard: Helps prepare public-use datasets for analysis, such as the National Health Inventory Survey data from the Centers for Disease Control and Prevention (CDC)

Easier collaboration with colleagues: Easily share sampling and analysis plans

More accurate analyses: Enables you to take up to three stages into account when analysing data from a multistage design

A more precise picture of your data: Unlike traditional statistics, subpopulation assessments take other subpopulations into account

Use the following types of sample design information with IBM SPSS Complex Samples:

• Stratified sampling: Increase the precision of your sample or ensure

a representative sample from key groups by choosing to sample within

subgroups of the survey population. For example, subgroups might be

a specific number of males or females or contain people in certain job

categories, people of a certain age group and so on.

Clustered sampling: Select clusters, which are groups of sampling units, for your survey. Clusters can include schools, hospitals or geographic areas with sampling units that might be students, patients or citizens. Clustering often helps make surveys more cost-effective.

Multistage sampling: Select an initial or first-stage sample based on groups of elements in the population; then create a second-stage sample by drawing a sub-sample from each selected unit in the first-stage sample. By repeating this option, you can select a higher- stage sample. For example, in a face-to-face survey, you might sample individuals within households and city blocks.

Accurate analysis of survey data Sampling Plan Wizard Plan files
Accurate analysis of survey data
Sampling Plan
Wizard
Plan files

Analyse dataanalysis of survey data Sampling Plan Wizard Plan files Results Analysis Preparation Wizard The accurate analysis

Results
Results
Analysis Preparation Wizard
Analysis
Preparation Wizard

The accurate analysis of survey data is easy in IBM SPSS Complex Samples. Start with one of the wizards (which one depends on your data source) and then use the interactive interface to create plans, analyse data and interpret results.

In particular, use the wizards to specify a sampling scheme (when collecting data) or to explain how a sample was drawn (when working with public-use data).

Then utilise the procedures that have been specifically developed for use with complex samples to predict numerical, ordinal and categorical outcomes or time to a specific event.

specifications

Key Features

Complex Samples Plan (CSPLAN):

Provides a common place to specify the sampling frame to create a complex sample design or analysis design used by procedures in IBM SPSS Complex Samples�

Complex Samples Selection (CSSELECT): Selects complex, probability-based samples from a population� It chooses units according to a sample design created through the CSPLAN procedure�

Complex Samples Descriptives (CSDESCRIPTIVES): Estimates means, sums and ratios, and computes their standard errors, design effects, confi- dence intervals and hypothesis tests�

Complex Samples Tabulate (CSTABULATE): Displays one-way frequency tables or two-way cross- tabulations and associated standard errors, design effects, confidence intervals and hypothesis tests�

Complex Samples General Linear Model (CSGLM): Enables you to build linear regression, analysis of variance

(ANOVA) and analysis of covariance (ANCOVA) models�

Complex Samples Ordinal (CSORDINAL): Performs regression analysis on a binary or ordinal poly- tomous dependent variable using the selected cumulative link function�

Complex Samples Logistic Regression (CSLOGISTIC): Performs binary logistic regression analysis, as well as multi- nomial logistic regression analysis�

Complex Samples Cox Regression (CSCOXREG): Applies Cox propor- tional hazards regression to analysis of survival times – that is, the length of time before the occurrence of an event�

System requirements

Requirements vary according to platform

Available on the following platforms:

Windows n Mac n Linux

according to platform Available on the following platforms: Windows n Mac n Linux View UK prices

View UK prices and purchase

29

IBM ® SPSS ® Statistics Server 19

Analyse “big data” and data in dispersed organisations

With IBM SPSS Statistics Server you can:

Analyse massive data files faster

Support distributed offices in performing analytics efficiently

Improve analyst productivity

IBM SPSS Statistics Server offers all the features of IBM SPSS Statistics but with faster performance because the processing

is centralised on the server machine. This

eliminates the need to transfer data over the network, which saves time, improves

productivity and enhances security. Analysts can work in their application of choice without any disruption while waiting for

a long-running job to complete and can

initiate multiple jobs at the same time.

Faster performance

IBM SPSS Statistics Server provides faster performance across your enterprise when working with large datasets having multiple predictors. There is no limit on the number of CPUs or cores that an analytical procedure can use, and no limit on the number of threads that can be used for multithreaded procedures. Operations like sorts and aggregates can be pushed back to the database, where they can be performed faster. Temporary files created by analytical procedures can be striped over multiple disks, which also speeds analytical processing. And with this latest release,

creating pivot tables in the output is now two to three times faster than before. In addition, you can now run IBM SPSS Statistics Server on your IBM System z machines using Linux for more powerful, enterprise-wide analysis.

Better security and standardisation

Because data is stored in a central location, standard processes can be enforced to ensure that all analysts are using the latest versions of a syntax or data file. Network administrators can use a single Administrative Utility for working with IBM SPSS Statistics Server, IBM SPSS Modeler Server and IBM SPSS Collaboration and Deployment Services. Centralisation also enhances security, helping you to better protect sensitive data and intellectual property.

Architected for scalability

When integrated with IBM SPSS Collaboration and Deployment Services, Statistics Server can be clustered to provide network load balancing and failover protection. This ensures that it can seamlessly scale

protection. This ensures that it can seamlessly scale Download datasheet and full specification from meeting the

Download datasheet and full specification

from meeting the analytic needs of a single department to meeting those of hundreds and even thousands of users across the enterprise.

When you combine the strength of world- class analytical tools and techniques with the flexibility and speed of server functionality, you have a powerful solution for supporting better decision-making throughout your enterprise.

“ IBM SPSS Statistics represents another important step to satisfy the needs of business clients with innovations that help describe, explain and predict.”

– Alexander Vinogradov, PhD Associate Professor Social Sciences and Social Technologies Faculty National University - Kyiv-Mohila Academy Ukraine

specifications

System requirements Operating system

Microsoft Windows Server 2008 or 2003 (32-bit or 64-bit); SunSolaris9 or 10 (SPARC 64-bit machine); IBM ® AIX ® 5�3 or 6�1; IBM zSeries running Linux, 64-bit only (PowerPC); HP-UX™ 11i v3 (64-bit Itanium); or Red Hat ® Enterprise Linux 4x or 5 (32-bit and 64-bit), Advanced Platform (32-bit and 64-bit) or Advanced Server 4x (64-bit)

Hardware