0% found this document useful (0 votes)
67 views50 pages

SQL Server 2022 Revealed: A Hybrid Data Platform Powered by Security, Performance, and Availability 1st Edition Bob Ward Instant Download

The document is a promotional overview of the book 'SQL Server 2022 Revealed: A Hybrid Data Platform Powered by Security, Performance, and Availability' by Bob Ward, which discusses the features and capabilities of SQL Server 2022. It highlights the book's focus on hybrid data management, security, performance enhancements, and availability. Additionally, it provides links to download the book and other related resources.

Uploaded by

lutfegordim
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
0% found this document useful (0 votes)
67 views50 pages

SQL Server 2022 Revealed: A Hybrid Data Platform Powered by Security, Performance, and Availability 1st Edition Bob Ward Instant Download

The document is a promotional overview of the book 'SQL Server 2022 Revealed: A Hybrid Data Platform Powered by Security, Performance, and Availability' by Bob Ward, which discusses the features and capabilities of SQL Server 2022. It highlights the book's focus on hybrid data management, security, performance enhancements, and availability. Additionally, it provides links to download the book and other related resources.

Uploaded by

lutfegordim
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
You are on page 1/ 50

SQL Server 2022 Revealed: A Hybrid Data Platform

Powered by Security, Performance, and


Availability 1st Edition Bob Ward pdf download

https://ebookmeta.com/product/sql-server-2022-revealed-a-hybrid-
data-platform-powered-by-security-performance-and-
availability-1st-edition-bob-ward/

Download more ebook from https://ebookmeta.com


We believe these products will be a great fit for you. Click
the link to download now, or visit ebookmeta.com
to discover even more!

Pro SQL Server 2022 Wait Statistics: A Practical Guide


to Analyzing Performance in SQL Server and Azure SQL
Database, 3rd Edition Thomas Larock

https://ebookmeta.com/product/pro-sql-server-2022-wait-
statistics-a-practical-guide-to-analyzing-performance-in-sql-
server-and-azure-sql-database-3rd-edition-thomas-larock-2/

Pro SQL Server 2022 Wait Statistics: A Practical Guide


to Analyzing Performance in SQL Server and Azure SQL
Database, 3rd Edition Thomas Larock

https://ebookmeta.com/product/pro-sql-server-2022-wait-
statistics-a-practical-guide-to-analyzing-performance-in-sql-
server-and-azure-sql-database-3rd-edition-thomas-larock/

Expert Performance Indexing in Azure SQL and SQL Server


2022 Fourth Edition Edward Pollack

https://ebookmeta.com/product/expert-performance-indexing-in-
azure-sql-and-sql-server-2022-fourth-edition-edward-pollack/

Teaching Women s Studies in Conservative Contexts


Considering Perspectives for an Inclusive Dialogue 1st
Edition Cantice Greene (Editor)

https://ebookmeta.com/product/teaching-women-s-studies-in-
conservative-contexts-considering-perspectives-for-an-inclusive-
dialogue-1st-edition-cantice-greene-editor/
Official ABCs of Love 1st Edition Georgia Coffman

https://ebookmeta.com/product/official-abcs-of-love-1st-edition-
georgia-coffman/

Electrode Materials for Energy Storage and Conversion


1st Edition Mesfin A. Kebede

https://ebookmeta.com/product/electrode-materials-for-energy-
storage-and-conversion-1st-edition-mesfin-a-kebede/

Translating Cultures An Introduction for Translators


Interpreters and Mediators Third Edition David Katan

https://ebookmeta.com/product/translating-cultures-an-
introduction-for-translators-interpreters-and-mediators-third-
edition-david-katan/

Cracks in the Windshield After I Do 1st Edition Bree


Weeks

https://ebookmeta.com/product/cracks-in-the-windshield-after-i-
do-1st-edition-bree-weeks/

Automated Guided Vehicle Systems A Guide With Practical


Applications About The Technology For Planning 2nd
Edition Günter Ullrich

https://ebookmeta.com/product/automated-guided-vehicle-systems-a-
guide-with-practical-applications-about-the-technology-for-
planning-2nd-edition-gunter-ullrich/
Colored Television American Religion Gone Global 1st
Edition Marla Frederick

https://ebookmeta.com/product/colored-television-american-
religion-gone-global-1st-edition-marla-frederick/
SQL Server 2022
Revealed
A Hybrid Data Platform Powered by
Security, Performance, and Availability

Bob Ward
Foreword by Rohan Kumar
SQL Server 2022 Revealed
A Hybrid Data Platform Powered
by Security, Performance,
and Availability

Bob Ward
Foreword by Rohan Kumar
SQL Server 2022 Revealed: A Hybrid Data Platform Powered by Security,
Performance, and Availability
Bob Ward
North Richland Hills, TX, USA

ISBN-13 (pbk): 978-1-4842-8893-1 ISBN-13 (electronic): 978-1-4842-8894-8


https://doi.org/10.1007/978-1-4842-8894-8
Copyright © 2022 by Bob Ward
This work is subject to copyright. All rights are reserved by the Publisher, whether the whole or part of the
material is concerned, specifically the rights of translation, reprinting, reuse of illustrations, recitation,
broadcasting, reproduction on microfilms or in any other physical way, and transmission or information
storage and retrieval, electronic adaptation, computer software, or by similar or dissimilar methodology now
known or hereafter developed.
Trademarked names, logos, and images may appear in this book. Rather than use a trademark symbol with
every occurrence of a trademarked name, logo, or image we use the names, logos, and images only in an
editorial fashion and to the benefit of the trademark owner, with no intention of infringement of the
trademark.
The use in this publication of trade names, trademarks, service marks, and similar terms, even if they are not
identified as such, is not to be taken as an expression of opinion as to whether or not they are subject to
proprietary rights.
While the advice and information in this book are believed to be true and accurate at the date of publication,
neither the authors nor the editors nor the publisher can accept any legal responsibility for any errors or
omissions that may be made. The publisher makes no warranty, express or implied, with respect to the
material contained herein.
Managing Director, Apress Media LLC: Welmoed Spahr
Acquisitions Editor: Jonathan Gennick
Development Editor: Laura Berendson
Coordinating Editor: Jill Balzano
Cover image designed by Freepik (www.freepik.com)
Distributed to the book trade worldwide by Springer Science+Business Media LLC, 1 New York Plaza, Suite
4600, New York, NY 10004. Phone 1-800-SPRINGER, fax (201) 348-4505, e-mail orders-ny@springer-sbm.
com, or visit www.springeronline.com. Apress Media, LLC is a California LLC and the sole member (owner)
is Springer Science + Business Media Finance Inc (SSBM Finance Inc). SSBM Finance Inc is a Delaware
corporation.
For information on translations, please e-mail [email protected]; for reprint,
paperback, or audio rights, please e-mail [email protected].
Apress titles may be purchased in bulk for academic, corporate, or promotional use. eBook versions and
licenses are also available for most titles. For more information, reference our Print and eBook Bulk Sales
web page at http://www.apress.com/bulk-sales.
Any source code or other supplementary material referenced by the author in this book is available to
readers on GitHub (https://github.com/Apress). For more detailed information, please visit http://www.
apress.com/source-­code.
Printed on acid-free paper
I’ve said this before, but it is worth repeating because it matters.
This book is dedicated to the #sqlfamily, the most dedicated and
passionate and largest technical community I know of
in the world. Without you we would not have a product.
Table of Contents
About the Author��������������������������������������������������������������������������������������������������� xiii

About the Technical Reviewer���������������������������������������������������������������������������������xv

Acknowledgments�������������������������������������������������������������������������������������������������xvii

Foreword����������������������������������������������������������������������������������������������������������������xix

Introduction������������������������������������������������������������������������������������������������������������xxi

Chapter 1: Project Dallas Becomes SQL Server 2022����������������������������������������������� 1


Project Dallas�������������������������������������������������������������������������������������������������������������������������������� 1
Becoming SQL Server 2022���������������������������������������������������������������������������������������������������������� 3
Announcing SQL Server 2022������������������������������������������������������������������������������������������������������� 7
Private to Public Preview������������������������������������������������������������������������������������������������������������ 11
The Path to General Availability�������������������������������������������������������������������������������������������������� 14
Introducing SQL Server 2022������������������������������������������������������������������������������������������������������ 14
Built on a Foundation������������������������������������������������������������������������������������������������������������ 15
Wheel of Power��������������������������������������������������������������������������������������������������������������������� 16
Cloud Connected������������������������������������������������������������������������������������������������������������������� 17
Built-In Query Intelligence����������������������������������������������������������������������������������������������������� 18
Industry-Proven Database Engine����������������������������������������������������������������������������������������� 18
Data Virtualization and Object Storage���������������������������������������������������������������������������������� 19
Enhancing T-SQL for Developers������������������������������������������������������������������������������������������� 19
Getting Started with SQL Server 2022���������������������������������������������������������������������������������������� 20
How to Get SQL Server 2022������������������������������������������������������������������������������������������������� 20
Installing SQL Server 2022���������������������������������������������������������������������������������������������������� 20
Learn All the Features and Editions��������������������������������������������������������������������������������������� 21
Learn About Pricing and Licensing���������������������������������������������������������������������������������������� 21
Get Training on SQL Server 2022������������������������������������������������������������������������������������������� 21
v
Table of Contents

Go Deeper with Our Blog Series�������������������������������������������������������������������������������������������� 21


Download Book Code and Samples��������������������������������������������������������������������������������������� 21
A Cloud-Connected, Intelligent, and Industry-Proven Data Platform������������������������������������������ 22

Chapter 2: Install and Upgrade������������������������������������������������������������������������������� 25


How to Install SQL Server 2022�������������������������������������������������������������������������������������������������� 25
Prerequisites������������������������������������������������������������������������������������������������������������������������� 26
What Is Different for SQL Server 2022?��������������������������������������������������������������������������������� 26
Other Installation Methods���������������������������������������������������������������������������������������������������� 31
Setting Up the Azure Extension for SQL Server��������������������������������������������������������������������������� 32
What You Should Know First�������������������������������������������������������������������������������������������������� 33
Providing Values for the Feature Setup��������������������������������������������������������������������������������� 36
Connect to Azure After Setup������������������������������������������������������������������������������������������������� 40
Removing the Azure Extension for SQL Server���������������������������������������������������������������������� 41
Deploying on Other Platforms����������������������������������������������������������������������������������������������������� 41
Side-by-Side and Multi-instance Installations���������������������������������������������������������������������������� 42
How to Upgrade to SQL Server 2022������������������������������������������������������������������������������������������ 42
The Importance of dbcompat������������������������������������������������������������������������������������������������������ 42
Configuration������������������������������������������������������������������������������������������������������������������������������ 43
Easy to Install and Upgrade�������������������������������������������������������������������������������������������������������� 43

Chapter 3: Connect Your Database to the Cloud����������������������������������������������������� 45


The Hybrid SQL Server���������������������������������������������������������������������������������������������������������������� 45
What Is Hybrid?��������������������������������������������������������������������������������������������������������������������� 46
SQL Server Hybrid Over the Years����������������������������������������������������������������������������������������� 46
The SQL Server 2022 Hybrid Lineup�������������������������������������������������������������������������������������� 47
Managed Disaster Recovery with Azure SQL Managed Instance������������������������������������������������ 49
Project Chimera and DAG������������������������������������������������������������������������������������������������������ 50
The Link Feature for Azure SQL Managed Instance��������������������������������������������������������������� 52
How It Works������������������������������������������������������������������������������������������������������������������������� 52
Using the Link Feature for Offline Disaster Recovery������������������������������������������������������������ 55
Keep in Mind These Details��������������������������������������������������������������������������������������������������� 71
The Future for the Link Feature for Azure SQL Managed Instance���������������������������������������� 72
vi
Table of Contents

Azure Synapse Link for SQL Server�������������������������������������������������������������������������������������������� 73


What Is Synapse Link for SQL Server?���������������������������������������������������������������������������������� 74
How Does Synapse Link Work?��������������������������������������������������������������������������������������������� 74
Try Out Synapse Link for SQL Server������������������������������������������������������������������������������������� 77
More Details About Synapse Link���������������������������������������������������������������������������������������� 108
Synapse Link Could Change Analytics for You��������������������������������������������������������������������� 113
Azure Active Directory (AAD) Authentication����������������������������������������������������������������������������� 113
How Does AAD Authentication Work?���������������������������������������������������������������������������������� 114
Setting Up and Using AAD Authentication��������������������������������������������������������������������������� 115
Microsoft Purview Policy Management������������������������������������������������������������������������������������� 124
How Do Purview Access Policies Work?������������������������������������������������������������������������������ 125
Using Purview Access Policies�������������������������������������������������������������������������������������������� 128
Connecting SQL to the World���������������������������������������������������������������������������������������������������� 144

Chapter 4: Built-In Query Intelligence������������������������������������������������������������������ 145


Built-In Query Intelligence in SQL Server 2022������������������������������������������������������������������������� 147
The New Query Store���������������������������������������������������������������������������������������������������������������� 149
On by Default����������������������������������������������������������������������������������������������������������������������� 150
Query Store Hints���������������������������������������������������������������������������������������������������������������� 152
Query Store Support for Secondary Replicas���������������������������������������������������������������������� 154
Store for IQP������������������������������������������������������������������������������������������������������������������������ 158
IQP Nextgen Defaults���������������������������������������������������������������������������������������������������������������� 159
Approximate Percentile������������������������������������������������������������������������������������������������������� 159
Optimized Plan Forcing������������������������������������������������������������������������������������������������������� 160
IQP Nextgen with dbcompat 140+�������������������������������������������������������������������������������������������� 169
Memory Grant Percentiles��������������������������������������������������������������������������������������������������� 172
Memory Grant Feedback Persistence���������������������������������������������������������������������������������� 172
The Intelligent Query Processor������������������������������������������������������������������������������������������������ 183

Chapter 5: Built-In Query Intelligence Gets Even Better��������������������������������������� 185


Parameter-Sensitive Plan (PSP) Optimization��������������������������������������������������������������������������� 185
What Is a Parameter-Sensitive Plan?���������������������������������������������������������������������������������� 186
How It All Started����������������������������������������������������������������������������������������������������������������� 188
vii
Table of Contents

How Does PSP Optimization Work?������������������������������������������������������������������������������������� 189


Let’s See PSP Optimization in Action����������������������������������������������������������������������������������� 192
What Other Details About PSP Optimization Should I Know?���������������������������������������������� 208
PSP Optimization Is a Powerful Innovation�������������������������������������������������������������������������� 211
Cardinality Estimation (CE) Model Feedback���������������������������������������������������������������������������� 211
What Is the CE Model Problem?������������������������������������������������������������������������������������������ 212
Try Out an Exercise�������������������������������������������������������������������������������������������������������������� 214
Limits and More Details About CE Feedback����������������������������������������������������������������������� 224
Degree of Parallelism (DOP) Feedback������������������������������������������������������������������������������������� 225
Why Did We Build DOP Feedback?�������������������������������������������������������������������������������������� 226
What Is DOP Feedback?������������������������������������������������������������������������������������������������������ 227
How Does DOP Feedback Work?����������������������������������������������������������������������������������������� 227
Let’s Try DOP Feedback������������������������������������������������������������������������������������������������������� 230
What Else Should I Know About DOP Feedback?���������������������������������������������������������������� 240
An Engine That Works for You��������������������������������������������������������������������������������������������������� 242

Chapter 6: The Meat and Potatoes of SQL Server������������������������������������������������� 245


The Core Engine Is the Meat and Potatoes of SQL Server�������������������������������������������������������� 245
Security������������������������������������������������������������������������������������������������������������������������������������� 246
Ledger for SQL Server��������������������������������������������������������������������������������������������������������� 246
Encryption Enhancements��������������������������������������������������������������������������������������������������� 265
Security Permission Enhancements������������������������������������������������������������������������������������ 269
Performance and Scalability����������������������������������������������������������������������������������������������������� 272
Columnstore and Batch Mode Improvements��������������������������������������������������������������������� 273
Scalability Improvements���������������������������������������������������������������������������������������������������� 275
“Hands-Free” tempdb��������������������������������������������������������������������������������������������������������� 278
More Concurrency Improvements��������������������������������������������������������������������������������������� 288
Availability��������������������������������������������������������������������������������������������������������������������������������� 289
Contained Availability Groups���������������������������������������������������������������������������������������������� 289
Other AG Enhancements������������������������������������������������������������������������������������������������������ 302
Recovery Enhancements����������������������������������������������������������������������������������������������������� 304

viii
Table of Contents

Backup/Restore Enhancements������������������������������������������������������������������������������������������ 306


Multi-write Replication�������������������������������������������������������������������������������������������������������� 311
Other Engine “Stuff”����������������������������������������������������������������������������������������������������������������� 312
XML Compression���������������������������������������������������������������������������������������������������������������� 312
In-Memory OLTP Memory Management������������������������������������������������������������������������������ 313
Auto-drop Statistics������������������������������������������������������������������������������������������������������������� 313
Resumable Add Table Constraints��������������������������������������������������������������������������������������� 313
New Wait Types������������������������������������������������������������������������������������������������������������������� 314
New Extended Events���������������������������������������������������������������������������������������������������������� 314
An Industry-Proven Engine������������������������������������������������������������������������������������������������������� 314

Chapter 7: Data Virtualization and Object Storage����������������������������������������������� 317


Data Virtualization in SQL Server 2022������������������������������������������������������������������������������������� 318
REST, Azure Storage, and S3����������������������������������������������������������������������������������������������� 318
Project Gravity Becomes Polybase v3��������������������������������������������������������������������������������� 319
Polybase v3 File Formats���������������������������������������������������������������������������������������������������� 321
Using the New Polybase v3������������������������������������������������������������������������������������������������� 322
Try Out Polybase v3������������������������������������������������������������������������������������������������������������� 324
What Else Do You Need to Know������������������������������������������������������������������������������������������ 344
Backup and Restore with S3 Compatible Object Storage��������������������������������������������������������� 344
How Does It Work?�������������������������������������������������������������������������������������������������������������� 345
Let’s Look at an Example����������������������������������������������������������������������������������������������������� 346
Migration from AWS������������������������������������������������������������������������������������������������������������� 348
SQL Server Is a Data Hub���������������������������������������������������������������������������������������������������������� 350

Chapter 8: New Application Scenarios with T-SQL����������������������������������������������� 351


JSON Functions������������������������������������������������������������������������������������������������������������������������� 352
Prerequisites for Exercises�������������������������������������������������������������������������������������������������� 353
JSON_ARRAY and ISJSON��������������������������������������������������������������������������������������������������� 353
JSON_OBJECT and JSON_PATH_EXISTS����������������������������������������������������������������������������� 357
T-SQL Surface Area������������������������������������������������������������������������������������������������������������������� 358
Prerequisites����������������������������������������������������������������������������������������������������������������������� 359
DATETRUNC������������������������������������������������������������������������������������������������������������������������� 359
ix
Table of Contents

WINDOW Clause������������������������������������������������������������������������������������������������������������������� 361


GREATEST and LEAST���������������������������������������������������������������������������������������������������������� 363
IS [NOT] DISTINCT FROM����������������������������������������������������������������������������������������������������� 366
STRING_SPLIT��������������������������������������������������������������������������������������������������������������������� 369
TRIM Function Extensions��������������������������������������������������������������������������������������������������� 370
Bit T-SQL Functions������������������������������������������������������������������������������������������������������������� 372
Time Series������������������������������������������������������������������������������������������������������������������������������� 379
Prerequisites����������������������������������������������������������������������������������������������������������������������� 380
DATE_BUCKET��������������������������������������������������������������������������������������������������������������������� 380
GENERATE_SERIES�������������������������������������������������������������������������������������������������������������� 382
Gap Filling with FIRST_VALUE and LAST_VALUE����������������������������������������������������������������� 385
T-SQL Is Alive and Well�������������������������������������������������������������������������������������������������������������� 388

Chapter 9: SQL Server 2022 on Linux, Containers, and Kubernetes��������������������� 389


SQL Server 2022 on Linux��������������������������������������������������������������������������������������������������������� 390
What’s New for SQL Server 2022���������������������������������������������������������������������������������������� 390
Deploying SQL Server 2022 on Linux���������������������������������������������������������������������������������� 392
How to Connect and Use SQL Server 2022 on Linux����������������������������������������������������������� 394
Optimizing SQL Server 2022 on Linux��������������������������������������������������������������������������������� 396
HADR for SQL Server 2022 on Linux������������������������������������������������������������������������������������ 397
SQL Server 2022 Containers����������������������������������������������������������������������������������������������������� 399
Why Containers?������������������������������������������������������������������������������������������������������������������ 399
Using SQL Server 2022 Containers�������������������������������������������������������������������������������������� 401
Building a Customized Container Image������������������������������������������������������������������������������ 404
The Container Switch Method��������������������������������������������������������������������������������������������� 404
SQL Server 2022 on Kubernetes����������������������������������������������������������������������������������������������� 405
Why k8s������������������������������������������������������������������������������������������������������������������������������� 406
Deploying SQL Server on k8s���������������������������������������������������������������������������������������������� 406
Connecting and Using SQL Server on k8s��������������������������������������������������������������������������� 410
High Availability with k8s���������������������������������������������������������������������������������������������������� 410
Helm Charts������������������������������������������������������������������������������������������������������������������������� 412
The World of Linux, Containers, and k8s����������������������������������������������������������������������������������� 412

x
Table of Contents

Chapter 10: SQL Server 2022 on Azure Virtual Machines������������������������������������ 413


What Is SQL Server on Azure Virtual Machines?����������������������������������������������������������������������� 413
Planning for Deployment����������������������������������������������������������������������������������������������������������� 414
The SQL Server IaaS Agent Extension��������������������������������������������������������������������������������������� 418
What Is the SQL Server IaaS Agent Extension?������������������������������������������������������������������� 418
IaaS Agent Extension Modes����������������������������������������������������������������������������������������������� 419
Installing the Agent Extension��������������������������������������������������������������������������������������������� 419
Details of the Agent Extension��������������������������������������������������������������������������������������������� 420
Deploying SQL Server on an Azure Virtual Machine������������������������������������������������������������������ 421
Prerequisites����������������������������������������������������������������������������������������������������������������������� 421
Steps to Deploy������������������������������������������������������������������������������������������������������������������� 421
Other Deployment Methods������������������������������������������������������������������������������������������������� 431
Exploring and Connecting to the SQL Server Virtual Machine�������������������������������������������������� 432
Exploring Your Azure Virtual Machine in the Portal�������������������������������������������������������������� 432
Connecting to the Azure Virtual Machine����������������������������������������������������������������������������� 436
Exploring a SQL Virtual Machine in the Portal��������������������������������������������������������������������� 438
Connecting to SQL Server on an Azure Virtual Machine������������������������������������������������������ 439
Migrating to SQL Server on an Azure Virtual Machine�������������������������������������������������������������� 440
Optimizing Performance����������������������������������������������������������������������������������������������������������� 441
Virtual Machine Size������������������������������������������������������������������������������������������������������������ 441
Storage Performance���������������������������������������������������������������������������������������������������������� 442
SQL Best Practices Assessment������������������������������������������������������������������������������������������ 443
High Availability������������������������������������������������������������������������������������������������������������������������� 445
Built-In Fault Tolerance with Azure�������������������������������������������������������������������������������������� 445
Failover Cluster Instance (FCI)��������������������������������������������������������������������������������������������� 446
Always On Availability Groups��������������������������������������������������������������������������������������������� 447
Disaster Recovery��������������������������������������������������������������������������������������������������������������������� 449
Storage Fault Tolerance������������������������������������������������������������������������������������������������������� 449
Backup Database Options��������������������������������������������������������������������������������������������������� 449
Using Azure Backup������������������������������������������������������������������������������������������������������������� 450

xi
Table of Contents

Monitoring��������������������������������������������������������������������������������������������������������������������������������� 450
Azure Monitor���������������������������������������������������������������������������������������������������������������������� 451
Virtual Machine Metrics and Logs��������������������������������������������������������������������������������������� 451
Azure Insights���������������������������������������������������������������������������������������������������������������������� 452
Azure Is the Best Cloud for SQL Server������������������������������������������������������������������������������������� 452

Chapter 11: SQL Edge to Cloud����������������������������������������������������������������������������� 455


Develop Once, Deploy Anywhere����������������������������������������������������������������������������������������������� 455
Azure SQL Edge������������������������������������������������������������������������������������������������������������������������� 457
When to Use Azure SQL Edge���������������������������������������������������������������������������������������������� 458
SQL Server�������������������������������������������������������������������������������������������������������������������������������� 458
When to Use SQL Server������������������������������������������������������������������������������������������������������ 459
Azure Arc–Enabled SQL Server������������������������������������������������������������������������������������������� 460
Linux, Containers, and Kubernetes������������������������������������������������������������������������������������������� 460
When to Use SQL Server on Linux��������������������������������������������������������������������������������������� 460
When to Use SQL Server Containers����������������������������������������������������������������������������������� 461
When to Use SQL Server on Kubernetes����������������������������������������������������������������������������� 462
SQL Server on Azure Virtual Machines�������������������������������������������������������������������������������������� 463
When to Use SQL Server on Azure Virtual Machines����������������������������������������������������������� 464
Azure SQL Managed Instance��������������������������������������������������������������������������������������������������� 465
When to Use Azure SQL Managed Instance������������������������������������������������������������������������� 466
Azure SQL Database������������������������������������������������������������������������������������������������������������������ 467
When to Use Azure SQL Database��������������������������������������������������������������������������������������� 469
Azure Arc–Enabled SQL Managed Instance������������������������������������������������������������������������������ 470
When to Use Azure Arc–Enabled SQL Managed Instance���������������������������������������������������� 472
SQL Is Everywhere You Need It������������������������������������������������������������������������������������������������� 472

Index��������������������������������������������������������������������������������������������������������������������� 475

xii
About the Author
Bob Ward is a principal architect for the Microsoft Azure
Data team, which owns the development for all of SQL and
Azure SQL. Bob has worked for Microsoft for 29 years on
every version of SQL Server shipped from OS/2 1.1 to SQL
Server 2022, including Azure SQL and Azure Arc. He is a
well-known speaker on SQL Server and Azure SQL, often
presenting talks on new releases, internals, and specialized
topics at events such as the PASS Summit, SQLBits, Azure
Data Conference, VSLive, Microsoft Build, Microsoft Inspire,
Microsoft Ignite, and many other events. You can follow him
on Twitter at @bobwardms or LinkedIn at https://linkedin.com/in/bobwardms.
Bob is the author of Apress books Pro SQL Server on Linux, SQL Server 2019 Revealed,
and Azure SQL Revealed.

xiii
About the Technical Reviewer
Erin Stellato is a senior program manager on the SQL
Experiences team, helping advance tools that customers
use daily with Azure SQL. She is passionate about data and
chocolate, not always in that order. She previously worked
as a consultant and was a Data Platform MVP and has been
an active member of the SQL Server community as both
a volunteer and speaker. Her areas of interest within the
engine include Query Store, Extended Events, statistics, and
performance tuning, and she also enjoys helping accidental/
involuntary DBAs figure out how SQL Server works.

xv
Acknowledgments
I want to first thank God for sending his one and only son, Jesus, for the forgiveness of
my sins, the immersion of the Holy Spirit, and the promise of everlasting life. Without my
faith, I am nothing.
I want to thank my wife, Ginger, so much for her love and dedication to me for all of
our years together but especially during the writing of this book. You patiently allowed
me to work crazy hours and situations even on some of our vacation time to bring this
book to life. I also want to thank my sons Troy and Ryan, and Ryan’s wife Blair, my
daughter-in-law. You all give me hope for the future as I see you all grow and exhibit
grace and truth in everything you do.
This is my fourth book with Apress, and I want to personally thank Jonathan Gennick
from Apress. Jonathan gave me my first chance at authoring when many publication
companies turned me down. Thank you, Jonathan, for always supporting me and letting
me “write my way.” I also want to thank Jill Balzano from Apress whom I’ve never met in
person but who always is so kind and professional despite all the crazy deadlines we try
to meet.
I asked Erin Stellato to be my technical reviewer because she is one of the deepest
experts on SQL and just an incredible person. Turns out it was a blessing when she
joined Microsoft during the writing of the book as we could discuss confidential
information. Erin, thank you for your thoroughness and great attitude as we all put
immense pressure on you to review so many chapters late in the cycle.
There were so many people at Microsoft who supported my work on this book.
But, first, I want to thank Joe Sack, Pedro Lopes, and James Rowland-Jones who no
longer work at Microsoft but were a huge part of helping me craft the story of
SQL Server 2022, which you see in this book.
At Microsoft I want to thank my leaders who support all of my efforts including
Rohan Kumar, Peter Carlin, Asad Khan, and Sanjay Mishra. The true heroes of this
book are all the people at Microsoft in the engineering team who helped me with all
my questions and the review of complex topics and for giving me some great quotes.
This list is long, but I have to call out everyone who helped because each of them

xvii
Another Random Scribd Document
with Unrelated Content
The work out of which this monograph grew was begun through the
encouragement of Judge Edward F. Waite of the Hennepin County
Juvenile Court. His earnest co-operation and my interest in the field
of mental testing has led me to continue the study. Judge Waite's
insight into his court problems resulted in the early organization of a
Juvenile Court clinic (153, 170) in Minneapolis. The clinic is in charge
of Dr. Harris Dana Newkirk, who has contributed materially to this
study by his thorough medical examination of each of the cases
brought to him. To the staff at the probation office I am also much
indebted.
The earnest help of Superintendent D. C. MacKenzie, of the Glen
Lake Farm School for the juvenile delinquents of Hennepin County,
made a close study of our most interesting group of boys much
more profitable personally than I have shown here. For detailed
expert work in tabulation and in examinations I wish to express my
thanks to my advanced students, a half dozen of whom have
contributed materially to the data of this book.
James Burt Miner.
Carnegie Institute of Technology
Pittsburgh, Pa.
CHAPTER I. INTRODUCTION

As an interpretation of the results which have been obtained with


mental tests, this book lies between the topics of deficiency and
delinquency. It is an attempt to discover the significance of objective
measurements of ability in connection with both of these fields. The
pressing practical problem was to find out what positions on a scale
for testing mental development were symptomatic of social
deficiency. After working out a percentage method for conservatively
indicating these borderlines for tested deficiency, it was then
possible to reinterpret the test records of over 9000 delinquents who
have been examined with some form of the well-known Binet Scale.
The size of the problem of the deficient delinquent has thus been
determined on a significant scientific plan. The outcome is a new
basis for judging the current statements about this problem by those
who have used the Binet scale. Scores of investigators by their
tireless energy have provided data which may now be compared for
many types of delinquents and in many parts of the country. Some
sixty studies of deficient delinquents have been thus summarized
from the point of view of psychological tests.
Closely related to the problem of the frequency of feeble-
mindedness among delinquents is the question of the cause of
delinquency. This has further been considered in the light of the
most important scientific studies, especially those using the method
of correlation. Among these researches stands out the fundamental
investigation of the causes of criminality by Goring, a work which
has received very inadequate attention in this country, although it
involved ten years study of a group of 3000 convicts by the best
quantitative methods. The careful study of these objective
investigations should take the question of the relation of deficiency
and delinquency out of the realm of opinion and theory. It may be
expected to have an important influence upon the social handling of
these problems. In this connection I have added a chapter of
suggestions which have grown out of my year's study of the
education of deficients and delinquents in European schools and
institutions.
To determine the size of the problem of dealing with deficients,
especially deficient delinquents, is a task of first importance. In spite
of our more conservative basis for judging the results with tests, the
necessity of caring for the feeble-minded remains the most vital
problem connected with social welfare. The movement for more
individual training in our schools, which has been gaining such
headway, may also be encouraged by the evidence that
maladjustment to school work is also definitely related to
delinquency.
It is essential that we should have objective data for determining the
borderline of tested deficiency among adults. To meet the present
serious lack of knowledge on this point, new data were collected
which for the first time afford the means of determining, by the use
of a randomly selected group what is a conservative borderline of
tested deficiency for those intellectually mature. These data include
the Binet test records for all the 15-year-old children who resided in
seven school districts in Minneapolis and who had not graduated
from the eighth grade.
The urgency of plans for indefinitely segregating certain types of the
feeble-minded, especially deficient delinquents, has placed a new
emphasis on those quantitative aids to diagnosis. The difficulty of
establishing feeble-mindedness before a court has been called to
attention by both Supt. C. A. Rogers (173)[1] of the Minnesota School
for Feeble-Minded, and Supt. Walter E. Fernald (104) of the
Massachusetts School. Both of these men recognize that
psychological tests are the most hopeful way of improving this
situation.
A fundamental feature of the diagnosis of deficiency is the plan here
advocated for designating the borderlines on a scale on the basis of
a percentage definition of tested deficiency. This involves the
distinction of intellectual deficiency from certain rare volitional forms
of feeble-mindedness, which the tests do not at present detect. This
percentage definition seems to afford the best approach to a test
diagnosis. It is apparent that the data are insufficient for finally
establishing such a quantitative description of the lower limit for
passable intellects on a mental scale. The plan, however, may be
easily adjusted to new data, and meanwhile avoids some of the
serious current misinterpretations of test results.
While the idea of a quantitative definition of the borderline of
deficiency is not new, the percentage method seems to have certain
fundamental advantages over either the “intelligence quotient” of
Stern (188), the “intelligence coefficient” of Yerkes (226), or the
description in terms of deviation, mentioned by Norsworthy (159)
and Pearson (164, 166, 167). Several investigators, including Terman
(57) and Yerkes (226), are utilizing the percentage method indirectly
for describing the borderline of feeble-mindedness, but have
inadequately distinguished it from the ratios. While ratio and
deviation methods are possibly more serviceable for certain
purposes, they are especially faulty near the borderline of deficiency,
since they are affected by variations in the units of measurement
and in the form of distribution from age to age. My paper on a
percentage definition and the detailed plan for determining the
borderline in the Binet scale, which was read at the meeting of the
American Psychological Association in 1915, seems to have been
contemporaneous with a similar suggestion by Pintner and Paterson
(44). They, however, would restrict the term “feeble-mindedness” to
tested deficiency, while I advocate the use of percentage borderlines
on a test scale as symptomatic of one form of feeble-mindedness,
much as excess of normal temperature on a clinical thermometer is
symptomatic of disease.
Although no system of objective tests will ever dispense with the
need for expert interpretation in diagnosing individual cases, still
there are few who would doubt that it is desirable to reduce the
option of expert judgment as much as we reasonably can. This is the
scientific method of procedure. The borderline cases, however, which
are often most troublesome in their delinquencies, are just those
which will longest defy rigid rules. The diagnostician who wants to
be as free as possible from external restraint will find in this border
field of mental capacity a happy hunting ground. His scientific
instincts should make him eager to discover when he leaves the
mundane sphere and sallies forth into uncharted realms where he
bears the full responsibility of his own opinion. Let me hasten to add
that reasoning from objective data in the mass to the diagnosis of an
individual case may lead to serious mistakes, unless one keeps alert
to detect the exception from the general rule, and unless one
understands the numerous sources of error entering into an
examination. On the other hand the test results when properly
interpreted afford the most important criteria on which to base a
prognosis if they are considered in relation to the history of the case
and the medical examination.
By the use of more conservative borderlines for raising the
presumption of deficiency and also by designating a doubtful
position on the scale, on the plan advocated herein, it is possible to
make scales for testing mental capacity more serviceable both to the
clinician and to the amateur tester. The latter may use the scales for
his own information or may wish to discover whether an examination
by an expert in mental development is desirable, without attempting
to make a diagnosis himself. The scale may thus take a place in the
study of child mentality analogous to the familiar Snellen chart in the
testing of vision. For every teacher familiarity with a development
scale may thus become as essential and desirable as the knowledge
of the chart for eye testing. It should find a place in all progressive
schools which do not have the services of a clinician.
The Binet system of tests was used for obtaining new data on
groups of juvenile delinquents in Minneapolis and Pittsburgh. The
use of this scale, around which the discussion centers, grew out of
the necessity for immediate practical results for the clinic at the
Minneapolis Juvenile Court which I was called upon to serve. In
1912, when that work began, there was practically nothing
approaching norms with children for any other scale of tests. Even
today it is plain that there is more data available for interpreting
results with the Binet scale than with any other system of tests.
While my experience would make me unwilling to advocate the Binet
tests as an ideal method for building up a measuring scale, I still feel
that it remains the most useful method at present for discovering the
fundamental symptoms of intellectual deficiency. The percentage
method, here advocated, as the best way available for determining
the borderlines with a scale, would be quite as serviceable, however,
with any other testing system. It has been my aim to contribute to
the interpretation of the results of the tests as they are, not to
perfecting the arrangement or details of the separate tests.[2] It
happens that one of the main objections which has been raised to
the Binet scale, the inadequacy of its tests for the older ages, loses
its force so far as the diagnosis of feeble-mindedness is concerned
for those who accept the borderlines described in this paper.
Some diagnosticians may hesitate to use the Binet scale because of
the criticisms it has received. Yerkes and Bridges state: “Indeed, we
feel bound to say that the Binet scale has proved worse than useless
in a very large number of cases” (226, p. 94). So far as this
objection arises from the attempt to use the descriptions of the
borderline of feeble-mindedness published with Binet scales, it will
meet with a wide response. The difficulty is hardly less, as I shall
show, with other scales. The definition of the borderline is certainly
the vital point with any objective method for aiding diagnosis. Only
by improving methods for determining the borderline can this
weakness be attacked. The central contribution of this paper is
directed, therefore, to this problem of the interpretation of the
borderline, so that objective scales may be made more reliable for
purposes of diagnosis.
In Part Two I have added an intensive discussion of the
measurement of development and a comparison of the different
objective methods for describing the borderline. This may well be
omitted by those who are not interested in the technical aspects of
these questions. To those who care only for accounts of individual
lives, let me say that I am contributing nothing herein to that
important field which has been covered in authoritative form by Dr.
Healy (27) and by Dr. Goddard (112). They will find instead, I hope,
the fascination of figures, a picture book in which probability curves
take the place of photographs and biographies, in which general
tendencies are evaluated and attention is focussed upon the problem
of properly diagnosing deficiency and upon plans for the care of the
feeble-minded, whether they be potential or actual delinquents.

1. Numbers in parenthesis indicate the references in the bibliography at the


close of the book.

2. Those concerned with other features of the Binet scale will find an admirable
bibliography by Samuel C. Kohs, Journal of Educational Psychology, April,
May and June, 1914, and September, October, November, and December,
1917. Other references are contained in the Bibliography by L. W. Crafts (9).
PART ONE
PRACTICAL CONSIDERATIONS
CHAPTER II. THE FUNCTIONS OF A SCALE IN
DIAGNOSIS

A. The Meaning of Intellectual Deficiency.

Whatever form the definition of feeble-mindedness may take, in this country at


least[3] the concept has become quite firmly established as describing the
condition of those who require social guardianship, because, with training, they
do not develop enough mentally to live an independent life in society. The
feeble-minded are socially deficient because of a failure to develop mentally.
They are proper wards of the state because of this mental deficiency. Goddard
says, they are “incapable of functioning properly in our highly organized
society” (112, p. 6). The most generally quoted verbal description of the upper
line of social unfitness is that of the British Royal Commission on Feeble-
Mindedness: “Persons who may be capable of earning a living under favorable
circumstances, but are incapable from mental defect existing from birth or from
an early age (a) of competing on equal terms with their normal fellows; or (b)
of managing themselves and their affairs with ordinary prudence.” It is clear
that the intention is to distinguish mental deficiency from senile dementia, from
hysteria and from insanity, in which there is a temporary or permanent loss of
mental ability rather than a failure to develop. Feeble-mindedness may,
however, arise from epilepsy or from other diseases or accidents in early life as
well as from an inherent incapacity for development. Moreover, mental
deficiency, or feeble-mindedness, (I use the terms interchangeably) does not
imply that the social unfitness is always caused by intellectual deficiency. Mind
is a broader term than intellect, as we shall note in the next section.
This definition of the feeble-minded is the main idea expressed by Witmer
(221), Tredgold (204), Pearson (164), and Murdock (164). The historical
development of the concept is traced by Rogers (172) and Norsworthy (159).
It is criticized by Kuhlmann (140) as impractical and indefinite. The
indefiniteness is indicated by such terms as “under favorable circumstances,”
“on equal terms,” and “with ordinary prudence.” This objectionable uncertainty
as to social fitness can be considerably relieved for those types of feeble-
mindedness which involve the inability to pass mental tests, since this result
can later be correlated with subsequent social failure and predictions made
during childhood on the basis of the tests. Attempts to make the concept of
feeble-mindedness more definite have, therefore, naturally taken some
quantitative form in relation to objective tests. Binet and the French
commission in 1907 (77) called attention to the method in use in Belgium for
predicting unfitness objectively on the basis of the amount of retardation in
school at different ages. With the appearance in 1908 of the Binet-Simon
revised scale for measuring mental development, quantitative descriptions
began to be concerned with the borderlines of mental deficiency on scales of
tests.
While the quantitative descriptions of tested deficiency do not include all forms
of feeble-mindedness, as I shall show in the next section, they have made the
diagnosis of the majority of cases much more definite. Nobody would think of
returning to the days when the principal objective criteria were signs of
Cretinism, Mongolianism, hydrocephalus, microcephalus, epilepsy, meningitis,
etc., which LaPage (141) has shown are not found among more than 9% of
784 children in the Manchester special schools. The impossibility of agreeing
upon subjective estimates of mental capacity without the use of objective
criteria is well shown by Binet's methodical comparison of the admission
certificates filled out within a few days of each other by the alienists for the
institutions of Sainte-Anne, Bicêtre, the Salpêtreire and Vaucluse. These
physicians gave their judgments as to whether a case was an idiot, imbecile or
higher grade. Binet says: “We have compared several hundreds of these
certificates, and we think we may say without exaggeration that they looked as
if they had been drawn by chance out of a sack” (77, p. 76).
The rapid accumulation of data with psychological tests has made it possible to
take our first halting steps in the direction of greater definiteness in diagnosis
by a larger use of objective methods. This increase in significance of the
concept of deficiency is fruitful at once in estimating the size of the social
problem and planning means for undertaking the care of these unfortunates.
We can discover something of the error in the previous subjective estimates of
the frequency of feeble-mindedness. We can bring together and compare the
work of different investigators, not only in our country, but throughout the
world. We can discover, for example, how important the problem of deficiency
is among different groups of delinquents, knowing that the differences are not
to be explained by differences in expert opinion. Furthermore, we can now
determine, with considerable accuracy, whether the diagnosis made by a
reliable examiner is independent of his personal opinion.
If we disregard the natural antipathy of many people to anything which tends
to limit the charming vagueness of their mental outlook, we may endeavor to
chart this horizon of tested deficiency with something of the definiteness of
figures, which shall at the same time indicate a range of error. As soon as our
aim comes to be to plot the borderline on a measuring scale of mental ability,
we find that the borderline must be so stated that we can deal with either
adults or children. Two sorts of limiting regions must be described, one for
mature minds and one for immature minds. The latter will be in the nature of a
prediction as to what sort of ability the children will show when they grow up.
We must keep in mind, therefore, that we should attempt our quantitative
definition for both growing and adult minds. As soon as the growing mind
passes the lower limit for the mature it is then guaranteed access to the social
seas although it may never swim far from shore nor develop further with
advancing years. In seeking greater definiteness, our aim should then be to
describe both the limit for the mature individuals and the limit for the immature
of each age. In this paper the definition will be restricted to intellectual
deficiency, i. e., tested deficiency. It will take the form of describing the
positions on a scale below which fall the same lowest percentage of intellects.
This percentage definition of intellectual deficiency offers such a simple method
of consistently describing the borderlines for mature and immature that it is
surprising so little attempt has previously been made to work it out for a
system of tests. Although the principle on which the definition is based
depends upon the distribution curve of ability, it is concerned only with the
lower limit of the distribution. Since the exact form of this distribution is
uncertain I have preferred to call it a percentage definition of intellectual
deficiency rather than to state the limits in terms of the variability of ability.
Moreover the lowest X per cent. in mental development requires no further
explanation to be understood by the layman.

B. Forms of Mental Deficiency Not Yet Discoverable by Tests.

The first broad conclusion that impresses those who try to use mental scales
for diagnosing feeble-mindedness is that the lower types, the idiots and
imbeciles, can be detected with great accuracy by an hour's testing. The
difficulties pile up as soon as the individual rises above the imbecile group. The
practical experience of those in institutions for the feeble-minded here
becomes of fundamental importance. They are able to supply the history of
exceptions that should make us cautious about our general rules. Certain
people whom they have known for years to be unable to adjust themselves
socially because their minds have not reached the level of social fitness will yet
be able to pass considerably beyond the lower test limit for mature minds. The
mental scales can only detect those feeble-minded who cannot succeed with
our present tests. This is the basal principle in using any system of tests.
Stated in another way, this first caution for anybody seeking the assistance of a
mental scale is that tests may detect a feeble-minded person, but when a
person passes them it does not guarantee social fitness. The negative
conclusion, “this person is not feeble-minded,” can not be drawn from tests
alone. Mental tests at present are positive and not negative scales. This fact
will probably always make the expert's judgment essential before the discharge
of a suspected case of mental deficiency. When a subject falls below a
conservative limit for tested ability a trained psychologist who is familiar with
the sources of error in giving tests, even without experience with the feeble-
minded, should be able to say that this person at present shows as deficient
development as the feeble-minded. To conclude however that any subject has
a passable mind requires in addition practical experience with feeble-minded
people who pass the tests. It is very much easier to state that the tests do not
detect all forms of feeble-mindedness than it is to give any adequate
description of the sort of feeble-mindedness which they do not as yet detect.
This distinction between the feeble-minded who do well with test scales and
those who do not, is well known in the institutions for the feeble-minded. Binet
sought to distinguish some of the feeble-minded who escaped the tests by
calling them “unstable,” or “ill-balanced,” individuals as Drummond (77)
translates the term. To use the historical distinctions of psychology, their minds
seem to be undeveloped more on their volitional and emotional sides than on
their intellectual side. Weidensall (59) has described another type as “inert.”
She found that quite a number of the reformatory women might slide through
the tests but fail socially from the fact that “their lives and minds are so
constituted that they feel no need to learn the things any child ought to know,
though they can and do learn when we teach them.” Again, it seems to be a
disturbance of will through the feeling, rather than an intellectual deficiency.
Many of the so-called “moral imbeciles” are probably able to pass intellectual
tests lasting but a few minutes. Like the unstable or inert they are not failures
because of a lack of intellectual understanding of right and wrong, but because
of excess or deficiency of their instinctive tendencies especially in the
emotional sphere. Such weakness of will may arise either from abnormality of
specific instinctive impulses or inability to organize these impulses so that one
impulse may be utilized to supplement or inhibit another. We may call all this
group of cases socially deficient because of a weakness in the volitional, or
conative, aspect of mind.
The discrimination of mental activities which are predominately emotional and
conative from those in which intellect is mainly emphasized is also well
recognized by those who have been making broad studies of tests in other
fields than that of feeble-mindedness. Hart and Spearman (123), for example,
call attention to the fact that tests passed under the stimulus of test conditions
represent what the subject does when keyed up to it rather than what he
would do under social conditions. We cannot be sure that speed ability as
tested will represent speed preferences. The subject may be able to work
rapidly for a few minutes, but in life consistently prefer to work deliberately.
Regarding the eighteen tests which they studied with normal and abnormal
adults they say: “These tests have been arranged so as to be confined to
purely intellectual factors. But in ordinary life, this simplicity is of rare
occurrence. For the most part, what we think and believe is dominated by what
we feel and want.” Kelley (130) finds by the regression equation that the factor
of effort amounts to two-thirds of the weight of that of the intellectual factor in
predicting scholarship from teachers' estimates. Webb (217) thinks that he
finds by tests a general conative factor comparable to Spearman's general
intellective factor.
With the change in point of view that has come from the adoption of the
biological conception of the mind the discrimination of the different forms of
feeble-mindedness must be recognized as a distinction in the emphasis on
intellectual, emotional and conative processes, not a distinction between
actually separable forms of mental activity. On account of the organic nature of
the mind it is well established that various mental processes are mutually
dependent. Any disturbance of the emotional processes will tend to affect the
thinking and vice versa. Even if we believe that emotions are complex facts,
involving vague sensations as well as feelings, and that terms like emotion,
memory, reasoning and will are names for classes of mental facts rather than
for mental powers, it still remains important to distinguish between feeling,
intellect and will, as well as to recognize the interdependence of the mental
processes. Common sense seems to agree with psychological descriptions in
regarding mind as a broader term than intellect, and feeble-mindedness as a
broader term than intellectual feebleness.
Since tests at present tend to reach the intellectual processes more surely than
the emotional, we describe those who fail in them as intellectually deficient.
The term “intellect” seems to be better than “intelligence” because the latter
seems to include information as well as capacity, while the aim of measuring
scales has been to eliminate the influence of increasing information with age.
To be thoroughly objective, of course, one should talk about “feebleness in
tested abilities;” but we would then fail to point out the important fact about
our present scales that they detect mainly intellectual deficiency, that they do
not reach those forms of feeble-mindedness in which the weakness in such
traits as stability, ambition, perseverance, self-control, etc., is not great enough
to interfere with the brief intellectual processes necessary for passing tests.
Intellectual deficiency will be used hereafter to refer to those social deficients
whose feebleness is disclosed by our present test scales.
In the opinion of Kuhlmann these cases of disturbed emotions and will which
shade off into different forms of insanity should not be classed as feeble-
minded at all, although he recognizes that they are commonly placed in this
group. He regards them as an intermediate class between the feeble-minded
and the insane. He says: “They readily fail in the social test for feeble-
mindedness and because of the absence of definite symptoms of insanity are
often classed as feeble-minded. In the opinion of the present writer they
should not be so classed, because they require a different kind of care and
treatment, and have a different kind of capacity for usefulness” (140). So long
as this group of what we shall term “conative cases” is discriminated from the
intellectually deficient it matters less whether they be regarded as a sub-group
of the feeble-minded or as a co-ordinate class. In grouping them with the
feeble-minded we have followed the customary classification. An estimate of
the size of this group will be considered later in Chapter III.

C. Doubtful Intellects Accompanied by Delinquency Presumed Deficient.

Conative forms of feeble-mindedness are perhaps the most serious types in the
field of delinquency. They are the troublesome portion of the borderland group
of deficient delinquents about which there is so much concern. It is important
to remember that it is just among these cases that the test judgment is least
certain. In this dilemma one principle seems to be sound enough
psychologically to be likely to meet with acceptance. I should state this
principle as follows: A borderline case which has also shown serious and
repeated delinquency should be classed as feeble-minded, the combination of
doubtful intellect and repeated delinquency making him socially unfit. This will
relieve the practical situation temporarily until tests are perfected which will
detect those whose feebleness is specialized in those phases of volition
centering around the instinctive passions, control, balance, interest and
endurance. The principle recognizes that mental weakness is sometimes
emphasized in the volitional processes of the mind.
The principle is apparently in conflict with the rule advocated by Dr. Wallin.
Referring to the mental levels reached by individuals, he says: “We cannot
consider X-, XI-, or XII-year-old criminals as feeble-minded because they
happen to be criminals and refuse to consider X-, XI-, and XII-year-old
housewives, farmers, laborers and merchants as feeble-minded simply because
they are law abiding and successful” (214, p. 707). At another place he insists
“that the rule must work both ways” (215, p. 74). Logically it would seem at
first that it was a poor rule which did not work both ways. Further
consideration will show, I believe, that there has been a confusion of feeble-
mindedness with tested deficiency. If all the feeble-minded tested deficient
intellectually then the tested level should determine whether or not they were
feeble-minded. This, however, is not a correct psychological description of the
facts. I prefer, therefore, to allow for those in a defined narrow range of weak
intellects to be classed as deficient provided their weakness also manifests
itself pronouncedly in the conative sphere.
The principle that all mental deficients need not show the same low degree of
intellectual ability is clearly recognized in perhaps the most important legal
enactment on deficiency which has been passed in recent years, the British
Mental Deficiency Act of 1913. It states regarding “moral imbeciles” that they
are persons “who from an early age display some permanent mental defect
coupled with strong vicious or criminal propensities on which punishment has
had little or no deterrent effect.” It specifically distinguishes them from the
group of feeble-minded which require guardianship because of inability to care
for themselves.

3. In Great Britain the term is restricted to those above the imbecile group.
CHAPTER III. THE PERCENTAGE DEFINITION OF
INTELLECTUAL DEFICIENCY

A. The Definition.

In order to direct attention to the quantitative description of intellectual


deficiency which is here proposed, let us state the percentage definition in its
most general form. Individuals whose mental development tests in the lowest
X per cent. of the population are PRESUMABLY INTELLECTUALLY DEFICIENT, unless their
deficiency is caused by removable handicaps. Above these is a group of Y per
cent. within which the diagnosis of intellectual deficiency is uncertain on the
basis of our present tests. The size of the presumably deficient X group is to be
determined by the number of intellectually weak which society is at present
justified in indefinitely isolating. The doubtfully deficient Y group should include
all those who are so intellectually deficient as to be expected to need
assistance indefinitely. The feeble-minded, or MENTALLY DEFICIENT, are those who
require social care indefinitely because of deficiency in mental development.
They include the X group, that portion of the doubtful Y group which is found
to require isolation, guardianship or social assistance, and any others not
detected by the tests but requiring prolonged social care on account of their
failure to develop mentally. Under the principle which we stated at the close of
the last section the combination of Y ability and persistent serious delinquency
brings the case within the group presumed to be feeble-minded.
Besides the greater definiteness and significance of such a definition of
intellectual deficiency, it affords the simplest practical criterion for determining
the borderline of passable intellects with a scale of mental tests. A detailed
comparison of the percentage plan with other forms of quantitative definition
will be found in Part Two. We may note here, however, that it guards against a
number of the absurdities of current descriptions of the borderline with
measuring scales. It is a criterion which may be consistently applied to the
borderline of both the immature and the mature. It may be adapted with
comparative ease to any system of tests. It aids in comparing the frequency of
intellectual deficiency among different groups, for example, among different
types of delinquents, regardless of whether the investigators have used the
same series of tests, provided only that each series has been standardized for
similar random groups.
Any form of quantitative definition, on the other hand, involves certain
assumptions which must be defended before it can claim to be of advantage
for practical purposes.

B. The Assumptions of a Quantitative Definition.

(a) Deficiency is a difference in degree not in kind.

Fortunately the tendency to describe the feeble-minded person as if he were a


different species from the normal has been definitely attacked by two
noteworthy researches, that of Norsworthy (159) and that of Pearson and
Jaederholm (164) (167). In these two investigations mentally deficient children
either in special classes or in institutions have been compared with groups of
normal children from the same localities on the basis of objective tests. The
results are uniformly supported by numerous other studies of deficient and
normal groups with the Binet and other tests. The conclusion is, therefore,
thoroughly established that there is no break in the continuity of mental ability.
It grades off gradually from average ability, and continually fewer and fewer
individuals are to be found at each lower degree of ability. The borderline of
deficiency will, therefore, not be a mental condition which clearly separates
different kinds of ability, but a limiting degree of capacity to be decided upon
by social policy in attempting to care for those who most need social
guardianship. Since ability changes gradually in degree it is necessary to
indicate a doubtful border region of degrees of ability on which expert
judgment must supplement the test diagnosis. Below the doubtful region the
diagnosis is clearly supported by objective test criteria, so that the only
question to raise is whether the condition is caused by removable handicaps.
The percentage definition thus strictly conforms to the best objective studies of
mental deficiency in treating deficiency as a difference in degree.
It should, perhaps, be said that this view is in direct conflict with the opinion
that mental deficiency is accounted for as a Mendelian simple unit character.
The opposing view has been advocated by Davenport (95, p. 310) and others
in the publications of the Eugenics Record Office, and accepted by Goddard
(112, p. 556). It has been so fully answered by Pearson (164) and Heron of
the Galton Laboratory (127) and by Thorndike (198) that there is no occasion
to take up the question in detail. We seem to be reaching an understanding so
far as our present problem is concerned. If the explanation of the inheritance
of mental ability is through Mendelian characters, nevertheless intellectual
ability is the result of such a complex combination of units that it may best be
thought of in connection with the unimodal distribution of ability adopted in
this study. No random measurement of mental ability has ever shown any
other form of distribution.
The attempt has also been made by Schmidt (179) to find qualitative
differences between normal and feeble-minded children by means of tests, and
by Louise and George Ordahl (162) to find qualitative differences between
levels of intelligence among feeble-minded children. While these studies are
very suggestive in pointing out the tests which most clearly indicate differences
between individuals, they seem to me to fall far short of showing that the
qualitative distinctions are anything more than larger quantitative distinctions.
It is not clear that the authors intended them to mean anything more than this,
so these studies do not seem to conflict seriously with our assumption that
intellectual ability grades off gradually and uninterruptedly from medium ability
to that of the lowest idiot.

(b) As to the variation in the frequency of deficiency at different ages.

A quantitative definition of intellectual deficiency would certainly be much


simpler if it could be assumed that the percentage of deficients at each age is
practically constant during the time when a diagnosis of deficiency is most
important, say from 5 to 25 years. Otherwise the objection might be raised
that it is impracticable to determine different percentages for each year of
immaturity or to formulate our borderlines of ability for a particular age. When
the general instinctive origin of intellectual deficiency is considered along with
the incurability of the condition, we seem to be theoretically justified in
assuming that the variation will be slight from one year of life to the next. This
assumption is tacitly made by all those who use Stern's quantitative description
of deficiency in terms of the mental quotient. On the other hand, there is a
feeling among some of the investigators that there is a sudden influx of feeble-
minded at particular ages and this position should be examined. Probably more
important than this possibility of increase is the question of a decrease in
frequency with age on account of the excessive death rate among the
deficients.
It is a natural supposition that there is a sudden increase in the proportion of
feeble-minded at adolescence. On account of the increased rate of growth at
this period we might expect to find greater instability for a few years. It may
well be that there is a rather sudden influx of the unstable type of feeble-
mindedness at this period. Such an increase may occur without being detected
by a series of brief intellectual tests such as the Binet scale. It would be of the
conative type of feeble-mindedness that cannot at present be diagnosed by
objective tests, the type that requires diagnosis by expert opinion. It is to be
noted, however, that Binet, who paid much attention to the unstable type,
says: “Since the ill-balanced are so numerous at ten years of age, and even at
eight, we conclude that in many cases the mental instability is not the result of
the perturbation which precedes puberty. This physiological explanation is not
of such general application as is sometimes supposed” (77, p. 18).
Only when an emotional disturbance is so great as to be detectable by mental
tests will this influx need to be taken into consideration in stating the
borderline for objective tests. The evidence that few cases of feeble-
mindedness are not detectable until after ten years of age is all the other way.
With the Stanford measuring scale, Terman and his co-workers did not even
find a noticeable increase in the variability of the groups at the ages of
adolescence (57, p. 555). It is to be remembered also that we are not
concerned here with mere instability which corrects itself with more maturity,
such as has been described by Bronner among delinquents. This does not, of
course, amount to an incurable conative deficiency and is not classified under
feeble-mindedness.
Goddard has suggested that possibly the moral imbecile group comes into our
class of feeble-minded suddenly with a common arrest of development at
about the stage reached by the nine-year-old. He notes that “of the twenty-
three cases of this sort picked out for us (at Vineland) by the head of the
school department, fifteen are in the nine-year-old group, five in the ten-year-
old, two in the eleven, and one in the twelve” (113). He regards this evidence,
however, as meager and only suggestive. Doll has given evidence of late
appearance of retardation in rare cases (100 and 99).
It is to be noted that if a sudden change is found in the percentage of children
falling below a certain test standard it is perhaps more likely to mean that
there is a change in the difficulty of the tests at that point. For example our
Table V shows 1.3% of the nine-year-olds test two or more years retarded,
while 18.9% of the ten-year-olds are retarded two years or more. This
presumably indicates a change in the relative difficulty of the tests for VII and
VIII rather than a change in the frequency of retardation at ages nine and ten.
When we turn to Goddard's norms for VII and VIII we find that 81% of the
seven-year-old children pass the norm for VII while only 56% of the eight-
year-old children pass the norm for VIII.
The Jaederholm data (167) obtained by applying the Binet tests to pupils in the
regular school classes and in special classes for the retarded may suggest a
possible influx of intellectual deficiency at about 12 years of age or else “more
mental stagnation in the intellectually defective” at this life-age and after. If
one were to define intellectual deficiency in terms of the standard deviation of
the regular school children, this data suggests that there is a marked increase
in the number of children sent to the special classes at 12 years of age who
are -4 S. D. or lower. Roughly speaking it amounts to 36 children at 12 years of
age, 36 at 13, and 21 at 14, as compared with 11 at 11 years and 13 at 10
years. On the other hand, this may as well mean that intellectual deficiency
becomes greater in degree rather than in frequency at these ages. The latter
interpretation is adopted by Pearson for the Jaederholm data, so that it is
perhaps not necessary to consider this evidence further. On the average the
pupils in the special classes fall about .3 S. D. months further behind regular
school children with each added year of life from 5 to 14 inclusive. A third
possible interpretation of the greater number showing the degree of deficiency
measured by -4 S. D. with the older ages should be mentioned. It is possible
that 1 S. D. has not the same significance for 5-year-olds as for 12-year-olds.
The distribution of abilities at succeeding ages may be progressively more and
more skewed in the direction of deficiency. We shall return to this point in Part
Two as showing the advantage of the percentage definition over a definition in
terms of the deviation. In connection with the Jaederholm data on special
classes one should also consider the fact that younger children are not as likely
to be detected by the teachers and sent to the special classes. It is possible
also that the difference in difficulty of the tests for different age groups is
somewhat obscured by using a year of excess or deficiency as a constant unit
as Pearson has in treating this data. The bearing of this difference in difficulty
was pointed out above for Goddard's data.
The investigations by Pearson of children in the regular school classes indicate
that there is no important shift with maturity in the frequency of those with
different degrees of ability, when the ability is measured either in terms of
years of excess or deficiency with the Jaederholm form of the Binet scale or in
terms of estimates of ability relative to children of the same age (166 and
167). In both these studies the correlation of ability with age was shown to be
almost zero. For tested ability for 261 school children “r” was .0105, P. E.
.0417; with the estimated ability, the correlation ratios were for 2389 boys,
.054, P. E. .014; for 2249 girls, .081, P. E. .014. Until we have better data this
is certainly the most authoritative quantitative answer to the question of the
shift with age in the frequency of the same relative degree of mental capacity.
The best method of empirically settling this question of the early appearance
and constancy of deficiency would be to test the same group of children again
after they had reached maturity and find out how many of those who tested in
the lowest X per cent. still remained in the same relative position. This is, of
course, not possible at present, but it certainly should be done before we are
dogmatic as to the permanent isolation of the lowest X percentage at any age.
The nearest approach to this sort of evidence is Goddard's three annual
testings of a group of 346 feeble-minded children with the Binet scale (117, p.
121-131). Among these 109 showed no variation, 123 gained or lost 0.1 or 0.2
year, 18 lost 0.3 or more, and only 96 gained 0.3 or more of a year. With so
small a change in absolute tested ability the probability of a change in position
relative to normal children seems to be slight. Only one of the 76 who had
tested in the idiot group gained as much as a half year in tested age in three
years.
It is not possible to settle this question of the constancy of the percentage of
intellectual deficiency from one life-age to the next by considering the
frequency of different ages of children among those who are sent to special
classes for retarded pupils. This is evident from the fact that these classes
contain a considerable proportion of those who are feeble mentally mainly
because of conative disturbances. These would not be detected by our present
tests and would not be classed as intellectually deficient. In the second place
the pupils for the special classes are usually selected mainly on the advice of
their teachers, who cannot, of course, without tests select those who are
intellectually deficient except by trying them for a number of years in the
regular school classes. This means that a smaller percentage of pupils in the
special classes at the younger ages is to be expected.
The figures of the U.S. Census as to the ages of inmates of the institutions for
feeble-minded are also of little significance in connection with the question of
the variation from age to age. That the number of inmates at the different
ages is affected most largely by the pressure of necessity for shifting the care
from their homes to the institution is shown by the fact that three-fourths of
the admissions are of persons over 10 years of age. It is also indicated by the
fact that for the period from 15 to 19 the males are over 20% more frequent
than females, while from 30-34 the females are nearly 20% more frequent.
Considering those ages most frequently represented in the institutions, 10-24
years, the average variation for the three five-year periods in the percentage of
the population of the corresponding ages who are in these institutions is only
0.01%. The middle five-year period has the most, but even if there were a
cumulation of feeble-mindedness with age, which is not shown, we would
anticipate a change of not more than 0.05% for these 15 years. This would be
clearly negligible in considering the general problem.
That little allowance for the variation from age to age need be made for the
number of cases not discoverable at the beginning of school life is further
indicated by report of the Minnesota State School for Feeble-Minded. It shows
that in only 247 out of its 3040 admissions was the mental deficiency known to
commence after six years of age (154). If the number of feeble-minded who
should be isolated were found to increase after school age less than one in
10,000 of the population, as this suggests, it would surely be better to neglect
this variation from age to age than to emphasize it in dealing with the problem
of objective diagnosis and social welfare.
How rare is the onset of feeble-mindedness after five years of age is also
shown by the frequency of hereditary causes. In his study of the 300 families
represented at Vineland, Goddard places only 19% in his “accidental” group
and 2.6% in the group for which the causes are unassigned. The rest are
either in the hereditary group, probably hereditary, or with neurotic heredity.
Half of the cases in the “accidental” group are due to meningitis. His histories
show that only 9 of the “accidental” and unassigned groups were unknown at 5
years of age. This is only 3% of his total feeble-minded group. To these might
be added, perhaps, a few from the hereditary groups who did not show their
feeble-mindedness at so early an age, but so far as I can judge these would
not be of the intellectually deficient type that would be detectable by the Binet
scale at any age. They would test high enough intellectually to pass socially
and require expert diagnosis to be classed as feeble-minded.
Certain diseases, epilepsy and meningitis, are undoubtedly causes of feeble-
mindedness. The evidence, however, seems to be that they are so rare
compared with the mass of mental deficiency that after 5 years they may well
be offset by the excessive death rate among the feeble-minded. That
recoveries from feeble-mindedness are insignificant is generally agreed. Among
the 20,000 in institutions in 1910 only 55 were returned to the custody of
themselves. This is further evidence of the fundamental, if not congenital,
nature of the deficiency.
While the evidence submitted above makes it seem fair to assume that the
increase in the frequency of a certain degree of intellectual deficiency with age
is probably negligible, it is not clear that the decrease with age in the
proportion of feeble-minded caused by an excessive death rate may be
neglected even for the test ages 5 to 25. By searching the literature it has been
possible to assemble the records for nearly 3500 deaths among the feeble-
minded in institutions in this country and Great Britain distributed by ages in
ten-year periods. This evidence is presented in Table I. The number of cases
under five years of age living in the institutions is so small that the deaths
under five years are certainly misleading. They have, therefore, been omitted
from the table and the distribution calculated for those five years or over (123,
154, 204, 205). Comparison is made with a similar distribution of the total
deaths for a period of five years from 1901 to 1904, inclusive, within the area
of the United States in which deaths are registered, compiled from the special
mortality report of the Bureau of the Census (206). This registration area has a
population of about 32,000,000. The general agreement of the distribution of
deaths among the four different groups of institutional inmates seems to make
it reasonable to assume that the United States group of institutional deaths for
the year 1910 is a conservative description of excessive death frequency at the
early ages among the feeble-minded in institutions.

Table I. Age Distribution of Deaths in the General Population and Among


Feeble-Minded in Institutions.

Population Ages
5-14 15- 25- 35- 45- 55 &
24 34 44 54 over
Gen'l—U. S. 1,897,492 6.1% 9.6% 12.8% 13.0% 13.6% 44.9%
in death
registration
area
F. M. 1910 840 26.6 33.0 18.9 9.1 45 &
in over
Institut'ns 12.3
in U. S.
F. M. British 997 34.3 41.1 10.4 6.5 3.5 55 &
(Earlswood) over
4.2
F. M. British 613 34.7 46.8 9.5 35 &
(Barr) over
9.0
F. M. 982 27.6 38.0 16.1 8.6 3.5 55 &
Faribault over
Minnesota 6.2

Table II. Mortality of Institutional Deficients in the United States Compared with
the General Population, Showing its Possible Effect on the Frequency of
Deficiency at Different Ages.

Ages
5 10 15 20 25 30 35 40
General population 1000 983 972 956 934 903 872 835
Deficients in Institut'ns 1000 795 696 606 503 428 349 290

Per cent. deficient if 1% at 1.40 1.11 1.00 .75


age 15
Fig. 1. Mortality among Feeble-Minded in
Institutions Compared With the General
Population

A comparison of the death rates of the feeble-minded and the general


population at different ages is of prime importance in connection with all
attempts at quantitative descriptions of deficiency. Heretofore this has been
completely neglected. Fig. 1 and Table II have been prepared to provide a
roughly adequate estimate, on the basis of the above data for the United
States, as to the survival of 1000 institutional cases of feeble-minded 5 years
of age for successive age periods compared with 1000 people in the general
population. In constructing this table it was necessary to assume, since the
facts were not given, that the age distribution in the registration area of the
general population was the same as for the United States as a whole (census
of 1910) and that the number of feeble-minded in the institutions at the
various age periods was equal to the number enumerated on the first of
January plus the admissions during the year 1910, disregarding the number
discharged since they are not distributed by ages. The average annual death
rate among the institutional cases of feeble-minded 5 years of age and over in
the United States in 1910 was 35.19 per thousand, while the corresponding
death rate in the general population of the registration area for the five years
1901-1904 inclusive was 13.56. Assuming that the death rates are uniform
within the five-year periods, the decline in the proportion of institutional feeble-
minded from 5-25 years of age as the result of excessive mortality is indicated
by the last line in Table II, after allowing for the mortality in the general
population. That this effect of excessive mortality upon the percentage of
feeble-minded cannot be neglected between 5 and 25 years of age is apparent
unless the mortality among institutional cases is much greater than it is among
the deficient generally. As the figures stand the proportion of feeble-minded
would be reduced nearly one-half between ages 5 and 25. Only a small part of
this reduction probably would be compensated for by new cases developing
from accident or disease. On the other hand there is little doubt that the
institutions contain an excessive proportion of low grade cases among whom
the mortality is much greater. The mortality among institutional cases is,
therefore, probably not typical of that among the feeble-minded generally.
Nevertheless it is so great that any quantitative definition of deficiency which
neglects it entirely is open to serious objection. We shall, therefore, keep this
variation in mind in connection with the discussion in the next chapter of the
percentage which is deficient, and in the adaptation of the definition to a
measuring scale. It is clear that the percentage should be so chosen as to
allow best for the possible large effect of excessive mortality among the
deficients. Finally, it should be said that the percentage definition of feeble-
mindedness might be modified to meet a varying percentage from age to age
should that ever become desirable.

(c) As to the number of deficients not detected by tests.

If most of the feeble-minded for whom society should provide were of the type
which is only conative and not detectable by our present objective tests, a
quantitative definition would be abortive. We must, therefore, study our
assumption that it is worth while to direct our attention to those who are
intellectually deficient. We shall attempt to discover how frequent are the
primarily conative types.
Before examining the quantitative evidence we may note that it is in
conformity with two prominent recent tendencies in psychology to subordinate
specialized abilities, as compared with abilities which function commonly in
many situations. The first of these tendencies is represented by the
fundamental researches of Hart and Spearman (123) (185). This is not the
place to set forth the technical work on which their conclusions are based. It
may be said, however, that, with 17 different psychological tests, they were
unable to discover any important specific mental weakness which distinguished
adults who were suffering with any one of various mental abnormalities,
including imbecility, manic-depressive insanity, dementia praecox, paranoia,
and general paralysis of the insane. This may have been the fault of the tests,
but it seems to be more likely that the fault lies in the custom of emphasizing
special abilities and disabilities, at least from the point of view of tested
capacities. On the other hand, all of these mental abnormalities showed a
weakness in general intellectual ability. This is true whether this general ability
be regarded, as it is by Hart and Spearman, as due to a general fund of brain
energy, or whether general ability be taken to refer to the common recurrence
of many specific abilities in much of our mental life. Its significance for this
study is that a series of varied tests, such as that of Binet, may be expected to
give a good estimate of general ability, and its failure to disclose specific
disabilities is thus less important.
The second influence in psychology tending to emphasize average tested ability
is the establishment of the biological conception of the mind which recognizes
the mutual interdependence of the mental processes, organically united
through the activity of the brain. So long as intellectual, emotional and
volitional processes are all mutually dependent, a disturbance of one aspect of
mental life is bound to affect the others. In considering the mutual dependence
of the mental processes, it is important to weigh carefully the striking examples
which Bronner[4] has brought together, illustrating special abilities and
disabilities. She has made an admirable start toward a differential diagnosis of
special defects in number work, language ability and other mental activities.
The degree of special deficiency which results in social failure could be placed
upon an objective basis, but the rarity of special deficiencies as compared with
general deficiency will make this a slow task. In the meantime we may rely
upon the mutual dependence of the organic processes as a point of view which
emphasizes the common spread of deficiency to many activities. Knowledge of
a single case of specific disability is sufficient to make us recognize that such
cases do occur. On account of the rarity of those cases and the absence of
objective criteria, it seems necessary to leave the further differentiation to the
future, considering here only those cases which may be grouped together as
conative, as contrasted with those detected by our general intellectual tests.
Whether the group of primarily conative cases is of any considerable size can
be only very roughly estimated at present, since the diagnosis of such cases of
feeble-mindedness rests at present almost exclusively on the subjective opinion
of the examiner. Before their diagnosis is put upon an objective basis we must
have a different form of test directed at such traits of will as initiative,
perseverance, stability and self-control. These probably center on the mental
side around the instinctive emotional background of interest and the passions,
while, on the physical side, they raise the question whether the subject's
energy is adequate to endure the strain of competition or whether it shows
itself only in sudden bursts.
If the diagnosis of conative cases could be determined objectively, it is possible
that most forms of social unfitness would be found highly correlated with
intellectual deficiency. On the other hand, when the diagnosis of unfitness for
school or social life depends merely upon the opinion of experts or teachers,
the inaccuracy of the diagnosis may show a wide discrepancy between the so-
called conative and intellectual types of deficiency. Binet, on the basis of his
acquaintance with the pupils in special classes, suggested that the number of
unstable children is probably equal to the number of those who are
intellectually unsuited for the ordinary schools or institutions (77). Since he
then places the total number of the two classes at four or five per cent., it is
apparent that he is discussing a higher type of ability than is usually included
under the term feeble-minded. We can get somewhat better evidence on this
question by studying the results of Binet tests applied to children cared for in
special classes or in institutions for the feeble-minded. Chotzen (90) presents a
table of 280 children in the Hilfsschule in Breslau, only 201 of whom, however,
he himself diagnosed as feeble-minded, i. e., debile or lower. Of these only 51
were intellectually deficient as indicated by the Binet tests when we include the
doubtful cases according to the criteria we have adopted in this study. If we
suppose that, in addition to those in the special classes, there would be one
intellectually deficient child in an institution for feeble-minded for every child
testing deficient, we would then guess that only 40% of the feeble-minded
children in Breslau were intellectually deficient. This sort of estimate seems to
agree with Binet's belief that half of the children requiring special care, at least
during school ages, are cases which are primarily conative.
Pearson has approached the same problem in another way (164) (167). He has
used the results of the psychological tests applied by Norsworthy to children in
New York in special classes and institutions for feeble-minded compared with
those in the regular school classes, and the results of Jaederholm obtained
with the Binet tests applied to 301 children in Stockholm in the special classes
compared with 261 others selected from the regular classes. He found that
“70.5% of normal children fall into the range of intelligence of the so-called
mentally defective; and 60.5% of so-called mentally defective children have an
intelligence comparable with that of some normal children” (167, p. 23). On
the statistical assumption that those in the normal classes would distribute
according to the Gaussian normal probability curve he estimates that, with the
Binet tests, among those in the special classes “10% to 20%, or those from 4
to 4.5 years and beyond of mental defect, could not be matched at all from
27,000 children” (164, p. 46). Another 20 to 30% could be intellectually
matched by those in the regular classes having from 3 to 4.5 years of mental
deficiency, but they would be matched very rarely. On the assumption that 1%
of the children were feeble-minded, not more than about two children in a
thousand of this regular school population would be expected to be 3 or more
years retarded and thus overlap those of like deficiency in the special classes
(167, p. 30). Considering the results of Norsworthy's study he says on similar
assumptions: “It seems, therefore, that a carefully planned psychological test,
while not sufficing to differentiate 50 to 60% of the mentally defective from the
normal child, would suffice to differentiate 40 to 50%” (164, p. 35). Again we
come back to the estimate that psychological tests may well be expected to
select nearly half of the children at present found in special classes for retarded
pupils. Moreover, a considerable part of the overlapping of intellectual
deficiency in the regular classes with that in the special classes which he found
may be accounted for by the inadequate methods of selection of pupils for the
special classes by teachers or examiners who have used no objective tests.
Some who were left in the regular classes should undoubtedly have been
transferred to special classes and vice versa. There seems to be nothing to
indicate that less than half of those properly sent to special classes would be of
clear or doubtful intellectual deficiency. If the tests served to select even a
smaller proportion of those assigned to special instruction, the “school
inefficients” as Pearson calls them, their value as an aid to diagnosis would be
demonstrated.
Among groups of delinquents, where we would expect the purely conative
cases to be more common, we find that a careful diagnosis of feeble-
mindedness on the basis of test data, medical examination and case history
indicates that conative cases without serious intellectual deficiency are much
rarer than intellectually deficient delinquents. At least this is the evidence of
one study where such information is available. Kohs at the Chicago House of
Correction found among 219 cases over 16 years of age, which he diagnosed
as feeble-minded, only 28 tested XI and there were only 52 who did not test
either presumably deficient or uncertain intellectually according to our criterion.
Another bit of evidence is that collected at the Clearing House for Mental
Defectives in connection with the New York Post-Graduate School of Medicine,
where 200 consecutive cases (108 males) were examined by Miss Hinckley. Her
graphs show that only 15% tested X or above with the Binet revised scale, i.
e., above those presumably deficient in intellect. The cases were from 13 to 42
years of age. The clearing house provides an opportunity for social workers to
have suspected deficients examined and the few cases over X seems to
indicate that the purely conative type is not very commonly met with among
the social workers.
When we turn to the institutions for the feeble-minded we find that they are
today caring for few solely conative cases. Although I can find no tables which
give both the life ages and mental ages of the individual inmates, we can at
least be sure that few test so high as X, or above with the Binet scale. This
means that only a few have as yet reached the threshold for passable adult
intellects, which should be attained by 15 years of age. At the Minnesota state
institution for the feeble-minded in Faribault among 1266 inmates, excluding
epileptics, 41 tested X; 28, XI; 12, XII; and 8, XIII, a total of 7% (154). At
Vineland, N. J., Goddard reported among 382 inmates, 14 tested X; 5, XI; and
7, XII, about 7%. Some of the children who were under 15 in life-age might
later develop above the limit for intellectual deficiency. Of the 1266 at the
Minnesota institution, however, 508 were 15 or over at the time of their
admission, so that at least 82% of the 508 were clearly intellectually deficient.
Eight per cent. more tested X and were in the doubtful group in intellectual
ability according to the criteria we have adopted. This suggests that not more
than about 10% of those who are at present isolated in institutions are there
for feebleness of will alone. It seems to confirm our presumption that the
intellectually deficient discovered by tests form the great majority of the social
deficients who need prolonged care or assistance.

(d) Allowance may be made for variability.

The quantitative definition of intellectual deficiency must be made with careful


allowance for irregularities among different mental processes, among different
individuals, and among different groups. Theoretically it is possible to place the
borderline so low that a case with that degree of deficiency and without
removable handicaps would be clearly feeble-minded. The chance that the
diagnosis would be mistaken could be reduced to any minimum desired. Above
this a wider region of doubtful deficiency could then be stated in similar form.
This is the plan that we suggest in attempting the percentage definition.
Practically, however, the plan assumes that a suitable allowance can actually be
made for these variations and raises a number of problems as to variability
which should be considered. Four of these sources of variation are discussed
below: (1) the variation due to a limited sample of individuals measured, (2)
the variation among different communities, (3) the variations arising from sex,
race and social differences, (4) the variation of the same individual from one
mental process to another. We do not have the problem of neglecting these
variations, but of adequately allowing for them both in the percentage of
presumably deficient and in the doubtful region.
(1) Variation among Samples of Individuals Measured. The error introduced by
the fact that measurements are made on a limited rather than an unlimited
number of individuals, in establishing the standards with a system of tests, can
be taken care of statistically fairly well by applying the theory of probability as
to the error of a percentage in a single sample. The range of the error can
then be indicated on the measurement scale. This supposes, however, that
each sample to be measured is taken from a random group and not from a
selected group. Allowance for this error of sampling is therefore complicated by
the fact that the usual test data have been obtained from groups of school
children, even when there has been no further selection within the school
group. Data on school children are certainly reliable only within the years of
compulsory school attendance. Ordinarily in this country, they are not reliable
for children of 14 years of age or over. Moreover, the point of the scale which is
reached by the lowest X percentage of school pupils will exclude a slightly
larger percentage of all children of corresponding ages, since the idiots and
some imbeciles are not sent to the ordinary schools. This slight discrepancy
should be kept in mind. The problem of avoiding selected samples among

You might also like