Wanted: Researchers and Authors

1. Violence Prediction Scale

The first available project consists of standardizing the Violence Prediction Scale (VPS) constructed by Harold V. Hall, PhD, ABPP, and the late David Pritchard, PhD, ABPP. The standardization population must reflect a cross-section of different races, cultures and ethnic groups in order to produce a culturally sensitive violence prediction instrument.

The following contains explicit instructions for the cross-validation of the VPS.

Participants will respond to the VPS, which assesses their history of violence, triggers to violence, opportunity factors, such as the availability of weapons, and inhibitions to inflict harm on others. Deception scales are built into the VPS to correct for under-reporting and/or over-reporting of violence-related factors. The deception analysis, along with an extensive evaluation of previous aggression for each individual, is a unique and powerful feature of the VPS.

The Pacific Institute has taken preliminary steps toward creating and standardizing the VPS. Responses have been collected from approximately 850 adult subjects of different ethnic and cultural minorities from a variety of colleges and agencies in Hawaii and on the mainland. The item pool for the VPS was systematically reduced from 400 to 300 items. The pre-validation phase of the project involved the testing of several hundred more subjects, again utilizing large numbers of subjects from diverse ethnic and racial groups. An internally consistent and reliable preliminary measure of violence potential was created. Along with age and gender, cultural differences were consistently obtained.

Statistical analysis revealed that the majority of subjects reported previous consummated or threatened violence resulting in varying degrees of harm, suggesting that the VPS was indeed sensitive to historical aggression. Two multiple regression analyses of the scales predicted reported past violence. The first analysis was a step-wise regression analysis and the second was a hierarchical regression analysis. Both analyses yielded the same result: Excluding subjects who had an infrequency score greater than 5, a linear combination of the trait anger scale, the alcohol abuse scale, and sex predicted reported past violence with a multiple correlation of .64, a most promising result.

In general, final validation of the VPS will require (1) the administration of the scale to a large sample of subjects, at least some of whom are considered to be at high risk for committing violent acts in the future; (2) follow-up of these subjects for at least one year to determine who among them actually commits a new violent act; and (3) analysis of the data.

Each of these three parts of the study involves several related issues:

Part 1: It is important that the sample contain persons who are at high risk for future violent behavior. In general, such persons may be found among groups of individuals who have committed violent acts in the recent past (for example, spouse and child abusers, persons reported by significant/knowledgeable others to have committed violent acts which did not result in arrest, persons recently convicted of a violent crime but not sentenced to prison). The proportion of such at-risk persons in the total sample studied should be no less than 40%, as a smaller percentage of the total sample would likely result in too few violent acts at the time of follow-up. The remaining 60% of the sample can be drawn from groups which represent the population to which the VPS is intended to be administered in the future. For example, if the VPS is being validated for use with spouse and child abusers, the sample should consist of persons drawn from a population with no fewer than 40% of the sample having a history of violent behavior associated with spouse and child abuse. Similarly, if the VPS is being validated for use with adolescents, the sample should be representative of this group, with no fewer than 40% having a history of violent behavior. Because many populations will be tested by VPS investigators, the measure will have a wide utility within most ethnic groups and both genders.

One of the critical questions for each investigator in selecting the appropriate sample is the definition of past violence. Past arrests or convictions for violent acts could be used to assure that no less than 40% of the sample has a history of violence. However, other indications of prior violent behavior may also be used (for example, treatment at hospital emergency rooms for aggression-related injuries, statements from significant others). Finding the best definition of past violence will depend on what records or referral agencies are available.

Once an appropriate sample has been selected, care must be taken to assure that the subjects understand the items in the VPS. Poor reading skills, mental incapacity, or lack of cooperation among subjects will reduce the reliability of the VPS and thus reduce its validity. The preliminary version of the VPS is designed for a minimal reading level of grade 6.6, which is acceptable for must settings.

After the selected subjects have been screened for their ability to respond to the VPS, the test may be administered in one sitting, to one subject at a time or in small groups. The VPS booklet contains the actual 292 questions to be asked of each individual, with responses to be marked on the answer sheet. All VPS protocols will be scored and tabulated for analysis.

Part 2. All of the responding subjects will be monitored to determine whether or not they engaged in any violent behavior after completing the VPS. At a minimum, follow-up should occur one year after the administration of the VPS. However, if time and resources permit, the study may also include additional follow-up periods of three months, six months and nine months following the administration of the VPS.

At each follow-up point, each subject should be evaluated to determine the frequency and severity of any violent behavior which occurred after the administration of the VPS. This evaluation may involve an individual interview with each subject and/or checking available records for "markers" of violent behavior (such as arrests, convictions, hospitalizations).

The choice of an outcome measure will be determined by resources and the availability of records. However, once an outcome measure has been selected, it should be used consistently for all subjects and for all follow-up periods. The data cannot be analyzed if the outcome measure changes from subject-to-subject or from period-to-period.

For confidentiality purposes, each subject's outcome must be coded so that it can be linked to that person's VPS. With each individuals' consent, the subject's full name or some identifying number can be used so that the results of each period's outcome can be correlated with the results of the VPS administration.

Part 3. The analysis of the data will depend on the number of subjects for whom VPS and outcome data are available and on the type of outcome data collected. If the number of subjects is small (e.g., less than 100), then simple correlational methods will be used; if the number of subjects if large, then multivariate methods will be used. If the outcome data is ordinal in nature, multivariate methods could be used; if the outcome data is categorical, cross tabulation methods must be used.

Interested persons should send a letter of interest, proposed research plan and vita to the Pacific Institute. The booklet and answer sheets will be sent to approved researchers and/or organizations after a contract has been signed by relevant parties.

2. Book Publication

The Pacific Institute is seeking authors who have violence-related findings from meta-analysis or from a large normative base. Works which address mathematical models or decision path analysis will also be considered. The Pacific Institute pays 25% of royalties from net profits or 50% of assigned royalties if the book is published through a publishing house.

To reach an informed publishing decision about a proposed book, the Pacific Institute needs certain essential information. The following guidelines describe the type, order, and approximate amount of information we need to properly evaluate your proposal.

1. Title and Authors/Editors: The title may change before the book is published, but we need one by which to refer to your book. Will you have any co-authors (for a monograph) or co-editors (for an edited book)? If so, provide their names and affiliations. There is a maximum of three authors/editors for a single monograph or book.

2. Prospectus: Summarize, in one or two paragraphs, what your book is about and what approach you will take in presenting the information. Describe the need for such a book and its intended audience.

3. Special Features: Describe any special features of your book that would be of interest to readers. For example: "Dr. John Monahan, an international authority on risk analysis and violence prediction, has agreed to write a foreword for the book" or "The book will include easy-to-follow flow diagrams of forensic protocols" or "Each chapter will provide not only a summary but also an analysis of recent observations obtained by computer modeling."

4. Competition: List competing books (author/editor, title, publisher, year) currently on the market if you are aware of any. How would your book differ from or improve upon each of them?

5. Preparing the Manuscript: Using American Psychological Association style, approximately how many double-spaced, typed pages will your manuscript be (including references)? Estimate the number and type of illustrations (20 line drawings, 30 photographs, 7 tables, etc.). How long will it take you to complete the manuscript once you have a contract? There is a one-year limitation on completing your book. Under extenuating circumstances, a contract may be extended an additional year.

6. Table of Contents (tentative): Provide a detailed outline of chapters, with suggested author(s) for each (if the book is an edited treatise). Under each chapter heading, summarize the topic in a sentence or two, noting any special considerations (such as "This chapter will contain numerous photo-micrographs"). Use no more than five pages to outline your book.

7. Curriculum Vitae: Each author/editor should enclose a current curriculum vitae with the proposal.
   

Copyright © 2005, Pacific Institute for the Study of Conflict and Aggression, Kamuela, Hawaii  96743.
Email Contact