authority and a grid for comparing instruments within the domain. The review provides a discussion of the state of research and practice within the particular stage of the treatment process. It offers guidance on the clinical utility of particular instruments for assessing the domain and identifies specific issues on which additional research is especially needed.

The grid facilitates comparison of measures appropriate for use within that domain of treatment assessment. Administrative characteristics of each instrument are noted, including elements such as populations for which it might be particularly appropriate, time to administer and score, and availability of computerized formats. This grid contains information on all instruments in the Guide that have been identified as potentially appropriate for use in that particular stage of the treatment assessment process.

Instruments

The last section of the Guide presents copies of the instruments arranged alphabetically, each immediately preceded by a fact sheet that synopsizes administration, scoring, and interpretation and notes copyright status and how to obtain a copy of the scale. Although most details in the fact sheets were obtained directly from the instrument's author or an expert on the measure, minor editing was done by the panel members to assure consistency in tone and format across scales as well as to elaborate on items not fully addressed by the instrument's proponent. In a few instances, the expert independently prepared the fact sheet.

The opinions expressed in the fact sheets are intended to faithfully represent views of the instrument authors, but neither the

National Institute on Alcohol Abuse and Alcoholism (NIAAA) nor members of the panel certify accuracy of the data provided. Details on the fact sheets should be considered in conjunction with information obtained from original sources to determine the suitability of an instrument for a particular task.

The instruments are reproduced in their entirety when possible, but length and copyright concerns prohibited full reproduction of some. In most cases, sample items are provided when the full instrument is not available in order to provide a "flavor” of instrument tone and format. Users are reminded to secure the permission of the authors or copyright holders before using any instrument.

Online Availability

Information on instruments obtained from this volume is expected to be available in the online data base of Health and Psychosocial Instruments (HAPI). This data base can be accessed through Ovid Technologies or by calling Evelyn Perloff, Ph.D., the data base administrator, at 1-800-950-2035.

Updating the Guide

NIAAA hopes to periodically update the Guide. Users' assistance in identifying new instruments is urged. Information on candidate measures can be provided to:

Treatment Research Branch National Institute on Alcohol Abuse and Alcoholism

TREATMENT ASSESSMENT

INSTRUMENTS

6000 Executive Boulevard, Suite 505 Rockville, MD 20892-7003

Accurate and comprehensive assessment

is fundamental to both treatment of and research on alcohol problems. Although each of these activities is advanced by informed use of psychometric instruments, needs of professionals in the two endeavors differ. Most notably, the practitioner is primarily concerned with the clinical utility of the measure, particularly how well it identifies the needs of a given client and guides treatment planning. The researcher is likely to explore a broader range of variables that may quantify and explain the overall impact of an intervention (Connors et al. 1994). These variables may or may not be directly related to client care.

Psychometric properties of measures, especially validity and availability of relevant norms, are of considerable interest to the clinician. While such statistical information is not irrelevant to researchers, often it is less critical. In a treatment study, contrasts usually are between a control group and an experimental group or before and after treatment functioning among a given group of subjects. Since scores derived from measures with lower validity include a large component of error variance, their use may necessitate recruitment of larger numbers of subjects or inclusion of additional scales to in some way correct for measurement error. External norms may be of lesser immediate concern to the researcher.

Particularly in the current environment

of stringent controls on health care costs and service utilization, the clinician also is deeply concerned about issues such as ease of administration, scoring, and interpretation of the instrument as well as cost, time, and acceptability of the measure to clients (Allen et al. 1992). In research projects, however, subjects typically are reimbursed for their participation, and sufficient technical resources are usually available for administering measures and quantifying results.

Researchers seem to place a much higher premium on formal assessment than do many practicing clinicians, who appear to rely more heavily on interviews, review of past records (Nirenberg and Maisto 1990), or clinical impression. While such procedures can provide helpful information, psychometric techniques offer unique and very important advantages. Their standardization permits uniformity in administration and scoring across interviewers with diverse experience, training, and treatment philosophy. The measurement properties of formal assessment procedures, including their strengths and weaknesses, are known.

The large number and variety of formal techniques also allow such measures to respond to a broad range of client management questions. To their credit as well, formal measures are economical in terms of cost, clinician time, and effort required to succinctly and clearly communicate with other clinical staff treating the client.

Finally, results thus derived may well have more credibility, and thus influence, with clients than conclusions based on less formal procedures (Allen 1991).

Failure to fully appreciate and employ formal, validated assessment procedures is regrettable in the field of alcohol treatment practice. We continue to believe that:

While better assessment of alcoholic patients does not ensure more specific or more effective treatment, chances for successful rehabilitation are clearly enhanced if specific patient needs can be more accurately identified and if treatment can be tailored accordingly. (Allen 1991, p. 183)

As a greater variety of interventions are introduced into the alcoholism treatment system and as we more fully appreciate the treatment implications of differences among subtypes of alcoholics, the role of assessment in clinical practice will greatly expand. We hope that the current volume will enrich the contribution of assessment to alcoholism treatment both by apprising clinicians of the wide array of instruments available and by assisting them to make well-informed decisions about which are most helpful for serving their clients.

Although our primary audience is clinical practitioners, we hope that the volume will prove valuable to the alcoholism research community as well. In choosing instruments and developing the format for this text, we have tried to keep the needs of both readership groups in mind.

Elements in Instrument Selection

When choosing an instrument to help determine a client's treatment needs, the primary concern is: Is the instrument appropriate for the client? Several parameters should be considered in answering this question.

Purpose/Clinical Utility

In this volume, instruments are assigned to chapters according to their primary role in informing sequential decisions that direct the course of treatment (e.g., screening, diagnosis, triage, treatment planning). While some of these stages, such as screening and diagnosis, are narrowly defined, measures that assist in treatment planning or that assess the treatment process may answer questions very different from those resolved by other scales within the same domain. The purpose and prior use of an instrument should be consistent with its intended use.

Assessment Timeframe

Measures differ according to the period of client functioning that they encompass. For example, certain measures and tests are appropriate when the concern is recent drinking patterns, whereas others reflect long-term, chronic alcohol use. Similarly, screening and diagnostic scales are designed to evaluate either lifetime or current conditions.

Age or Target Populations

In choosing an instrument, it is important to consider its suitability for a given client. Most alcohol measures have been developed for adult populations. Of late, however, several useful adolescent scales have been constructed. This advance in the field is clearly welcome, since alcohol problems in adolescents often are manifested differently and lead to dissimilar consequences than in adults. Attention of test developers has recently been focused on needs of more specific subgroups, such as screening for alcohol problems among pregnant women and among the aged.

Examples of Groups Used

The field of alcohol assessment has emphasized development of a wide variety of instruments, to some extent in lieu of

efforts to refine existing instruments and to determine their particular applicability to subpopulations of individuals with alcohol problems. When choosing an instrument, it is helpful to know what groups have successfully been evaluated with the instrument.

Norms Available

Norms allow the test performance of a given client to be compared with that of a large, relevant group of individuals. While norms are essential to describe a single case or a sample by comparison to a larger group, they are less important, for example, in contrasting pretreatment and posttreatment behavior in an individual.

In other instances, too, norms are not of key concern. For example, screening measures are judged primarily on their ability to predict diagnosis irrespective of how an index case compares with others on the scale. In short, while some measures are interpreted normatively, others are interpreted ipsatively. In ipsative analysis, individuals are actually compared with themselves, such as their functioning before and after treatment or the relative strengths of various expectancies that the individual maintains for effects of drinking. Although normative instruments may often be interpreted in an ipsative manner, the converse is rarely true.

In determining the importance of normative information, the clinician should be concerned about whether norms are available that would assist in making clinical decisions in a particular case. Phrased differently, would the demographic characteristics of a client affect interpretation of the score and influence the choice of treatment?

As with other personality measures (Sackett and Wilk 1994), few scales in the alcohol field have ethnicity-based norms. Separate norms for males and females, however, are available for some alcohol measures. In that problem drinking and alcohol dependence are exhibited somewhat differently in men

and women, gender-based norming of measures for screening, alcohol use, and adverse consequences of drinking is generally desirable. It remains to be seen, however, if gender-based norming would significantly augment the utility of most treatment planning measures, which are often ipsative in nature. The more challenging issue may be whether or not the fundamental dimensions differ so greatly that different measures, rather than separate norms, are needed for various subgroups. Research on this topic remains in an early stage.

Administrative Options

An active area of investigation in instrument development has been alternative ways of administering the measure. These include written, interview, computer, and collateral inquiry formats. Alternative administration procedures may decrease clinician time, more effectively engage clients in the assessment process, and heighten accuracy of responding. Although most of this research has been on screening and measuring alcohol consumption rather than on variables associated with treatment planning, in general, results from computerized assessments seem similar to those of face-toface administration (Bernadt et al. 1989; Malcolm et al. 1990; Gavin et al. 1991).

The topic of collateral interviews for screening and measuring alcohol consumption was reviewed recently by Maisto and Connors (1992). In at least one instance, alcoholism screening was successfully performed by interviewing the spouse rather than the client (Davis and Morse 1987). Several projects also suggest that spouses can provide meaningful information on whether a client has been drinking, although their judgments of specific level of consumption and frequency of drinking usually are less reliable.

Training Required for Administration

While procedures for administering