The adolescent's family is an important factor in the adolescent's involvement in and treatment for substance use disorders. Therefore, it is critical to form a therapeutic alliance with the family to the fullest extent possible and to involve the family in the assessment process. If there is evidence that the adolescent is being abused at home, the family should still be questioned about the matter. It is important to pursue what is known about possible abuse from the parents, even the abusing parent, as well as other family members (e.g., siblings). Of course, the reporting requirements for professionals regarding evidence of abuse must be disclosed to each individual being interviewed (see Chapter 4 for details).

The assessment should not be considered complete until there has been time to assess the traditionally defined family and others identified by the court as legal custodians who can speak for the best interests of the adolescent, as well as the family that is defined by the young person. The assessor must determine who the "family" is as perceived by the adolescent and by legal considerations (that is, the person or entity able to legally represent the interests of the adolescent).

The assessment of an entire family requires a specific set of skills in addition to those needed to assess an individual (Szapocznik et al., 1988). Such assessments require people who are highly skilled and trained to interpret family dynamics, strengths, weaknesses, and social support systems. Assessors must also be able to identify key family structures and interrelationship patterns in which the adolescent's substance use disorder is enmeshed. It is also essential for the assessor to elicit previous treatment experiences, as well as previous attempts by the family to address the substance use problem and to ascertain the family's feelings about the adolescent. Do the family's responses to questions about this indicate the desire to help the adolescent, or do they suggest that the family sees the adolescent as "the problem?" These responses are useful in determining how to best proceed in working with the adolescent and the family.

Of course, the absence of a traditional family can be a barrier for adolescents seeking treatment. At-risk adolescents may be homeless or on the verge of homelessness. Some youth may go from shelter to shelter and have no address. In some States, a minor cannot gain access to any services unless an adult signs for her. Potential assistance can be obtained by initiating procedures to help the adolescent achieve emancipation or become a temporary ward of the State.

Key sources other than family members include adult friends, school officials, surrogate parent advocates in school-related issues, court officials, Court Appointed Special Advocates, social service workers (especially when the youth has been involved with the child welfare system), previous treatment providers, and previous assessors. Contacting these additional sources of information, with the client's consent, may be necessary to support or supplement the information that the adolescent provides in the comprehensive assessment.



Assessment Instruments

The Panel emphasized the importance of two methods for use when assessing adolescent substance use disorders: self-report questionnaires, and structured and unstructured interviews. (Laboratory testing, described in detail in Appendix C, is considered more relevant to the screening procedure.)

The use of well-designed questionnaires and interviews can yield an accurate, realistic understanding of the teenager and the problems he is experiencing. The information derived can also provide important insights into the young person's motivation and readiness to make use of and benefit from treatment.

Appendix B describes recommended instruments and their purpose, content, administration, time required for completion, training needed by the assessor, how to obtain them, their cost, and persons to contact for further guidance. All the instruments met the two most important criteria in the evaluation of any measurement instrument: reliability and validity. It is important to briefly discuss these psychometric concepts.

Reliability

Reliability refers to the relative freedom of a measure from error. One indicator of favorable reliability in a test is high consistency of item responses. Two types of consistency are involved: internal consistency and temporal stability. Internal consistency represents the expectation that the client's responses to various items are congruent to each other. For example, if the response to one question is that drugs are used "daily," it would be consistent for the client to say, in response to another question, that he uses drugs frequently. Temporal or "test-retest" consistency is based on repeated use of the measurement and refers to how the person's responses compare over a short time period, that is, from day to day or even from week to week. Thus, if the instrument is administered a second time to the individual shortly after the initial administration and the results for the two occasions correlate highly with each other, then evidence for the instrument's "test-retest" consistency is demonstrated.

Validity

Validity refers to the extent or degree to which the assessment instrument measures what it is intended to measure. Of course, a test can be valid only to the degree that it is reliable--a result with a wide amount of error cannot measure exactly what it is intended to measure. Good reliability, however, does not guarantee validity. Descriptions of assessment instruments often mention four kinds of validity.

One is content (or face) validity. This is, based on logical reasoning, the extent to which the test items are judged, "on the face of it," to deal with information, questions, or problems related to the stated objectives of the test. Content validity is often assessed by developing in advance a table of specifications that describes all the domains and characteristics that should be included in a test, and then having experienced judges rate their content relevance. A drug abuse test might gather evidence for face validity by obtaining ratings of relevance of test items from experts in the field. Some effective tests eschew content validity because they seek items whose content cannot be recognized by the subjects.

Concurrent or criterion validity is the extent to which the results of an instrument are statistically consistent with a measure intended to address the same trait or domain. The concurrent validity of a test being developed can be measured by comparing it to an already established test. For example, the Wechsler Adult Intelligence Scale has been demonstrated to be effective in assessing the thinking, memory, and learning capabilities of adults, and it has established validity as a test of intelligence. If a group of researchers developed another instrument, such as one that requires a person to solve linguistic and graphic puzzles, they might administer the two tests to a group of adults. The group would have evidence that the new test reflects intelligence if each individual scored at about the same level on both tests. That is, there would be evidence that the new test measures the same construct of intelligence that is measured by the Wechsler test by virtue of it concurring with the validity evidence associated with the established scale.

Predictive validity deals with the effectiveness with which an assessment instrument predicts how people will function or behave in the future. Thus, a criminality instrument could be used on a group of people to predict whether they will actually become criminals. In this regard, they would be followed for several years after completing the questionnaire and checked for evidence of criminality. The instrument would be considered to have predictive validity if a high correlation (for example, a correlation of .50 or higher) was determined between the results on the instrument and the later incidence of illegal behavior.

A complex type of validity is construct validity. This refers to whether the results derived from a test are consistent with and reflect the underlying theoretical notion it is intended to measure. This can be determined by assessing the extent to which the results obtained are in line with what the theory claims. For example, the developer of an assessment instrument may theorize that people who are likely to commit crimes are without clear-cut values of honesty, social conformity, or sympathy for other people and are not thoughtful about their actions. The developer then organizes a questionnaire containing items related to these traits. The questionnaire is administered to a group of known criminals and to a group known not to be criminals. When the questionnaires are scored, construct validity is present if the criminals and noncriminals are successfully distinguished from each other to a statistically significant degree.

Validity evidence can be reported in the form of correlations. Generally, validity coefficients tend to be lower than reliability coefficients. They may range between .30 and .80 or even higher, depending on whether they refer to concurrent validity (in which case coefficients tend to be higher) or to predictive validity (in which case coefficients tend to be lower). Also, as the complexity of what is being evaluated is great, as in the assessment of personality makeup, the validity coefficients are likely to be lower. Another form of reporting validity evidence is with between-group difference tests. The user of the instrument should examine the data available on validity to determine whether they represent the type of validity that fits the purposes for which the test is to be used.



Other Test Features

Norms, which are provided by the author of an assessment instrument, represent the scores or results that the types of people who are to be assessed by the instrument tend to obtain. No psychological instrument is useful for all people. Therefore, the author of the instrument reports the types of individuals for whom its use is appropriate. This report should refer to such client characteristics as the age, sex, ethnicity, educational achievement, socioeconomic level, and medical and psychological status of the population on which the original measurements were made.

Norms are often provided as tables that show how the scores are distributed for key characteristics, such as the sex or age of the population. The central tendency, or the average, of the scores is shown, along with the range from highest to lowest scores. These normative tables can be very useful to the counselor in determining the extent to which a client's functioning is within normal or abnormal limits. Often, as a test is used more extensively, norms are expanded, and the instrument becomes appropriate for increasingly larger and differing types of client populations.

Conditions for administration of any test or assessment instrument should be clearly spelled out in a manual prepared by the author of the instrument. The manual for the instrument should describe how the test was constructed and should reportavailable information on its reliability, validity, and norms. It should also describe the content and structure of the instrument, as well as how it relates to similar instruments.

Of great importance to the user is the author's description of how the instrument is to be administered, scored, and interpreted. Specific statements should include

1. The purpose or aim of the test
2. For whom the test is and is not appropriate
3. Whether the test can be administered in a group or only on an individual basis
4. Whether it can be self-administered or if it must be given by an examiner
5. Whether training is required for the assessor, and, if so, what kind, how much, and how and where it can be obtained
6. Where the test can be obtained and what it costs

Consideration of the above practical issues and of the conditions for administration should enable program staff to select the instruments that are most applicable and useful for its program and clients. Once selected, the tests should be administered in the manner recommended by the authors. No substitutions should be made for any test items and no items should be eliminated or modified. For structured interviews, the interview format and item wording should be strictly followed. If this rule is not followed, the results obtained from the test cannot legitimately be interpreted in terms of the norms provided in the test manual. Changing the test in any way makes it, in effect, a different test, so that the reliability, validity, and norms reported for the test no longer apply, thus making it difficult to know how to interpret the results. However, not all assessment tools are tests. The more descriptive instruments may have more flexibility in terms of adaptation to the individual and the situation.



Written Report

Depending on the setting, the assessor should prepare a detailed report based on information gathered using assessment instruments and personal observation. The complexity of adolescence requires that the individual being assessed never be reduced to a test score. A child's range of strengths and problems can best be evaluated with both quantitative and qualitative procedures. The aim is to assess the strengths and competence, as well as the limitations, of the child (see Figure 3-2). After the information from the different sources has been assembled, the assessor writes a report of what he has learned about the adolescent in terms that can be understood by all concerned, including the adolescent. The written report captures the adolescent's range of problems, strengths, and sources of support, as well as those of the youth's family.

To maintain continuity with previous workups and interventions, to make efficient use of all information available, and to spare the adolescent (and the party paying for the assessment) unnecessary duplication of effort, the assessor should be actively involved in determining if organized, accurate information on the adolescent already exists. When appropriate, that information should be integrated into the current written report. In particular, historical information can provide an indication of the progression of symptoms and problem severity. However, the assessor's report, along with providing immediate direction for treatment and other interventions, has the potential to follow the young person for years and be a central factor in shaping decisions about the adolescent. Therefore, it is important not to include opinions and descriptions from previous reports unless that information is currently accurate. The report should deal with such issues as (1) the way the adolescent processes information most effectively and how this will affect treatment, (2) how the adolescent's past experiences will affect his reaction to certain treatment interventions, (3) specific treatment placement recommendations and justifications, and (4) counselor recommendations. As the field has many different levels of professionals, it is important that these reports be written with specific treatment recommendations that can be understood by all.

The report should be distributed on a need-to-know basis to those service providers who will be working with the adolescent. Adolescents and their parents or guardians often request reports or assessment findings. One practice is to write the report to the parents of a youth under 18 years of age and directly to the young adult if he is over 18, with a copy to the parents who may be paying for the assessment. However, in keeping with the requirements regarding confidentiality, information often cannot be released without the young person's approval and signature on the proper consent forms. Refer to Chapter 4 for further elaboration on the laws regarding release of information.

The report should specify recommendations for treatment placement and posttreatment support services, although the latter issue may require knowledge of treatment progress. The report should also contain a plan for use by a case manager or other responsible party for monitoring services provided to the youth.

Source: The National Clearinghouse for Alcohol and Drug Information
DHHS Publication No. (SMA) 99-3282