Handbook Of Measurements For Marriage And Family Therapy

Robert Sherman, Ed.D., Norman Fredman, Ph.D.

• Publisher: Routledge
• Year: 2013
• ISBN:
  9781134846658
• Edition: 1
• Topic: Psychology
• Subtopic: Mental Health in Psychology
• Index: Psychology

Published in 1987, Handbook of Measurements For Marriage And Family Therapy is a valuable contribution to the field of Family Therapy. The purpose of this handbook is to provide a single convenient source to which practitioners, researchers, and trainees can turn in order to learn how to use marriage and family instruments and to find descriptions of instruments suited to their needs.

1

Introduction to Marriage
and Family Testing

Why do we need tests about marriage and families? As a matter of fact, we are testing informally all the time. The therapist asks questions and makes judgments at both the initial meeting and the subsequent meetings with the couple. After all, “What’s going on?” is constantly being asked, silently or aloud. All right, then: Why do we need formal, objective tests about marriage and families? When Browning asked: “How do I love thee?/Let me count the ways,” she did not come up with a T-score and a standard error of measurement.

The question implies that there is something chilling about the very touch of testing that makes it inappropriate to marital warmth—inflexible plastic in a world of skin. It is true that tests are used in schools, but even there they can interfere with a close relationship between teacher and child. Can you see Carl Rogers sneaking a look at a child’s IQ score? If Goldman (1972) can speak of an uneasy marriage between testing and school counseling, what would he say of testing and marriage therapy?

Formal testing is usually appropriate in institutions like the army or big business where the institution is ignorant about the test taker, relates to him* in a formal manner, and affords him opportunities of promotion and change based on test results. Even schools can follow this three-point model: ignorance about the test taker, objective formal relations, and an opportunity for promotion or change. But surely not marriages. Husbands and wives, parents and children know everything worthwhile knowing about each other. Indeed, the Bible even calls the intimacy between husband and wife “knowledge.” Family relationships are warm. And marriage is forever.

Ignorance

At one level, questionnaires can be useful to eliminate fantasies about one’s mate. The low correlations between husbands’ and wives’ scores on measures of family adaptability and cohesion (Olson et al., 1985) might mean that they do not know what the family looks like in the eyes of their mates. Dunn and Dunn (1978) have found that couples in 72 intact marriages had a tendency to have quite different environmental preferences in such important, yet ignored, areas as keeping the room warm or cool, light or dark, noisy or quiet, organized or “lived in.” It may have always been so. Even nursery rhymes speak of the complementarity necessary to keep thin Jack Sprat and his corpulent wife efficient. Awareness of all differences for many couples can increase tolerance of those differences.

Formality

Some marriages are too informal without being warm. There are moments in a relationship in which objectivity can be a step towards better communication and ultimately more warmth.

Change

Nowadays, the possibilities of radical changes in the family not only can affect the husband-wife relationship but the parent-child relationship as well. Legally, a child can literally gain a new mother or father. Of course, on a less drastic level, change is what family therapy is about. Formal testing is now appropriate for families in distress and families that would like to avoid distress.

To the degree that marriage therapy can be taught, that progress is possible, that the next generation of therapists can learn from us (or at least from our mistakes), to the degree that marriage therapy is a science rather than an art, some answer must be attempted to the question “Does therapy work?” Objective formal testing tries to answer that question: Has the therapy had any effect? Insurance payments and court decisions mean that effectiveness of treatment will require more than clinical assertion.

Edward Thorndike’s statement “That which cannot be measured may not exist” seems a bit extreme, and we are tempted to respond, “That which can be measured may be irrelevant.” Yet the reality of self-delusion that we spot so well in others just might infect us as well. Measurement can put some limits on subjectivity and open the practitioner to alternative therapeutic techniques. It can help identify the problem, providing a quick and efficient first screening. Hours of interview time can be saved by appropriate measurement. Sometimes tests can even monitor the improvement or deterioration of family relationships while it is happening. Even when the test as a whole lacks respectable reliability and validity, individual items can suggest clinical questions and techniques that can influence the family relationship.

We live increasingly in a multimedia environment. However, much of therapy is an auditory experience. One of the purposes of our volume, Handbook of Structured Techniques in Marriage and Family Therapy, (Sherman & Fredman, 1986) was to add visual, kinesthetic, and tactile dimensions to the therapy session. It is important not to dismiss a test printout as “just intellectual.” It is a visual experience that often inspires confidence not granted to the spoken word. And by comparing the family’s results with the results of a sample of other families, the family can move beyond the myth of uniqueness. At the very least, many clients view a properly administered and interpreted inventory as “doing something” about the marriage conflict.

Problems of Marriage and Family Inventories

Measuring marriage is not quite like measuring reading. Cronbach (1984) made the distinction between trying to measure best performance and trying to measure typical performance. Some people advocate reserving the word “test” for measures of best performance. They use words like “scale,” “inventory,” or “questionnaire” for measures of typical performance.

A paper-and-pencil test has a tough enough time trying to rate a group of students trying to do their best. Personality evaluations usually demand honesty and insight that may be lacking in the general population, let alone among emotionally involved clients.

If someone ran pupil responses to an arithmetic test through a computer, she would probably find that all the questions were measuring the same thing: arithmetic. If someone were to run client responses to an anxiety questionnaire through a computer, however, she would probably not find that all the questions were measuring the same thing. Or she might find that they were measuring the same thing that could not really be labeled “anxiety.”

Some people, when given the choice of rating feelings, refuse to check extreme scores, like “always” or “never”; others check extreme scores regularly. Some people tend to agree with attitudinal questions with surprisingly little regard for their contents; other provocative souls disagree with the same disregard for the actual content.

Even assuming that the people filling out questionnaires are cooperative, honest, and insightful and do not have the habit of answering questions in a given pattern that slightly ignores well-thought-out directions, it is still technically difficult to total personality-based items into a score. The goal of a chemistry test is to get 100. The goal on any question is to get it right. Test makers can make sure to throw out any item on which unsuccessful students for some reason (like guessing or ambiguity) did better than successful students.

But take a test about anxiety or assertiveness or marital cohesion. Who says that a perfect score is ideal? A high score in cohesion may signal an enmeshment problem rather than good adjustment. Just as a perfect score on the total test is not desirable, it is not necessarily desirable to answer any single item “correctly.” Two people who produce the same cohesion score may have checked very different items. The relationship between two items is such that you cannot say that those who answered yes to one item will tend to answer yes to another item. If items do not function in such a way, it becomes hard to compute a total score that has much meaning.

It is extremely difficult to measure change scores, even when measuring knowledge (Thorndike & Hagen, 1978). It is even harder to measure change of individual emotions, and therapists are not interested in selection and description; they are interested precisely in change—emotional and behavioral change.

Thus it happens that not only are practitioners afraid that the use of tests can cool therapist-client relationships, but also researchers are afraid that personality measurements may be giving bogus answers to the important questions “Does therapy work?” and “What changes should be made?”

In addition, marriage inventories present a unique problem. The body to be measured is not the individual or even the group; it is the system and the relationships. Tests, by the definition used by test makers, are essentially measures of individual differences. Tests are written and rewritten to maximize the consistency of differences between individuals and to demonstrate that the difference has some meaning or use. When group scores are used, these individuals’ scores are averaged and the spread of scores described in a popularly agreed-upon fashion. But you cannot simply average a wife-and-husband’s score. We suppose you could if you were interested in their average weight to predict wear-and-tear on furniture. But if you are interested in the relationships, in the system, an average score, even accompanied by statements about how the wife differs from the husband, will not suffice.

The closest thing that now exists to a survey that measures systems are Classroom Interaction Scales. An outside observer marks (usually once every three seconds) who in the class said or did what to whom. Was the statement a question or a response, emotionally positive or negative, directed to a student, a peer, or the teacher? Modifications of this technique are not appropriate for marriage and family therapy, except, perhaps, by training institutions evaluating counselor trainees. For one thing, an outsider can sit in the back of a room without being too obtrusive when one teacher and 25 students are learning algebra. But when a therapist and a couple are discussing alcohol abuse, how unobtrusive can an outsider be? Moreover, the classroom observed is the arena being evaluated when we measure classroom interaction. Except when we are measuring the therapeutic relationship, the observed therapy session is usually not what we really want to evaluate in family therapy; we usually want to measure the family system outside the therapy session.

Observational techniques have been reported by Filsinger (1983a,b). These techniques generally require special equipment and special training to improve a commonly reported fault: low agreement between observers. For first-time users, the authors of the Couples Interaction Scoring System “… estimate that it can take 24 hours to code an hour of videotape” (Filsinger, 1983b, p. 126). Videotape for therapy insight is increasingly used and is described as a technique in the Handbook of Structured Techniques in Marriage and Family Therapy (pp. 198, 233–236). As a practical measurement tool, observational techniques await a breakthrough in efficiency. (Chapter 2 of this book will briefly describe some of these observational systems.)

The measurements included in this volume are self-report questionnaires dealing with values, attitudes, adjustment problems, and crisis management. All but one of the inventories we discuss are less than a decade old. On the negative side, this means that most have not had time to develop a history of successful clinical use. On the positive side, this means that the authors have had the technical advantage of using computers and the practical advantage of decades of testing and family therapy research.

How Inventories Are Made

Although the vast majority of marriage and family inventories located by Straus and Brown (1978) were constructed to answer a specific research problem, inventories increasingly are made to help the clinician. Nevertheless, universities remain the major source of new inventories. When family centers are associated with universities or church organizations, or are desirous of government funds, these centers must prove that they are effective.

When a presenting problem arises consistently at the family center such as wife abuse or quarrels over child rearing, or a grant has been funded, some clinicians may suggest using a questionnaire to expedite intake interviews, enlighten clients, or monitor therapeutic progress. A computer search through the literature may be disappointing for various reasons. Marriage and family therapy is such a rapidly expanding field that the available inventories might be inconsistent with the theoretical framework of the center. It might be felt that certain specific areas crucial to the needs of the clients have not been thoroughly covered.

If a new paper-and-pencil questionnaire needs to be constructed, the authors will first collect statements from the literature that are consistent with both their theoretical framework and their information-gathering needs. Hundreds of statements will be gathered. Then they will be rewritten in a style that is very readable. Seventh grade vocabulary is the goal. Jargon is weeded out. All statements are usually rewritten in a similar form. “Experts” will evaluate the statements for readability and appropriateness. If the questionnaire is long, a true-false or yes-no answering format will be used. A shorter questionnaire can use what is similar to a multiple choice format. There are often five choices varying from “always” to “never” or from “strongly agree” to “strongly disagree.”

In questionnaire construction, the easiest step to describe is the hardest to do: obtain a representative sample, first to try out the inventory, then to serve as a comparison group. This is a weak spot in every test. The sample is made up of volunteers, and volunteers tend to be nonrepresentative. Kinsey discovered this when he reported that his sample of men had more sex with women than his sample of women had with men (Kinsey et al., 1948, 1953). The samples used by achievement tests number in the tens of thousands. Yet even these samples are so nonrepresentative that clever superintendents can change the results of their pupils simply by switching to a test that uses a different population as a comparison group.

For the first run-throughs a truly random or even representative sample is unnecessary. The authors just need to find the items that are not “functioning.” Marriage and family inventories tend to use students and their willing friends meeting a class assignment.

The next step sounds very difficult, but thanks to the computer is actually quite simple (and fantastically impressive): run an item analysis. The authors “tell” the computer to select the items that “go together,” that usually produce the same results as the total inventory. The assumption is that the total inventory is closer to the truth than any single item. Those items that do not “function” (that do not produce the same results as the total inventory) are identified. In baseball, a line drive can be just an impressive out; a strikeout is an embarrassment. Managers will yank the players who strike out and try again with the playe...

Table of contents

• Cover
• Halftitle
• Title
• Copyright
• Dedication
• Contents
• Acknowledgments
• Preface
• Chapter 1. Introduction to Marriage and Family Testing
• Chapter 2. Clinically Popular Tests
• Chapter 3. Observational Measurements
• Chapter 4. General Marital Satisfaction and Adjustment Scales
• Chapter 5. Marital Communication and Intimacy Scales
• Chapter 6. Special Family Assessment Scales
• Chapter 7. Minnesota Family Inventories
• Chapter 8. Epilogue