More Than a Thousand Words

Graphics for Clinical Practice

Picture this: A third-grade boy is referred to the school social worker for behavior problems in the classroom, including stabbing another child with a pencil and kicking his teacher. Digging a bit deeper, the social worker discovers the boy has not yet learned to read or solve the simplest of arithmetic problems. He often misses school because of asthma. Observing in the classroom, the worker notes that the child's teacher tends to use a high level of power-assertive and aversive-control strategies. Calling the boy's home, the worker finds that he lives with his grandmother; the child later tells the worker his father was shot and killed three years earlier and he has not seen his mother, a crack and heroin user, for several months. The child lives in an inner-city housing project that has a serious crime problem but that has a reputation for resident pride and involvement in management. The boy's grandmother speaks little English and works as a home health aide in the evenings, so the boy is often home alone.

What is the problem in this case (which reflects the realities of much contemporary clinical social work practice)? What is the cause? What is the best intervention? Clearly, none of these questions has a single or simple answer, although the worker may be tempted to invent one based on theoretical or methodological preference, or selective attention. Would play therapy focused on loss be best? Or social skills training in a group? Environmental intervention of one kind or another? Social work cases have always been complex, and are increasingly so today. This complexity only intensifies when the clinical social worker faces a large caseload reflecting diverse interrelated patterns.

Finding ways to pull all this information together, to make sense of it, to clarify not only one issue at a time but ever-changing transactions among persons, systems, and issues is a crucial challenge to clinical practitioners. Words fail us here. Words are linear, and must be funneled down so only one emerges at a time, isolated from its context. Although abstract language distinguishes humans from other species and is of critical importance, other specialized resources are also available. Throughout human history, people have used images for perceiving and communicating complex information holistically. The earliest cave paintings had both symbolic and practical functions. Maps and globes guide us to our destinations despite physical and temporal distances. Artistic images allow us to tap into our deepest emotions.

The brain is a massively parallel processor. It can absorb and integrate a tremendous amount of information simultaneously, if the incoming channel is wide enough. This is the great advantage of images; the bandwidth is much wider than for verbal expression. Visual images, therefore, play increasingly important roles in disciplines and professions dealing with complex data. In clinical social work, images permit workers to see the often circular transactions occurring in a case, to integrate case data holistically, and to expect that causation is reciprocal and multiple, rather than linear and singular. This perspective, this view of things, profoundly affects assessment, intervention, and practice evaluation. The remainder of this chapter examines the increasing use of graphic visualization techniques in a variety of fields, reviews a range of graphic tools and techniques currently in use in clinical practice, and touches on empirical evidence supporting use of the approach and principles important for enhancing its effectiveness. Subsequent chapters examine the use of existing graphic tools and suggest enhancements of them for work with individuals and families, groups, organizations, and communities, providing examples from clinical practice and practice research.

Visualizing Complex Systems

Graphic images and models are proving not only useful, but crucial in many contemporary fields dealing with complex data (Gleick, 1987; Markoff, 1988; Mundie, 1989). Only by graphically combining millions of discrete data items, for example, can meteorologists track weather systems, including hurricanes that may threaten thousands of lives. Theoretical physicists depend on images to think about events like collisions of cosmic strings at 75 percent of the speed of light, and chaos theorists require pictures to find order within apparent disorder. Computer-aided design systems are critical to the design and manufacture of an extensive range of contemporary products, from silicon chips to automobiles, in fields ranging from space technology to clothing design.

Computer systems, on which we all increasingly depend, are also becoming more and more visually oriented. Graphical user interfaces are the emerging standard, and graphic tools are the heart of contemporary systems analysis (Yourdon, 1989), identifying and preventing critical errors and omissions that might otherwise not be evident until the system is in daily use. Graphic computer-aided systems engineering (CASE) tools are now used to build other computer systems. Recent research indicates that visual computer programming produces better comprehension and accuracy than do traditional programming languages based on words (Cunniff, 1988), and commercial visual programming packages are now on the market.

Medical research and practice offer many examples of the use of graphic tools and images. Conceptualizing the deoxyribonucleic acid (DNA) double helix permitted dramatic advances in genetic research years before the structure could actually be seen. Computerized imaging systems like computerized tomography (CT) and magnetic resonance imaging (MRI) have produced dramatic improvements in the diagnosis and treatment of serious illness, and a project compiling a three-dimensional cross-section of the human body provides a new approach to the study of anatomy. X-rays, venerable medical imaging tools, are now being combined with expert systems to help physicians identify other cases similar to those they are handling, suggesting additional diagnostic and treatment information relevant to patients ("Yale Developing," 1990).

Sociologists and social psychologists use graphic tools extensively in their research programs. They often turn to sociograms and other visual tools to present and explain concepts extracted from complex statistical analyses and to identify meaningful patterns in the data (Knoke & Burt, 1983; Marsden, 1982; Moreno, 1934). Graphic depiction of exchange networks (for example, Cook, 1982) permits the study of changes among groups over time. Another useful approach is Bales's Systematic Multiple Level Observation of Groups (SYMLOG), which provides a three-dimensional graphic representation of friendliness, instrumental-versus-expressive orientation, and dominance in small groups.

Graphic visualization has demonstrated utility for organizing information effectively and coherently in a broad range of fields dealing with complex data. Social work deals with similarly (and sometimes more) complex patterns and contextual situations, and, in fact, social work and related disciplines have discovered the utility of images for conceptualizing and communicating about clinical practice.

Graphic Practice Tools and Techniques in Current Use

Social workers began to focus on the potential for graphic aids, including ecomaps, social network maps, and genograms, in the 1970s. Meyer (1976) discussed assessment, as follows:

The assessment of the unit of attention is in itself a process, the beginning of intervention. It can be concretized as the making of a true map of a situation, a map that should be drawn with the client himself. As an assurance of the contract, a mutual definition between social worker and client of the situation confronting them, the drawing of a map could elicit a joint understanding of the salient and relevant features of the client's story.
(p. 177)

In subsequently describing how an ecological/systems approach might change the way a case is handled, Meyer described a mother and daughter "literally drawing the case for themselves" and showed what the map looked like. Meyer (1983) also emphasized the conceptual use of circles rather than lines, favoring a contextual, ecosystemic approach over the linear for clinical assessment.

Ann Hartman has made substantial contributions to the development and use of graphic aids for practice. In an important early paper, Hartman (1978) discussed the utility of ecomaps, graphic tools for portraying family-environment transactions at a point in time, as well as genograms, a form of graphic family tree used to assess and coherently conceptualize the intergenerational life of a family. Hartman and Laird (1983) expanded on the use of these tools and demonstrated the use of family mapping to portray the inner workings of the family system, further developing Minuchin's (1974) structural mapping of families. Each of these tools, and a number of others in current use, will be briefly examined in this chapter. Certain of these techniques, in particular ecomaps and family mapping, are also examined in greater detail in later chapters.

Perhaps the first major applications of images to clinical work emerged from psychology. In projective tests, like the Rorschach and Thematic Apperception Test, a patient's perceptions of ambiguous shapes or drawings are interpreted. In a third popular projective test, the Draw-a-Person test, diagnostic information is extracted from a drawing completed by the patient. Personality tests like the Minnesota Multiphasic Personality Inventory (MMPI) produce scores on several scales; contemporary interpretation of these tests is not based primarily on scores on individual scales, but on the graphic profile of scores, emphasizing interactions among the dimensions. Personality tests can be of some utility, especially if the tester is familiar with recent empirical work. Still, a major limitation of these measures is that they are oriented to producing a personality description based on relatively enduring traits, an approach to assessment with serious limitations (Mischel, 1968; Stuart, 1970).

Nevertheless, these early efforts offer a number of important insights. For example, while the Draw-a-Person test may have only limited utility as a measure of personality, a worker can certainly develop genuinely useful clinical hypotheses from clients' drawings. For example, a 10-year-old girl, who was living in a foster home after being removed from her mother's care because of physical abuse, drew a picture of a young boy crying. She said it was her four-year old brother. When asked what made him cry, she said, "He's crying because he's being hit and burned a lot." After further gentle probing and a protective services investigation, the girl proved to be correct. In this case, clearly, the client used an image to communicate.

Profiles

Another concept drawn from early psychological testing approaches is the profile. Profiles simultaneously portray ratings on several variables at the same time, allowing clinicians to examine variables individually and also look at their interrelated patterns. One example is the Structural Profile developed by Lazarus (1981). Lazarus's approach to clinical work is to examine the BASIC ID (an acronym for "Behavior, Affect, Sensation, Imagery, Cognition, Interpersonal Relationships, and Drugs" -- under the last of which he subsumes physical health). Lazarus suggests that either clients, or clinicians in concert with clients, draw charts indicating the extent to which clients see themselves as being primarily "doers," "feelers," "people-persons," and so forth and that the resulting profiles may be useful for selecting treatment techniques likely to have a good fit for a particular client. For example, the profiles of two clients in Figure 1-1 suggest that the first client, an active, behavior-oriented individual, may respond very well to task-focused interventions, but may not respond well to rehearsing alternative behaviors in imagination. The second client is likely to be most comfortable with more intellectual, cognitive approaches, but because this modality is so out of balance with the others, the profile also suggests an emphasis on expanding other types of activities and relationships. Lazarus suggests that developing profiles like these may be particularly useful in couples work, to highlight potential areas of similarity and difference that may be important to intervention.

Another type of profile with immediate applicability to clinical social work is the Pie Analysis (Figure 1-2), developed by Kanfer and Schefft (1988). In this approach, a client divides two pie circles into segments relevant to assessment and intervention planning. The first pie is divided into life sectors on the basis of the client's current investment of energy and effort in each. The second pie is divided to show an ideal distribution among the sectors. The worker and client can develop an individualized list of sectors, if desired, or can use those suggested by Kanfer and Schefft, which include Work, Family, Health, Social Contacts, Self-Care, Moral-Religious, Economic, Education, Recreation, Sex, and Personal Achievement.

Kanfer and Schefft suggest that a single activity can fit into more than one sector (for example, going to college may relate to education, social contacts, and other sectors) and that the level of effort indicated may be based not only on time actually spent in activities related to the sector, but also on time the person spends planning, thinking about, or worrying about the activity. An advantage of this technique is that relationships among the sectors are evident; to increase attention to one, a person must often decrease attention to others. The client whose profiles are shown in Figure 1-2, for example, thinks relatively too much effort is spent on work and achievement and not enough on family, social contacts, or health. Interventive goals can be extracted from these client-illustrated relationships.

Time Lines

Among other graphic tools social workers have implemented in practice are time lines. Quam and Abramson (1991) present simple time lines and slightly more complex life lines for use with mentally ill elderly clients and their families. Time lines depict significant events across time on a single dimension. For example, Figure 1-3 shows a time line for an adolescent in residential treatment. Developing a simple time line representation with a client can help a worker quickly form a picture of the client's life history and communicate understanding and interest to a client for whom action-oriented approaches may be more effective than strictly verbal ones. Figure 1-4 demonstrates the use of the life line, on which significant life transitions are listed over time on the horizontal line and client ratings on the vertical line. Transitions rated positively by the client are plotted above the line and those rated negatively, below it. Notice that in this example from Quam and Abramson, the client's moves to an apartment and a group home were rated as positive transitions, while hospitalizations, moving in with a sister, and the mother's death were rated as negative.

Time lines and life lines are useful for general life review, and they ensure that potentially salient events (such as losses and moves) are discussed. They also may be useful for identifying types of situations that are particularly fulfilling or problematic for a client. The advantages of these techniques, and most of the others discussed here, are that they require a minimum of time and materials and no artistic talent -- important points about which to reassure clients (as well as oneself).

A variation on the time line approach is the Life History Grid, developed by Anderson and Brown (1980) and shown in Figure 1-5. Anderson and Brown suggested that life history grids are especially useful for child and adolescent clients. Significant events occurring during each year of life are recorded on the grid, with horizontal lines listing age, location, and several life areas of particular developmental importance. The grid approach ensures consideration of each potentially salient area. As Williams (1981) suggested, it is more likely that social workers will examine factors that are specifically structured into the assessment process. Completing a grid instrument may take time, which could be helpful during the engagement process because it structures a way of getting to know the client. Under other circumstances, the time factor could be more problematic. Because a child's memory may be limited, especially for more distant events, the use of other sources may be required. For some child and adolescent clients, obtaining missing information from relatives or other significant persons can be a useful between-sessions task.

Social Support Mapping

The level of social support available to clients is of critical importance in many fields of clinical practice, including, for example, work with severely and persistently mentally ill patients, work with aging clients, and work in health care settings. Social workers have developed and used a number of techniques to visualize and map social support. The Social Network Map (Figure 1-6), as adapted by Gaudin (1988), is one of the simplest. Beginning with a blank map, individuals who are in some way important to the client are plotted in the appropriate quadrant with a dot and name or initials. The closeness of the relationship is depicted by distance from the self in the center. Types of support (emotional, tangible, advice, social) can be portrayed by lines of different colors connecting with the self. Gaudin suggests that connections among network members also be drawn, providing data regarding the density and interconnectedness of the network. He indicates that social networks tend to consist of two or three tightly connected clusters of persons.

A social network map provides a holistic depiction of a client's support situation at a glance and often suggests immediate treatment implications. The importance of such transactional, contextual pictures for integrating important clinical material cannot be overstated. The temptation for social workers to attend to only some aspects of a complicated case is strong, and conceptual integration is never easy. Exchanges with multiple systems often have aggregate results, however, and are not simply independent factors. Including everything in a single picture conveys this message clearly, even when the worker decides to focus on one sector at a time.

With some clients, including certain neglecting families or homeless mentally ill persons, the networks portrayed may be extremely sparse. This is critical treatment information, but it needs to be handled sensitively with clients because it may only confirm negative perceptions of their life situation. Another limitation of the approach is that it tends to focus on individuals, when in some cases larger systems are the most important actors. Other instruments capture such complexities more comprehensively. Nevertheless, social support maps are relatively easy to prepare and clinically quite useful.

A related instrument is the Social Support Network Map, developed by Cheers (1987) and shown in Figure 1-7. This tool, which is completed by the client, is somewhat more complex than Gaudin's approach, specifying additional sectors for examination and including data like age and gender. Cheers's map has the advantage of portraying an individual's situation in an immediate and holistic way, simultaneously examining exchanges among important persons. Cheers has reported use of the map only with social work students, who found the instrument very helpful. For other populations, different sectors may be required. Cheers notes that repeated use over time may be particularly productive as a way of tracking changes and progress. This is an important observation, also suggested by Hartman for ecomaps, because this approach has potential for tracking multivariable clinical situations more holistically than do other existing methods. This potential will be explored in chapter 2.

At a slightly higher level of complexity are the Social Network Map and accompanying Social Network Grid, developed by Tracy and Whittaker (1990). The map is constructed first (see Figure 1-8). As is true for other social-support instruments, the Tracy-Whittaker map is divided into sectors (in this case, Household, Other Family, Work/School, Formal Services, Friends, Neighbors, and Clubs/Organizations/Church). The social worker and client consider each sector in turn, noting important persons in each. The most important persons (up to 15) are listed on the Social Network Grid (see Figure 1-9). Data items are completed for each individual listed on the grid, including the sector to which the person belongs; ratings of concrete, emotional, and advice-giving support; direction of help; degree of closeness; and others. Each identified individual is listed on a card, and the client sorts the cards into piles -- for example, into categories of "not very close," "sort of close," and "very close." The results are listed on the grid. This method takes time but permits the gathering of a wealth of important data while perhaps deepening engagement with a client. Although the transactional, contextual interrelationships within a social network may not be as immediately apparent as with other social-support mapping techniques that are more graphically intensive, the level of information collected is high. For some populations, especially those who receive services over a long period, developing the map and grid may be time well spent. Note that this approach focuses on a client's perception of the situation, rather than the worker's.

Genograms

Genograms are graphic tools, rather like family trees, on which intergenerational family relationships are portrayed (McGoldrick & Gerson, 1985). They are used in social work practice, as well as in family medicine. Symbols used in developing genograms (see Figure 1-10) are relatively standardized. The level of detail used to complete a genogram ranges from simple to very complex. The genogram shown in Figure 1-11, taken from Hartman and Laird (1983), is quite detailed and is appropriate for practice situations in which considerable family history is relevant. Large amounts of data may not be required or appropriate for many practice situations. For example, a family coping with a crisis is unlikely to be effectively engaged by developing a complex genogram. In adoption cases, however, an adoptive child in later years will likely treasure every bit of information listed on a genogram, and the birth mother often feels that the time she spends developing a genogram is a gift to the child. Genograms are also useful in many types of family practice, including work with the aging.

Hartman and Laird (1983) suggest the need for judgment about the types of settings and client situations in which genograms are used and about the amount of detail to include. The strength and quality of relationships, occupations, ethnicity, and other factors may or may not be included for clinical reasons. For example, when dealing with complex blended families, a relatively simple diagram like Figure 1-12 may be useful, if only to help a worker identify persons discussed by a client and to keep relationships straight. The three-generation household shown in Figure 1-12 includes a single developmentally disabled sibling of the mother, a stepfather, a boyfriend of one of the daughters, and several small children. An older son has been emotionally cut off from the family. Any or all these persons may be involved in the problems being experienced by the primary client, the 16-year-old boy.

Genograms or similar figures may help prevent situations in which a social worker discovers, perhaps several weeks late, that a critical person (grandmother or live-in partner, for example) has been overlooked in treatment planning. Detailed health-status genograms can be invaluable in health and genetic counseling. A genogram included in a case record also can save time if multiple members of a team are working with the same family, so each professional does not need to rediscover family relationships independently.

Family Mapping

Family mapping focuses on current patterns and relationships within a household. A common way of diagramming family structure uses Minuchin's (1974) symbols (see Figure 1-13). Minuchin stated that constructing a family map enables a family therapist to organize complex material obtained over the course of contacts with the family and to identify repetitive patterns that characterize the family. Examples of family mapping using Minuchin's symbols are presented in Figure 1-14. Panel a of Figure 1-14 shows a triangulated situation in which conflict between parents is detoured toward the child, reducing marital strain but increasing stress on the child. Panel b shows a household in which a grandmother assumes the bulk of parental functioning and the single mother and her children are overly enmeshed, with quite diffuse boundaries. A healthy single-parent family in which an older "parental child" has taken on some executive authority is portrayed in Panel c. Minuchin used these diagrams to present a wide variety of structural family arrangements, and his work merits careful study. Family maps are used primarily by clinicians to formulate cases for themselves, although in some instances they could be used with a family during sessions.

Other Depictions of Family Functioning

Clinicians have used a number of other graphic instruments to portray families and their functioning. For example, some clinicians suggest that simply drawing a floor plan of a home, including sleeping arrangements, can provide useful information about crowding, boundaries, and relationships (Meyer, in press). On a more abstract level, Olson (1986) developed the Circumplex Model, which is associated with a measure of family functioning, the Family Adaptability and Cohesion Evaluation Scale (FACES III). Although they posited three critical dimensions of family functioning (cohesion, adaptability, and communication), only cohesion and adaptability are plotted on the two-dimensional graphic model of family functioning derived from responses to the instrument (Figure 1-15). Sixteen types of family systems can be identified using these dimensions. Note that in the Circumplex Model, the healthiest families are "balanced," falling in the midrange on each dimension. Any extreme is regarded as indicative of pathology. Available research provides some empirical support for this model (Olson, 1986), but other data are less convincing (Daley, Sowers Hoag, & Thyer, 1991; Green, 1989; I. W. Miller, Epstein, Bishop, & Keitner, 1985).

Beavers and Hampson (1990) presented a different two-dimensional model, on which family style (centripetal to centrifugal) is plotted against family health or competence. Although extremes of centripetal or centrifugal style are generally seen as negative, competence is regarded as a linear dimension ranging from severely dysfunctional to optimal. Figure 1-16 depicts the Beavers Model, including types of offspring commonly associated with each family type. Beavers and Hampson reported substantial empirical support for the validity and reliability of the instruments on which their model is based and for its overall clinical utility. Although abstract models necessarily simplify phenomena to some extent and can only yield hypotheses that require clinical verification, such instruments can provide useful clinical guidance.

Graphic tools also are useful for work on specific types of family issues. For example, Stuart (1980) devised the powergram, a means of examining and reallocating decision-making authority for couples (Stuart, 1980). In the powergram (see Figure 1-17) each individual first plots the allocation of authority for certain decisions presently handled by the couple and then plots an ideal authority pattern. With the therapist's help, the couple can then negotiate a more acceptable pattern. (Only four of the 16 areas Stuart suggests are illustrated in the figure.)

Ecomaps

An ecological or ecosystems perspective, in which clients' transactions with environmental systems are taken into account in assessment and intervention, is a hallmark of social work practice (Germain & Gitterman, 1986; Meyer, 1976, 1983; Whittaker, Schinke, & Gilchrist, 1986). Although many of the graphic tools and techniques considered thus far have substantial utility for realizing this perspective, the ecomap was developed specifically for this purpose. Hartman (1978) discussed the advantages of this simulation tool for assessing and conceptualizing clinical cases holistically and for organizing a large amount of data in a nonlinear, contextual fashion. The ecomap is particularly useful for capturing information important to systemic formulations, including boundary issues, and the direction, rate, and mutuality of resource exchanges, while concurrently increasing the client's level of involvement in assessment and intervention planning. Hartman and others have observed that collaborative use of this tool with clients can dramatically decrease clients' defensiveness and that clients often feel the social worker using it is making a genuine effort to understand their life circumstances. The use of ecomaps reduces the amount of narrative required to capture a case situation, portraying relationships with extended family, formal systems like schools and public welfare offices, informal friendship networks, and many other relevant factors. It also provides a way of examining each of these factors in the context of the others. Ecomaps are, therefore, deeply contextual, systemic tools that are useful for all fields of practice.

Hartman suggests that workers keep a supply of blank ecomaps on hand, which can be relatively standardized for a particular setting. Figure 1-18 is an example of a blank ecomap (note the key at the lower left). Whereas some systems that are commonly salient to cases in the example are listed, other circles are left blank to be completed as needed. Some social workers develop specialized ecomaps (see Figure 1-19) in which additional detail beyond that commonly found on more standardized ecomaps is provided. Often there are trade-offs between the available time and level of detail, so deciding on the extent of individualization and detail to be depicted requires clinical judgment.

Hartman suggests using ecomaps to examine exchanges with systems outside the household and using family mapping separately to examine exchanges within the family. It can be effective to combine the two, however, to achieve a more integrated picture of the case situation when this can be done without making the image too confusing. Most social workers use variations of Hartman's approach in which connections between systems represent relationships -- positive, stressful, or tenuous -- and arrows reflect the flow of energy or resources (for example, Grieve, 1983). It also is possible to represent theoretically important exchanges in other ways, such as modeling networks of mutual reinforcement and aversives (for example, Mattaini, 1990a).

Ecomaps have the potential for wide use in social work practice, providing a generic means of modeling and simulating transactional exchanges that define an ecosystems perspective. They can be useful for work at multiple-system levels, from individual clinical work to community organization and across a broad range of practice approaches. Ecomaps also can be computerized, reducing the level of effort required to use them. A variety of enhanced ecomapping techniques are examined in later chapters.

Graphic Visualizations of Larger Systems

Graphic modeling techniques also have been used in case management and in group, organizational, and community practice. As social work practice becomes increasingly generalist, clinical social workers are often concurrently involved on multiple systemic levels, so these tools may be of significant utility to workers. For example, in a recent paper, Bar-On (1990) used graphics to visually portray contacts of a typical social worker in England with multiple clients and other systems in the course of day-to-day work. This approach, which is explored in chapter 4, is particularly suited to examining and monitoring case-management activities.

In group work practice (with treatment, task, and socialization groups), the level and distribution of involvement in group activities are often crucial factors. Rose (1977) suggested tracking the distribution of verbal participation by placing a mark in the sector associated with each group member for each statement, regardless of length, as shown in Figure 1-20. The social worker can share the graphic with group members; the resulting discussion will often resolve serious imbalances. If not, Rose presented other techniques for achieving a more equal level of participation. He and other group workers have developed a variety of graphic approaches to monitoring and modifying group interaction and working with individuals in groups, which are examined in chapter 3.

At organizational- and community-practice levels, several graphic techniques have proved helpful. One example is data mapping, in which occurrences of a social problem of interest are literally plotted on a map. The resulting graphic can be directly useful for service planning, for example, in determining locations of offices and levels of service required to meet the most serious needs. By examining such maps over time, it is possible to track changes in overall incidence rates and discover changes related to permutations of social conditions. This approach has proved useful for problems ranging from the suicide of youths in rural Alaska to child abuse in New York City. As is discussed in chapter 4 on organizational and community practice, graphic visualization techniques are valuable for direct grass-roots community organizational analysis, as well as for social planning.

Modeling Clinical Processes

Most graphic visualization approaches discussed thus far are attempts to model individual or systemic functioning, often at a single point in time. Graphic tools are also useful for other purposes, however. W. J. Reid and Smith (1989) called for "well-explicated" practice models, in which connections between client characteristics and situations, interventions, and outcomes are clear. When such connections are simple (for example, empathic responding will produce a decrease in anxiety), such models can be straightforward. When the theoretical model is more complex, however, graphic models can be very helpful for explaining clinical processes. For example, Berlin and her colleagues (Berlin, Mann, & Grossman, 1991) presented a task analysis of cognitive therapy for depression, grounded in a program of qualitative research. One outcome they regarded as crucial is the generation by the client of more constructive appraisals of interpersonal situations, which, in turn, leads to increased feelings of freedom and self-efficacy. The process through which they observed clients typically move in achieving this goal is diagrammed in Figure 1-21. The flow chart models the process much more clearly than would a verbal description of alternatives and permits the depiction of alternatives and recursive relationships.

Certainly, other models and interpretations of the steps involved in resolving serious depression are common; Berlin's approach, however, helps frame issues and generate testable hypotheses. Such models are extremely useful to help clinicians think clearly and precisely and they also can help, sometimes in simpler forms, to clarify clinical concepts. For example, the uncomplicated figure on the dedication page of this book can be useful for couples work, to clarify the question of closeness versus distance, which often requires attention. The utility of visual models for explaining conceptual frameworks is the primary focus of chapter 7 and is discussed by Carol H. Meyer in chapter 9.

Graphic Visualization in Practice Monitoring, Evaluation, and Research

Contemporary clinical social workers recognize the need to monitor practice on an ongoing basis and are increasingly involved in more formal clinical research. In fact, social workers are increasingly recognizing that research that is truly useful and accessible generally has strong practice roots. Graphic tools can inform this work at all stages, from conceptualizing hypotheses and gathering data through the analysis and final presentation. Practice monitoring (discussed in chapter 5) often is conducted using variations of single-case designs, including simple line graphs, to track clinical progress. For example, Figure 1-22 (from Slomin-Nevo & Vosler, 1991) shows the results of an intervention used to reduce the level of pathological jealousy experienced by a married woman toward her husband's deceased first wife. The graph is based on a simple self-anchored scale on which the client was asked to rate her level of jealousy on a 1-to-10 scale.

Although extremely useful for cases with simple and unidimensional treatment goals, single-variable line graphs may be too limited for monitoring complex cases (Gordon, 1983; Kagle & Cowger, 1984). More-contextual approaches are often required, which allow the concurrent tracking of multiple variables and permit workers to think through interactions among the variables. A number of such approaches are presented in chapters 2 and 5.

For larger-scale clinical research (discussed in chapter 6), many types of graphs and figures are used to present data, including pie charts, bar and line graphs, and data maps. Such figures often aggregate thousands of data points in ways that enable viewers to identify and follow patterns over time and are more accessible than complex tables of numbers -- although these also have their place. Such graphs and figures have been used for presentation purposes for many years, and their utility for data analysis has increased dramatically with the availability of low-cost computer systems that can rapidly display dynamic views of data. As these systems become increasingly user-friendly, their use by clinical and generalist practitioners will expand.

The Utility of Graphic Visualization for Practice

The foregoing review of the wide variety of graphic techniques currently used in social work practice indicates that many social workers and clients find them helpful. In her original article presenting ecomaps and genograms as practice tools, Hartman (1978) reported that both workers and clients responded positively to them, and substantial anecdotal evidence supports this conclusion. There also are other confirming data available.

Beginning with more basic research, there is evidence that concrete and diagrammatic elaborations of abstract materials lead to improved comprehension and remembering (Larkin & Simon, 1987; Mayer, 1976). Larkin and Simon discussed the advantages of images for maintaining information about the relationships among components, as compared with more linear textual presentation. There is also evidence that pictures are remembered better than are words (Anderson, 1980). Although common wisdom suggests that some people are more visual (have greater visual aptitude) than are others, Cunniff (1988) found that comprehension is more rapid and accurate for computer programs presented graphically than for those presented in traditional textual-programming languages, for nearly all subjects whatever their visual aptitude.

In some fields, in other words, graphic representation appears to be genuinely superior for almost every one. Although it is too early to be certain, this seems true also for the complex contextual situations dealt with in social work practice. There is some evidence that this is the case. For example, Tracy and Whittaker (1990) reported that all 23 practitioners with whom they tested their Social Network Map planned to continue to use it because they found it helpful for clarifying the specifics of stressors, supports, and resources for assessment and for providing a means of engaging and empowering clients. In a program of developmental research reported in chapter 8, early results indicate that when social work students complete an ecomap, they tend to formulate cases more transactionally and systematically and to look more evenly at intrapersonal, interpersonal, and environmental issues. Both heuristic and empirical evidence, although still preliminary, thus far support the use of graphic visualization to enhance clinical practice.

Effective Visualization

Given that visual techniques are valuable for social work practice, developing the most effective tools is clearly important. In general, approaches are sought that facilitate contextual integration of as much information as possible while making representations clear and comprehensible. Several categories of graphic techniques have been identified and principles for effective presentation have been explained.

Black and his colleagues (Teachers College, Columbia University, in association with the Visual Arts Program at the Massachusetts Institute of Technology, 1989) identified five categories of visual representation. The first is spatial representation, which provides a relatively direct image of the physical world like a map. Characteristics regarded by the preparer as essential are preserved and those that are not relevant are eliminated. For example, a worker may draw the floor plan of a client's life space. Other examples include projects in which workers map the physical space in social agencies as a way of studying services to clients (Seabury, 1971). The second category, pictorial representation, refers to more artistic renderings that portray things in the real world. They often seek to capture abstract or emotional qualities as well. Children's drawings of their families or their homes, for instance, may suggest clinically relevant hypotheses. The third category is narrative representation, in which multiple images are juxtaposed to explain changes and tell a story over time. An example is Sunday newspaper cartoons that often incorporate words. Sequential ecomaps, discussed in chapter 2, are in part narrative representations.

Perhaps most broadly useful for clinical practice is the fourth category, graphic representation, by which Black and his colleagues mean "synthetic visual representations of abstracted properties and relationships" (Teachers College, 1989, p. 61). Statistical graphs are a familiar example. Many of the graphic tools discussed in this chapter, like genograms and ecomaps, fit into this category. They can be primarily qualitative like a genogram; primarily quantitative like a graph of a client's score on a rapid assessment instrument over time; or, increasingly, both qualitative and quantitative, as shown in the quantified ecomaps presented in chapter 2.

The last category of visual representation is simulation, dynamic models that are used, for example, to change the value of one or more variables and follow the results over time. Applications of simulation techniques in social work practice currently are somewhat limited, but there are some examples, including computerized models for examining alternative service systems for severely and persistently mentally ill clients. The potential for modeling systemic behavior, while varying specified parameters over time, is an intriguing challenge for the future.

Whatever category of visual representation is used, some principles of effective graphic design have been formulated and are applicable. Bertin (1983) presented a detailed technical analysis of effective visualization techniques, exploring alternative approaches in depth. Edward Tufte (1983, 1990) presented outstanding synopses of critical principles and techniques for envisioning information. One such principle is "Show the data." Although this principle seems obvious, Tufte demonstrated that it is easy to produce distracting graphics in which the actual information is difficult to extract because of "chart junk," decoration, or meaningless detail. Tufte advocated lean graphics in which a high proportion of ink is used to actually show the data. Some compromises are necessary in the real world, however. For example, some authors limit the use of arrowheads on ecomaps to situations in which resources flow primarily one way and assume reciprocal flows if no arrows are shown (see Figure 1-19, for example), but workers and clients may find graphs heuristically more communicative if double arrows are used to model reciprocal exchanges.

Tufte suggested that excellent graphics should include as much information on multiple dimensions in as small a space as possible; induce readers to attend as much as possible to the data, rather than to the mechanics of presentation; and serve a clear substantive purpose. Tufte discovered that similar techniques have been used to produce these results for hundreds of years. One is the use of micro/macro displays, in which data can be examined at several levels of detail. For example, an ecomap can be examined to form an overall impression of the level of positives and negatives having an impact on a client or household, as well as to determine particular exchanges with specific systems. A conflict in an extended family may have different implications for a family that has substantial resources available from friends, work, recreation, schools, and other systems than such a conflict has for a very isolated family.

Other common approaches extracted by Tufte from the historical record include the use of layering, the use of transparent overlays to display different aspects of data that can be combined and separated visually. The use of a number of small related images that can be juxtaposed and compared -- small multiples -- is another common strategy, and it is applied frequently in the following pages. Graphic techniques that increase dimensionality (like the use of perspective or multidimensional models) can help viewers metaphorically "escape flatland," in Tufte's words, to capture more complex data on the same essentially two-dimensional plane. Narratives of space and time were among the earliest methods of envisioning information. Last, Tufte indicated that the effective use of color can significantly enhance the information-carrying capacity of visual representations, although the poor use of color can be distracting. (An advantage to using only black and white is that workers can prepare graphs and figures with just a pen or pencil and paper. For some purposes, however, such as work with children or with groups, color can be very effective.)

In the following chapters, graphic visualization for clinical practice is explored in depth. Principles of effective design are applied to a broad range of practice situations, and enhanced tools that are practical for day-to-day practice are presented. Readers are invited to use the materials as a starting point for further explorations of this dynamic landscape.