Attitudes toward the mentally ill: a review

Abstract

This review of the literature investigates how attitudes toward the mentally ill affect their involvement in the community and what can be done to enhance such involvement. The author draws conclusions from research on these issues and makes recommendations for policy based on the findings.

More than 20 years ago, Star reported a national study of public attitudes about mental illness. She found that the public regarded behavior as proving the existence of mental illness when the behavior exhibited the following three interrelated characteristics:

1. A breakdown in intellect … a loss of reason.

2. A loss of self-control, usually to the point of dangerous violence against others, and certainly to the point of not being responsible for one’s acts.

3. Behavior which is inappropriate—that is, neither reasonable nor expected under the particular circumstances in which the person finds himself [Star, 1955, p. 5],

The following report is from another ambitious study, sponsored by the Institute of Communications Research at the University of Illinois and conducted over a six-year period from 1954 to 1959:

There is a strong “negative halo” associated with the mentally ill. They are considered, unselectively, as being “all things bad.” … The average man generalizes to the point of considering the mentally ill as dirty, unintelligent, insincere, and worthless. Such unselectively negative attitudes are probably due in part to a lack of information about mental illness and a failure to observe and learn about mental illness phenomena in daily life

[Nunnally, 1961, p. 233],

Given these perspectives on mental illness and the mentally ill and 15 years of experience with an active, nationally sponsored community mental health program, how does the public now perceive the mentally ill? How does it view mental illness? How does it act on these perceptions? And what types of governmental action or programming should be recommended on the basis of public attitudes toward the mentally ill? Recommendations for action are contingent on an in-depth understanding of the answers to the first three questions.

The public’s perception of the mentally ill and mental illness and the way it acts on these perceptions can be summarized in five statements derived from research in this area, which are listed and discussed in the sections below. The discussion of the first two statements provides a deeper understanding of how the public views mental illness and the mentally ill today. The other three statements focus on the public’s actions based on their views of the mentally ill.

RANGE OF BEHAVIOR

A broader range of behavior is viewed by the public today as “mental illness” than was the case 15 years ago.

Review of the research on public attitudes toward the mentally ill allows us to draw such conclusions, albeit cautiously, from replicated studies over time, largely as a result of the following two innovations in research methodology in the area: (1) the use of standardized case descriptions of behavior that might be diagnosed as mental illness and (2) the use of social distance scales to measure the degree of contact the public will accept with the mentally ill. (The latter point will be discussed later.)

The first innovation was the development by Star (1955) of six fictitious case descriptions illustrating different types of mental disorders as conceived by the mental health professions. These case descriptions, called Star vignettes, were developed with psychiatric consultation to depict the type of persons that might be given the diagnoses of paranoid schizophrenia, simple schizophrenia, alcoholism, anxiety neurosis, juvenile character disorder, and compulsive phobia. The utility of these case descriptions was verified not only in Star’s original research, but also in subsequent research endeavors (Dohrenwend and Chin-Shong, 1967). When asked to assess the accuracy of these case descriptions, psychiatrists generally agreed that they did in fact portray the type of behavior that would correspond to each of the diagnostic categories.

The Star vignettes were used in seven studies, reviewed by Dohrenwend and Chin-Shong (1967) and Bentz, Edgerton, and Kherlopian (1969), in which respondents were asked whether, in their opinion, the individual described in each of the six vignettes were mentally ill. The proportion of the public that was likely to identify a vignette as describing a mentally ill person became progressively smaller from the first to the last case description. Thus, more individuals described the paranoid as mentally ill than were willing to describe the simple schizophrenic as such, still fewer described the alcoholic in this way, and so forth, through the anxiety neurotic, person with a juvenile character disorder, and compulsive phobic.

Although the same pattern of identification held for all seven studies considered, the subjects in the more recent studies showed an increasing tendency to identify more of the vignettes as illustrating problems of mental illness. For example, in two of the earlier studies, conducted in 1950 and 1951, respectively, the only case identified by a majority of the respondents as being mentally ill was the paranoid schizophrenia. In studies done a decade later, from 1960 to 1964, at least three of the vignette characters were identified as being mentally ill—that is, the paranoid schizophrenic, the simple schizophrenic, and the alcoholic. In one study conducted in 1960, a majority of the population of community leaders identified the juvenile character disorder as a mental illness in addition to the other three cases just mentioned. This finding, however, is consistent with the observation that individuals who have more education are more likely to classify behavior as mental illness (Rabkin, 1974).

In 1969, Bentz, Edgerton, and Kherlopian reported on a study conducted using only four of the Star vignettes (excluding the paranoid schizophrenic and the compulsive phobic). In this study, at least 50 percent of the subjects labeled each category as describing someone who was mentally ill. Assuming that a majority of the group would have so labeled the paranoid schizophrenic had it been included in the study, this sample would have labeled at least five of the Star vignettes as representing mentally ill persons.

It thus appears that there is an increasing tendency in the population to identify a broader spectrum of behaviors as mental illness. This conclusion must, however, be tempered by the observation that the apparent historical shift may be “more one of superficial labeling than of conviction, for there are still some important differences in the way psychiatrists and the public view these cases” (Dohrenwend and Chin-Shong, 1967, p. 511). Also, the study samples differed in size and the studies were conducted in different locations. Carl D’Arcy and Joan Brockman (1976; and forthcoming) failed to find any differences in public recognition of mental illness between their 1974 replication of a study completed in Saskatchewan, Canada, and the original study done by John and Elaine Cummings in 1951. They attributed differences found in other studies to differences in the demographic characteristics of the population surveyed in different areas and to differences in study methodologies.

In Dohrenwend and Chin-Shong’s study (1967) the authors asked their psychiatrists to rate the six Star vignettes in order of their seriousness. Psychiatrists ranked the Star vignettes in a manner exactly corresponding to the order in which the general public recognized mental illness, that is, the psychiatrists tended to see the paranoid schizophrenic and the simple schizophrenic as being the most serious problems, followed by the alcoholic and the juvenile character disorder, and so on. However, when Dohrenwend and Chin-Shong’s sample of general population was asked if the problem portrayed in the vignette was serious, regardless of whether or not they thought the fictitious individual was mentally ill, they expressed a priority different from the psychiatrists. Respondents in the general population “were most likely to see as serious those cases that threatened others: the paranoid, the alcoholic, and the juvenile character disorder” (p. 512). Although psychiatrists emphasized the psychodynamics of the mental illness, as evidenced by their ranking of the simple schizophrenic as the second most serious vignette, the public eschewed that concept of mental illness and tended to rely more on the criterion of dangerousness in rating the seriousness of the problem. The general public rated the simple schizophrenic as the fourth most serious case.

D’Arcy (1975) found a similar discrepancy between the responses to the Star vignettes when individuals in the general public were asked whether the people were mentally ill as opposed to being asked whether they had “something wrong with them.” D’Arcy’s subjects also tended to find “something wrong” with the fictitious characters according to a criterion of dangerousness rather than mental illness per se.

The trend to define more behavior as mentally ill is better established by replicated studies than the findings regarding the public’s diagnosis of particular behavior patterns. Taken together, however, the results indicate that Star’s conception of how people view the mentally ill is still viable. Star (1955) pointed out, on the basis of her study of 3500 people representative of a cross section of the American public, that people categorize individuals into those with “nervous conditions” and those who are “insane.” The latter category—“true” mental illness in the public eye—involves the attributes of unpredictability, impulsiveness, loss of control, extreme irrationality, and legal incompetence, or such symptoms as violent behavior, incomprehensible talk, delusions, or hallucinations (Star, 1955, p. 3). The classification “insane” is usually the source of the public’s negative attitudes.

Only one study that the author is aware of raises serious questions about the relationship between negative attitudes toward the mentally ill and the public’s conception of them as violent and unpredictable. Schwartz, Myers, and Astrachan (1974) pointed out in their study of 124 relatives of former mental patients that the type of rejection they found was due primarily to neurotic impairments and to a lesser extent to the psychotic type of behavior described by Star. It should be pointed out, however, that this population of relatives of former patients expressed less rejection of the mentally ill than is evidenced by other samples taken from the general population.

In a consideration of five studies, Aviram and Segal (1973) found that on a 6-point scale of acceptance-rejection, the average amount of acceptance, taken across studies, was 3.26, indicating a willingness on the part of the public to live next to a former mental patient, but not to room with one. Computing a comparative acceptance-rejection score from the sample from the Schwartz, Myers, and Astrachan study yields an average score of 5.5 (assuming a perfect Guttman Scale). This score implies a willingness to allow one’s child to marry someone who had been mentally ill. This is a higher level of acceptance than that obtained based on a reworking of Kirk’s (1974) study of 864 community college students, using a similar scale and offering their opinions regarding association with a “normal” person. It thus appears that the type of rejection explained by Schwartz, Myers, and Astrachan is one that occurs within the family as opposed to the response of the general public to a mentally ill outgroup.

These studies indicate that it may be the public’s judgment that “true” mental illness involves dangerousness and unpredictability that leads to their rejection of the mentally ill population. The tendency of people today, however, to include a broader spectrum of behavior (including “nervous conditions”) under the rubric of mental illness may result in a diluting of negative responses toward the mentally ill, lessening negativity toward some while extending it over a broader segment of the population.

EFFECTS OF BEHAVIOR

The behavior itself or the pattern of behavior, is the major determinant of the positive or negative character of the public’s attitude toward mental illness.

Sociologists, particularly Sarbin and Mancuso (1972), have tended to be pessimistic about current efforts to educate the public about the nature of mental illness. They have pointed to the negative impact of the label of “mental illness” on the reaction of the public to individuals so defined, stating that “if the semantic tag, mental illness, is attached to a particular behavior, the public will tend to reject and to advocate isolation of the persons who are thus labeled” (p. 161).

“What seems to be important is the impact the label of mental illness may have when used in the absence of an actual person—and therefore in the absence of a particular behavior pattern.”

Sarbin and Mancuso rely on the work of Phillips (1963, and 1964) and other sociologists using similar methodologies for their pessimism about educating the public. Interestingly, however, a careful consideration of the work of Phillips and his followers tends to offer some evidence to contradict this pessimistic view. Instead of indicating that the label “mental illness” is an overriding factor in people’s reactions, the behavior of the individual in question was found to be most influential in evoking expressed attitudes of social rejection toward that individual.

Phillips (1963) presented housewives with four of the Star vignettes and one vignette of his own devising that described a normal person. In presenting these vignettes, he systematically altered the source of help that the fictitious person was described as using, such as clergyman, physician, psychiatrist, or mental hospital. Using a measure of social rejection, Phillips found that the closer to the psychiatric profession the person described moved in seeking aid, the more likely they were to be rejected, especially when they sought help from a mental hospital. Phillips also found, however, that the overt behavior of the fictitious person in these cases exerted a much more powerful influence on social rejection than the source of help.

Schroeder and Erlich (1968) replicated Phillips’s study, using psychiatric nurses as subjects. They found that the behavior variable was even more important in influencing responses of social rejection than among Phillips’s housewife subjects and that the source of help was of far less influence. Bord (1971), using a sample of college students, also found that the behavior of the mentally ill person was the prime predictor of social rejection and that the source of help seemed unrelated to level of social rejection.

In a more recent study, Kirk (1974) attempted to define the independent influence of behavior, labels (such as “mental illness, “situational stress,” “wickedness,” “moral deficiency,” or “normal stress”), and the labeler (that is, the person who interpreted the cause of the behavior) on the perception of mental illness. The labelers Kirk used as the providers of information in his study were “self,” “family member,” “other people,” and “psychiatrist.” Three vignettes were used—Star’s paranoid schizophrenic and anxiety neurotic and the normal case developed by Phillips. Kirk’s study revealed no significant interactions among the behavior described, what the behavior was called, and who interpreted the cause of the behavior. The label itself and the persons offering the label had no significant effect on rejection scores; only the behavior of the individual was important in influencing social rejection.

These findings bring into question the impact of the label of mental illness itself, especially when that label can be associated with a specific pattern of behavior. What seems to be important is the impact these labels may have when used in the absence of an actual person—and therefore in the absence of a particular behavior pattern. Without evidence about an individual’s actual behavior, mental illness labels may have a significant impact on social interactions (Kirk, 1974), The negative impact of such “disembodied” labeling is substantiated in the work of Nunnally (1961) and Simmons and Chambers (1965). As noted above, expansion of the term mental illness to include a broader spectrum of behavior may ultimately dilute the negative response pattern precipitated by the label. However, a continuing public concept of mental illness based on evidence of dangerous behavior would appear to make predictions of social rejection more contingent on situational factors such as the frequency and seriousness of media reports about violent acts committed by the mentally ill or the general conception of how dangerous the neighborhood in which a person lives is. These situational factors offer a cue as to the specific behavior pattern people respond to. Lack of specificity of such cues makes predictions of social rejection more variable.

SOCIAL DISTANCE

There is less reported social distance between the mentally ill and the public today, although situations requiring a greater degree of personal involvement still result in significant efforts at avoidance on the part of the public.

During the past 25 years several studies have attempted to assess the degree of closeness the general public will accept in their relations with the mentally ill. Despite some structural differences, each of the studies made use of a social distance scale as the dependent measure, the second research innovation mentioned earlier. These scales consist of statements concerning the degree of closeness with the mentally ill that a person would tolerate. The respondent agrees or disagrees with statements such as the following:

1. I would not hesitate to work with someone who had been mentally ill.

2. I would be willing to sponsor a person who had been mentally ill for membership in my favorite club or society.

3. If I owned an empty lot beside my house I would be willing to sell it to a former mental hospital patient.

4. I would be willing to room with someone who had been a patient in a mental hospital.

5. I would strongly discourage my children from marrying anyone who had been mentally ill.

6. I can’t imagine myself falling in love with a person who had been mentally ill [Bentz and Edgerton, 1971, p. 31].

Aviram and Segal (1973) compared the scores from five of these studies, spanning a period from 1957 to 1971, in terms of the relationship between mental illness and response tendency on a social distance scale of acceptance or rejection. Their results indicated that the public is in fact, becoming more accepting toward the mentally ill. It should be noted, however, that the dividing line on the scale for the average respondent is still between “selling a lot beside my house to a former mental hospital patient” and “rooming with such a patient.” The average person will accept the former, but not the latter. Aviram and Segal concluded:

Assuming that all that is needed is the willingness to live near a former patient to enable his reentry into the community, we can still see that a large group will not tolerate this amount of contact and that the possibility of any greater contact than this would be viewed with rapidly increasing distaste

[p. 127].

ATTITUDES AND ACTIONS

There is little evidence of a direct relationship between negative attitudes held by the public and their behavior toward the mentally ill; yet, recent public actions focused on keeping the mentally ill out of local communities imply such a relationship and require governmental intervention.

Rabkin (1974), reviewing several studies relating attitudes of the public and their behavior toward the mentally ill, points out that this relationship is not well defined owing to the pressures exerted on individual people in individual situations. The Opinions about the Mentally Ill (OMI) Scale, designed by Cohen and Struening (1962), has been widely used to assess attitudes and related behavior of volunteers and professional staff toward the mentally ill. The scale itself has five general attitudinal categories or subscales:

1. Authoritarianism—indicating a tendency to view the mentally ill as an inferior class requiring coercive handling.

2. Benevolence—indicating a kindly, paternalistic, and humanitarian attitude toward mental illness.

3. Mental hygiene ideology—indicating a positive orientation and acceptance of modern mental hygiene concepts. (This scale emphasizes the capability of the patient in situations of independent functioning.)

4. Social restrictiveness—indicating a tendency to perceive the mental patient as a threat to society and in need of restriction in social functioning.

5. Interpersonal etiology—indicating acceptance of the belief that mental illness arises from interpersonal experience [Pryer, Distefano, and Marr, 1969, p. 253].

In a study of hospital staff’s attitudes toward mental illness and patients’ history of treatment Cohen and Struening (1964, and 1965) found that if the staff scored high on the authoritarianism and social restrictiveness subscales, their patients tended to spend less time in the community between hospitalizations than patients in hospitals whose staff scored low on these two subscales. Ellsworth (1965) related the attitudes of 65 aides and nurses in a Veterans Administration hospital, as reported on the OMI Scale, to average ratings of their behavior made by the patients on their wards. His major finding was that aides and nurses who scored high on the OMI social restrictiveness subscale were more likely to have their behavior characterized as rigid, inconsiderate, domineering, and lacking in understanding, trust, and responsiveness to patients.

The studies of Cohen and Struening and of Ellsworth offer only tentative evidence of a relationship between attitudes and reaction of the public toward the mentally ill. These studies support the hypothesis of the existence of a negative relationship between an authoritarian belief system and its behavioral consequences for the mentally ill. These attitudes, however, are carried beyond the hospital into the community. Community programs throughout the country in the past five years, have had results that, although unsubstantiated by systematic research, indicate a very significant relationship between negative public attitudes toward the mentally ill and action taken against them.

Aviram and Segal (1973), for example, report on the use of exclusionary techniques, such as restrictive zoning ordinances, fire safety regulations, and simple bureaucratic red tape with respect to the establishment of small sheltered-care facilities for the mentally ill in local communities. Perhaps the most blatant illustration of such an effort to exclude the mentally ill from the community is found in a city ordinance passed in a New York municipality that attempted, in effect, to stop any former psychiatric patient from registering in a local hotel. Although this particular ordinance has been declared unconstitutional, restrictive zoning throughout the country continues to exclude the mentally ill from many communities.

NORMAL ROLES

If the mentally ill assume roles in which they can be perceived as “normal,” they will be evaluated and treated as such.

There are several indications that if the public is allowed to interact with the mentally ill in situations that allow the mentally ill to be perceived as “normal” or as “real human beings,” the attitudes of the public will shift in a positive direction. Several studies have been made relating to the impact of volunteer programs and student training programs on the attitudes of their participants. In these studies, generally using the OMI as an evaluative index, consistent changes in a positive direction have been detected (Pryer, Distefano, and Marr, 1969; Holzberg and Gerwitz, 1963; and Lewis and Cleveland, 1966). It should be emphasized that simple contact with the mental patient was not sufficient to change attitudes; in most cases it was necessary to provide additional education, emphasizing the humanist aspects of dealing with mental illness (Holmes, 1968; and Pryer, Distefano, and Marr, 1969).

“Contact with former mental patients can be a double-edged sward. … Groups evidencing characteristics that would reinforce the public’s negative attitudes, such as dangerous, violent, or bizarre behavior, are not the type of individuals with whom public contact should be promoted.”

It should also be pointed out, however, that contact with former mental patients can be a double-edged sword. Amir (1969) in reviewing the literature on attempts to promote contacts between racial and ethnic groups to reduce prejudice, listed a large number of favorable and unfavorable circumstances under which such contact should be conducted. The discussion is too detailed to reproduce here, but suffice it to say that groups evidencing characteristics that would reinforce the public’s negative attitudes, such as dangerous, violent, or bizarre behavior, are not the type of individuals with which public contact should be promoted. The majority of individuals who fall into the category of former mentally ill patients do not, however, evidence such behavioral characteristics.

In an experimental approach to the problem, Farina, Felner, and Boudreau (1973) attempted to document the reactions of male and female workers to job applicants of the same sex, half of whom were described to the interviewers as former mental patients. Although this status did not seem to have an effect on the assessments by the female workers, male workers were more rejecting of former patients than of “normals.” In several previous studies, males have been found to receive higher rejection scores (Phillips, 1963) and to be viewed as more dangerous (Levinson and Zan York, 1974). This type of reaction to mentally ill males might offer a very plausible explanation of the results obtained by Farina, Felner, and Boudreau. Certainly, this explanation would be congruent with the public conception of mental illness and its relation to dangerous behavior.

In reporting the results of a 20-year follow-up study of hospitalized schizophrenics, Clausen and Huffine (1975) indicated that over time,

if a former mental patient manages to function well on the job and in the family—and many do—such feelings of stigma as were originally engendered will diminish and neither patient, family or friends will be disposed to think in terms of former patienthood or mental disorder

[p. 415].

Such findings emphasize the importance of providing normal roles for former mental patients.

RECOMMENDATIONS

These findings suggest five specific recommendations to mental health professions and concerned governmental agencies for future policy and action. The first recommendation is that these groups emphasize educational programs that involve selective contact with the mentally ill. This recommendation rests on the observation that reactions to the mentally ill are based primarily on their behavior. As indicated above, however, contact can be a double-edged sword.

Although this article has shown that the concept of mental illness is broader today than in the past, the public still views a subgroup of the seriously mentally ill as objectionable, dangerous, and largely unpredictable. Attitudes that are not responses to a specific set of behaviors or to a specific group of mentally ill individuals tend to range from positive, in regard to the broader category of mental illness, to negative in regard to the smaller group regarded as seriously ill. Certainly, a distraught neighbor arriving at the door with a petition to block the entrance of a sheltered care facility into the neighborhood might well precipitate a negative response in an individual with otherwise positive attitudes toward the mentally ill. The literature demonstrates that contact with the mentally ill is not in itself sufficient to induce positive changes in attitude. Only selectively planned contacts coupled with educational efforts will begin to approximate this goal.

Second, the mental health professions and concerned governmental agencies should encourage the development of programs that place former mental patients in more normal roles. Many formerly mentally ill individuals have skills to offer and can make positive contributions to voluntary community service activities. Allowing their participation enables the public to see them in more responsible positions. Programs emphasizing patient activism, such as New Opportunities through Voluntary Action (NOVA), should be given priority because of the high level of responsibility and competence developed by and imputed to individuals participating in such programs. NOVA, for example, is an organization of former mental patients living in board and care homes. It has been effective in achieving better living conditions for residents of such homes and has been instrumental in insuring these residents’ rights.

Third, it is important to encourage a broad educational program on the interpersonal nature of mental illness and especially on the role of the environment in causing mental illness. The focus of such a program would be to eliminate the impersonal, disembodied stereotype of mental illness. Emphasis should be placed on specifying the applicability and the strengths and weaknesses of the medical model as it relates to each different diagnosis. Mental illness, like physical illness, is an inclusive rubric describing many syndromes.

The impact of the past 15 years of community mental health programs has been to broaden the concept of mental illness and to move the public’s perception of the mentally ill toward a more positive focus. Although this has by no means solved the problems of community reaction toward the mentally ill, it has made overt rejection of them more difficult and has increased the public’s willingness to associate with them.

The fourth recommendation is to develop community resource personnel such as operators of family care homes as potential community educators. These individuals are not only aware of their community but are usually able to relate well to its members, to understand their need, and to help promote change when they are made aware of the necessity of these actions to enhance the environment of the mentally ill population they serve. This approach may have already contributed to the process of changing attitudes in communities, as evidenced in a recent survey conducted by Hazleton, Mandell, and Stern (1975). The community studied initially exhibited a strong negative reaction to the development of sheltered-care facilities for the mentally ill but appeared to have shifted to a positive orientation as the result of locally sponsored community education.

Finally, the mental health professions and concerned governmental agencies must promote and fund research specifically directed to determining the link between attitudes and behavior toward the mentally ill. This link is the weakest aspect of the literature, yet, it is perhaps the most important area for future policy action.

One particularly important question for future research is what degree of acceptance is necessary to say that the mentally ill have been included in the community. Certainly no individual or group is perfectly accepted in society by everyone else. Do 95 percent of the people in the community have to accept the mentally ill as next-door neighbors for this subgroup to lead a satisfactory life, or will a substantially smaller percentage of the population suffice? The problem with attitude studies is that there is no baseline for assessing what degree of acceptance is necessary for people to live a fulfilling life in this society.

CONCLUSION

In the past 25 years society has begun to view the mentally ill in a more positive light. The general public is now in the position of having to interact with mentally ill individuals as they move into the community on an increasing basis to take up long-term residence. Educated contact with the mentally ill has proved beneficial in changing attitudes and in enabling the public to reduce the amount of social distance between themselves and the mentally ill. Yet the basic concept of mental illness prevailing in the mind of the public is that of a serious, unpredictable, dangerous disorder. That such a concept may have some basis in fact cannot be ignored, and is itself in need of research. However, it is essential that social work direct future efforts to helping the public find some specific basis for its attitudes toward the mentally ill, so that the unjustified negative implications of a stereotype can be avoided and ultimately eliminated.