Invited Symposium: Development of Social Phobia



Materials & Methods


Discussion & Conclusion



INABIS '98 Home Page Your Session Symposia & Poster Sessions Plenary Sessions Exhibitors' Foyer Personal Itinerary New Search

Toward an Improved Nosology of Social Phobia: Dimensional or Latent Class?

Contact Person: Jonathan M Oakman (jmoakman@watarts.uwaterloo.ca)


The DSM-IV1 defines social phobia as a "marked and persistent fear of one or more social or performance situations in which the person is exposed to unfamiliar people, or to possible scrutiny by others. The individual fears that he/she will act in a way, or show anxiety symptoms that will be humiliating or embarrassing". Situations typically avoided by social phobics include public speaking; speaking at meetings or in small groups; giving a presentation; meeting strangers; attending social gatherings; eating, writing or working in front of others; and, dealing with authority figures.

The core features associated with social phobia are a fear and avoidance of situations that have a potential for either scrutiny or evaluation by others. However, the clinical presentation of social phobia varies widely with respect to the number of feared situations, severity of phobic avoidance, and degree of impairment attributable to the symptoms24. There are currently a number of different ways of understanding this diversity. It has been suggested that the number of different situations feared by individuals with social phobia may be characteristic of two subgroups of social phobic patients55. This distinction is reflected in the specification in the current DSM system of a generalized subtype of social phobia that is thought to be quite different from a more circumscribed or "nongeneralized" type. There has also been considerable interest in the degree to which fears of different types of situations might characterize different subgroups of social phobia. Several researchers have suggested an important distinction between concerns related to performance and concerns related to social interaction34,36,62. Alternatively, Schlenker and Leary54 have proposed a distinction between fear of social situations in which people's behaviours follow from the responses of others (contingent interactions) and fear of social situations in which people's responses are primarily guided by internal plans or scripts (noncontingent interactions).

In addition to the subtypes of social phobia based on number and types of reported fears, a number of other distinctions may be possible. For example, reporting on a direct interview family study of social phobia, Fyer et al.17 noted that first degree relatives of people with both social phobia and comorbid panic disorder had a risk profile similar to that of relatives with panic disorder only. Relatives of people with both social phobia and panic disorder were at increased risk for panic disorder but not social phobia. These results suggest that social phobia with comorbid panic disorder may confer a different risk to family members than social phobia without comorbid panic disorder. If such is indeed the case, there may be some utility in making a distinction between "somatic" (social phobia with panic) and "cognitive" (social phobia without panic or social phobia with obsessive compulsive disorder) forms of social phobia.

Social phobia is reported to have an early age of onset as compared to other anxiety disorders and it would appear that the mean age of onset for social phobia is in mid-adolescence55. However, Schneier et al.55 reported peaks at the interval 0 to 5 years and at age 13. There may be interesting and important differences between "early" as compared to "late" onset social phobia. Finally, it may be useful to consider a distinction between social phobia which is the result of direct traumatic conditioning, in which fear develops as the result of exposure to a traumatic event and social phobia that results from chronic exposure to a variety of vicarious learning and information transfer experiences42,44,45.

Types of Shyness

The various typologies of social phobia described above are at present rather speculative. However, there is converging work from the field of developmental psychology on different types of shyness that may help to further our understanding of extreme social anxiety of the kind reported by people who have the DSM-IV diagnosis of generalized social phobia.

Buss and Plomin's temperament theory9,10 provides one possible explanation for understanding social behaviour generally, and social phobia in particular. Buss and Plomin's9 original model proposed that there were four temperaments: Emotionality, Activity, Sociability, and Impulsivity. The current model no longer includes Impulsivity. According to this conceptual scheme, shyness is a derivative of the temperament of Sociability, but is also highly correlated with fearfulness (a component of Emotionality). Most important to the present purpose, shyness can be divided into two main types: fearful shyness and self-conscious shyness.

Fearful shyness is thought to develop during the first year of life7,8, and be based on wariness of strangers. Fear of strangers is extremely common in childhood, and typically develops at about the age of eight months56. While fear of strangers seems to fade for some children, some adults continue to experience fear when meeting unfamiliar people.

Understanding of the fear of strangers has been advanced by those researchers who investigate the broader construct of "behavioral inhibition." Kagan and his colleagues49 have described the temperamental construct of behavioral inhibition to the unfamiliar as "a temperamentally based disposition of children to react consistently to unfamiliar events, both social and nonsocial, with initial restraint," (p.301). When confronted with novelty, some children react cautiously, perhaps withdrawing or behaving in a fearful way. This pattern of response is called behavioural inhibition. Other children react to novel situations quite differently, either showing no change in previous behaviour, or actually approaching novel situations. This pattern of response is referred to as uninhibited behaviour.

As might be expected of a temperamentally based disposition, behavioural inhibition is reported to be quite stable across childhood20,21,26,48,50,52. In addition, Kagan has found several diverse physiological indicators of behavioural inhibition, including a high and stable heart rate, laryngeal muscle tension resulting in low variability in vocal pitch, high norepinephrine levels, high salivary cortisol levels, and pupillary dilation30. Such a constellation of physiological signs may be suggestive of a greater than normal sympathetic nervous system reactivity30. In summary, there appears to be a remarkable stability of anxious behaviour in the face of novelty and its associated physiological markers. This has prompted several researchers to suggest a genetic predisposition to inhibited behaviour which provides a basis for a possible link between behavioural inhibition and anxiety disorders6,51.

Self-conscious shyness is thought to differ from fearful shyness in several important ways. It is thought to develop much later than fearful shyness because it is based on the capacity for reflexive self-awareness. Fearful shyness develops in early infancy, prior to the development of a sense of self, and is essentially an alarm reaction to social novelty or intrusive approach. In contrast, self-conscious shyness can only develop after children acquire the meta-representational capacity to be aware of themselves as as they might appear to others, a capacity which is thought to develop as late as the fourth or fifth year of life. According to Buss7,8, public self-awareness involves a focus on aspects of the self that are publicly observable, such as ones appearance, actions, manners, speech, odour, and so on. Self-conscious shyness is an outgrowth of public self-awareness, and it occurs as a result of chronic awareness of oneself as a social object accompanied by negative thoughts about the self. When public self-awareness is acute, the person feels conspicuous and embarrassed.

As the result of a series of investigations on types of shyness in children and adults Asendorpf2,3,4 has described two different types of childhood shyness - one related to social novelty, and one related to social evaluation. In a four-year longitudinal study of children, Asendorpf3 found that shyness with strangers is largely unrelated to shyness with familiar peers, and that the correlation between stranger fears and social-evaluative shyness decreased with age. Interestingly, fear of social evaluation is associated with internalizing difficulties in childhood, while fear of social novelty (strangers) is not3. In a series of studies on adults, Asendorpf2 systematically varied two properties of a social situation: the novelty of an interaction partner, and the degree to which the situation was social-evaluative in character. Asendorpf2 found that social-evaluative shyness and shyness towards strangers were additive in their ability to predict the experience and behavioural manifestations of social anxiety. Taken together, these results imply that stranger fears and social-evaluative shyness are meaningfully distinct constructs that are additive in their ability to predict social anxiety in adults.

At first glance, Asendorpf's3 two component model of social anxiety seems quite similar to Buss and Plomin's10 two types of shyness and it is tempting to simply identify stranger fears with fearful shyness and social-evaluative shyness with self-conscious shyness. However, Asendorpf3 is careful to note the differences between the two models. Buss7,8 makes a distinction between social evaluation and public self-awareness and argues that public self-awareness could be caused by something as simple as being conspicuous, whereas social evaluation has more to do with evaluative situations such as dating or a job interview, where the relevance of interpersonal scrutiny is more clear. As a result, Buss7,8 assigns social evaluation to the super-ordinate category of fearful shyness, and proposes that self-conscious shyness is the result of public self-consciousness. In contrast, according to Asendorpf3 social evaluative shyness is the result of sensitivity to social evaluation or to public self-consciousness, whereas fearful shyness is equivalent to sensitivity to social novelty.

From Subtypes to Typologies

There is currently no agreement about the relationship between shyness and social phobia. Although they appear to be very similar - both being defined by social discomfort and inhibition - there may be something distinctly different about people who develop social phobia from those who are simply shy. However, in a review of the available evidence regarding somatic responses, cognitive characteristics, behavioral responses, daily functioning (and its opposite - impairment), clinical course, and onset Turner et al.61 noted several striking similarities between shyness and social phobia. According to Turner et al.61, the main differences between shyness and social phobia concern "onset and course," and the importance of avoidance behaviour. Considerable evidence suggests that shyness develops early in life, and people may simply "grow out" of it64. In contrast, the onset of social phobia is generally in mid-adolescence, and its course tends to be chronic and unremitting. Finally, phobic avoidance is one of the defining features of social phobia; however, interview studies of shy people indicate that behavioral avoidance is rated as one of the least important aspects of shyness46.

While there are some important features that distinguish shyness and social phobia, such differences could be parsimoniously interpreted as a matter of degree47. It is quite possible, for example, that avoidance is characteristic of only very shy individuals who would be diagnosed as socially phobic. Similarly, the considerable evidence that social phobia is a far more impairing condition than shyness could be interpreted as simply the result of more extreme shyness. It is also possible, however, that there is something that makes social phobics distinctly different from highly shy people. This is the nature of the dimensional versus typological distinction - is the difference between people with social phobia and those slightly less shy a matter of degree, or is the difference a more typological difference in kind?

The results of research in developmental psychology described above may be useful in furthering a typological understanding of the extreme social anxiety reported by people who have the DSM-IV diagnosis of generalized social phobia. For example, two processes, either fearful and self-conscious shyness8 or fear of social novelty and fear of social-evaluation3 could combine in an additive, dysjunctive, or synergistic way to produce social phobia. If social anxiety is the result of two distinct processes that interact to produce more than the sum of their parts (synergy), then social anxiety might be best conceptualized as a typological variable. It would then follow that people high on both types of shyness would be importantly (and quantitatively) quite distinct from those high in only one type of shyness. While this idea is provocative, Asendorpf3 clearly favours an additive model, and his results are inconsistent with either a dysjunctive or synergistic model.

Alternatively, Buss & Plomin10 and Asendorpf3 have suggested behavioural inhibition as the foundation stone of fearful shyness. Kagan's27 work implies that behavioural inhibition may be typological, insofar as people who exhibit stable behavioural inhibition are different from other people. This distinction may be subserved by fundamental neuropsychological differences between people involving the septohippocampal system. If behavioral inhibition is typological, then the best measurement model for social anxiety may be one that accounts for a dichotomous (typological) latent trait reflecting fearful shyness, and a continuous latent trait reflecting social-evaluative shyness.

It is possible to consider a variety of other typological understandings of social phobia. For example, according to Asendorpf's3 conceptual scheme it is possible that people with social phobia could be high on either stranger fears, social-evaluative fears, or both. However, given Asendorpf's3 findings regarding the apparent lack of association between stranger fears and internalizing problems in childhood, we might expect that people with social phobia are either high on social-evaluative shyness, or high on both types of shyness. In contrast, the two-component model of Buss7,8 is consistent with the idea that there may be a type of social phobia based largely on behavioural inhibition (the fearful subtype), and a second type that is largely the result of high self-conscious shyness (the self-conscious subtype). Given Buss's8 theory it may also be possible for a person with social phobia to be high in both fearful and self-conscious shyness. Finally, Kagan's27 work on behavioural inhibition, and the extension of it to clinical anxiety problems by Rosenbaum51 and Biederman6, suggests that people with social anxiety may be either high on fearful shyness, or high on both types of shyness. In summary, there is a range of possibilities for the relationship between a two-component model of shyness and the clinical diagnosis of social phobia. Asendorpf's3 model implies that social phobia could be a disorder of painful self-consciousness, or both self-consciousness and fearful shyness. Buss's7,8 theory implies that social phobia could be the result of either fearful shyness or self-conscious shyness, or both, and Kagan's27 work implies that social phobia is a derivative of fearful shyness.

Distinguishing Types from Dimensions: Taxometric Methods

What does it mean to propose that social phobia may be a taxonomic class which is distinct from extreme shyness? In descriptive terms, we might say that such a hypothesis supposes that people with social phobia are qualitatively different from people who are very shy, and they are different from others in more ways than simply being more shy. In other words, having social phobia has implications beyond simply being more shy than people who are simply very shy. These descriptive statements can be readily translated into a measurement model for a taxonomic class. According to Golden23, the measurement model for a typology involves a latent variable (factor) that can assume only two values (whereas conventional latent variables are continuous). The two values of this latent variable are 1 and 0 (type member versus type nonmember).

While the measurement model for a typology is quite different from the measurement model for dimensional personality traits, in practice it is very hard to distinguish between the two. Intuitively, it would seem that a distribution of scores from a measure of a typological trait would be bimodal, while a distribution of scores of a dimensional (continuous) trait would be unimodal. Unfortunately, a bimodal distribution can be produced from a measure of a dimensional trait, and a measure of a typological trait can be unimodal5. The shape of a distribution of scores may be suggestive of a typology or a dimension, but is actually indeterminate.

Conventional methods used for numerical taxonomy such as factor analysis and cluster analysis are also unsuited to distinguishing between scores derived from a typological versus a dimensional measurement model38,41. Instead, Meehl39,40 has advanced a set of statistical procedures that are suited for this sort of typological problem. These methods operate on a set of presumed indicators of a typology, detecting whether the indicators are based on a typology or a dimension. Meehl's39,40 method begins with a conjectured latent typology. The question is not how many groups or clusters there may be, but rather whether there are two groups or just one. The critical distinction is between a latent typology versus a latent dimension as being responsible for a distribution of scores. If a dimension underlies the scores, then the method should report two groups - a group of type members, and a group of type nonmembers. These new taxometric methods provide a means for evaluating the typological question at hand. It is possible, as many researchers suggest13,33,57, that social phobia is the same thing as extreme shyness. Alternatively, dimensional measures of shyness (useful in the general population) may obscure a latent typology, where generalized social phobia is a clinical condition of shyness that is typologically distinct from simply being very shy, as implied by Turner et al.61and the DSM-IV system1.

In order to evaluate the relative utility of a dimensional as compared to a typological model of social phobia, data from the Revised Cheek and Buss Shyness Scale (RCBSHY)12 and a self-report version of the Liebowitz Social Anxiety Scale (LSAS)34,53 were analyzed using Meehl's MAXCOV-HITMAX taxometric procedure37,39,41. The RCBSHY was selected because it is a well-validated and frequently used instrument for measuring shyness. The LSAS was selected because it is a well-validated and widely used instrument for assessing social phobia25. If generalized social phobia identifies a latent typology characterized by extreme shyness, then the method should report a typological result for both the RCBSHY and LSAS data. If generalized social phobia simply identifies an extreme on a dimensional trait of shyness, the method should not report a latent typology on either of the measures. Alternatively, it possible that shyness is dimensional and social phobia is typological. If such is the case, one would expect the method to report a latent typology for the LSAS but not for the RCBSHY.

Back to the top.

<= Abstract INTRODUCTION Materials & Methods =>

| Discussion Board | Next Page | Your Symposium |
Oakman, JM; Van Ameringen, M; Mancini, C; Farvolden, P; (1998). Toward an Improved Nosology of Social Phobia: Dimensional or Latent Class?. Presented at INABIS '98 - 5th Internet World Congress on Biomedical Sciences at McMaster University, Canada, Dec 7-16th. Invited Symposium. Available at URL http://www.mcmaster.ca/inabis98/ameringen/oakman0804/index.html
© 1998 Author(s) Hold Copyright