There are two clinical scales that are particularly relevant to the assessment of anxiety symptoms: Scale 7 (Psychasthenia or Pt) and Scale 0 (Social Introversion or Si). The former scale was developed by selecting items that differentiating a group of patients with Psychasthenia symptoms—a psychiatric syndrome characterized by obsessiveness, tension, and anxiety, from a group of nonpatient individuals (McKinley & Hathaway, 1942). Because the Psychasthenia group was small, Hathaway and McKinley (1942) also correlated the candidate scale with all remaining items in the MMPI pool to ensure a longer and more internally consistent scale. Later research supported that Scale 7 did indeed measure symptoms of anxiety, distress, tension, and self-doubt, but also a substantial amount of general maladjustment (e.g., Dahlstrom, Welsh, & Dahlstrom, 1972; Graham, 2011).

Scale 0 (Si) is a 70-item scale that was developed by Drake (1946) to assess social introversion/extroversion. Later research has supported the use of Scale 0 for assessing the individuals’ experiences in social situations, especially as it relates to introversion and social maladjustment, including anxiety (e.g., Sieber & Meyers, 1992; Ward & Perry, 1998). Scale 0 has three subscales intended to assist in clarifying clinically significant elevations, including Si1 (Shyness/Self-Consciousness), Si2 (Social Avoidance), and Si3 (Alienation—Self and Others; Ben-Porath et al., 1995), with the Si1 subscale being most sensitive to social anxiety (cf. Sieber & Meyers, 1992).

There are several RC scales germane to assessing anxiety symptoms. One of these scales is RC7 (Dysfunctional Negative Emotions), which focuses on core negative emotionality markers, such as anxiety, fear, guilt, and anger. Although RC7 is the scale that is considered most useful in the general assessment of anxiety symptoms, because they measure broad affective dimensions relevant to unipolar mood and anxiety disorders, RCd (Demoralization) and RC2 (Low Positive Emotions) should also be considered in the assessment of some forms of anxiety symptomatology. These scales were developed by Tellegen et al. (2003) who recognized the methods of scale creation used for the MMPI clinical scales resulted in substantial saturation of those scales with general maladjustment and distress common to most psychiatric disorders. Tellegen et al. (2003) suggested demoralization could be isolated and accounted for if conceptualized through Tellegen’s (1985) framework of positive and negative emotionality, where demoralization corresponds to the pleasantness–unpleasantness vector between these two orthogonal affective dimensions. Tellegen’s model links depression to low positive emotionality and anxiety to high negative emotionality. Therefore, as a first step in RC scale construction, demoralization markers were identified based on factor analyses of Clinical Scales 2 and 7. Demoralization items were then factor analyzed with item from respective clinical scales to derive distinct core components for each of them. A final set of nonoverlapping RC scales were developed from these core components.

Butcher et al. (1990) developed four content scales that are relevant to the measurement of anxiety symptomatology: Anxiety (ANX), Fears (FRS), Obsessiveness (OBS), and Social Discomfort (SOD). The FRS and SOD scales have two content component scales (FRS₁: Generalized Fearfulness and FRS₂: Multiple Fears; SOD₁: Introversion and SOD₂: Shyness) that can assist with further delineation (Ben-Porath & Sherwood, 1993). Among the SOD component scales, Shyness is more sensitive to social anxiety (Ben-Porath & Sherwood, 1993). Each of these scales were rationally derived by identifying target constructs, nominating test items to measure each construct, and having judges rate the degree to which items indeed reflected the intended construct. Statistical approaches were also used to refine the scales’ internal reliability.

There are two supplementary scales, originally developed for the MMPI and later revised for the MMPI-2, that were developed to assess anxiety. Welsh’s Anxiety (A) scale (Welsh 1965; Welsh and Dahlstrom 1956) assesses the largest factor emerging from factor analytic studies of the MMPI, which is generalized maladjustment. Despite the scale’s label, it is not a specific measure of anxiety at all, but rather captures the demoralization component described earlier. Indeed, the correlation between Welsh’s A and the RCd scale is greater than 0.90 in most clinical samples (e.g., Rouse, Greene, Butcher, Nicholas, & Williams, 2008). It is not recommended for use in the assessment of anxiety. The second scale is Keane, Malloy, and Fairbank’s (1984) PTSD scale (PK), which was developed in a Veterans Affairs sample by identifying items that differentiated a group of carefully diagnosed psychiatric patients with PTSD from those without. Notably all PTSD patients had some form of combat trauma.

There is one PSY-5 scale that is relevant to the assessment of anxiety: Neuroticism/Negative Emotionality (NEGE). This scale was designed to measure a dispositional tendency to experience a wide range of negative emotions, including anxiety and fear, particularly with a pathological range (Harkness et al., 1995). These authors developed the NEGE scale using a combination of rational-replicated selection, where lay raters were asked to deduce which MMPI-2 items reflected the PSY-5 constructs (including NEGE), and a series of rationale and statistical refinements.

In the following sections, we provide conceptual and empirically informed guidelines for how to use the MMPI-2, and in particular, the scales just described, in the assessment of anxiety symptoms. Rather than strictly adhering to the categorical perspective outlined in the American Psychiatric Association’s current edition of the Diagnostic and Statistical Manual (DSM-IV-TR, 2000), we group anxiety symptoms in accordance with a variety of empirically supported models of anxiety symptoms. In general, these models emphasize a broad general negative emotionality/demoralization factor, in addition to specific anxiety factors (e.g., Brown, 2007; Watson, 2005; Zinbarg & Barlow, 1996). Further, some research has indicated that depression and general anxiety are genetically indistinguishable (Kendler, 1996) and that they are structurally (both genotypically and phenotypically) separate from other anxiety disorders, particularly those with a fear-based component (e.g., Kendler, Prescott, Myers, & Neale, 2003; Krueger & Markon, 2006). Thus, the domains we emphasize here are negative affect/generalized trait anxiety, post-traumatic stress, social anxiety, obsessive-compulsivity, and phobic fear. Although an argument could easily be made for a separate panic or interoceptive anxiety domain, the MMPI-2 instruments are unlikely to capture these types of symptoms specifically.

The MMPI-2-RF is a 338-item true/false self report inventory, which is conceptually and empirically linked to contemporary models of personality and psychopathology. The test uses the non-gendered version of MMPI-2 normative sample (Ben-Porath & Forbey, 2003). The standard scales of the MMPI-2-RF include 8 Validity Scales, 3 Higher-Order (H-O) Scales, 9 Restructured Clinical (RC) scales, 23 Specific Problem (SP) scales, 2 Interest Scales, and 5 revised Personality Psychopathology Five (PSY-5) Scales. A majority of the scales of the MMPI-2-RF are organized in a hierarchical fashion with higher-order scales at the top, RC scales in the middle, and SP scales at lowest level.

The eight validity scales consist of seven revised versions of scales from the MMPI-2, as well as one new scale designed to assess over-reporting of somatic symptoms. The three higher-order scales represent measurement of the broad domains of psychopathology related to internalizing, externalizing, and thought disturbance that have been consistently identified in the empirical literature, including large epidemiological studies (e.g., Kotov, Gamez, Schmidt, & Watson 2010; Krueger & Markon, 2006; Vollebergh et al., 2001). The nine restructured clinical (RC) scales are identical to their MMPI-2 counterparts described earlier. The 23 specific problems (SP) scales were developed primarily not only to assist in clarifying H-O and RC scale interpretation but also to measure clinical and personality domains not sufficiently covered by the RC scales. The SP scales include five Somatic/Cognitive, nine Internalizing, four Externalizing, and five Interpersonal scales. The two interest scales measure two distinct core components derived from the original MMPI-2 clinical Scale 5 (Masculinity/Femininity)—Physical/Mechanical and Aesthetic/Literary interests. Finally, the MMPI-2-RF includes a set of revised PSY-5 scales (Harkness & McNulty, 2007; Harkness et al., 2002) that are identical in domain coverage to their MMPI-2 counterparts.

The scales within the internalizing hierarchy would be most useful to assessing anxiety symptoms and are listed and described in Table 10.1. This table also includes reliability information derived from the MMPI-2-RF Normative sample. Because the RC scales on the MMPI-2-RF are identical to those featured on the MMPI-2, and the PSY-5 scales are very similar, we will only describe the Specific Problems (SP) scales associated with anxiety symptoms here. However, applied recommendations are provided for all MMPI-2-RF scales applicable to the assessment of anxiety later in this chapter.

It should be noted that the MMPI-2-RF is a relatively new instrument, and with the exception of the RC scales, there are few peer-reviewed studies on its scales that are focused on the measurement of anxiety symptoms. However, the MMPI-2-RF Technical Manual (Tellegen & Ben-Porath, 2008) provides extensive data supporting the validity of MMPI-2-RF scale scores using samples from diverse settings (e.g., outpatient and inpatient clinical, forensic, medical) with various criterion modalities (e.g., self-report, therapist ratings, etc.). In the following, we use the same framework for applied recommendations as we did for the MMPI-2.