Synopsis of the Derogatis Stress Profile (DSP)®
Derogatis, Leonard R.
Derogatis Stress Profile (DSP®)
Clinical Psychometric Research Inc., 1228 Wine Spring Lane, Towson, MD 21204 and MAPI Research Trust, 27 rue de la Villette 69003 Lyon, France
Year Developed: 1980
Primary Measurement Constructs: Stress, based on an interactional stress theory paradigm which holds stress to be a phenomenon arising from a dynamic interaction between environmental events, personality characteristics and emotional responses.
The DSP® is a 77-item self-report inventory derived from interactional stress theory. According to this model, stress is comprised of three interactional components: environmental events, personality mediators, and emotional responses. These three components are the specific referents in interactional theory for the more general categories of environmental stimuli, person mediating characteristics, and individual responses. Some stress researchers have attempted to operationally define stress in terms of only one of these components (e.g., life events, Type A personality, psychological symptoms); however, interactional theorists have effectively argued that these domains must be appreciated and measured interactively to achieve a meaningful definition of the stress construct.
The design of the DSP® uses a hierarchical model to represent stress as an interactive construct: eleven primary stress vectors (dimensions) are subsumed under three principal stress components (domains), which cumulatively provide a quantitative overall summary estimate (global stress score) of the respondent’s current stress level. Figure 1 provides a graphic representation of the DSPÂ®’s hierarchical construct structure.
The DSP® paradigm holds that stress vectors from the three primary stress-inducing life event areas (i.e., domestic, vocational and health) are mediated by five critical personality characteristics and the quality of the individual’s emotional response, which itself becomes an aspect of the stimulus environment, to result in the phenomenologic experience of stress. Stress vector burdens arising from the environment are either magnified and enhanced or deflected and diminished by personality mediators and the nature and magnitude of the emotional response (i.e., affective, cognitive and physiologic) that they ultimately elicit.
Although the precise nature of the relationships between the principal stress domains has not been definitively established, it is conservatively represented as additive in the DSP®. This strategy enables, after appropriate statistical scaling, summation across appropriate stress dimensions to achieve the three principal DSP® domain scores, and subsequently to sum across domain scores to achieve a single global definition of stress– the Total Stress Score (TSS). The DSP® also features a Subjective Stress Score (SSS), measured in analogue fashion, to provide an estimate of the respondent’s conscious appreciation of his/her current stress level.
A distinct advantage associated with the DSP® is that it operationally defines stress at three distinct but related construct levels: primary stress dimensions, secondary stress domains and as a tertiary global stress construct. The secondary domain constructs are those specified by interactional theory, and represent the fundamental dynamic elements comprising the DSP®’s interactional definition of stress. The 11 primary stress vectors all represent constructs consistently identified as stress-inducers, stress-mediators or stress-indicators, and were developed based on comprehensive research in the areas of life events, personality and psychopathology (Derogatis, 1987; Derogatis & Coons, 1993).
Definition of the Primary DSP® Dimensions
Vocational Environment- Vocational environment focuses on the individual’s work environment as one of the three primary stress-inducing areas of life. Although typically a significant source of self-worth and self-esteem, the work environment is a frequent source of stressful life events which may be of an acute or chronic nature. Success or failure in the work role plays a significant part in most individual’s feelings of well-being.
Domestic Environment- The Domestic Environment represents the second principal life experience that is critical to healthy psychological adjustment and feelings of well-being. Like the vocational sphere, domestic life carries the potential for extremely disruptive interpersonal conflicts and high stress-induction. Relationships with spouses, parents, children, neighbors and extended family all possess the capacity for significant conflict, and resultant high levels of stress. Specific developmental epochs tend to share common difficulties e.g.,”adolescent rebellion”, “midlife crisis”, “geriatric decline”, as well as those that are specific to the particular epoch at hand.
Health Environment- The Health Environment probably possesses the potential for the most profound and pervasive influence on our psychological adjustment and sense of well-being than any other aspect of life experience. Disease and illness carry with them significant burdens for coping and adjustment, since all serious illnesses degrade quality of life and have high potential for stress-induction. Although we do not possess definitive knowledge of health-optimizing behaviors, evidence does exist concerning the impact of some behaviors and attitudes on health. By assessing an individual relative to these behaviors, postures and practices the Health Environment dimension is designed to measure stress arising from this important and central aspect of life.
Time Pressure- The concept of “time pressure” has become synonymous in contemporary culture with performance, evaluation and task deadlines, and thereby, with stress. Time pressure induces stress by functioning as a multiplier for the demand characteristic of the situation: Not only must the individual accomplish the task accurately, but must do so in a conspicuously limited, perhaps insufficient, amount of time. The imposition of a time constraint “ups the ante” in terms of the physical, psychological and emotional resources that must be brought to bear to accomplish a task.
Driven Behavior- Frequently seen as a concomitant personality characteristic of time pressure, driven behavior is nonetheless a distinct personality trait. Driven behavior represents a compulsive need to be involved in activities that the individual views as broadly constructive, and the pattern of behaviors which he/she employs to satisfy that need. The self-imposed, demanding criterion of constant, tangible accomplishment is required by these individuals to achieve feelings of consistent well-being and self-worth.
Attitude Posture- Attitude posture is highly related to the “achievement ethic”. The idea that “good” members of our society achieve tangible accomplishments is interwoven throughout our social fabric. Some individuals extend the concept further, however, behaving as though if tangible accomplishment is equated with “good”, then more accomplishment must make them “better”. For these people there can never be enough productivity. Unfortunately, these individuals can rarely enjoy the fruits of their labors, having to quickly focus on the next set of potential achievements. This characteristic promotes the feeling of being “caught in a ratrace”, from which the person can never seem to escape.
Relaxation Potential- As the label implies, Relaxation Potential refers to the individual’s capacity to become involved in healthy diversions from the daily demands of work, family and community. Sometimes referred to as “hedonic potential”, this characteristic captures a person’s capacity to engage in stress-deflecting activities that are perceived as “fun”. Contemporary research has consistently shown that the greater and varied the number of activities an individual enjoys participating in, the more resistant to cumulative stress burdens that individual will be.
Role Definition- Role Definition, as the intrapsychic and interpersonal representation of self-concept, is a personality characteristic comprised of both private and public components. Role-defining personal characterizations probably exist on a stress-related continuum; however, a number of consistently stress-inducing definitions have been identified. Of these, the most clearly defined is the “atlas syndrome”. These individuals define themselves as the maximum “can do” person; they delegate little authority to others and represent themselves as “indispensable”. They are “all business” and rarely request help from others, perceiving such actions as a sign of personal weakness and a fundamental betrayal of their identity. Individuals high on this personality trait tend to be lonely, alienated, chronically fatigued and unhappy, having never learned how to share either life’s demands or life’s pleasures.
Hostility- The Hostility dimension is the first of the three affect/symptom stress vectors that define the Emotional Response domain. Like all of the emotional response dimensions, Hostility reflects affective experiences arising from one of the fundamental human emotions–in this case anger. The items comprising the Hostility dimension were selected to measure thoughts, feelings and behaviors that are characteristic expressions of anger, and items were developed to reflect all three modalities of expression. In addition to explicit experiences of anger, the Hostility dimension is designed to capture other shadings of this emotional complex such as irritability, resentment and aggression.
Anxiety- The Anxiety dimension represents the second in the DSP®’s triad of Emotional Response measures. Its items are designed to reflect mild-to-moderate manifestations of anxiety. Items reflect characteristics such as tension, free-floating anxiety, worry, nervousness and apprehension. These expressions of anxiety were chosen because they represent relatively moderate, less clinically remarkable, manifestations. This strategy is consistent with our view of stress as a transitional phenomenon which can progress to more dramatic clinical states (e.g., formal anxiety disorders). Thus elevations on the emotional response vectors can signal an “at riskness” for the possibility of more serious emotional problems.
Depression- Clinical depression is well-established as the most prevalent manifestation of psychiatric disorder (Derogatis & DellaPietra, 1994; Derogatis & Wise, 1989), and depressed affect (i.e., “feeling down”) is a very common occurrence. The items comprising the Depression dimension reflect a broad spectrum of depressive manifestations, and like the other Emotional Response dimensions, are drawn from the “mild-to-moderate” range of symptoms. The Depression dimension assesses such experiences as fatigue, loss of interest, feelings of loneliness and lowered self-esteem. Suicidal ideation is included because of its paramount importance; however, more dramatic signs and symptoms of profound clinical depression (e.g., vegetative signs) are absent. The Depression dimension is not designed to enable clinical diagnosis; however, like the Anxiety subscale, high scores do signal a potential “at riskness” for more formal psychiatric conditions.
Fundamentally, norms provide interpretive points of reference regarding the meaning of psychological test scores. They deliver information on an individual’s relative standing in comparison to the standardization sample, and in the case of multidimensional tests, enable a meaningful comparison of a respondent’s scores on multiple attributes (e.g., levels of depression vs anxiety). Norms for the DSP® are based on approximately 1,000 community respondents who ranged in age from 18 to over 70. Slightly more men than women contributed to the sample, and all were employed at the time of evaluation. All DSP® norms have been developed in terms of Area T-scores, which represent normalizing area transformations. Area T-scores, unlike their linear counterparts, deliver accurate centile equivalents, and thereby enable precise actuarial representation and meaningful clinical interpretation of score profiles.
Clinical/Research Instrument, Self-Report
The DSP® is comprised of 77 items, seven (7) items for each of the 11 primary dimensions. The scale takes approximately 12 to 13 minutes to complete under normal conditions, although some individuals make require a few minutes longer. Each item of the DSP® is rated on a 5-point scale ranging from 0=”not-at-all true of me”, to 4 =”extremely true of me”.
Reliability of the DSP®
In the self-report modality of measurement, the two most indicators of reliability are measures involving consistency of items (homogeneity) and stability through time (repeated measures). The DSP® has been evaluated concerning both forms of reliability and has demonstrated very acceptable reliability coefficients. Test-retest coefficients are based on a small sample of employees (N=34) who presented to a corporate medical office with what were judged by the physician in charge to be stress-related disorders. The nature of presenting complaints ranged from general anxiety to headaches and gastrointestinal disorders. Patients were medically evaluated and assessed with the DSP® twice, 7 days apart. Test-retest coefficients ranged from a high of .92 for Time Pressure, to a low of .72 for Hostility, with the majority of values falling in the mid to high .80’s. The Total Stress Score (TSS) had a coefficient of .90.
Internal consistency coefficients for the DSP® were established on a large sample of 867 individuals, who were employed by twelve commercial companies. The majority of respondents held middle management positions, although individuals’ positions ranged from clerical staff to several CEO’s. Item homogeneity was demonstrated to be at acceptable levels for all DSP® primary stress dimensions, ranging from a high of .93 for Time Pressure
to a low of .79 for Vocational Environment. Internal consistency for the three principal domain scores was also quite good, ranging from .83 to .88.
Validity of the DSP®
Internal Relationships Among DSP® Scales: An ideal of multidimensional measurement is an instrument which demonstrates relative orthogonality among primary measurement dimensions, while simultaneously revealing moderate-to-high correlations between the dimensions and the test’s total score. Such a pattern of relationships among test subscales is highly desirable because it confirms the goal of designing relatively unique, uncorrelated components of the superordinate construct being assessed (e.g., stress). It also helps insure minimum measurement redundancy while increasing measurement breadth and sensitivity. In the case of the DSP®, we have approached such a pattern. Interscale correlations are relatively high between primary stress dimensions and their corresponding domains, and much more modest between dimensions and non-corresponding domains. As examples, the average r among the 5 Personality Mediator dimensions with the domain score was .71, while the average correlation of this set of measures with non-corresponding domains was .41 . The mean coefficient for Environmental Events dimensions was .70 with the corresponding domain, but only .39 with divergent domains. Dimensions comprising the Emotional Response domain showed a mean correlation of .80 with the Emotional Response score, but only .41 with other domains. At each level of the measurement hierarchy a reasonably non-redundant pattern of relationships was observed.
Confirmation of Dimensional Structure: An essential step in the construct validation of a multidimensional psychological measure is confirmation of its proposed dimensional structure. Such confirmation is generally achieved through factor analysis. For the DSP®, Derogatis (1987) reported a factor analysis of the instrument, at the level of the primary stress dimensions, based on a sample of 867 respondents. Hypothesized structure called for three factors, corresponding to the three principal stress domains of the instrument. Four factors were identified accounting for approximately 70% of the variance in the matrix. The initial factor was loaded almost exclusively by the five dimensions of the Personality Mediators domain, while the second factor clearly represented the Emotional Response domain. Hostility, Anxiety and Depression all showed substantial loadings on Factor II. The third factor identified had significant correlations with Vocational Environment and Domestic Environment. The third Environmental Event measure, Health Environment, did not correlate with this factor, instead, forming a unique, separate factor on which it revealed a highly saturated loading and accounted for approximately 8% of the variance in the matrix. In general, this analysis provides a substantial degree of corroboration for the hypothesized dimensional structure of the DSP®: 10 of the 11 dimensions conform to the designed structure of the test. It remains unclear as to why the Health Environment dimension did not load with the other Environmental Event measures. However, theorists currently posit health status to be one of the seven major dimensions of outcomes assessment (Docherty & Streeter, 1996), and it may turn out that elements of the perception of health are more orthogonal to other dimensions of life experience than previously believed.
Criterion Validity: Although contemporary validity theory (Messick, 1995) has reassigned criterion-oriented validation to the external aspect of construct validity, when most investigators and clinicians use the term “validity” it is this aspect of validation they are referring to. Criterion validity refers to the programmatic series of experiments which demonstrate patterns of correlations (i.e., empirical relationships) between test scores and external criteria (which themselves may be other test scores) that are consistent with and confirm the theory of the construct being measured. In the case of the DSP®, there is a growing body of research demonstrating empirical relationships with external criteria which confirm the test as a valid interactional measure of stress. The Principal Citations which follow contain references to much of this work.
How to Obtain
The DSP® is co-distributed by Clinical Psychometric Research, Inc. 1228 Wine Spring Lane, Towson, MD 21204, Phone: (410) 321-6165 and MAPI Research Trust, 27 rue de la Villette 69003 Lyon, France, eprovide.mapi-trust.org
Copyright & Trademark Owner
Derogatis Measurement Assessments, LLC
Derogatis, L.R. (1984). The Derogatis Stress Profile DSP®: Preliminary Administration & Scoring Manual. Baltimore, MD, Clinical Psychometric Research.
Derogatis, L.R. (1987). The Derogatis Stress Profile DSP®: Quantification of psychological stress. In G. Fava & T. Wise (Eds.), Research Paradigms in Psychosomatic Medicine, (pp 30-54). Basel, Karger.
Derogatis, L.R. (1982). Self-report measures of stress. In L. Goldberger & S. Breznitz (Eds.), Handbook of Stress: Theoretical and Clinical Aspects. (pp 270-294), New York, MacMillan.
Derogatis, L.R. & Coons, H.L. (1993). Self-report measures of stress. In L. Goldberger & S. Breznitz (Eds.), Handbook of Stress: Theoretical and Clinical Aspects, (Second Edition). New York, Free Press/MacMillan.
Derogatis, L.R. & DellaPietra, L. (1994). Psychological tests in screening for psychiatric disorder. In M. Maruish (Ed.), Psychological Testing in Treatment Planning and Outcomes Assessment. New York, Lawrence Earlbaum Associates.
Dobkin, P.L., Pihl, R.O. & Breault, C. (1991). Validation of the Derogatis Stress Profile using laboratory and real world data. Psychotherapy & Psychosomatics, 56, 185-196.
Duquette, R.L., DuPuis, G. & Perrault, J. (1991). A new approach for quality of life assessment in cardiac patients. Cardiovascular Medicine, 10, 106-112.
Lewandowski, A., Byl, N., Franklin, B., Gordon, S., Timmis, G.C. & Beaumont, W. (1987). Relationship of the Derogatis Stress Profile to staff perceptions of adjustment and prognosis of cardiac patients. Journal of Cardiopulmonary Rehabilitation, 7, 502-506.
Solis, S. (1991). Psychosocial Stress in Marine Corps officers. Military Medicine, 156, 223-227.
Sturdevant, J.R., George, J.M. & Lundeen, T.F. (1987). An interactional view of dental Student stress. Journal of Dental Education, 51, 246-249.