Related Pages -
- Who can use the DAS?
- Psychometrics of the DAS24
- Psychometrics of the DAS59
- Example items of the DAS24 and DAS59
There have been four distinct phases in the DAS story -
- Development of the original, 136 item version (DAS136),
- Refinement of this to produce a factorial scale (DAS59),
- Development of the short form (DAS24),
- Study of a structured sample of the general population to determine prevalence and types of concerns about appearance and data for norm tables.
Design and development of the initial DAS136 experimental scale
Initially, we designed an experimental scale (DAS136) composed of 136 items sub-divided into three sections. Items were based on phraseology of respondents to an autobiographical study of a large and representative sample of plastic surgery patients with disfigurements and aesthetic problems of appearance (Harris, 1982), relevant literature and clinical experience.
In a postal survey, a sample of untreated plastic surgery patients from the NHS waiting list, Derriford Hospital, Plymouth returned completed DAS136 scales twice with a three-month interval. The three sections of the new scale intercorrelated strongly and good criterion validity was demonstrated by moderate, positive correlations (.74 to .62) with social anxiety (Social Anxiety and Distress Scale, Watson & Friend, 1969). Good convergent construct validity was shown by positive correlations with state anxiety (0.54 – 0.43) and trait anxiety (0.58 0.48) (Strait-Trait Anxiety Inventory, Spielberger, Gorsuch, Vagg, & Jacobs, 1983), neuroticism (0.50 0.44) (EPQ; Eysenck & Eysenck, 1991) and depression (0.63 to 0.51) (Beck Depression Inventory, Beck & Steer, 1987). Divergent construct validity was shown by negative correlations with extraversion (-0.27 – -0.19) (EPQ) and by independence from hysteria (0.17 .013) (Crown Crisp Experiential Inventory, Crown & Crisp, 1979) (Carr, Harris & James, 2000).
Our study also demonstrated the validity of the new scale as a measure of the effectiveness of plastic surgery. Pre-operatively, mean DAS136 full-scale scores for plastic surgery patients were significantly higher than those for surgical control patients (p<.0001) and similar to mean scores of waiting list patients. Post-operatively, whilst mean scores of surgical controls and waiting list patients remained stable between T1 and T2, there was a large and significant reduction in mean scores of plastic surgery patients (p<.0001).
These results clearly indicated that the new scale had excellent potential to provide an appropriate instrument to meet our aims. Feedback from the plastic surgery patients was strongly positive regarding the scales’ content, which they judged to encompass satisfactorily all relevant aspects of appearance-related distress and dysfunction. We therefore planned to refine the scale and produce two user-friendly versions:
A factorial scale to provide detailed description and measurement (the DAS59).
A short non-factorial scale to measure the same construct (the DAS24).
Development of the DAS59: The factorial version of the scale
We developed a large database of DAS136 item scores of 606 respondents to a postal survey of adult plastic surgery patients on the NHS (National Health Service) waiting list at Derriford Hospital, Plymouth. The sample was composed of patients awaiting treatment of aesthetic problems of appearance (eg. large/small breastedness, breast asymmetry, prominent ears, problems with the nose), patients awaiting treatment of congenital and acquired disfigurements and deformities (eg. cleft lip, haemangiomas, scarring from disease, injury or surgical therapies) and patients awaiting treatment of conditions not primarily concerned with appearance (eg. hand surgery).
Items of the three sections of the DAS136 were investigated first by within-scale item analyses, retaining items with item-whole correlations of 0.5 or higher and by within-scale factor analyses (principal components and varimax) retaining items that loaded higher than 0.5. From the retained items, those, which duplicated similar clinical information in different sections of the scale, were identified and those with the higher item-whole correlations and greater face validity on the basis of clinical experience were retained. Finally, the items, which had not been retained, were reviewed and some with particularly good face validity for a clinical population (N = 3), were identified for inclusion in the refined item set. Of the 136 original items, 57 were thus retained. Their scores were extracted from the original 606 data sets and their totals were found to correlate highly with totals of the original scale (r=0.98).
The refined set of 57 items was reorganized into a single scale. As many disfigurements can be associated with physical discomfort and difficulty, two new items were added (items 25 and 26) to assess these dimensions and to improve the acceptability of the refined scale for clinicians to produce the Derriford Appearance Scale 59 (DAS59) (Carr, Harris & James, 2000).
Development of the DAS24: A short form of the scale
The items were selected from the DAS 136, to include as broad a range of the behavioural and emotional responses to being visibly different as possible. Through a series of trials at Derriford Hospital, Plymouth, survivor items were selected for the DAS 24. This was on the basis of both their psychometric properties and inclusiveness of the range of clinically observed phenomena. In the trials below, 26 items were initially selected to as the best ‘survivors’ of previous intensive analysis, and as representatives of the factors evinced in the development of the DAS 59 (Carr, Harris and James, 2000. Two items did not perform adequately to enable their inclusion in the final version of the scale, hence twenty four items and the ‘DAS 24’.
Data were collected from a large scale multi centre trial in the UK, in which 535 participants provided data. These were recruited from plastic surgery and burns clinics. All participants were adults (18 years or over) who were literate in English. Two recruitment routes were selected. Approximately half the participants (n=271) were recruited in routine out-patient clinics. The remainder (n=264) were recruited through waiting lists. One hundred and forty seven (27.5%) were men, and 388 (72.5%) women. The range of aetiologies was fully represented, including congential malformation (e.g., cleft lip, haemangiomas), scarring from trauma and burns, disease (e.g., skin cancer, acne), developmental growth (e.g., breasts, nose), obesity, weight loss, effects of pregnancy and breast feeding, and facial ageing. Alongside the DAS 24, these samples also completed a range of appropriate scales to facilitate evaluation of the validity of DAS 24.
General population study
We randomly selected 6,000 males and females from the general population of South West Devon, UK (the locality of Plymouth) with constraints for age (18 years and over), gender (even representation) and socio-economic status (based on postal codes). We mailed 3,000 a copy of the DAS59 and 3,000 a copy of the DAS24. The response rate of completed scales was 40%. Thirty-five percent male respondents and 61% female respondents were concerned about an aspect of their appearances (see below). Among males, the prevalence of concern was highest in the age band 18-21 years (56%) and fell off progressively with increasing age. Among females, the prevalence of concern remained high (63%-69%) through to age 60 years. Even above the age of 60 years, a significant proportion of men and women were concerned about appearance (21%-33%). We found no association between prevalence of concern about appearance and either socio-economic status or living status.