The assessment of emotional distress experienced by people with an intellectual disability: a study of different methodologies

Abstract

The assessment of emotional disorders such as anger, depression and stress among people with an intellectual disability has traditionally used one of three methodologies: ratings by a significant other, a clinical interview or self-report. Despite the widespread use of all three methodologies, there is little research into their equivalence. This paper assesses the convergence among these three approaches for 147 people with a mild or moderate intellectual disability across the affective domains of anger, depression and stress. The results showed the overlap among the three methods to be consistently low, although limited convergence was found between self-report and clinical interview. Ratings by work supervisors discriminated least clearly between anger, depression and stress while self-report was the most discriminating between these three overlapping but conceptually distinct states. Suggestions are made for ongoing research into the methodologies of assessing affective states among people with an intellectual disability.

Introduction

The systematic study of emotional disorders among people with an intellectual disability is a relatively recent phenomenon with most studies reporting the presence of affective symptomatology. A number of studies in the 1980s found that people with an intellectual disability experience at least as high, and perhaps higher, rates of emotional disturbance than the general population. On the basis of a survey of the research literature, Benson and Reiss (1984) estimated emotional disorder prevalence rates among people with an intellectual disability to be somewhere between 15 and 20% of the population. Lund (1985) put the figure as high as 30%. Iverson and Fox (1989) reviewed the literature and found incidence rates of emotional disorders varied greatly, ranging from 13 to 47%. In their own sample of 165 intellectually disabled adults, 35% reported at least one psychopathological disorder, although Iverson and Fox noted that reliable measurement was difficult with this population.

The varying incidence rates reported for these disorders may be largely a consequence of the variation in assessment techniques and measures used by researchers. When dealing with a population where measurement has always been difficult, different methodologies and measures are not just a matter of convenience or preference - as they sometimes are with the general population. There is a need for research into the strengths and limitations of assessment methods when dealing with people who have an intellectual disability. Concerns over the validity of these techniques provided the motivation for the present study which examined the impact of three assessment methods—self-ratings, ratings by significant others (informant ratings), and ratings by trained clinicians—in the domains of anger, depression, and stress. We begin with a review of each of these three methodologies before describing instances of their application in the three domains.

The most common way to detect emotional disorders is probably through a visit to a clinician such as a family doctor or a psychologist who uses a clinical interview as the basis for a global diagnosis. This approach generally accesses high clinical expertise but, for people with an intellectual disability, it often involves a clinician with limited experience in the disability field. The second approach, having ‘a significant other’ provide information on behalf of the person, is more common in research among younger children and people with an intellectual disability. Use of a significant other, or informant rating, is a convenient approach but the accuracy of the ratings depends heavily on the experience and insight of the rater, as well as on how well he/she knows the person being rated. Little is known about the comparability of ratings made by different informants (Kazdin & Petti, 1982). In the third approach, emotional disorder is assessed directly by asking the individuals to report on their symptoms. The use of this technique among people with an intellectual disability is especially problematic given that they frequently have expressive and receptive language deficits and perhaps an inability to grasp the task demands in testing situations (Heal & Sigelman, 1995).

A search through the literature on anger, depression, and stress shows that researchers have used all three methods of assessment. Benson and Ivins (1992), for example, studied anger across different ranges of intellectual disabilities using an adapted form of the Children’s Inventory of Anger (Finch, Saylor & Nelson, 1983). They used self-report ratings on a 4-point Likert scale with the items read aloud to the participants. Reiss and Rojahn (1993) investigated the joint occurrence of depression and aggression in people with an intellectual disability using the Reiss Screen for Maladaptive Behavior (Reiss, 1988). This instrument uses ratings on behalf of the person with an intellectual disability by a significant other (informant rating) on alphabetically listed symptoms. Anger among a group of people with a mild intellectual disability was assessed by Baker and Bramston (1997) using self-reported anger scores on an adaptation of the State-Trait Anger Inventory (STAXI: Spielberger, 1988). These authors concluded that their results supported the validity of self-reported anger by people with mild to moderate intellectual disability.

Research into depression among people with an intellectual disability is characterized by this same readiness to use different assessment techniques. Matson, Barrett, and Helsel (1988) administered the Children’s Depression Inventory (CDI: Kovacs, 1985) within a clinical interview to compare the depression rates of groups with and without an intellectual disability. Meins (1993) used an informant rating version of the CDI to assess the depression levels of 798 people with an intellectual disability, the majority of whom had a mild or moderate disability. Direct care staff were used as informants and all 24 items on the scale were found to be relevant to a group with mild/moderate disability, leading the author to conclude that the informant version of the CDI is a suitable screening instrument for depression in adults with an intellectual disability. Benavidez and Matson (1993) used self-report to compare three depression scales, one of which was the CDI, on groups of adolescents with and without an intellectual disability.

In research on stress among intellectually disabled persons, Zetlin (1993) used clinical interviews, while Bramston and colleagues Bramston and Bostock 1994, Bramston and Fogarty 1995, Fogarty and Bramston 1997, Bramston et al 1999 relied upon the self-report Lifestress Inventory, designed specifically for this population (see description later).

The use of different assessment methods is not a problem in itself but it is important to examine the convergent validity of these three main assessment techniques, particularly when used among people with an intellectual disability. The small amount of research that exists on this topic suggests that there may not be a great deal of convergence. For example, in a study involving intellectually average children, a comparison between self-report and informant rating showed a correlation of only 0.05 on the CDI (Kazdin & Petti, 1982). Among people with an intellectual disability, a study by Benson and Ivins (1992) compared self-report with informant ratings of anger, depression, and self-concept. On the anger scale, the correlation between self-report and informant rating was 0.13 and on the depression scale it was 0.26, suggesting there was little overlap between the assessment methods. Nearly twice as often the informants viewed the participant as angrier than he/she self-reported. Anger and depression were found to correlate 0.14 on self-report and 0.13 on informant rating. More recently, Benavidez and Matson (1993) compared the responses of people with and without an intellectual disability to three different depression measures. On the CDI, they found correlations of 0.48 between self-report and informant rating for people with an intellectual disability and 0.19 for those without an intellectual disability.

Clearly, in research designed to assess convergence among measurement techniques, there are worrying signs that convergence is weak. There is an obvious need to determine equivalence across the various measurement techniques, particularly as we begin to further explore emotionality within people with an intellectual disability. The aim of the present study was to assess the degree of convergent validity among clinician ratings, informant ratings, and self-report ratings for people with mild/moderate intellectual disability across the affective domains of anger, depression, and stress. The multitrait-multimethod (MTMM) approach was used to address this question.