Evaluating the perceived quality of care by youths with mild to borderline intellectual disability: the influence of assistance in the assessment procedure on outcomes.
Martine Kool 11642300
Master thesis Forensische Orthopedagogiek
Date: 08-07-2019
Supervisor: Prof. Dr. Xavier Moonen Second supervisor: Machteld Hoeve
Table of Contents
Abstract ... 3
Abstract ... 4
Introduction ... 5
Mild to Borderline Intellectual Disability ... 6
Acquiescence ... 7 Submissiveness ... 7 Anti-submissiveness ... 11 Current study ... 11 Method ... 12 Participants ... 12 Procedure ... 13 Measures ... 13 Analyses ... 16 Results ... 17 Description Sample ... 17 Statistics ... 18 Discussion ... 23 References ... 27
Abstract
In satisfaction research with clients with a mild to borderline intellectual disability (MBID) the client is often assisted by the primary caregiver to clarify questions. The purpose of this exploratory study is to examine if clients are influenced by the presence of a person assisting them when completing a client satisfaction self-report instrument. Data were collected from youths between 12 and 23 years of age with an MBID who receive residential treatment in Koraal De Hondsberg and Gastenhof, two residential treatment facilities in The Netherlands. The participants (N=75, Mage= 16.75; 46.7% girls) answered questions about the quality of care unassisted (N=27) or assisted by the primary caregiver or an unfamiliar researcher (N= 48) using an online questionnaire called MyOpinion (in Dutch: MijnMening). Results show that reported client satisfaction was significantly higher in the assisted condition. No differences in total satisfaction scores were observed when assessment was assisted by a primary caregiver compared to assistance by an unfamiliar researcher. Exploratory analyses to investigate possible interactions with age and the quality of the relation with the primary caregiver yielded no significant results. It can be concluded that when assistance is provided, youth with MBID report higher satisfaction with care. Furthermore, based on analysis, it is likely that social desirability influences these outcomes more than dependency. This study provides evidence for carefully considering the pros and cons of adding assistance to clients with an MBID when assessing the satisfaction of the care they receive. In the discussion some weaknesses of this study are reported.
Keywords: Residential care, a Mild to Borderline Intellectual Disability, Assessment of
Abstract
In cliënttevredenheidsonderzoek onder jeugdigen met een lichte verstandelijke beperking (LVB) wordt zij bij het beantwoorden van (online) vragen vaak ondersteund door de persoonlijke begeleider om vragen te verhelderen. Het doel van dit onderzoek is om te onderzoeken of de ondersteuning door de persoonlijke begeleider van invloed is op de mate van cliënttevredenheid wanneer cliënten een digitale vragenlijst invullen. De data voor de huidige studie is verzameld onder jongeren tussen de 12 en 23 jaar met een LVB, die tijdelijk verblijven in een residentiële behandelsetting van Koraal, te weten De Hondsberg en Gastenhof. De participanten (N=75, Mleeftijd= 16.75; 46.7% vrouw) beantwoordden vragen over de kwaliteit van de ontvangen behandeling zelfstandig (N=27) of met begeleiding van hun persoonlijk begeleider of een onbekende onderzoeker (N=48), de jongeren werden at random toegewezen aan een conditie. Zij vulden de digitale vragenlijst MijnMening in. Uit de ANOVA analyse bleek dat de mate van tevredenheid significant hoger was wanneer de jongeren ondersteund werden. Dit suggereert dat het toevoegen van ondersteuning van invloed is op de rapportage over tevredenheid door jongeren met een LVB. Er werden geen verschillen gevonden in de mate van tevredenheid tussen de ondersteuning geboden door de persoonlijke begeleider of door een onbekende onderzoeker. Nadere analyses wijzen erop dat sociale wenselijkheid hier een belangrijkere rol bij speelt dan afhankelijkheid. De ANOVA analyse naar hoofd en interactie- effecten van leeftijd en de kwaliteit van de relatie met de persoonlijke begeleider lieten geen significante verschillen zien tussen de condities of in de invloed op de tevredenheid. De resultaten van dit onderzoek laten zien dat het belangrijk is om bij cliënttevredenheidsonderzoek de voor- en nadelen van het bieden van assistentie door een persoonlijke begeleider kritisch af te wegen . In de discussie worden de beperkingen van dit onderzoek besproken.
Sleutelwoorden: jeugdzorg, Lichte Verstandelijke beperking, cliënttevredenheidsonderzoek,
Introduction
In the last years, there is a growing tendency to include the views of people with disabilities when assessing the quality of care (Rapley, 2003). It is important to monitor the quality of care in residential care from several perspectives, including that of young persons placed in residential care for people with an intellectual disability (ID) themselves. It is important to directly assess their vision on the quality of care they receive instead of only using proxies because proxies tend to express other opinions than the person with an ID himself (White-Koning et al., 2005). Especially for youth staying in residential care it is important to speak about their experiences because certain aspects of residential care may violate the children’s rights (De Valk, Kuiper, Van der Helm, Maas, & Stams, 2016). The results of these assessments can be used to improve care and treatment (Orobio de Castro, Schuiringa, & Kooijmans, 2013).
In the care for people with an ID, often self-report assessment is used for this purpose. But people with an ID may struggle to answer questions independently due to their inadequate level of reading and impaired comprehension of texts (Douma, Moonen, Noordhof, Ponsioen, 2012; Kaal, 2013). The level of comprehension of texts also influences the extent to which the instructions for completing a questionnaire are understood (Douma et al., 2012), which may precipitate the adding of a personal assistant to help them through the process of answering questions. In daily practice, assistance is often provided by a primary caregiver. In the manual of the assessment procedure there is often no solid argumentation based on empirical evidence for this choice (Orobio de Castro, Schuiringa, & Kooijmans, 2013). Thus far, there is no clear evidence for positive or negative effects of assistance of a primary caregiver to a person with an ID when he or she assesses the perceived quality of care received.
Mild to Borderline Intellectual Disability
Around 15 percent of Dutch children and adolescences theoretically have a total IQ score between 50 and 85, of which about 70 percent is able to participate without special support in society (Serra, 2016). The definition of a mild intellectual disability (MID) is different around the world. In The Netherlands the distinction between MID and a Borderline Intellectual Functioning (BIF) is of less practical relevance (Seelen‐de Lang et al., 2019). These two groups often receive the same care, so in this study we will use the term Mild to Borderline Intellectual Disability (MBID). A person with an MBID is defined as: a person who has a Total IQ score between 50 and 85 and significant limitations in adaptive behavior (Douma, Moonen, Noordhof, & Ponsioen, 2012; Kaal, 2013; Greenspan, 2017). In most countries, only people with a Total IQ score between 50 and 70 are defined as having an mild intellectual disability (MID) (Moonen, 2017), although international authors recognize the need for specialized professional assistance of people with Total IQ scores between 71 and 85 (Greenspan, 2017). People with an MBID can experience problems with abstract thinking, problem solving, learning, understanding complex ideas and using experiences to learn (Van Hove & Van Loon, 2006). Information processing is slower for people with an MBID (Douma et al., 2012). People with an MBID also have problems in executive functions, metacognition and generalization of knowledge (Douma et al., 2012).
In addition to the intellectual problems, people with an MBID often demonstrate problems with social and cognitive skills (Douma et al., 2012). Research indicates that they have difficulty recognizing emotions in others. Moreover, it is often difficult for people with an MBID to view the world from another’s perspective (Ponsioen & Van der Molen, 2002). Adaptive behavior consists of conceptual, social and practical skills (Moonen & Wissink, 2015). Problems in social skills include the lower feeling of confidence, lower self-image, higher degree of naivety, more trust in others and vulnerability to manipulation (Douma et al.,
2012). People with an MBID experience problems in each of these domains, although the mix of problems differs from person to person.
Acquiescence
People with an MBID show more acquiescent response styles than normal intellectual functioning people (Beekman, 2018; Clare & Gudjonsson, 1993; Sigelman, Budd, Spanhel, & Schoenrock, 1981). Acquiescence is the tendency of a respondent to agree with or say yes to an item irrespective of the content of that item (Billiet & Mc Clendon, 2000; Finlay & Lyons, 2002). There is a strong relation between intellectual functioning and acquiescence that is that people with lower intellectual functioning show more acquiescence (Gudjonsson, 1990). Furthermore Finlay and Lyons (2002) found evidence for acquiescence in interviews with people who have a mental retardation. A ‘nay-saying’ effect was also identified – the disposition to say ‘no’ regardless of the content of the question (Heal & Sigelman, 1995). This naysaying apparently reflects a desire to present oneself in a positive way by denying any association with taboo subjects.
Submissiveness
Some writers use the term acquiescence to indicate a general submissiveness (Finlay & Lyons, 2002). Submissiveness is the tendency to conform to the opinions of people with authority. Submissiveness can be a personality trait or a learned response (Finlay & Lyons, 2002). Researchers suggested that submissiveness may be more adaptive for people living in institutions, such as group homes (Finlay & Lyons, 2002). They suggested that it cannot be attributed to the intellectual disability itself, but that it is influenced by the environment in which they live. However, other studies found evidence for the submissiveness and acquiescence effect outside institutions (Finlay & Lyons, 2002).
People with an MBID can have problems answering questions independently because of their insufficient level of reading comprehension. The level of reading comprehension also
influences the extent to which the instructions for completing a questionnaire are understood (Douma et al., 2012). In situations where it is apparent that a person does not have the necessary reading or comprehensions skills, it is suggested to provide assistance. In the field of MBID research, this assistance is often provided by a primary caregiver. Garton and Copland (2010) showed that the nature of the relationship between the interviewer and the interviewee influences outcomes. Normally, the assessment procedure is in a researcher-informant format. Sometimes the interviewer is a colleague, friend or caregiver whereby the participant has a prior relationship with the interviewer that has evolved through contexts other than research. Garton and Copland (2010) stressed that interviewers with prior relationships must be cognizant of their influence and take this into account in the research. On the other hand they also identify some benefits, including the fact that the interviewers can acquire more specific information and can easily cross-reference the information given by the interviewee (Garton & Copland, 2010). The mechanisms of influence and how to take the influence into account are not clear yet. For instance is the influence of age on submissiveness still unclear. The acquiescence bias seems to increase with age as older people (above 50) report higher client satisfaction scores compared to younger people (Morales-Vives, Vigil-Colet, Lorenzo-Seva & Ruiz-Paies, 2014). Conversely, Soto, John, Gosling and Potter (2008) stated that submissiveness reduces with age. Another study states that it is not the age, but the level of education that is related to submissiveness (Costello & Roodenburg, 2015). We do not know the influence of age on submissiveness for youths in general, let alone for youth with MBID.
Five causes of submissiveness are identified in research.
1. Social desirability. Social desirability influences the results of the questionnaire; people adapt their answer to present a positive image (Van de Mortel, 2008). Social desirability is the tendency to present a positive image of oneself (Van de Mortel, 2008). Scales to assess the degree of social desirability can be used to correct for the tendency to report in a social
desirable way in questionnaires (Van de Mortel, 2008). A social desirability scale typically consist of a number of questions about behavior which one may assume that everyone tends to show. Social desirability biases vary by gender, dependent of what is social desirable according to the gender (Hebert et al., 1997). Moreover social desirability increases when face-to-face interviews are used as a substitute for computer assisted instruments, particularly when the questions are about highly sensitive behavior (Richman, Kiesler, Weisband, & Drasgow, 1999). So it seems the level of social desirability is higher when another person is present while answering questions.
2. Dependency. Youth in residential care are dependent of their caregivers. The professional associations for professionals working in residential care in the Netherlands recognize the dependency and role of power in the client- caregiver relationship and have issued some guidelines for these professionals (Berger & Zwikker, 2010). Despite the value that is placed on autonomy of the clients, professional caregivers often decide for their clients (Verzaal, 2002). Because of standard procedures and rules in residential care, the professional does not always follow up on the ideas of a client (Verzaal, 2002). Youth in residential care seem to have limited possibilities to participate “meaningfully” in decision‐ making (Ten Brummelaar, Harder, Kalverboer, Post, & Knorth, 2018). Youth are aware of the dependency in the relationship with their primary caregiver and this may translate to more positive evaluations if the caregiver is present when a youth is asked to give his or her opinion on quality of care issues.
3. Content sensitivity. The acquiescence effect is mainly found with regard to taboo topics (Heal & Sigelman, 1995). When the content of an item is sensitive, there is a greater chance of social desirable answering (Krumpal, 2013). When asked on sensitive items, people have the tendency to response in what they believe to be an desirable answer (Krumpal, 2013). An item is ‘sensitive’ when a certain answer might pose a significant threat (Lee & Renzetti,
1993). The threat can be negative feelings of embarrassment or shame or fear of negative consequences. Giving a non-desirable answer increases the possibility of sanctions or effects the relationship with others, like the primary caregiver (Krumpal, 2013). Another dimension of content sensitivity is cultural, where the content within a given culture is taboo or more private. Bias because of the sensitivity of the content is quite common (Tourangeau & Yan, 2007). People misreport when the respondent has anything embarrassing or out of the ordinary to report. They avoid embarrassment in the presence of an interviewer by adapting the answers (Tourangeau & Yan, 2007).
4. Pleasing. People with an MBID show a tendency to 'please' an interviewer; they will answer the question in a certain way because they think that is the 'right' answer the interviewer would like to hear (Rapley & Antaki, 1996). In this context, it is also important to acknowledge that people with an MBID have often become insecure due to repeated failure experiences (Moonen, De Wit, & Hoogeveen, 2011). In addition, they are often accustomed to having someone else control their lives and are therefore inclined to conform (Stalker, 1998). Youth with an MBID often participate in treatment programs to meet the expectations of the people with authority and they learn to conform to their primary caregivers (Perlman, 1994). This leads to an increased chance of accepting suggestions and questions, especially when these questions are more complex, and to a greater acceptance of external advice (Moonen, De Wit, & Hoogeveen, 2011). Pleasing might be related to the degree of dependency.
5. Suggestibility. People with an MBID are more suggestible than those with average intellectual abilities (Everington & Fulero, 1999; Henry & Gudjonsson, 1999). Suggestibility can be defined as ´the extent to which people come to accept messages communicated during formal questioning as a result of which their subsequent behavioral response is affected´ (Gudjonsson, 1990, p. 227). Individuals with an ID also have poorer
memory than those with average intellectual abilities (Everington & Fulero, 1999). There is a greater likelihood of suggestibility when uncertainty or doubt is created (Gudjonsson, 1986). A further problem with orally presented questionnaires is the likelihood of the primary caregiver providing too much information related to the questions (Chester et al, 2015). The primary caregiver can unconsciously give some suggestions or change the sensitivity of the question (Antaki, 1999). Therefore, it is important to provide clear instructions to the primary caregivers about the assessment procedure.
Anti-submissiveness
There is also some evidence for anti-submissiveness (Raplay & Antaki 1996). This means that respondents resist the pressure to change their answers. In research by Raplay and Antaki (1996) in adults with an ID, no influence of adding an assistant to the procedure on the way clients answered was found. The authors conclude that, according to this evidence, the traditional view of submissive, influenceable people needs to be replaced with a more complete view that includes the competence of people with an ID (Raplay & Antaki, 1996). More recent examples of anti-submissiveness are also found in police interviews. The respondents with an ID frequently proved to disagree with or challenged the suggestions of the interviewer (Antaki, Richardsoe, Stokoe, & Willott, 2015).
Current study
The purpose of the current study is to examine if a clients’ assessment of satisfaction with his or her care is influenced by the presence of a person assisting him or her with completing a client satisfaction self-report instrument. The central research question of this thesis is
‘what is the influence on answering by adding assistance to a person with an MBID when he or she rates the quality of care?’
In the current research we formulated some expectations, following the theoretical
framework. We formulated the expectations based on the studies with adults with MBID, although for youth with an MBID the effects of adding assistance is still unclear.
H1: Youth with MBID will report higher total satisfaction scores when the assessment is
assisted, compared to unassisted self-report.
H2: Youth with an MBID will report higher total satisfaction scores when the assessment
procedure is assisted by a person with whom he/she is in a dependent relationship (i.e. a primary caregiver), as opposed to a person who has a more neutral position (i.e. an unfamiliar researcher).
H3: The magnitude of the effect of condition on satisfaction scores will be greater for more
sensitive subjects.
In this study, we will explore the question if and how client satisfaction scores are influenced by the conditions under which the assessment takes place, for youths with MBID. More specifically, we will study whether assessment assistance by a primary caregiver or an unknown researcher leads to higher scores compared to unassisted completion of the same questionnaire. There is a general consensus that it is best to question clients directly about their satisfaction, as opposed to external informants. It is important that these results are valid. The current study may contribute to valid assessment of client satisfaction.
Method Participants
The study sample included N=75 youths with an MBID, receiving residential treatment in Dutch Koraal treatment facilities Gastenhof and De Hondsberg. Having an MBID is an admission criterion for care in either of these treatment facilities. An MBID was defined
according to the Dutch definition as having a Total IQ score between 50 and 85 and having additional adaptive problems. Participants were between ages 12 and 23.
A power analysis using the Gpower computer program (Faul & Erdfelder, 1998) indicated that a total sample of 101 people would be needed to detect a medium effect (d=.5) with 80% power using an ANOVA with alpha at .10.
Procedure
After the ethics committee of the University of Amsterdam (2019-CDE-10133) and the Koraal client board approved the research plan, participants were recruited by Koraal. Only participants with an MBID, between 12 and 23 years of age, were included. In Koraal, the MyOpinion (in Dutch: Mijn Mening) questionnaire is routinely administered approximately one month prior to each bi-annual care plan evaluation. Results are used as input for this care plan evaluation. After consent to participate of the youth and their parents was received, assessment of the MyOpinion questionnaire was planned, whereby participants were assigned to one of three research conditions: unassisted self-report, assisted reporting by the primary caregiver or assisted reporting by a researcher that was not familiar to the participant. Within each care group, participants were randomly assigned to one of three conditions. All participants in all conditions received a small reward when they had completed the questionnaire. The primary caregivers were instructed about the procedure.
Measures
Independent variables: self-report or assisted self-report. In the self-report condition,
the participants answered the questions without assistance. In the second condition the assisted self-report version, the participants were assisted by either their primary caregiver or an unfamiliar researcher. The assistant verbalized every question and the participant was subsequently invited to answer the question. After the question was answered there was a possibility for additional comments by the client. The assistant was instructed not to elaborate
on the formulation of the original questions. If the client did not seem to understand the question as intended, the assistant suggested to move to the next question.
Dependent variable: total score on the MyOpinion questionnaire regarding client satisfaction with the received care. The MyOpinion questionnaire measures a client’s
satisfaction of the received care. The MyOpinion questionnaire consists of 17 items. All eight Quality of Life (QoL) domains as defined by Schalock et al. (2002) are included, as well as questions on more physical attributes that contribute to client satisfaction (e.g. ‘quality of meals provided by the service organization’). Assessment results in scores on the subscales Safety, Autonomy, Support and Leisure & Physical Environment and in a Total Satisfaction Score. The questions are presented to the participant in a digital format. Responses are recorded on a 5-point Likert scale. The questionnaire MyOpinion is approved by the Dutch Association of Service-Providers for People with Disabilities (VGN) for measuring client experiences in care (Orobio de Castro, Schuiringa & Kooijmans, 2013). The internal consistency of the MyOpinion questionnaire is satisfactory (Cronbach’s α = .83). The convergent validity was assessed by comparing MyOpinion results to the results of the so-called C-test (in Dutch: C-toets; Franssen & Jurrius, 2015) in a sample of youths with an MBID. MyOpinion was found an appropriate instrument to measure client satisfaction for youths with an MBID (De Meyer, Van Dam, & Delsing, 2016).
Dependent variables: sensitive and non-sensitive item scores. To determine if results
would vary across more or less sensitive items, MyOpinion items were rated as ‘more sensitive’ or ‘less sensitive’ by two raters independently. The more sensitive items are more related to the functioning of the primary caregiver. Cohen's κ was calculated to determine if there was agreement between two raters on whether items were more or less sensitive. According to the guidelines of Altman (1991) the interrater reliability was very good; κ = .880, p < .001. The raters disagreed on the Ambiance item. This was solved through a discussion between raters,
resulting in consensus that this item should be classified as ‘less sensitive’. The variables sensitive and non-sensitive are the means of the sum of the items. The items that were categorized as ‘more sensitive’ all proved to pertain to the client-caregiver relation. For these items, it was expected that the presence of the caregiver (i.e. the ‘subject’ of the question) would impact highly on the score given on this item. Less sensitive items addressed more contextually based items, such as the quality of meals and housing (see Table 1).
Additional questions. To collect more information about social desirability and on the
relation between the participant and the primary caregiver, we added two questions were added to the MyOpinion questionnaire. In all conditions these two questions were answered anonymously, without assistance. The two questions were: 1. “How do you rate your primary caregiver” in a ‘star rating format (1 to 5 stars whereby 5 would be the maximum positive answer); 2. “If I had the choice, I would complete the questionnaire” (a. on my own, b. together with my primary caregiver, c. with someone I do not know). These questions were not validated in prior research but were considered as explorative questions.
Table 1: Items of the MyOpinion Questionnaire Categorized More Sensitive or Less Sensitive
Items More sensitive Less sensitive
Do caregivers keep to agreements? Do you think the caregivers are honest?
Do the caregivers help you if you want them to? Do the caregivers listen to you?
Do the caregivers have enough time for you? Can you tell your problems to someone? Are you afraid at night?
Is it peaceful at your group home?
How is the ambiance at the group home?
Do you like the meals served at your group home? Do you have a say on the group home rules? Do you have a say on your own appointments?
Do you have enough contact with relatives and friends? Are there enough activities for you to do in your spare time? Do you like the activities you do with group members and staff?
How do you like the interior decoration of your group home? How do you like your bedroom or apartment?
x x x x x x x x x x x x x x x x x Analyses
Factorial between groups analysis of variance (ANOVA) was used to assess differences in the total satisfaction scores under two conditions (self-report and assisted self-report). To assess if there were differences between the two assisted conditions, differences in the total
satisfaction scores were also examined under the three conditions (self-report, assistance by a primary caregiver and assistance by an unfamiliar researcher). The independent variable was the condition and the dependent variable was the aggregated client satisfaction score.
To assess differences in the magnitude of the effect of condition on satisfaction scores between more and less sensitive subjects, a multivariate analysis of variance (MANOVA) was conducted. The two dependent variables were sensitive and non-sensitive and the independent variable was the condition (self-report and assisted self-report).
Furthermore, we conducted some analyses to examine possible interaction effects for age and the rated quality of the primary caregiver with the conditions on the total satisfaction scores. These analyses were exploratory, because of lack of research on these subjects or inconclusive results from prior research. Results of these exploratory analyses can lead to new insights.
Results Description Sample
Of the 75 participants, 53.3 percent were male. The mean age of the youth with an MBID was 16.75 years (SD=3.3) (see Table 2). The mean Total IQ-score was 72.07 (SD=10.5, min= 54, max=85). 27 Clients completed the client satisfaction assessment without assistance (36%, Mage=17.07, 44.4% male), 23 clients were assisted by their primary caregiver (30.7%, Mage=16.30, 47.8% male) and 25 clients were assisted by an unfamiliar researcher (33.3%, Mage=16.80, 68% male).
Table 2: Sample Characteristics of the Conditions and in Total
Self-report Assisted PC Assisted RS Total (N = 27) (N = 23) (N = 25) (N = 75) Age Mean and SD Gender Male Female IQ Mean and SD 17.07 (3.3) 12 (44.4) 15 (55.6) 76.44 (4.9) 16.30 (2.9) 11 (47.8) 12 (52.2) 68 (13.2) 16.80 (3.6) 17 (68) 8 (32) 75.88 (11.5) 16.75 (3.3) 40 (53.3) 35 (46.7) 72.07 (10.5)
Note. Reported means are the Estimated Marginal Means from the ANOVA analyses.
Statistics
An ANOVA was conducted to check for randomization. Participants in the three conditions (self-report, assisted by their primary caregiver and assisted by an unfamiliar researcher) did not differ significantly with respect to their age, gender, IQ and the anonymously rated quality of the primary caregiver (see Table 3).
Because of uneven sample sizes the estimated marginal mean scores (EMMS) were calculated. The total EMMS of all respondents was 2.65 (SD=.6) (see Table 4). The EMMS for sensitive content items was 2.71 (SD=.7) and the EMMS for less sensitive content items was 2.58 (SD=.6). By applying a Pearson correlation test the relationship between age and the total satisfaction scores was tested. A significant positive correlation was found between age and the total satisfaction score, r (75) =. 274, p(two-tailed) =.018, meaning that when age increased, clients reported higher satisfaction scores related to the care received.
Table 3: ANOVA Effects Differences between Conditions in Sample Characteristics Variable df F η p Age Gender IQ Rating Caregiver 2 2 2 2 .342 1.656 1.500 .378 0.01 0.04 .130 .010 .712 .198 .247 .687
Table 4: Estimated Marginal Means of the Dependent Variables for each Condition and Total Self-report Assisted PC Assisted RS Total
Variable EMMS (SE) EMMS (SE) EMMS (SE) EMMS (SE)
Total Satisfaction Sensitive Items Non-sensitive items 2.39 (0.1) 2.63 (0.1) 2.27 (0.1) 2.82 (0.1) 2.98 (0.1) 2.74 (0.2) 2.78 (0.1) 2.81 (0.2) 2.76 (0.2) 2.65 (0.6) 2.71 (0.7) 2.58 (0.6)
Fifty-six percent answered they would want to complete the client satisfaction assessment without assistance if they had the choice. Thirty two percent reported they wanted to do this together with a primary caregiver and twelve percent reported they wanted to do this with an unfamiliar assistant. The vast majority rated the quality of the primary caregiver with three stars or more (1=1,3%, 2=4,0%, 3=26,7%, 4=37,3%, 5=30,7%).
Differences in total client satisfaction scores by two conditions
A factorial between groups analysis of variance (ANOVA) was conducted to compare the main effect of the condition (self-report and assisted self-report) and possible interaction
effects between age, gender and the rated quality of the primary caregiver with total satisfaction scores (N=75). For the purpose of the ANOVA, three age categories were designed (12-14, 15-18, 19-23). Before conducting the ANOVA the data were examined using SPSS Statistics to ensure all if its underlying assumptions were met. Univariate normality was assessed with a Shapiro-Wilk test and Levene’s test was used to evaluate the assumptions of normality and homogeneity of variance respectively. Neither assumption was violated.
As all the underlying assumptions were supported by the data, an ANOVA was conducted (see Table 5). Findings showed that the main effect of the two conditions (self-report and assisted self-report) on total satisfaction scores was significant, F (1, 74) = 6.727, p < .05,
η2 = .11. Youth who were assisted (M =2.82, SE =.1) (see Table 6) reported significantly higher total satisfaction scores than youth in the unassisted self-report condition (M =2.39, SE =.1). Hypothesis 1 was confirmed; Youth with MBID give higher total satisfaction scores when the assessment is assisted, compared to unassisted self-report.
There was no significant interaction between condition (unassisted vs. assisted self-report) and age and the rated quality of the primary caregiver, respectively. This means that the differences in total satisfaction scores between the two conditions did not depend on age or the rated quality of the primary caregiver.
Table 5: ANOVA Effects on Total Satisfaction Scores under Two Conditions Source df F η p Corrected Model Condition Age Rating Caregiver Condition*age Condition*Rating Caregiver 20 1 2 3 2 3 1.278 6.727 1.093 .844 .486 1.546 .321 .111 .039 .045 .018 .079 .234 .012* .342 .476 .618 .213
Note. *Significant at the p<.05 level.
Table 6: Means Total Satisfaction Scores in Different Categories
Variable Mean Total Satisfaction Score (SE)
Condition Self-report Assisted Age 12-14 15-18 19-23 RatingCaregiver 1 or 2 stars 3 stars 4 stars 5 stars 2.39 (0.1) 2.82 (0.1) 2.46 (0.2) 2.48 (0.1) 2.90 (0.1) 2.31 (0.3) 2.52 (0.2) 2.73 (0.1) 2.82 (0.1)
Differences in total client satisfaction scores by three conditions
Findings showed that the main effect of the three conditions (self-report and caregiver-assisted and caregiver-assisted by an unfamiliar researcher) on total satisfaction was not significant. This means that there was no significant difference between the total satisfaction scores among conditions. Hypothesis 2, Youth with an MBID will report higher total satisfaction scores when the assessment procedure is assisted by a person with whom he/she is in a dependent relationship (i.e. a primary caregiver), as opposed to a person who has a more neutral position (an unfamiliar researcher), was rejected.
Content sensitivity
Another MANOVA (see Table 7) was conducted to examine differences in main effects of sensitive content satisfaction and less sensitive content satisfaction under the two conditions (self-report and assisted self-report). Findings showed that the main effect of the two conditions on sensitive content satisfaction was not significant. The main effect of the two conditions on less sensitive content satisfaction was significant, F (1,74) = 9.184, p <0.05, η2=.15. Youth who were assisted reported significantly higher satisfaction on less sensitive content (M =2.78, SE =.2) than youth in the unassisted self-report condition (M =2.27, SE =.1). This means that hypothesis 3, the difference in satisfaction scores between the assisted vs non-assisted condition will be greater when the items relate to more sensitive subjects, is rejected. The differences are, conversely, more present when the content of satisfaction was less sensitive.
Table 7: MANOVA Effects content sensitivity by two conditions
Variable df F η p
Sensitive content Less sensitive content
1 1 1.192 9.184 .022 .145 .280 .004*
Discussion
In the present study the effects of adding a primary caregiver to assist a client who is assessing his or her satisfaction with care is examined. The results of the study show that there is an effect of adding an assistant to the assessment procedure. Based on the present study it can be concluded that youth with an MBID who were assisted, reported significantly higher satisfaction with the quality of care than youth who were unassisted. It is plausible that the presence of an assistant influenced the results of the assessment, but to draw firmer conclusions replication of these finding is needed. The effect of the influence was stronger when the question content was less sensitive. There were no differences found between assistance from the primary caregiver, compared to an unfamiliar person, which suggests that social desirability is a better explanation than dependency.
There are several possible explanations for the influence of adding an assistant to the assessment procedure. First, youth with an MBID are more vulnerable for submissiveness. They also exhibit lower feelings of confidence and a more negative self-image (Douma et al., 2012). Furthermore, their degree of naivety is higher, same as the trust in others (Douma et al., 2012). This causes vulnerability to (unintended) manipulation and susceptibility for suggestions (Douma et al., 2012).
A second possible explanation is that socially desirable answering increases when someone is interviewed. The clients who were assisted might have adapted their answer to present a positive image. A third possible explanation is that people with an MBID show a tendency to 'please' an interviewer (Rapley & Antaki, 1996) and give answers they think the interviewer wants to hear. So, the clients might be more positive because they think the assistant wants to hear positive answers.
Furthermore, the primary caregiver could have given (unintended) suggestions to the clients or provide extra information so clients adapt their answers. Youth in care learn to conform to primary caregivers. This learned behavior to conform increased the chance of
accepting suggestions and external advice (Everington & Fulero, 1999; Henry & Gudjonsson, 1999; Moonen, De Wit, & Hoogeveen, 2011).
A final explanation is the dependency in the relation between the youth and his or her their primary caregiver. Being negative about the quality of care or about the primary caregiver could possibly have negative consequences for the young person and this increases the chance of answering more socially desirable i.e. more positive about the quality of care. Because there was no differential influence found for the primary caregiver as opposed to the unfamiliar researcher, this explanation does not seem all that plausible: the clients did not have a dependent relationship with the unfamiliar researcher, but were influenced nevertheless.
It was expected that the influence of adding an assistant on rating satisfaction would be stronger when the question content was more sensitive, because it reflects on the client- caregiver relationship. Results of the present study did not support this expectation. Conversely, the influence was higher when the question content was less sensitive. This supports the view of a more general concept of submissiveness, not of submissiveness for sensitive subjects. A methodological explanation is that the variances of sensitive content were more skewed then the less sensitive content, which may have influenced the results.
Possible interaction effects of age and the condition of assessment on the total satisfaction score were examined exploratively. There was no evidence found for an interaction effect of age. Thus, age did not significantly increase or decrease satisfaction scores. An explanation for not finding any differences is that it’s not age, but education level which causes more submissiveness (Costello & Roodenburg, 2015). However, in the present study we only included youth with an MBID. Follow-up assessments are needed to test for the effect of the education level on submissiveness in a population of youth varying in education levels from high to low. The effect of intelligence on submissiveness is already found in adult populations (Beekman, 2018; Clare & Gudjonsson, 1993; Gudjonsson, 1990).
Further the interaction effect of the rated quality of the primary caregiver was tested in the two conditions (self-report and assisted self-report) on the total satisfaction score. There were no significant differences found between the two conditions regarding the total satisfaction scores in relation to the rated quality of the primary caregiver. This means that the influence of assistance did not differ between youth who rated the quality of their primary caregiver high or low. An explanation for this last finding can be found in the little deviation in the rated quality of their primary caregiver. This may cause the finding was not significant.
This study has several limitations. A first limitation is that the needed number of participants needed to report a medium effect was not met because of sample difficulties. A low power increase the likelihood of false negatives or decrease true positives. The low power might have influenced the significance of effects. A further limitation of the present study is that the actual assessment procedures in the assisted self-report conditions were not recorded or observed. Because of this is was not possible to appraise the level of suggestibility or manipulation during the assessment. Follow-up studies are needed to examine the processes that causes the influence of adding an assistant to an assessment procedure.
A third limitation is that it was assumed that the satisfaction in the self-report condition is the true perceived satisfaction. It was assumed that when people assess satisfaction alone, there are no processes of suggestibility, pleasing, dependency and social desirability. It was not checked in a qualitive way if this score was a representation of the true satisfaction.
A fourth limitation of the study is that it was unknown how long the youth and the primary caregiver knew each other and thus no controlled for a possible effect is possible. It is possible that the processes of pleasing and dependency are more common when people have a relationship with the primary caregiver for a longer period of time. Further research is needed on this subject.
Fifth; some youth included in this study were living in a crisis group of Koraal, so the relationship with the primary caregiver had just started. In this situation there might be less influence of the assistance of the caregiver on assessing quality of care.
A final limitation is that the division of items in ‘sensitive’ and ‘non-sensitive’ was experimental, and only partly grounded in sound scientific theory. Only the inter-rater reliability (Cohen's kappa) of the scales sensitive and less sensitive satisfaction was tested for. On the basis of interrater agreement alone, it cannot be concluded that the scales are valid.
If the same results as found in this study would be found in replication studies, this would have several implications for the assessment of a clients’ satisfaction in a population of youth with an MBID. Firstly, because most youth report they want to do the assessment alone, this should be the basic mode. This also diminishes the possibility of social desirable answering, suggestion or manipulation. Assistance by an external unfamiliar person is no alternative for the assessment, since the influence does not differ from that of a primary caregiver. Following the results of this study, unassisted assessment can lead to more reliable assessment of a clients’ satisfaction with care. Reliable assessment can lead to a better quality of care.
In summary, the present study is the first randomized study that examines the effect of adding assistance to an assessment procedure in youth with an MBID. It was found that adding an assistant leads to higher reported client satisfaction. Additional research is needed to replicate the findings from this study. Further studies should include more participants. Further studies should also include observation and or recording of the assessment or additional interviewing. For assessing and enhancing the quality of (residential) care it is important to be critical about the assessment methods used!
References
Altman, D.G. (1991). Practical Statistics for Medical Research. London: Chapman and Hall. Antaki, C. (1999). Interviewing persons with a learning disability: How setting lower standards may inflate well-being scores. Qualitative Health Research, 9(4), 437-454.
Antaki, C., Richardson, E., Stokoe, E., & Willott, S. (2015). Can people with intellectual disability resist implications of fault when police question their allegations of sexual assault and rape? Intellectual and Developmental Disabilities, 53(5), 346–357.
Beekman, H. (2018). Acquiescence in adolescents with and without Mild Intellectual
Disabilities or Borderline Intellectual Functioning. Internal Structure and Reliability of
the Adapted Amsterdam Acquiescence Scale (AAAS). Amsterdam: Universiteit van
Amsterdam; master thesis.
Berger, M., & Zwikker, N. (2010). Professionalisering van de jeugdzorg. Jeugd En Co Kennis,
4(2), 38-48.
Billiet, J., & Mc Clendon, M. (2000). Modelling acquiescence in measurement models for two balanced sets of items. Structural Equation Modelling, 7(60), 8-628. Chester, V., McCathie, J., Quinn, M., Popple, J., Ryan, L., Loveridge, C., & Spall, J. (2015). Clinician Experiences of Administering the Essen Climate Evaluation Schema (EssenCES) in a Forensic Intellectual Disability Service. Advances in Mental Health
and Intellectual Disabilities, 9(2) 70-78.
Clare, I., & Gudjonsson, G. (1993). Interrogative suggestibility, confabulation, and acquiescence in people with mild learning disabilities (mental handicap): Implications for reliability during police interrogations. British Journal of Clinical Psychology,
32(3), 295-301.
Costello, S., & Roodenburg, J. (2015). Acquiescence Response Bias: Yeasaying and Higher Education. The Australian Educational and Developmental Psychologist, 32(2), 105- 119. doi:10.1017/edp.2015.11
De Meyer, R., Van Dam, C., & Delsing, M. (2016). Rapportage psychometrisch onderzoek
MijnMening. Nijmegen: Praktikon.
Douma, J., Moonen, X., Noordhof, L., & Ponsioen, A. (2012). Richtlijn diagnostisch
onderzoek LVB: Aanbevelingen voor het ontwikkelen, aanpassen en afnemen van
diagnostische instrumenten bij mensen met een licht verstandelijke beperking.
Utrecht: Landelijk Kenniscentrum LVG.
Everington, C., & Fulero, S. M. (1999). Competence to confess: Measuring understanding and suggestibility of defendants with mental retardation. Mental Retardation, 37(3), 212- 220.
Finlay, W., & Lyons, E. (2002). Acquiescence in interviews with people who have mental retardation. Mental Retardation, 40(1), 14-29.
Franssen, J., & Jurrius, K. (2005). De C-toets getoetst. Ervaringen, ontwikkelingen en plannen rondom de C-toets. Nederlands Tijdschrift voor Jeugdzorg, 9, 293-296.
Garton, S., & Copland, F. (2010). ‘I like this interview; I get cakes and cats!’: The effect of prior relationships on interview talk. Qualitative Research, 10(5), 533-551.
Greenspan, S. (2017). Borderline intellectual functioning: an update. Current Opinion In
Psychiatry, 30(2), 113-122.
Gudjonsson, G. (1986). The relationship between interrogative suggestibility and acquiescence: Empirical findings and theoretical implications. Personality and
Individual Differences, 7(2), 195-199.
Gudjonsson, G. (1990). The relationship of intellectual skills to suggestibility, compliance and acquiescence. Personality and Individual Differences, 11(3), 227-231.
Heal, L., & Sigelman, C. (1995). Response biases in interviews of individuals with limited mental ability. Journal of Intellectual Disability Research, 39(4), 331-340.
Hebert, J., Ma, Y., Clemow, L., Ockene, I., Saperia, G., Stanek III, E., ... & Ockene, J. (1997). Gender differences in social desirability and social approval bias in dietary self-report. American Journal of Epidemiology, 146(12), 1046-1055.
Henry, L., & Gudjonsson, G. (1999). Eyewitness memory and suggestibility in children with mental retardation. American Journal on Mental Retardation, 104, 491-508
Kaal, H. (2013). Ongewoon moeilijk. Jeugdige delinquenten met een LVB als (voortdurende)
bron van zorg. Den Haag: Boom Lemma.
Krumpal, I. (2013). Determinants of social desirability bias in sensitive surveys: a literature review. Quality & Quantity, 47(4), 2025-2047.
Moonen, X. (2017). (H)erkennen en waarderen: Over het (h)erkennen van de noden, mensen met licht verstandelijke beperkingen en het bieden van passende ondersteuning.
Nederlands Tijdschrift voor de Zorg aan Mensen met Verstandelijke Beperkingen, 43(2), 163-176.
Moonen, X., De Wit, M., & Hoogeveen, M. (2011). Mensen met een licht verstandelijke beperking in aanraking met politie en justitie. PROCES, 90(5), 235-250.
Moonen, X., & Wissink, I. (2015). Signalering van kinderen die functioneren op het
niveau van een LVB in het basisonderwijs: screening met behulp van de SAF, onderzoek
met de VALT en een korte uitleg over de BSA-k en de BSA-j. Utrecht: Landelijk
Kenniscentrum LVB.
Morales-Vives, F., Vigil-Colet, A., Lorenzo-Seva, U., & Ruiz-Pamies, M. (2014). How social desirability and acquiescence affects the age–personality relationship. Personality and
Individual Differences, 60(3), 342-348.
Orobio de Castro, B., Schuiringa, H. & Kooijmans, R. (2013). Cliënttevredenheid meten met
Ponsioen, A., & Van der Molen, M. (2001). Cognitieve vaardigheden van licht
verstandelijke gehandicapte kinderen en jongeren: een onderzoek naar mogelijkheden.
Ermelo: Landelijk Kenniscentrum LVG.
Rapley, M. (2003). Quality of life research: A critical introduction. Thousand Oaks: Sage. Rapley, M., & Antaki, C. (1996). A conversation analysis of the ‘acquiescence’ of people with learning disabilities. Journal of Community & Applied Social Psychology, 6(3), 207- 227.
Richman, W. L., Kiesler, S., Weisband, S., & Drasgow, F. (1999). A meta-analytic study of social desirability distortion in computer-administered questionnaires, traditional questionnaires, and interviews. Journal of Applied Psychology, 84(5), 754.
Seelen‐de Lang, B. L., Smits, H. J., Penterman, B. J., Noorthoorn, E. O., Nieuwenhuis, J. G., & Nijman, H. L. (2019). Screening for intellectual disabilities and borderline intelligence in Dutch outpatients with severe mental illness. Journal of Applied
Research in Intellectual Disabilities.
Serra, M. (2016). ‘Hip’onderzoek. Kind & Adolescent Praktijk, 15(4), 5-5.
Sigelman, C., Budd, E., Spanhel, C., & Schoenrock, C. (1981). When in doubt, say yes: Acquiescence in interviews with mentally retarded persons. Mental Retardation,
19(2), 53.
Soto, C., John, O., Gosling, S., & Potter, J. (2008). The developmental psychometrics of big five self-reports: Acquiescence, factor structure, coherence, and differentiation from ages 10 to 20. Journal of Personality and Social Psychology, 94(4), 718.
Stalker, K. (1998). Some ethical and methodological issues in research with people with learning difficulties. Disability and Society, 13(1), 5–19.
Ten Brummelaar, M. D. C., Harder, A. T., Kalverboer, M. E., Post, W. J., & Knorth, E. J. (2018). Participation of youth in decision‐making procedures during residential care: A
narrative review. Child & Family Social Work, 23(1), 33–44.
https://doi-org.proxy.uba.uva.nl:2443/10.1111/cfs.12381
Tourangeau, R., & Yan, T. (2007). Sensitive questions in surveys. Psychological Bulletin,
133(5), 859.
Van de Mortel, T. (2008). Faking it: social desirability response bias in self-report research.
Australian Journal of Advanced Nursing, 25(4), 40.
Van Hove, G. & Van Loon, J. (2006). Personen met een verstandelijke beperking. In E. Broekaert & G. van Hove (Red.), Handboek Bijzondere Orthopedagogiek (pp. 17-43). Antwerpen/ Apeldoorn: Garant.
Verzaal, H. (2002). Empowerment in de jeugdzorg : onderzoek naar empowerment :
bevorderend gedrag van hulpverleners. Universiteit Amsterdam.
White-Koning, M., Arnaud, C., Bourdet-Loubère, S., Bazex, H., Colver, A., & Grandjean, H. (2005). Subjective quality of life in children with intellectual impairment–how can it be assessed?. Developmental Medicine and Child Neurology, 47(4), 281-285.
