Interactional patterns between staff and clients with borderline to mild intellectual disabilities

Interactional patterns between staff and clients withborderline to mild intellectual disabilitiesjir_1515 1..14

E. Reuzel,1,2 P. J. C. M. Embregts,1,3,4 A. M.T. Bosman,5 M. van Nieuwenhuijzen6 &A. Jahoda7

1 Tilburg University,Tilburg, Noord-Brabant,The Netherlands2 J.P. van den Bentstichting, Deventer, Overijssel,The Netherlands3 HAN University of Applied Sciences, Nijmegen, Gelderland,The Netherlands4 Kennisn@Dichterbij, Oostrum, Limburg,The Netherlands5 Behavioural Science Institute, Radboud University, Nijmegen, Gelderland,The Netherlands6 VU University of Amsterdam, Amsterdam, Noord-Holland,The Netherlands7 University of Glasgow, Glasgow, UK

Abstract

Background Client-centred models of care implythat clients should have a collaborative relationshipwith staff providing support. This study investigateswhether dialogues between staff and clients in natu-rally occurring contexts reflect this collaborativeideal.Methods Nineteen staff members video recorded asocial interaction with one of their clients. The topicof the interaction concerned an aspect of theirsupport needs. The recordings were transcribed andanalysed using the Initiative Response Analysisdesigned by Linell et al.Results Staff were more dominant than clients,albeit the level of asymmetry in the dialogues wasrelatively small. However, a different pattern of turnswas used by staff and clients. Staff asked more direct

questions and sometimes neglected meaningfulclient contributions. Clients, on the other hand,provided more extended turns in response to staffmembers’ questions, thereby helping to maintain thedialogue. However, in a notable minority of commu-nicative turns, the clients failed to link with the staffmember’s contribution.Conclusions The interactional patterns found inthis study suggest that staff and clients can facedifficulties establishing collaborative dialogues onshared topics. Future research should take accountof what staff and clients want to achieve in dia-logues, along with the nature of their non-verbalcommunication.

Keywords collaborative relationship, intellectualdisabilities, interactional patterns, socialinteractions, staff–client interactions

Introduction

A number of studies have highlighted that interac-tions between clients and staff members are often

Correspondence: Ms Ellen Reuzel, J.P. van den Bentstichting, POBox 361, 7400 AJ Deventer, The Netherlands (e-mail: [email protected]).We thank clients and care staff of J.P. van den Bentstichting in theNetherlands for their participation.

Journal of Intellectual Disability Research doi: 10.1111/j.1365-2788.2011.01515.x1

© 2012 The Authors. Journal of Intellectual Disability Research © 2012 Blackwell Publishing Ltd

asymmetrical (e.g. Prior et al. 1979; Cullen et al.1983; Markova 1991; Antaki et al. 2002). Thismeans that clients can have limited choice andcontrol in their lives (Goble 1999; Goodley 2000).However, in recent decades there has been increas-ing recognition of the agency of people with learn-ing disabilities, and that they should be supportedto make decisions as active citizens (Wehmeyer &Schwartz 1998; Antaki et al. 2006, 2008). This, inturn, has major implications for their relationshipswith support staff. A more client-centred approachmeans that staff members would be expected toactively seek the opinion of the client, and the clientshould be able to talk openly and ask questions toreach shared decisions (Barry 1999). Even if staffand professionals remain in a more powerfulsupport role, a cooperative pattern of dialogueshould mean that the power in the relationshipbetween partners is more balanced.

Jahoda et al. (2009) stated that in addition tocommunicative ability per se, communication stylemay be a barrier for the development of a collabo-rative relationship. Early research on the communi-cation style of staff in institutional settings foundthat they tended to use ‘controlling’ and directivespeech rather than being engaged in social conver-sations (Prior et al. 1979). Even more recent studieshave indicated that staff tend to favour the use ofdirectives and questions, and can fail to adjust theirlanguage to the client’s level of understanding(Jingree et al. 2006). Hence, clients would seem tohave few opportunities to engage as equal partnersin conversations with staff (Leudar 1981; McConkeyet al. 1999).Yet different forms of communicationdo appear to encourage more positive engagement.Mirenda & Donnellan (1986) found that whenadults use a facilitative rather than a directive (orquestion-based) style, adolescents with intellectualdisabilities (IDs) initiated a higher proportion oftopics and produced more spontaneous commentsand questions in their conversational exchanges.

A drawback to the research on communicationbetween staff and clients is that it has usuallyfocused on the abilities of the clients or on thecompetence of staff (Dagnan et al. 2000). Withrespect to clients, one prevalent view is that peoplewith IDs tend to be passive in communication orprone to acquiescence with people of a higher socialstatus (Heal & Sigelman 1995). Finlay & Lyons

(2002) also pointed out that complicated questionsmay increase the likelihood that respondents withIDs will simply agree with the questioner or say yes.

There has been little work examining inter-actional patterns between clients with IDs and staff.Researchers often assume that interactions can beanalysed from a linear perspective, which viewscommunication as the contribution of individualacts (Linell & Markova 1993; Markova & Linell1996). In communication theory, on the otherhand, dialogue is viewed from an interactionalperspective (Collins & Markova 1995). It is not justabout who raises the topics being discussed or howmuch each individual talks, but rather about twopeople being able to sustain and actively contributeto a shared dialogue. The more symmetrical thedialogue is, the more staff and clients share theinteraction on equal terms; the more asymmetricalthe dialogue is, the more one individual dominates.Linell et al. (1988) developed an initiative-responsemethod of analysis, which allows dialogue to becoded from this interactional perspective.

Consideration of the extent to which staffworking with clients who have an ID are able tosustain a shared dialogue, provides insight into thecollaborative nature of the relationship (Jahodaet al. 2009). Collaboration does not just mean thatstaff are able to communicate effectively as experts,it also means that the client must feel that he orshe is properly heard and understood (Jahoda et al.2009). It follows that interactions between staff andclients with an ID are crucial to how successfulservices are in meeting people’s needs (SocialExclusion Unit 2005, p. 57). Several studies investi-gated the quality of the interactions between staffand clients with profound or severe IDs and/orcommunication problems (Markova 1991; Purcellet al. 1999; Bradshaw 2001; Edge 2001).Yet fewerstudies of this kind have been undertaken withyoung adults with mild to borderline IDs, who havegood verbal abilities and might be expected to bemore equal partners when interacting with staff.

The present study aims to capture the natureof dialogue between staff and clients with mild toborderline ID from an interactional perspective.These interactional patterns will be studied in natu-rally occurring contexts, and examine how coopera-tive dialogues between direct-care staff and adultswith borderline to mild IDs prove to be. Linell et al.

2Journal of Intellectual Disability Research

E. Reuzel et al. • The nature of staff–client social interactions


(1988) innovative method of interactional analysiswill be used to examine the pattern of interactionaldominance between staff and clients. Finally, thepattern of interaction will be examined in moredetail and the responsiveness of the staff and clientsto each other.

Methods

Participants

Staff

Nineteen staff members (3 men and 16 women)working at the JP van den Bent foundation, theNetherlands, participated in this study. They wereselected by a manager or psychologist and asked ifthey wished to volunteer to take part in the study.None of them refused. The staff members wereselected from different regions and work settingsin the Netherlands, so a broad range of contextswere included. Most staff members worked incommunity-based residential houses (13), othersprovided outreach care (3), and the rest worked incrisis care (3). The staff members had worked inservices for people with ID fro an average of7.1 years (range = 1–27 years).

Clients

Each staff member was asked to select a client withwhom they worked frequently. At this point, thepurpose of the present study was explained to theclients and their consent to participate was sought.The severity of clients’ ID was assessed by meansof the Wechsler Adult Intelligence Scale (WAIS-3).According to their WAIS scores, the clients’ cogni-tive functioning ranged from the mild (8) to border-line level (11). All clients were able to verballyexpress thoughts and feelings fluently. Their agesranged from 18 to 39 years (M = 25.1 years;SD = 6.1). Individuals with autistic spectrumdisorders were excluded from the study, as it wasthought that the associated communication difficul-ties might have a significant bearing on their inter-actions with staff members. Seven of the clientswere men and 12 were women and they were allliving in community settings and receiving supportfrom services; three clients were living alone in theirown apartment with outreach support, and seven

were living alone or with a partner with 24 hsupport available; six clients lived in staffed housesfor training purposes, and three clients were livingtemporarily in staffed houses (crisis care).

Procedure

For the selection of participants, the researcher firstobtained permission from the organisation toconduct the research. Managers and psychologistswere provided with information about the purposeof the study. All participants volunteered for thestudy and received an explanation of its purposeand what would happen. The researcher contactedeach staff member to explain the main goals of thestudy.

Participating staff video recorded a regular con-versation, which had already been scheduled, with aclient in his or her home. They were asked to putthe portable camcorder in a corner of the room, inorder to be unobtrusive and keep the situation asnormal as possible. The interactions were requiredto meet the following criteria: (1) the topic con-cerned an aspect of the participants’ support needs;and (2) it was a type of interaction that occurredon a regular basis, at least once a week. The averagelength of the video recordings was 14.6 min(SD = 6.2, range = 7.2–29.7 min). No instructionswere given to the participants other than to interactas usual. Ten different types of topics were dis-cussed: (1) establishing or refining a support actionplan (n = 5); (2) planning or evaluating client goals(n = 1); (3) planning household activities (n = 4); (4)planning other activities like a schedule for the weekor a visit to the doctor (n = 3); (5) discussing leisureopportunities (n = 2); (6) planning finances (n = 5);(7) reviewing clients’ work situation (n = 2); (8)discussing parenting problems (n = 1); (9) copingwith interpersonal conflicts (n = 6); and (10) findingsolutions for a range of other problems faced bythe clients (n = 3). All the topics of conversationrequired both clients and staff to listen actively toeach other in order to achieve a shared view abouthow to tackle the issue being discussed.

All videotapes were then transferred into ‘TheObserver XT’ (Noldus Information Technology,Wageningen, the Netherlands). The Observer XTis a professional event logging software that can beused for the collection, analysis, and presentation of




observational data and is developed by Noldus(2009). Next, all videotapes were analysed using themethod of coding dialogue developed by Linellet al. (1988; see below).

Measures

The aim, background and structure of Linell et al.(1988) approach to interactional analysis will bedescribed, before going on to outline the measuresthat can be derived from this Initiative ResponseAnalysis, which form the basis of the results.

Method of coding interactional analysis

This analysis aims to capture dominance and coher-ence in dialogue. Interactional dominance includesthe communicative actions, initiatives and responsestaken by the interlocutors. The dominant party isthe one who manages to direct and control theother party’s actions to the greatest extent and whoalso avoids being directed and controlled in her orhis own interactive behaviour. Coherence is createdby communicative actions which are relevant to andlink-up with the preceding discourse.1

According to Linell et al. (1988) an ‘ideal dia-logue’ includes the following conditions:• Condition 1: It is locally coherent, in that inter-locutors try to say things which are relevant to andcohere with the current topic;• Condition 2: Conversationalists are mutuallyresponsive, in that each of them links up with whatthe interlocutor has just said;• Condition 3: It is progressive, in that interlocutorstry to contribute the progression of the discourse byproviding new material;• Condition 4: It is coherent, in that interlocutorsstick to the main content of the discourse;• Condition 5: It is non-imposing, in that interlocu-tors refrain from imposing strong restrictions on thepartner’s responses; and• Condition 6: It is symmetrical, in that the inter-locutors are in control of (and are themselves con-trolled in) the dialogue on an equal basis.

Linell et al. (1988) maintain that the best way ofunderstanding a dialogue is by comparing it with a

chain, with the basic unit of analysis being eachinteractional turn. Each turn is coded for propertiesof response, or how it is linked to the previousturn, and initiative, which concerns how it links tothe next turn. Initiatives continue the dialogue byrequesting (soliciting or inviting) a response fromthe communicative partner and/or by the introduc-tion of a new topic by the speaker itself. Responsesensure coherence with the preceding discourse bylinking up to what the interlocutor or the speakerhas said. Each turn of the dialogue under analysis isassigned to a particular category. There are 18 cat-egories (plus three non-categories: turn miscar-riages, back-channel items, and inaudible turns).

The category system consists of a small set offunctions for initiative and response, which arebased on the conditions of what Linell et al. (1988)consider essential to an ideal dialogue. Thesedistinguishing features include:• The distinction between initiative and response:An initiative means that an interlocutor’s communi-cative turn will help the conversation to progressand is symbolised by > (strong initiative) or ^ (weakinitiative). A response means that an interlocutorlinks up with the preceding turn and is symbolisedby <. Both features, initiative and response can beused in one turn;• The strength and scope of initiatives: Initiativescan be divided into strong initiatives that are(explicitly) inviting or demanding, for example, byasking a question, which are symbolised by >, andweak initiatives (asserting or submissive), which aresymbolised by ^;• The adequacy of a response: When a response isadequate (accepted) it is symbolised by <. When aresponse is inadequate or partially accepted, theturn is treated by the interlocutor as not satisfyingthe turn demands of his own preceding initiative;• The focality of turns: Focal means focusing onthe main content of the other speaker’s turn. Whena response is focal it is an adequate response whichis symbolised by <. A non-focal link usually involvesremarking on or challenging the form or functionof the interlocutors’ preceding turn. A non-focalresponse is symbolised by : instead of < (adequateresponse);• Scope of links – local vs. non-local responses. Alocal response means linking up with an immedi-ately preceding turn, which is symbolised by <

1 The explanation of the Initiative Response Analysis method isderived from and, therefore, almost identical to the descriptionpresented by Linell et al. (1988) and Jahoda et al. (2009).




(adequate). A non-local response is a turn beinglinked to a specific nonadjacent turn further backin the preceding dialogue and is symbolised by ..instead of < (adequate); and• Alter or self-linked response. When an interlocu-tor is linking up with the other speaker’s precedingturn it is an adequate response, symbolised by <.When an interlocutor links up with his own preced-ing turn, it is symbolised by = instead of <.

The whole system of turn types is given, togetherwith brief definitions, in Appendix 1. The 18 catego-ries can be ordered on a six-point ordinal scalefrom the strongest initiative with no response prop-erties, to the weakest response without any potentialfor promoting the dialogue further (see Table 1).

Measures of initiative response analysis

For each dyad in this study, three measures werecomputed:

1. The level of asymmetry. For each dyad in thisstudy the degree of asymmetry is measured. Thelevel of asymmetry is assessed by computingthe difference between the level of dominance ofthe staff member and that of the client.

2. The level of dominance of clients and staff. Thelevel of dominance is derived from an initiativeresponse profile. This is a summary of the frequen-cies of the parties’ turn categories on the six-pointordinal scale. The level of dominance refers to themedian value of the scores on that scale.

3. Turn types coefficients used by clients and staff,which are required to examine interactional patterns.All conversational turns were coded according tofour different turn types as described by Linell et al.(1988). The frequencies of various turn types as apercentage of all turns used by each partner in thedyad, yield interaction coefficients.

Turn types coefficients

• Expanded reponses (B-coefficient, B = Balance):the number of expanded responses as a percentageof all turns in the dyad. This coefficient showshow often an individual responds to what has beensaid and provides sufficient initiative to allow thedialogue to continue on the same topic;• Direct questions (S-coefficient, S = solicitation):the number of questions (or other strong initiatives)as a percentage of all turns in the dyad. Thiscoefficient shows how often individuals explicitlysolicit their interlocutors into responding on theirinitiative;• Abrupt topic shifts (F-coefficient, F =fragmentation): the number of turns that break theinteraction into fragments by the introduction ofnew and unrelated topics as a percentage of allturns in the dyad. This coefficient indicates howoften parties perform abrupt topic shifts, thuscontributing to local incoherence or fragmentationof discourse; and• The implicit turns (O-coefficient, O =obliqueness): the number of turns involving selflinking responses or responses in which the form orfunction of the preceding turn is challenged as a

Table 1 Turn categories and interactional strengths on the six-point ordinal scale and the percentage of each interactional strength used bystaff and clients

.. > .. ^ .. <: > : ^ ( >< = > < = ^ < )

= > = ^ - >> ^ < > < ^ < -

Interactional strength 6 5 4 3 2 1Percentage used by staff 0.6 1.3 31.2 37.1 29.7 0.01Percentage used by clients 0.2 0.8 4.5 62.3 30.2 0.2

Turns independentand stronglyproactive

Turns totallydependent and notat all proactive




percentage of all turns in the dyad. This kind ofobliqueness is designed to capture how often actorsavoid linking up with the main content of theirinterlocutor’s adjacent turn, in spite of the fact thattheir contribution in question is locally related. Theimplicit turns can be subdivided in three categories:(1) holding monologues; (2) ignoring a meaningfulcontribution of the other speaker; and (3) challeng-ing the form or function of the other speaker’scontribution (Linell et al. 1988).

Inter-rater reliability. In this study a total of 5105

turns were coded. Despite the complexity of themethod, good inter-rater reliability was obtainedbetween the experimenter and a research assistant.Initial inter-rater reliability coding was carried outwith four videotapes encompassing 910 turns. Priorto assessment of formal reliability ratings, onevideotape with 314 turns was used for training pur-poses. There was an overall agreement of 83% forusing the same codes. Disagreements between raterswere discussed and consensus on how these turnsshould be coded was reached.

Results

The balance of power; the level of asymmetry

The level of asymmetry scores shows the extent towhich social interactions were symmetrical or domi-nated by the staff or clients. The results show thatthe mean level of asymmetry was 0.40 (SD = 0.31,Min = -0.07, Max = 1.07), which means that therewas a reasonable balance of power in the dialogue,when compared with previous findings concerninga variety of communicative partners (Linell et al.1988).

The level of dominance of clients and staff

The level of dominance refers to the global measureof domination or being controlled during socialinteraction. Linell et al. (1988) scored all turns on asix-point ordinal scale from the strongest initiative,with no attempt to respond to the other person’scontribution, to the weakest response, without anyattempt to elicit a response and maintain the dia-logue. Table 1 shows the percentage of all six cat-egories from both staff and clients, resulting in aninitiative response profile. The median value of the

scores of the interlocutors on the ordinal scale isthe level of interactional dominance. Table 2 showsthat staff scores were significantly higher than clientscores, which means that staff members were moredominant in these dialogues.

Interactional patterns

The overall asymmetry and dominance scores maskthe complexity of the interactions revealed by thissophisticated method of interactional analysis. Inthis section the pattern of interaction will be exam-ined in more detail, including the responsiveness ofthe staff and clients to each other. Interactional pat-terns were examined by calculating the associationsbetween: (1) turn type coefficients and the level ofasymmetry; (2) turn type coefficients and the levelof dominance of each partner; and (3) turn typecoefficients used by staff and turn types coefficientsused by clients. Before presenting the results ofthese three sets of associations, a description will begiven of the turn types used by staff and clients.

Descriptive statistics of turn types used by staffand clients

Turn types included expanded responses, directquestions, abrupt topic shifts, and implicit turns.With respect to these four different turn types,coefficient data were calculated for staff as well asclients. Table 2 presents the percentage of the turntypes for staff and clients. These coefficients repre-sent proportions of total turns used by each com-municative partner and paired-samples t-tests wereconducted to test for differences between staff andclients. Table 2 shows that expanded responses wereused most frequently by clients. These are turnswhere the individual responds to what has beensaid and provides sufficient initiative to allow thedialogue to continue on the same topic. Clientsused expanded responses significantly more oftenthan staff did. Implicit turns were the second mostcommon turn types for clients. Staff and clientsused similar numbers of implicit turns. These arethe turns in which the interlocutors fail to connectsmoothly to each other and consist of: (1) holdingmonologues; (2) ignoring a meaningful contributionof the other speaker; and (3) challenging the formor function of the other speaker’s contribution




(Linell et al., 1988). Table 3 shows that the majorityof the implicit turns consisted of one interlocutorproducing a monologue, and paying little attentionto the other speaker’s contribution. Both staff andclients produced a similar number of monologues,and there were also similar scores for both sets ofpartners ignoring meaningful contributions fromthe other.

Direct questions were the most frequently usedturn type by staff. A very small proportion of the

clients’ turns included direct questions. Staff scoreswere significantly larger than client scores, whichmeans that staff explicitly solicits clients intoresponding on their initiatives more often thanclients did.

Associations between the level of asymmetry and theturn type coefficients

This section will outline how the communicativeacts of staff and clients were related to the balance

Table 2 Mean, SD, range for the level ofinteractional dominance, all turn typesused by clients and staff

Client Staff t-value P-value

Level of dominanceM 2.7 3.1 -5.62 0.001**SD 0.21 0.18Range 2.3–3 2.8–3.6

Expanded responsesM 39.4 22.8 6.88 0.001**SD 8.5 7.9Range 20.5–54.3 10.2–41.2

Direct questionsM 4.6 32.4 -7.28 0.001**SD 5.8 11.3Range 0.8–23.3 12.9–66.7

Abrupt topic shiftsM 1.9 2.3 -1.51 0.16SD 1.4 1.8Range 0.55–4.5 0.4–6.7

Implicit turnsM 24.3 19.3 0.89 0.39SD 12.8 13.5Range 4.2–7.5 4.7–56.5

** Significant at a P < 0.01 level.

Table 3 Mean, SD and range of the threecategories of the implicit turns as apercentage of all implicit turns

Client Staff t-value P-value

Holding monologuesM 89.9 89.9 -7.49 0.46SD 15.8 14.7Range 40–100 50–100

Neglecting a meaningfulcontributionM 9.4 17.5 -0.16 0.88SD 10.4 15.8Range 1.6–40 1.4–50

Challenging form or functionM 22.9 – – –SD 26.8 –Range 2.4–53.3 –




of power in the dialogues. The level of asymmetrycorrelated significantly with the implicit turns usedby both clients and staff. Staff dominance reducedwhen clients failed to link up with the main contentof their interlocutor’s adjacent turn. Moreover, thedifference in dominance became larger when stafffailed to connect with what the client was saying. Asthe implicit turns mainly consisted of monologues,there was a strong correlation for both staff andclients between holding monologues and the levelof asymmetry of the interaction. In one social inter-action the use of implicit turns by staff was 47.7%(The level of asymmetry of that interaction was0.75, which is relatively high compared with theother interactions in this study). An extract of thisinteraction is shown in Fig. 1a.

This long extract shows how the staff member ledthe conversation and the client had little input. Theleast asymmetry between communicative partnerswas 0.02 and in this interaction the client used atleast three times more implicit turns than the staffmember. An extract of this interaction is shown inFig. 1b. Both staff and client influence the conversa-tion. Table 4 also shows that neither expandedresponses nor direct questioning by staff, the mostfrequent turn types used by the interlocutorsrespectively, correlated significantly with level ofsymmetry.

Associations between the level of dominance of staff andclients and the turn type coefficients

The associations between the level of asymmetryand the different turn type coefficients give usinformation on interactional patterns of the dyad ingeneral. In this section we will give more detailedinformation on what type of turns gave staff andclients control over the interaction. Associationswere calculated between the level of dominanceof each interactional partner and the various turntypes. Clients increased their dominance by usingexpanded responses (Table 4). Implicit turns werethe second most common turn types for clients andthere was a positive and significant correlationbetween clients holding monologues and their levelof dominance. In turn, clients’ dominancedecreased when staff held more monologues.

Staff also became more dominant by asking moredirect questions and neglecting meaningful contri-

Example a

Staff: In itself, the finances

are going well, in that,

when you need something

we discuss together what is

possible

Client: yes

Staff: Then we request the

money

Client: mm mm

Staff: That goes well

Clients; mm mm

Staff: I just think that when

everything becomes more

clear to you

Client: mm mm

Staff: For example during a

week, it will give you more

peace

Client: mm mm

Staff: Because, for

example, you get stressed

when you don’t have

enough money in the

supermarket.

Client: mm mm

Staff: When the groceries

appeared to be more

expensive then you thought..

^

<

= ^

<

= ^

<

= ^

<

= ^

<

= ^

<

= ^

Example b

Client: And then via via they

told me that he (boyfriend)

was cheating on me

Staff: mm mm

Client: And then I said;

“Listen, it is over now”.

Staff: Has it been like that

with all your boyfriends or

just a few and do you have

the feeling of being used by

everybody?

Client: Well, with Michael,

he is also my ex

Staff: mm mm

Client: I’ve been with him

for almost 2 years

Staff: mm mm

Client; And we had fun

together and with him I did

not have that feeling

Staff: You did trust him?

Client: Yes

^

<

= ^

< >

< ^

<

= ^

<

= ^

< >

<

Client: mm mm

Staff: Maybe we should

think about how we can get

more rest in that..

Client: Yes

Staff: Because you don’t

have much money you can

spend in a week

Client: mmm

Staff: Because when you

have extra things, like last

week you wanted to go to

the bingo…

Client: mm mm

Staff: You want to have

something extra

Client: mm

Staff: That makes sense. But

well, you simply don’t have

that.

Client: No

Staff: So we should have a

look… Well, you know how

I feel about the smoking…

Client: mmm

Staff: I mean, one packet of

cigarettes less and it saves

you 4 euro..

Client: mmm

Staff: Maybe you should

think about that

<

= ^

<

= ^

<

= ^

<

= ^

<

= ^

<

= ^

<

= ^

<

= ^

Client: Yes <

Figure 1 Examples of implicit turns in an unbalanced (a) andbalanced interaction (b).




butions from the other speaker. In the most unbal-anced interaction (level of asymmetry = 1.07), thestaff member asked direct questions in 67% of theturns. An extract of the interaction is provided inFig. 2.

Associations between turn types used by staff and turntypes used by clients

This section will describe which turn types fromone interlocutor tended to elicit particular responsefrom the other person. Although all possible corre-lations between the different turn types were calcu-lated, Table 5 only shows the associations that weresignificant.

Table 5 shows that there was a strong correla-tion between the number of direct questions giveby one speaker and the expanded responses of the

other. In turn, when clients used more expandedresponses then staff were more inclined to linkup with the main content of what the client wassaying. The more questions clients asked, the fewerquestions staff produced, and vice versa. When itcame to implicit turns, there was a negative asso-ciation between clients’ failure to link with thecontent of what the staff said, and staff responsesto what clients were saying. This finding is consis-tent with the fact that most implicit turns weremonologues. It is also noteworthy that there was astrong positive correlation between the number ofdirect questions produced by clients and abrupttopic shifts by staff. Additionally, Table 5 showsthat abrupt topic shifts by staff were associatedwith the same turn types by clients. An exampleof this pattern is illustrated by an excerpt of adialogue in Fig. 3.

Table 4 Associations betweenexperimental variables and the level ofdominance of clients and staff, theasymmetry level, respectively

Dominance

AsymmetryClients Staff

r P r P r P

Expanded responses clients 0.58 0.01 0.36 0.13 -0.18 0.47Direct questions clients -0.09 0.77 -0.28 0.34 -0.07 0.81Abrupt topic shifts clients -0.47 0.15 0.21 0.53 -0.49 0.12Implicit turns clients 0.7 0.001 -0.68 0.001 -0.87 0.001Holding monologues clients 0.65 0.002 -0.73 0.001 -0.87 0.001Neglecting clients 0.18 0.55 0.4 0.18 -0.09 0.77Expanded responses staff 0.01 0.98 -0.04 0.88 -0.03 0.92Direct questions staff 0.12 0.66 0.67 0.001 0.33 0.17Abrupt topic shifts staff -0.14 0.57 -0.001 0.99 0.09 0.71Implicit turns staff -0.62 0.005 0.40 0.09 0.66 0.002Holding monologues staff -0.66 0.002 0.37 0.12 0.66 0.002Neglecting staff 0.34 0.31 0.63 0.04 0.01 0.97

Note: The correlations in bold type font are significant.

Staff: Can you describe which feeling you have when you are feeling not so good? >

Client: I don’t know how to describe that < ^

Staff: No, that is difficult. But what do you feel on such a moment? < >

Client: I don’t know < ^

Staff: You don’t know, but do you know what you want to do on such a moment? < >

Client: Searching distraction < ^

Staff: Yes, in the past you choose to do other things, as last you went to buy a can of beer.

Can you tell me why you did that? < > Figure 2 Example of direct questions inan unbalanced social interaction.




In this example staff and client are discussingseveral topics, and is characterised by both of themmaking sudden changes of topic. The result is afragmented dialogue.

Discussion

The analysis found that recorded interactionsbetween staff and clients were dominated by staff.Staff dominance was characterised by the use ofdirect questions. This pattern is consistent withpast studies that have found staff favour the useof directives (Prior et al. 1979; McConkey et al.1999). Although this study did not investigate thetype of questions being used, a strong relationshipwas found between the number of questions askedby staff and the use of expanded responses byclients. Therefore, staff questioning did not alwaysinhibit dialogue but could elicit responses fromclients that helped to maintain dialogue on ashared topic. In other words, the questions asked

Table 5 Associations between the turn types used by clients and theturn types used by staff

r-value P-value

Expanded responses clients–Directquestions staff

0.51 0.03

Expanded responses clients–Implicitturns staff

-0.52 0.02

Expanded responses clients–Holdingmonologues staff

-0.56 0.01

Direct questions clients–Directquestions staff

-0.69 0.01

Direct questions clients–Expandedresponses staff

0.72 0.004

Direct questions clients–Abrupttopic shifts staff

0.66 0.01

Abrupt topic shifts clients–Abrupttopic shifts staff

0.84 0.001

Implicit turns clients–Implicit turnsstaff

-0.69 0.001

Holding monologues clients–Holdingmonologues staff

-0.7 0.001

Staff and client are planning their next meeting

Client: So it will be next Friday, ok..

Staff: Yes

Client: but.. I don’t know, with these games.. with Jason (client’s son)

Staff: Yes

Client: Is that useful?

Staff and client discuss the utility of play therapy for the client’s son.

Staff: And play therapy will be covered by your insurance..

Client: Did you call Jane?

Staff: No, I e-mailed her and asked if you could get compensated for the

clothes and if you could get money from the school fund.

Staff and client then discuss which persons were contacted and which

agreements were made.

Staff: I tried to contact Ziggo, but I don’t think he works on a Friday, so

I didn’t get that answer.. I will contact him on Monday from my home, if

you agree with that.

Client: Yes , please

Staff: Friday I will have to go to training.

Client: Well that is part of the job

< ^

<

^

<

= >

= ^

>

< ^

< ^

< ^

^

< ^ Figure 3 Example of a fragmentedinteraction.




by staff could enable clients to play an active rolein the interaction.

Although staff were usually dominant, the inter-actions between staff and clients were reasonablybalanced. Linell et al. (1988) investigated largesamples of different social situations using theinitiative-response analysis. They found interestingand characteristic differences between various con-versation types, such as doctor–patient interviews,criminal trials, police interrogations, radio chatprogrammes, language lessons, and informal con-versations between friends. Comparisons withthese previously published data showed that themean scores of the level of asymmetry of socialinteractions between staff and clients in thisstudy were akin to scores obtained from informalconversations between friends.

On closer examination, there were a wide varietyof scores concerning the level of asymmetry forthe participants’ dialogues. While 14 out of the 19

conversations in the study were in the same rangeas the scores for informal conversations betweenfriends, four were in the same range as dialoguesin radio chat programmes, and one in the samerange for doctor–patient interviews. Jahoda et al.(2009) also found that the level of asymmetry incognitive behavioural therapy sessions betweentherapists and clients with IDs was similar to thatfound by Linell et al. (1988) for informal conversa-tions between friends. However, it would be mis-taken to think that the staff members were treatingthe clients as friends, as they were the ones whowere asking the questions. As stated previouslythe balance was achieved by clients providingexpanded responses to the questions, therebyhelping to maintain the dialogue. However,another contributory factor helping clients todominate sections of dialoguewas by holding monologues and failing to payattention to the other speaker’s contributions.

The coding system did not reveal the reason whyclients used implicit turns. Holding monologuesmay have been a result of clients’ misunderstandingwhat was being said to them. Several studies haveinvestigated the communication difficulties peoplewith an ID may experience (Law & Lester 1991;Bartlett & Bunning 1997; Bott et al. 1997), and themismatch that can sometimes be found between theclients’ level of understanding and the language

used by staff (Blackwell et al. 1989; Enderby &Davies 1989; van der Gaag 1998; Purcell et al. 1999;Bradshaw 2001). Given the fact that staff used analmost equal number of implicit turns, anotherplausible explanation for the use of such turns mustbe considered; staff and clients with borderline tomild ID might want to talk about different things.The implicit turns may be related to competitionbetween the communicative partners (Linell et al.1988), and there did appear to be tensions aboutwho was in charge of the interaction. When clientstried to control the interaction by asking questions,staff performed more abrupt topic shifts, possiblytrying to bring the client back to their agenda. Asa reaction, clients may have increased their use offragmented turns. This suggests that there may havebeen a tension between the clients’ goals and thoseof staff during the dialogues. Studies concerningshared decision-making suggest that there can be adiscrepancy between staff and clients’ perceptionsabout who makes decisions (Carle 1986; Jenkinsonet al. 1992; French 1994; Antaki et al. 2002; Jingreeet al. 2006). However, as with the use of implicitturns, there are other possible explanations of whyclients and staff use fragmented turns.

The method of interactional analysis used in thisstudy shows promise in contributing to an under-standing of conversations as a mutual process interms of maintaining a dialogue. However, careneeds to be taken in the interpretation of the pat-terns found in this research. To understand whatthe patterns of interaction actually mean, it isnecessary to link the analysis more closely to whatthe dialogues are about (Linell et al. 1988). In thislight, Antaki’s and Finlay’s sophisticated investiga-tions of interactions between staff and clients withIDs combine ethnography with fine-grained conver-sation analysis to identify some conversational prac-tices that staff use to offer choices to clients withIDs (Antaki et al. 2006, 2008; Finlay et al. 2008).Taking greater account of the nature and contextof the dialogues being analysed would help toinform a more sophisticated analysis of staff–clientinteraction using Linell’s approach.

Another drawback to the present study is thatthe coding frame focused on verbal communicationper se. However, social interaction consists of bothverbal and non-verbal communication, includingfacial, vocal and postural transfer of information




(Winkielman et al. 2009). Consequently, non-verbalpatterns of interaction between staff and clients mayprovide important additional information abouthow they communicate and the power relationships.Studies of both verbal and non-verbal interactionalpatterns could inform training for front-line staff inservices for people with IDs, to enhance their com-munication skills and help foster a more client-centred approach.

References

Antaki C., Young N. & Finlay M. (2002) Shaping clients’answers; departures from neutrality in care-staffinterviews with people with a learning disability.Disability and Society 17, 435–55. doi: 10.1080/09687590220140368.

Antaki C., Finlay W. M. L., Sheridan E., Jingree T. &Walton C. (2006) Producing decisions in service-usergroups for people with an intellectual disability: twocontrasting facilitator styles. Mental Retardation 44,322–43.

Antaki C., Finlay W., Walton C. & Pate L. (2008) Offer-ing choices to people with intellectual disabilities: aninteractional study. Journal of Intellectual DisabilityResearch 52, 1165–75.

Barry M. J. (1999) Involving patients in medical decisions.How can physicians do better? Journal of the AmericanMedical Association 282, 2356–7. doi: 10.1001/jama.282.24.2356.

Bartlett C. & Bunning K. T. (1997) The importance ofcommunication partnerships: a study to investigate thecommunicative exchanges between staff and adults withlearning disabilities. British Journal of Learning Disabili-ties 25, 148–53. doi: 10.1111/j.1468-3156.1997.tb00030.x.

Blackwell C. L., Hulbert C. M., Bell J., Elston L., MorganW., Robertshaw B. A. et al. (1989) A survey of the com-munication abilities of people with a mental handicap.British Journal of Mental Subnormality 35, 63–71.

Bott C., Farmer R. & Rohde J. (1997) Behaviour prob-lems associated with lack of speech in people withdisabilities. Journal of Intellectual Disability Research 41,3–7. doi: 10.1111/j.1365-2788.1997.tb00671.x.

Bradshaw J. (2001) Complexity of staff communicationand reported level of understanding skills in adults withintellectual disability. Journal of Intellectual DisabilityResearch 45, 233–43. doi: 10.1046/j.1365-2788.2001.00318.x.

Carle N. (1986) Helping People to Make Choices: Opportuni-ties and Challenge. Talking Points No. 5. Campaign forPeople with Mental Handicaps, London.

Collins S. & Markova I. (1995) Complementary in theconstruction of a problematic utterance in conversation.

In: Mutualities in Dialogue (eds I. Markova, C. F. Grau-mann & K. Foppa), pp. 238–63. Cambridge UniversityPress, New York, NY.

Cullen C., Burton M., Watt S. & Thomas M. (1983)A preliminary report on the nature of interactionsin a mental-handicap institution. Behaviour Researchand Therapy 21, 579–93. doi: 10.1016.0005-7967(83)90050-5.

Dagnan D., Chadwick P. & Proudlove J. (2000) Towardsan assessment of suitability of people with mental retar-dation for cognitive therapy. Cognitive Therapy andResearch 24, 627–36. doi: 10.1023/A:1005531226519.

Edge J. (2001) Who’s in Control?: Decision-Making by Peoplewith Learning DifficultiesWho Have High Support Needs.Values Into Action, London.

Enderby P. & Davies P. (1989) Communication disorders:planning a service to meet the needs. British Journal ofDisorders of Communication 24, 301–31. doi: 10.3109/13682828909019893.

Finlay W. M. L. & Lyons E. (2002) Acquiescence ininterviews with people who have mental retardation.Mental Retardation 40, 14–29. doi: 10.1352/0047-6765(2002)040<0014:AIIWPW>2.0.CO;2.

Finlay W. M. L., Walton C. & Antaki C. (2008) Promot-ing choice and control in residential services for peoplewith learning disabilities. Disability and Society 23,349–60.

French S. (1994) The disabled role. In: On Equal TermsWorking with Disabled People (ed. S. French), pp. 47–60.Butterworth Heinemann, Guildford.

van der Gaag A. (1998) Communication skills and adultswith learning disabilities: Eliminating professionalmyopia. British Journal of Learning Disabilities 26, 88–93.doi: 10.1111/j.1468-3156.1998.tb00057.x.

Goble C. (1999) ‘Like the secret service isn’t it?’ Peoplewith learning difficulties’ perceptions of staff and ser-vices: mystification and disempowerment. Disability andSociety 14, 449–61. doi: 10.1080/09687599926064.

Goodley D. (2000) Self-Advocacy in the Lives of Peoplewith Learning Difficulties. Open University Press,Buckingham.

Heal L. W. & Sigelman C. K. (1995) Response biases ininterviews of individuals with limited mental ability.Journal of Intellectual Disability Research 39, 331–40.doi: 10.1111/j.1365-2788.1995.tb00525.x.

Jahoda A., Selkirk M., Trower P., Pert C., Stenfert KroeseB., Dagnan D. et al. (2009) The balance of power intherapeutic interactions with individuals who have intel-lectual disabilities. British Journal of Clinical Psychology48, 63–77. doi: 10.1348/014466508X360746.

Jenkinson J., Copeland C., Drivas V., Scoon H. & YapM. L. (1992) Decision-making by community residentswith an intellectual disability. Australia and New ZealandJournal of Developmental Disabilities 18, 1–8. doi: 10.1080/07263869200034761.




Jingree T., Finaly W. M. L. & Antaki C. (2006) Empower-ing words, disempowering actions; an analysis of inter-actions between staff members and people with learningdisabilities in residents’ meetings. Journal of IntellectualDisability Research 50, 212–26. doi: 10.1111/j.1365-2788.2005.00771.x.

Law J. & Lester R. (1991) Speech therapy provision in asocial education centre: is it possible to target inter-vention? Journal of the British Institute of MentalHandicap (APEX) 19, 22–8. doi: 10.1111/j.1468-3156.1991.tb00614.x.

Leudar I. (1981) Strategic communication in mental retar-dation. In: Communication with Normal and RetardedChildren (eds W. I. Fraser & R. Grieve), pp. 113–29.John Wright and Sons, Bristol.

Linell P. & Markova I. (1993) Acts in discourse: frommonological speech acts to dialogical interactions.Journal for the Theory of Social Behaviour 23, 173–95.doi: 10.1111/j.1468-5914.1993.tb00236.x.

Linell P., Gustavsson L. & Juvonen P. (1988) Interactionaldominance in dyadic communication: a presentationof initiative-response analysis. Linguistics 26, 415–42.doi: 10.1515/ling.1988.26.3.415.

Markova I. (1991) Asymmetries in group conversationsbetween a tutor and people with learning disabilities. In:Asymmetries in Dialogue (eds I. markova & K. Foppa),pp. 1–20. Harvester Wheatsheaf, Hemel Hempstead.

Markova I. & Linell P. (1996) Coding elementary contri-butions to dialogue: individual acts vs. dialogical inter-actions. Journal for the Theory of Social Behaviour 26,353–73. doi: 10.1111/j.1468-5914.1996.tb00297.x.

McConkey R., Morris I. & Purcell M. (1999) Communi-cations between staff and adults with intellectual

disabilities in naturally occurring settings. Journal ofIntellectual Disability Research 43, 194–205. doi: 10.1046/j.1365-2788.1999.00191.x.

Mirenda P. L. & Donnellan A. M. (1986) Effects of adultinteraction style on conversational behaviour in studentswith severe communication problems. Language, Speechand Hearing Services in Schools 17, 126–41.

Noldus (2009) The Observer XT. Noldus InformationTechnology, Wageningen.

Prior M., Minnes P., Coyne T., Golding B., Hendy J. &McGillivary J. (1979) Verbal interactions between staffand residents in an institution for the young mentallyretarded. Mental Retardation 17, 65–9.

Purcell M., Morris I. & McConkey R. (1999) Staff per-ceptions of the communicative competence of adultpersons with learning disabilities. British Journal ofDevelopmental Disabilities 45, 16–25.

Social Exclusion Unit (2005) Improving Services, ImprovingLives: Evidence and Key Themes. Interim report. Office ofthe Deputy Prime Minister, London.

Wehmeyer M. & Schwartz M. (1989) The relationshipbetween self-determination and quality of life for adultswith mental retardation. Education and Training in MentalRetardation and Developmental Disabilities 33, 3–12.

Winkielman P., McIntosh D. N. & Oberman L. (2009)Embodied and disembodied emotion processing:learning from and about typical and autisticindividuals. Emotion Review 1, 178–90. doi: 10.1177/1754073908100442.

Accepted 11 November 2011




Appendix 1

Legend of the 18 Categories as designed by Linell et al. (1988)

Symbol Description

< > Turn with clear properties of both response and initiative, the retroactive part (response aspect) being linked to themain content of the interlocutor’s preceding (adjacent) turn and the proactive (initiating aspect) involving a stronginitiative. By ‘strong initiative’ the speaker explicitly solicits or demands a response from the interlocutor.

< ^ Turn same as < > except that the proactive part is a weak initiative. By ‘weak initiative’ the speaker assertssomething or submits a proposal for comment without explicitly soliciting or demanding (but often inviting)a response from the interlocutor (prototypical expanded response).

> Turn involving a strong initiative on a new and independent topic (‘free initiative’ with no retroactive part).^ Turn involving a weak initiative on a new and independent topic (‘free initiative’ with no retroactive part).< Turn linked to the interlocutor’s adjacent turn and involving no initiating properties (minimal response).The turn is

treated by the interlocutor as satisfying the demands of (being conditionally relevant to) his own precedinginitiative (adequate response).

..> Non-locally linked strong initiative: turn with clear properties of both response and a strong initiative, theretroactive part being linked to a specific nonadjacent turn further back in the preceding dialogue.

..^ Non-locally linked weak initiative, turn with clear properties of both response and a weak initiative, the retroactivepart being linked to a specific nonadjacent turn further back in the preceding dialogue.

..< Turn linked to and treated as satisfying the demands of, a nonadjacent initiative and involving no initiating properties.This is a non-local minimal response to a nonadjacent initiative, for example to a question posed at an earlierpoint in the dialogue.

= > Turn linked to the speaker’s own preceding turn (rather than the interlocutor’s turn) involving a strong initiative.The turn is either merely a repetition or simple reformulation of the speaker’s preceding initiative or (in case theinterlocutor has only given or tried to give a minimal response) a continuation of this preceding turn.Typicallyoccurs when the interlocutor’s interjacent utterance is not accepted as an adequate response.

= ^ Turn linked to the speaker’s own preceding turn (rather than the interlocutor’s turn) involving a weak initiative.Theturn is either merely a repetition or simple reformulation of the speaker’s preceding initiative or (in case theinterlocutor has only given or tried to give a minimal response) a continuation of this preceding turn.

< = > Turn with clear properties of both response and strong initiative, the retroactive part being linked to the speaker’sown preceding turn and clearly ignoring an interjacent initiative (strong or weak) by the interlocutor.Ostentatiously self-linking initiative.

< = ^ Turn with clear properties of both response and weak initiative, the retroactive part being linked to the speaker’sown preceding turn and clearly ignoring an interjacent initiative (strong or weak) by the interlocutor.Ostentatiously self-linking initiative.

: > Turn with clear properties of both response and strong initiative, the retroactive part being non-focally linked to theinterlocutor’s preceding turn.A non-focal link usually involves remarking on, or challenging, the form and/orfunction of the interlocutor’s preceding turn.

: ^ Turn with clear properties of both response and weak initiative, the retroactive part being non-focally linked to theinterlocutor’s preceding turn.A non-focal link usually involves remarking on, or challenging, the form and/orfunction of the interlocutor’s preceding turn.

- Turn linked to, or at least possibly linked to, the interlocutor’s adjacent turn and involving no initiating properties.The turn is treated by the interlocutor as not satisfying the turn demands of, or as not even conditionallyrelevant to, his own preceding initiative.This is a (minimal and) inadequate response.

- > Turn linked to the interlocutor’s preceding turn but deferring rather than in itself providing the adequate responseto that turn.This type of contribution involves a very weak initiative, subordinated to the interlocutor’s precedingturn, and has no further initiating properties of its own.This is a deferring question asking for repetition,confirmation, or simple clarification of something contained in the interlocutor’s preceding turn.

( > Turn lacking substantial content but involving an initiative (such as a proposal) to open a new topic or subgame(the topic to be introduced in the speaker’s next turn).This is a preparatory initiative or preinitiative.

< ) Turn closing, or preparing to close, the current topic or subgame, and involving no further new initiatives.X Turn miscarriage, turn which does not count to the dialogue at all, for example when they are interrupted.




Documents

Interactional patterns between staff and clients with borderline to mild intellectual disabilities