Assessing and assisting prospective adoptive parents: Social workers’ communicative strategies in adoption assessment interviews

The assessment of prospective adoptive parents is a complex task for professional social workers. In this study, we examine the structure and function of professional social workers’ follow-up questions in assessment talk with adoption applicants. The analysis shows that adoption assessment through interviews involved a delicate and complex task that was accomplished by using a particular genre of institutional talk. This both invited the applicants’ extended and ‘open-ended’ responses and steered these responses and their development towards the institutionally relevant topics. Detailed interaction analysis demonstrates that social workers used a broad range of question types to steer and guide applicants’ responses, organising talk about specific assessment topics. On the basis of initial open-ended topic initiations and applicants’ responses, the social workers steered topic development by using follow-up moves such as polar questions and clarifying questions that asked for specification, challenged applicants’ ideas, confirmed their knowledge and encouraged self-reflection. These follow-up moves allowed social workers to achieve the progression of talk into relevant areas of investigation and constituted a central and characteristic feature of assessment interviews. We suggest that they allow social workers to accomplish two hybrid institutional goals: i) the assessment of applicants’ suitability and ii) applicants’ preparation for future parenthood.


Introduction
The assessment of prospective adoptive parents is a complex matter (Noordegraaf et al., 2008a). It involves both the assessment of whether applicants meet the practical criteria, and is also heavily reliant upon the outcome of an interviewbased evaluation of parental potential, a process that is constrained by normative notions of good parenthood (Lind and Lindgren, 2017). In Sweden, intercountry adoption (i.e. the adoption of a child from another country than your own, also referred to as international or transnational adoption) is preceded by meetings between professional social workers and prospective adoptive parents in interview-like conversations during which applicants' suitability as adoptive parents is assessed. Unlike in domestic adoption, the child(ren) who will be adopted and their specific needs are still unknown. Applicants therefore have to demonstrate and display their parenting capacity, the requirements of which are institutionally formulated and established, in order to 'pass' as eligible (cf. M€ akitalo, 2006). But, as with many institutional conversations, the assessment of prospective adoptive parents has dual, partly contradictory, aims (Van Nijnatten, 2010). On the one hand, the social workers' task is to scrutinise applicants' suitability, i.e. their knowledge and awareness of parenting issues (Noordegraaf et al., 2008a). On the other hand, the interviews have to serve as an opportunity for the prospective adoptive parents to develop a certain awareness of parental skills and the specificities of adoption (National Board of Health and Welfare, hereafter NBHW, 2009).
This twofold agenda constitutes a potential communicative dilemma for participants; social workers must steer the conversation into institutionally relevant areas of discussion but do so in ways that allow applicants to make their own contributions. This dilemma is sometimes discussed in terms of control and help (cf. Hall et al., 2014). Therefore, the social worker's role as both gatekeeper and counsellor requires different conversational responsibilities and communicative strategies, i.e., methods aimed at accomplishing specific communicative goals (Linell, 1998: 227). In this respect, adoption assessment interviews constitute a type of institutional communicative activity that is characterised by specific institutional norms and expectations as well as interactional organisation.
Previous research on institutional communication, including social work, has demonstrated that professionals' questions form a significant communicative strategy that is used to organise and achieve various, and at times divergent, institutional tasks. A question/answer format is common in social work practice and is used for multiple tasks . However, the communicative strategies (i.e., asking questions) used by social workers to pursue these multiple goals in adoption assessment interviews have thus far not received extensive research attention (but see Noordegraaf, 2008a).
In the present study, we examine social workers' communicative strategies for developing and guiding institutional topic discussions with prospective adoptive parents. By employing an interaction analysis of audio-recorded adoption assessment interviews in Sweden, the study focuses on the follow-up questions that social workers used in response to applicants' vague, digressing, or institutionally incongruent responses. We show how social workers steered the interviews towards specific areas of assessment and guided applicants to display their knowledge of institutionally appropriate aspects of adoptive parenthood, i.e. they supported applicants in articulating normatively acceptable ideas and views on adoption and parenthood. We argue that a close interaction analysis can provide crucial knowledge about the communicative effects of social workers' strategies, and, in turn, this knowledge will allow us to influence and improve the organisation of social workers' professional practice. The present study is informed by a theoretical perspective that views institutional interactions as situated communicative practices in which meanings and tasks are accomplished through dialogue (cf. Linell, 1998Linell, , 2009Sarangi, 2000).

The adoption assessment process in Sweden
The adoption process preceding intercountry adoption in Sweden 1 includes several steps: a mandatory preparatory parenting course, assessment interviews with an assigned social worker, the written report, and the final decision, which is made by the municipal social welfare committee. By means of this procedure, the Swedish state aims to ensure that children will be adopted by families who can care for the children's needs and support them properly (Lindgren, 2015;NBHW, 2009). During this process (i.e. the parenting course and assessment interviews), which is designed according to the Hague Convention on Intercountry Adoption (HccH, 1994), it is intended that the prospective parents will develop knowledge and gain a greater understanding of adoption, prepare for it at a psychological level, and reach "a well-founded decision on adoption" (NBHW, 2009: 30).
The assessment interviews require collaboration between the prospective parents and social workers, in which the prospective parents demonstrate "frankness and active participation" (NBHW, 2009: 39). Notably, assessment interviews have multiple goals: social workers must investigate applicants' suitability, contribute to their preparedness for adoption by providing an opportunity for them to gain greater insights into adoption and assisting them in their process of maturity, and prepare the way for a final report. Therefore, the social workers guide adoption assessment interviews through several topics/subject areas, including childhood experiences, relationships and everyday life, future parenthood, child-rearing, and their motives for adoption (NBHW, 2009). These interviews are also tied to the production of an institutional document, a final report (M€ akitalo, 2005), in which social workers are required to present applicants' views on specific issues (for instance, their reasons for adopting a child) (NBHW, 2009: 104-107). The recommendation as to whether or not the applicant(s) should be granted consent to adopt is made on the basis of information about the applicant that is important in relation to their suitability. Adoption assessment interviews are interactionally organised as topic-related question/answer sequences, and they constitute a multifaceted, hybrid practice of institutional communication (Wirz� en and Lindgren, 2020).

Professionals' communicative strategies in institutional interaction
A number of institutional practices have been shown to combine various institutional goals that position the professional as both a gatekeeper and a helper, whose institutional task is to prepare the client (i.e., assisting them to gain knowledge concerning an institutional procedure) and assess if and when the client is ready (e.g., has thought through his/her decision, Sarangi et al., 2004;Van Nijnatten, 2010). Such institutional practices are accomplished through professionals' deployment of larger communicative structures (e.g., topic discussions, reflective frame, etc.), and communicative resources (e.g., questions in different formats). In many types of hybrid institutional interaction, especially in counselling, professionals take a non-directive stance and strive to encourage the client's self-direction. They solicit the client to present his/her perspective before they give an assessment, advice, or information. In other words, professionals avoid offering advice or solutions, and instead encourage clients to find and articulate their own way (Vehvil€ ainen, 1999(Vehvil€ ainen, , 2003. During the first stage of institutional talk, the professionals can organise their question/answer sequences as "perspective display sequences", which aim to solicit the client's perspective before professionals present their own (Silverman, 1997: 30-31, on HIV counselling). Hence, inviting, exploratory questions are used to solicit and guide the client to present his/her perspective, rather than giving instructions about what to do or how to think. Similarly, Sarangi et al. (2004), in a study on genetic counselling, show that the assessment of clients' preparedness for an institutional procedure (genetic testing) is achieved by soliciting the clients' responses within a reflective frame. Thus, the counsellor performs as an active listener by offering the client the opportunity to speak and reflect upon a specific topic. In Sarangi et al.'s study, the reflective frame was established and sustained through various open-ended questions within the larger topic discussion; in this way, the counsellor explored the client's emotional state (e.g., anxiety) and preparedness, and the client was assisted to understand the implications that might follow any decision about testing. This institutional interaction was clearly related to the professional's role as a gatekeeper who was conducting an in situ exploration and assessment of the client's state, preparedness, and knowledge.
Similar studies on adoption assessment interviews (Noordegraaf et al., 2008a(Noordegraaf et al., , 2008b have shown that what can be seen as a perspective display sequence is used by social workers to initiate specific areas of investigation in order to collect or produce certain information. In Noordegraaf et al.'s study, such a communicative structure allowed social workers to discover what applicants considered important or relevant in relation to adoption and, on the basis of these responses, social workers were able to proceed with hypothetical questions that added complicating circumstances and deepened the discussion (Noordegraaf et al., 2008b). These hypothetical questions required applicants to demonstrate their skills and solicited information relevant to the social workers' assessment of their suitability and future parenting qualities (Noordegraaf et al., 2008a). It can be suggested that such communicative strategies provide methods for progressing the assessment interview and guiding interviewees' thinking and responses. Similarly, in an interview study with social workers about adoption assessment interviews, Wirz� en and Lindgren (2020) showed that social workers wanted the applicants to talk and describe their lives in particular, institutionally relevant, ways. To achieve this, they assisted the prospective adoptive parents by encouraging and guiding them to reflect upon their experiences and perspectives; thus, the institutional agenda (i.e. goals and tasks informing the activity) required active interactional work from the social workers.

Questions in institutional talk
Prior research has suggested that, even when institutional talk is conducted so as to facilitate the clients' self-direction, their responses are not free from the institutional agenda. Rather, by using various types of questions, professionals can guide the direction of clients' responses (V€ ahvilainen, 1999). In this respect, interactions in institutional encounters are informed by a predetermined agenda that is designed in situ between the participants (Vehvil€ ainen, 1999), and the professional's management of topic-specific questions is shown to be related to the talk agenda (Noordegraaf et al., 2008b). The ways in which questions are structured communicate the purposes of an activity and the specific institutional knowledge and agenda that characterise the practice (Arminen, 2005;Ehrlich and Freed, 2010;Heritage and Clayman, 2010;Linell, 1998;M€ akitalo, 2005;Vehvil€ ainen, 1999). Accordingly, asking questions is not a "neutral activity"; rather, questions constitute the institutional interactional context through their format (Heritage, 2002: 86).
Professionals' question design (i.e. how a question is formulated and what answer it requires) can project the appropriate response. For instance, questions that embody response preference, such as Yes/No (polar) ("are you married?") or declarative questions ("you're married?" or "you aren't married?") solicit the clients' Yes/No ratification of specific formulations and lead talk into particular areas of the interview (Heritage and Clayman, 2010: 142). In addition, "candidate answers" (Pomerantz, 1988: 278), which formulate and offer a version of the recipient's answer, can be used to guide "the recipient to respond in a certain way" (Arminen, 2005: 93) and serve as an information-seeking strategy. Communicative strategies that provide "cues" concerning responses are also common in learning situations, where they assist recipients in attuning to what is relevant in a specific subject and guide them towards a correct answer (on teaching "cues", see Arminen, 2005: 127).
Notably, the research on specific communicative formats shows that an analysis needs to be sensitive to the character and goals of the particular institutional context (Noordegraaf et al., 2008a). For instance, within certain institutional practices, Yes/No questions can receive extended responses that accomplish a different kind of work than simply eliciting minimal answers from clients (Hakulinen, 2001). By using extended responses, recipients can affirm the proposal and simultaneously do repair-work on previous talk; for instance, they can change and reformulate the posed question. Yes/No questions can also position the client to provide an extended account concerning their perspective or actions. As demonstrated by Aronsson (2018), parents in child custody examinations may respond with expanded responses even though the question projects a minimal Yes/No response. In Aronsson's study, negative interrogatives, such as "haven't you?" or "don't you?", were used by attorneys to move the examination "into a more adversarial direction" (Aronsson, 2018: 48). Instead of single Yes or No responses, the parents produced extended defensive accounts that explained or justified their actions. Overall, research on institutional interaction suggests that, in their answers, clients are oriented, not only to the local question, but also to the overall institutional task.

Data and analytical approach
This study is part of a larger research project investigating the assessment of prospective adoptive parents for intercountry adoption in Sweden. The data consists of audio-recorded assessment interviews between 11 prospective adoptive parents 2 and six social workers. It comprises 36 hours of recordings of 24 assessment interviews. The study has been approved by the Ethical Board Committee (Dnr 2015/ 111-31). All participants gave their informed consent to participate and all names and potentially identifying details have been anonymised.
The analysis is informed by conversation analytical methods that direct attention towards how social actions are accomplished through talk-in-interaction and how meaning is constructed between participants through their orientation to each other's turns at talk (Heritage and Clayman, 2010;Juhila et al., 2014). The analysis therefore focuses both on what is said in the interaction and how it is formulated, i.e., turn design (Noordegraaf et al., 2008a). Notably, we analyse social workers' and applicants' talk and communicative strategies within the context of the institutional activity, i.e. the process, goals, and practices of adoption assessment (Linell, 2009). In doing so, we aim "to identify and explicate the ways in which interactional activities contribute to the accomplishment of institutional tasks" (Arminen, 2005: 37).

Analytical procedure
The analytical process involved the first author transcribing all the audio-recorded interviews verbatim according to the Jeffersonian transcription system (Jefferson, 1984, see Appendix 1), which aims to represent the delivery of talk. The analysis involved repeated listening to the recordings, and re-reading of the transcripts. Conversational patterns that frequently recurred in the data at large were identified. The analysis shows that the social workers introduced topics on adoption and that they used a broad range of question formats to develop the assessment interview. On the basis of applicants' responses, the social workers progressed the talk by using various follow-up moves (i.e. strategies used to respond to prior talk, and to move the talk on). We became particularly interested in their institutional functions, and a collection of follow-up moves was made. Their communicative functions and grammatical formats were identified. Next, we analysed these interactional situations in detail: how social workers introduced follow-up questions, applicants' responses to such moves, and what interactional function they accomplished.
In the presentation of our analysis, we use extracts from several adoption interview cases to illustrate recurring forms of follow-up moves and typical patterns of social interaction. These extracts have been translated from Swedish into English for presentation to an international audience. 3 The analytical findings are based on our work with the entire data corpus. 4 We specifically focus on follow-up moves that were used to respond to applicants' vagueness or divergence from institutional expectations. In the extracts, we use abbreviations to refer to the social worker [SW] and to applicants as "prospective mother" [PM] or "prospective father" [PF].

Findings
The analysis shows that the adoption assessment interviews displayed a recurring structure that was closely linked to the discussion of adoption-relevant topics. Topic discussions were initiated by the social workers with the help of openended questions (Noordegraaf et al., 2008b) and the applicants were required to demonstrate relevant views and knowledge about parenting and children. Notably, the applicants' responses to the social workers' questions varied in terms of clarity, relevance, elaboration, and appropriateness in relation to the institutional expectations concerning their parental suitability and knowledge about adoption and parenting. In cases where the applicants' responses digressed, or in other ways failed to converge with the institutional agenda, social workers employed followup moves to steer the progression towards discussion of an assessment-relevant topic. In this respect, the social workers' interview questions highlighted a normative perspective (cf. Heritage, 2002) and provided a certain amount of guidance. These communicative strategies will be presented in the following four sections: i) clarification questions; ii) Yes/No (polar) questions; iii) juxtapositional summaries; and iv) teaching cues. These strategies were deployed in order to guide the applicants to present assessable, institutionally relevant, and appropriate information, to assist them to reflect from a particular point of view, and to assist them in developing preparedness for parenting.

Guiding applicants' responses through simple clarification questions
To initiate a topic, social workers used open-ended questions, e.g.: "what do you think about. . .?" or "what are your thoughts about. . .?" and invited applicants to respond with unconstrained talk and thinking; the applicants were invited to think aloud and display their understanding of a subject (Sarangi et al., 2004). Applicants' responses were usually broad; they touched upon various aspects of parenting, but also frequently lacked clarity. Social workers then used the information provided to guide applicants' responses: a simple strategy involved follow-up questions that requested clarification of a concept or of applicants' understanding. In this way, applicants' broad responses could be developed into more concrete, institutionally relevant forms, adoption-relevant issues were brought to the surface, and key notions were reflected upon and clarified (cf. Noordegraaf et al., 2008b).
In Excerpt 1, the social worker asks a female applicant to explain why she longed for a child (longed to become a parent).
Excerpt 1. Single female applicant, 1 social worker. (1.5) like experiences an' (1.1) fun things 15. with The applicant's response (lines 2-3) is a bit vague and rather general "well it's probably about wanting to ha-be a family". In her follow-up move, formulated as a "simple question" (Heritage, 2002: 62), the social worker recycles the concept of "family" and invites the applicant to clarify "an' what's in this 'being family'?" (line 7). The social worker requests an expanded, clarifying description of this rather broad but, for the adoption process, crucial concept. By introducing her follow-up move with "an'" (i.e. Swe: "˚a"), the social worker makes visible her agenda -guiding and transitioning the talk into a more concrete dimension, while sustaining the coherence of the interview topic development (Noordegraaf et al., 2008a;Vehvil€ ainen, 1999). In response, the applicant starts to unpack the meaning of "family" by developing and elaborating upon her thinking in situ. Such elaboration is important for the applicant's own self-reflection and for her to gain preparedness for adoption, i.e., becoming a "family". Overall, although the communicative strategy of clarifying follow-up questions (i.e. asking the applicant to explain their views on a concept) is seemingly simple, it accomplishes several institutional goals: it elicits information that is needed for the final assessment report, and engages the applicant in in situ reflection and reasoning about adoptionrelevant issues.

Guiding through social workers' yes/no questions
In cases where the applicants' responses diverged from, or in other ways did not suit, the institutional agenda, social workers used follow-up moves to guide them by exploiting their prior responses in order to highlight the incongruent and potentially normatively conflicting features. Repeatedly, Yes/No polar questions (e.g. "you don't mean a slap" or "is that what you intend to do?") were used as a communicative strategy to mildly challenge the meaning and/or implications of applicants' prior responses. Such questions positioned the applicants to confirm or reject the message formulated by the social worker. Simultaneously, the format of these Yes/No questions indicated the social workers' normatively preferred responses (Heritage, 2002). The topical progression in the assessment interview was achieved because such challenging questions, in addition to receiving a confirmation or rejection, resulted in the applicants engaging in extended responses and repair-work (Aronsson, 2018;Hakulinen, 2001;Linell, 2009) aimed at clarifying and adjusting their answers in the institutionally relevant direction. In Excerpt 2, the applicant's vague and complicated response about what it means to be a "loving" person ends with them formulating an ambiguous and potentially discrediting clarification: "someone has hard love" (line 22). As a follow-up move, the social worker asks a Yes/No question that exposes the normative inappropriateness of this response and requires the applicant to take a stance.
person (1.0) is lo-meaning well that someone 12.
is loving (0.6) being there for hugs an' being 13.
one is in that carin-loving thing .hh but also 15.
being loving ( After the applicant's ambiguous reflection on the concept of "loving" (line 22), the social worker takes over the conversational floor and, as a topic continuation, poses a Yes/No polar question that points out its problematic features: "an' here you don't mean a slap" (i.e. physical punishment, 5 line 25). This formulation clearly involves a sensitive issue, witnessed by the embarrassed laughter of both the applicant and the social worker, which indicates their alignment on the matter (lines 26-29). The social worker's polar question is formulated as a negative declarative that favours a confirmatory "no" response, which is subsequently produced by the applicant (line 26 "naeh"). Interestingly, this simple confirming response is not enough: the applicant herself furthers the development of this topic, expanding (Linell, 2009) and re-formulating her response into a more acceptable one: "'I love you but you still have rules'" (lines 32-33). Thus, by pointing out the inappropriate features of the applicant's earlier response, the simple, but challenging, polar question makes the applicant accountable for repairing and clarifying her previous talk about "loving" (Aronsson, 2018;Hakulinen, 2001). In line 37, the social worker launches yet another follow-up move that assists the applicant in formulating the institutionally relevant response. This time, her Yes/No declarative question invites a response that, instead of challenging the applicant's views, summarises, on the basis of her explanation (lines 32-33), an institutionally acceptable view on parenting style and offers it for the applicant's confirmation: "hard love" implies that "rules and structure are important" (line 37). In this way, the social worker contributes to formulating and co-constructing, through the applicant's acknowledgement, an institutionally appropriate answer. The applicant's description of a loving upbringing can then be assessed as suitable and the topic is closed.
As demonstrated above, on the basis of the applicant's responses, the social worker used Yes/No questions as a format that implicitly signalled institutionally relevant responses. By confirming or rejecting the version that indicated, through this polar formulation, what was at stake, the social worker assisted the applicant in the formulation of an institutionally preferred description. Interestingly, while the larger interactional organisation could be characterised as demonstrating a certain non-directiveness, the social worker's questions and candidate responses provided opportunities to clearly direct the applicants to attend to specific assessment concepts, and relevant responses. These institutional communicative strategies indicated a preferred stance and highlighted the relevant formulation and discussion issue. Simultaneously, they allowed the applicants to confirm or disconfirm it.

Guiding through social workers' juxtapositional summaries
By using juxtapositional summaries (i.e. putting two contrasting perspectives or views side by side) of applicants' previous talk, and offering them for the applicants' response, the social workers pointed out discrepancies between two perspectives; commonly, what is best for the child and what seems to be best, or least complicated, for the prospective parents. This juxtapositional communicative strategy did not explicitly criticise the applicants' views, but it did solicit their response by bringing into the open any inconsistencies or vagueness in relation to the institutional preferences, and made the applicants accountable for attending to and clarifying their responses.
Prior to Excerpt 3, the applicants were openly asked to describe their thoughts about parental leave. 6 The applicants highlighted their emphasis on the child's needs as the decisive principle in how they would share the parental leave, explaining that they would share it equally. The discrepancy arose when the PF said that his work situation would be the key factor instead. In Excerpt 3, it is these responses that the social worker is recycling in her juxtapositional summary (lines 1-6).
Excerpt 3. Couple applicants, 1 social worker. can see that myself (that) it's easier to let 27. go like okay you will have (0.6) hundred percent 28.
here instead of going in and have eighty 29.
percent an' then anyway [will have] to¼ 30. PM: [ mm ] 31. PF: ¼like make (i.e. work) that 20 percent The social worker highlights the discrepancy between two contradictory perspectives: she juxtaposes the child's perspective and the parents' perspective (their work situation), thereby highlighting the dilemmatic aspects of the applicants' responses and then leaving the conversational floor open (cf. Arminen, 2005) (lines 1-6). The self-repairs and re-formulations that follow indicate that she is dealing with a potentially morally sensitive issue. The prospective father takes the responsibility to respond and clarify his position by articulating an account in which the reason for these difficulties is beyond his control: he points out the responsibilities associated with his position as head of department/superior (lines 12-14, 16-17) (PF's account is supported by PM, line 15). There is, however, some vagueness in the prospective father's account, because it does not explicitly deal with what he will do with his parental leave. An institutionally acceptable response is then formulated with the social worker's assistance. With a subsequent declarative, the social worker articulates a candidate understanding of the applicant's answer: "but to be completely off you think it's not that which is the concern, rather it's more about if you would work part time at some point later" (lines 19-22). The social worker's follow-up move presumes and invites the applicant's alignment with her description, which is "suitable for the institutional framework" (Arminen, 2005: 93), namely, that being on full-time parental leave is not a problem and that the child's needs are the key factor in future parenting. This important information must be articulated by the applicants in order for them to be assessed as suitable parents, and PF continues to explain his position on the basis of the support provided by the social worker's candidate understanding.
Overall, in her follow-up moves, the social worker highlights the discrepancies through a juxtapositional summary of the applicants' talk, and then invites their responses. This juxtapositional format positions the applicants to face this inconsistency, indicating the negative features of their responses and, consequently, of their qualities as prospective parents. By presenting these versions for the applicants' confirmation or rejection, the social worker guides their response towards an answer that is suitable for this institutional context and can therefore be positively assessed.

Guiding through social workers' teaching cues
In cases when the applicants' responses were not consistent with institutional norms and preferences, and they indicated or displayed a lack of knowledge, the social workers could use communicative strategies that resembled teaching cues from educational settings (e.g. repeating fragments or asking a question that directs the focus towards something specific) (Arminen, 2005). These communicative strategies shaped the applicants' responses, thinking, and knowledge by guiding them, like learners, "to attune to the relevant dimension of the subject" (Arminen, 2005: 127). Likewise, the professionals reformulated their questions (for instance, replacing a word) by "self-editing" (Vehvil€ ainen, 1999: 104) to make it clear what to focus on and to avoid any misunderstanding by the applicants.
In Excerpt 4, the applicants had been asked to talk about their prospective organisation of parental leave and, although PM reasoned that they would "share it approximately equally", PF reasoned from the employers', and his own, perspective, suggesting that longer parental leave periods would be preferable (lines 1-2). Both applicants also indicated their insecurity and lack of knowledge (lines 6, 10). It is these dilemmatic features and inconsistency between the institutional perspective, i.e. what is best for the child, and the employers' perspective that were used by the social worker as a question-based teaching strategy to guide the applicants (lines 16-17, 20).

PF:
longer emp-longer periods are perhaps better 2.
for for the employer 3. PM: well but 4. PF: an' for oneself as well (xxx) well perhaps it 5.
your parenting course?
Instead of criticising the applicants' responses directly, the social worker spells out a discrepancy in PF's talk and formulates "but we think for the child", changing the perspective of the response (lines 16-17, 20) (Arminen, 2005). While such a juxtaposition of contrasting views can generate clarifications and accounts from applicants (see Excerpt 3), here, the applicants do not elaborate upon their responses, but simply acknowledge the social worker's turn (lines 21-22). The social worker then reformulates her prior question "e: an' wh-what can be best or what can one expect to be best for a child" (lines 23-24). Her follow-up move is characterised by "selfediting", which models the direction of the preferred response from the PF (Vehvil€ ainen, 1999: 104). With the help of a question that works as a teaching cue: "what can be best" (line 23), the social worker directs the applicants towards what to focus on in their response. The social worker's self-correction to a generic mode ("what can be best" -> "what can one expect to be best", lines 23-24) invites the applicants to engage in reflective talk and "thinking aloud" (Sarangi, 2000: 9), rather than displaying knowledge. However, the applicants simply claim their lack of knowledge ("I don't know", "naeh in fact I don't know either", lines 29-30). While such an outright acknowledgment of one's lack of knowledge could lead to the social worker producing the correct information, giving the right answer (cf. Vehvil€ ainen, 2003), in this case she maintains her non-directive stance and gives cues that guide the applicants' responses, referring to the preparatory parenting course (lines 32-33).
Overall, the social worker balances between a non-directive position, which invites the applicants to think and reflect (Sarangi et al., 2004;Vehvil€ ainen, 2003) and the position of a guide who points out the right direction or right answer (Arminen, 2005: 131, on learning formats). By steering the conversation through the use of elaborating questions, the social worker turns the adoption assessment interview into a learning activity: her questions give the applicants information about what it is important to pay attention to in this context, i.e. what is best for the child (cf. Silverman, 1997), and guides them into being able to know and produce the institutionally appropriate response by themselves.

Discussion
The assessment of prospective adoptive parents in Sweden is a practice that has several institutional goals, most obviously the examination of applicants' suitability and eligibility to adopt a child, as well as preparing them for parenting (NBHW, 2009). This study employed interaction analysis to examine the communicative strategies used by social workers to achieve these multiple goals. The study illustrates the concrete methods employed to progress the development of the assessment topic, and demonstrates how the applicants were guided to produce assessable, reflective responses that were appropriate for the institutional framework. The analytical focus was directed towards the ways in which the social workers responded to applicants' answers that were vague or diverged from the institutional agenda concerning their perspectives, plans, and knowledge about parenting and adoption. The findings from this study provide knowledge about practice that can inform the education and training of professional social workers.

Functions of social workers' follow-up questions
The study reveals a recurring interactional organisation of the assessment interview that involved the development of specific topics, accomplished through the social workers' question-based steering and guidance of assessment talk. As demonstrated, the social workers followed a well-established interview format (e.g. Juhila et al., 2014;Silverman, 1997) to invite applicants to engage in a specific topic with an initial open-ended question that set the agenda for talk. The social workers' questions were used as an invitation for the applicants to reflect, think, and present their knowledge and perspectives on adoptive parenting (cf. Sarangi et al., 2004). Notably, the study highlights a communicative dilemma for the prospective parents, related to the open-ended questions, which were rather broad (see Excerpt 1). While this open-endedness characterised the assessment interviews as non-directive, and such interactional moves allowed the social workers to gain information about applicants' perspectives and knowledge (e.g. Noordegraaf et al., 2008b), the study has revealed that responses to such questions could be cumbersome, rather broad, and characterised by vagueness (Excerpts 1, 2). Moreover, by developing the topic under discussion, the applicants sometimes presented reasoning and responses that not only digressed from the institutional agenda, but also indicated certain divergences and discrepancies within the views they articulated (Excerpts 3, 4). The social workers' follow-up moves, formatted as various communicative resources (i.e. various types of questions), guided the prospective parents, through their interactional design, to produce responses that were relevant to the institutional agenda, and therefore assessable.
The interactional analysis demonstrates several communicative strategies in detail. The question design (within the larger communicative structure of topic discussion) was significant in steering the direction of responses. The social workers guided the interviews by using simple questions that solicited clarifying responses from the prospective parents (Excerpt 1). These included asking the applicants to clarify and elaborate upon their perspectives and thinking about concepts that were crucial for displaying their knowledge and motives for adoption. Other formats more clearly guided the responses in the institutionally appropriate direction. For instance, contradictory responses from the prospective parents could be challenged and their normatively divergent features made visible by the use of Yes/No questions or juxtapositional summaries of what they had said (Excerpts 2-4). As demonstrated, the social workers extracted and summarised the relevant dimensions of the responses and offered them for the recipients to confirm or reject (Excerpts 2, 3). For instance, the social workers offered their candidate understandings of the applicants' answers, and solicited their confirmation of this version (Excerpts 2, 3). Through their design, these Yes/No questions indicated the preferred response: by confirming or rejecting the version that was at stake, the applicant was assisted in presenting a specific perspective or view. Such questions also mildly challenged the applicants and solicited their accounts, elaboration, and modification of their views and responses (Excerpts 2-4). In such cases, simple Yes/No responses were not enough (cf. Aronsson, 2018). In these ways, the applicants were implicitly assisted in formulating institutionally relevant responses that provided information for the final report and could be assessed appropriately.

Hybridity and non-directiveness
The hybridity of assessment interviews was also achieved through the social workers' use of questions as teaching cues when teaching and guiding the prospective adoptive parents' thinking (Excerpt 4). The guidance provided by social workers was oriented both towards the applicants' lack of knowledge in an assessment situation, and towards preparation for parenting. By providing some information, but also avoiding pointing out exactly what the appropriate responses were (Excerpts 1-4), the social workers balanced their roles of assessor and educator (cf. Noordegraaf et al., 2008a). The educational and preparatory aim of the assessment process called for communicative acts that positioned the social workers as educators and the applicants as learners; simultaneously, the significance of assessment made the social worker into a gatekeeper, who evaluated and reacted to what applicants said in the dialogue. Notably, the interactional organisation and hybridity of adoption assessment interviews can be interpreted in relation to the various goals of the assessment procedure (NBHW, 2009), including the task of producing a final written report that summarises the institutional argument for giving, or withholding, consent for the applicants to adopt a child (cf. M€ akitalo, 2005).
This study suggests that these institutional interactions, characterised by the professionals' questions, combined non-directiveness (in that the questions were used to solicit the prospective parents' perspectives) with, at times, rather direct and explicit guidance. While the open-ended questions provided interactional spaces for the applicants' rather unconstrained responses, such responses were not always sufficient to fulfil the hybrid institutional agenda, in which assessment and preparedness constituted the intertwined goal of the institutional practice. Through the design of their follow-up moves, the social workers actively guided and directed the applicants' responses. Although they did not present concrete pieces of knowledge for the applicants to repeat, and did not formulate the assessment, the design of the follow-up move could be used to both implicitly articulate the relevant institutional perspective, and to present it for the applicants' approval (in a way, ventriloquising the response). Responses were thus co-produced in dialogue, and with guidance from the social workers' questions (cf. Vehvil€ ainen, 1999).

Implications for practice
Social work practice fundamentally relies upon talk between professionals and clients. It is through talk that social work and its institutional tasks are accomplished. Hall et al. (2014: 178) suggest that analytical methods describing in detail what "actually happens in practice" can be used to demonstrate this complex area of professional work. Hence, knowledge gained on the basis of detailed interaction analysis can be used in social workers' professional development and training. This knowledge can be used to encourage social workers' own examination of and reflections upon their practice, and in formulating guidelines for interview practice. In such a way, collaboration between researchers and professionals can generate practice-relevant results and allow the development and organisation of institutional practice in ways that are beneficial to clients (O'Reilly et al., 2020). Overall, by demonstrating in detail how the social workers' communicative strategies were deployed in situ, this study adds to the contributions of previous work on adoption assessment (Wirz� en and Lindgren, 2020). Notably, the deployment of interaction analysis methods allowed us to show how adoption assessment is accomplished in concrete, situated ways. We suggest that a meaningful focus for future research could involve detailed attention to question formats from different social work practices and an analysis that takes into account how context-specific discourses and theories inform the questioning. Such work, together with insights from research on social workers' own notions about their complex institutional task, could contribute to a holistic understanding of professional social work practices, and create an opportunity to engage professional social workers in dialogue for practice development.

Declaration of conflicting interests
The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding
This work was supported by the Swedish Research Council for Health, Working Life and Welfare, FORTE [Forskningsr˚adet om H€ alsa, Arbetsliv och V€ alf€ ard] under Grant 2015-00542.

ORCID iD
Madeleine Wirz� en https://orcid.org/0000-0003-4531-0400 Asta Cekaite https://orcid.org/ 0000-0003-4580-3002 Notes 1. The Swedish state is responsible for the assessment and regulation of both intercountry and domestic adoption; only professionally educated social workers, employed by the state and working in state-financed social work offices, are permitted to assess prospective adoptive parents. 2. Four couples and three single parties.
3. The quality of the presentation of transcripts, and in particular translations of transcripts, is especially important for qualitative research (Nikander, 2008). Throughout the process, we have worked with the original (Swedish) transcripts and the translations into English have been done carefully in dialogue with researchers and translators. For lack of space, the original transcripts are not included in the article. 4. The phenomena in focus for this study should not be considered representations of general social phenomena but each one as a "phenomenon in its own right" (Per€ akyl€ a, 2016: 415). Every utterance is locally understood by the participant in interaction and it is through participants' responses that the local meaning is displayed. Rather than striving to make claims about the findings' generalisability beyond the research context, studies in institutional settings allow for comparison with other institutional practices (Per€ akyl€ a, 2016). 5. Physical punishment of children has been forbidden since 1979 and has legal consequences. 6. In Sweden, both parents are entitled to parental leave and are expected to share it.