Characteristic interviews,different strategies

Voluntary informed consent

Safeguarding the voluntary informed consent of informants is often cited as a methodological challenge in the use of people with intellectual disabilities as informants. Norwegian research studies in social science comply with guidelines given by The National Committee for Research Ethics in the Social Sciences and the Humanities in Norway (NESH, 2013). Voluntary informed consent includes three elements, all of which must be met to allow participation in a research study (Ellingsen, 2009). Voluntary participation also includes the right to withdraw at any time, without stating a reason. Consent depends on voluntariness and capacity to make decision based on the information provided. Therefore, participants must be given genuine opportunities to reserve themselves without encountering inappropriate pressure. In some countries, a valid consent is defined as a voluntary decision, which is given only in the absence of undue influence by a third party (Dennehy and White, 2012; Veenstra et al., 2010). Individuals, who do not have authority to consent to participate in research, must still provide their assent. Assent is an active affirmation of a desire to participate but differs from consent that is provided from an individual with legal authority to do so. NESH (2013) doesn’t differ between the concepts consent and assent. Nevertheless, the guidelines from NESH show specific directions if individuals are unable to grant informed consent:

Individuals unable to grant informed consent shall generally only be included in research that cannot perform on individuals who are able to grant consent, and can show it to be probable that the research in question is of direct substantial benefit to the individual or group being studied. (NESH, 2013: 16)

In Norway, parental consent is required when the studies including children under the age of 15. Otherwise, children’s own consent is necessary from the time they are old enough to express their own opinion. However, voluntary consent can be a particular challenge because people with intellectual disabilities often have a strong respect for authority and thus find it difficult to set their own limits. Consent depends on information. In this context, a participant should be able to understand information about the research project, the potential influence of the project, and his or her role in the project. Consent is also related to a participant’s competence to give consent as well as an assessment of the participant’s capabilities, which is linked to an evaluation of the areas to which this consent applies. Several studies show that in situations in which it is difficult to obtain satisfactory consent, this problem can be solved by other persons representing and providing consent on behalf of individuals with intellectual disabilities (Dusseljee et al., 2011; Veenstra et al., 2010).

In the author’s interview study, informed consent was obtained by students and parents together (Sigstad, 2013a, 2013b). The students were from 13 to 16 years old, thus a parental consent was required. Moreover, to ensure that the information was understood, the parents were involved. The children hopefully would find it easier to argue their views to the parents, if they did not wish to participate. In this way, their contribution could potentially reduce any imbalance of power. The teachers forwarded an informational letter to the students and parents involved in the project. The parents were asked to communicate simple information about the project to their children and to describe what their participation would entail. The parents and students provided permission for the interview by giving a response form to the teachers and by giving a response form directly to the research manager, including signatures by both students and parents. This means that both gave their consent.

Where there was a need for further information and clarification, the parents made direct contact with the research manager. Some parents availed themselves of this opportunity. More comprehensive information provided on the phone appeared to increase the decision to join the study. Difficult words in the information letter could be clarified, thus the parents were better able to explain to their children what it was about. Information about the project and questions concerning participation were repeated before the actual interview. The informants then had the opportunity to give their consent again verbally or, alternatively, to withdraw their consent. All of the informants consented to attend when they met for an interview.

The extent to which interviewees understand what the research process actually entails or the implications of their involvement appears to be particularly challenging in the process of obtaining voluntary informed consent. Previous studies have shown that people with intellectual disabilities may have difficulty fully understanding the main aims of the research, and some parts of the research process appear to be more problematic than others (Arscott et al., 1998; McCarthy, 1998). It appears to be relatively easier to understand the information that is needed to agree to meet for an interview. However, it may be more difficult for informants to give consent for the follow-up after the research interview is completed, perhaps because of a limited understanding of the risks and benefits associated with such a research project or because of the rights of informants to refuse to participate or to withdraw from the study (Arscott et al., 1998).

For example, it may be difficult to understand that communication in an interview will be analyzed, categorized, quoted, and published. In the author’s study (Sigstad, 2013a, 2013b), the information was not formulated in this way to the informants. Instead, the research manager chose to rely on consent given by the students and parents together based on their communication at home, so that together they could understand the consequences of attending and feel free to withdraw at a later stage, if necessary. We saw that parents were more capable of conveying information about the research project. They could explain, clarify misunderstandings, and thus provide the child the opportunity to refuse. One can never be quite sure if the parents actually have persuaded the children to participate or not. Therefore, in this study we emphasized the importance of repeating the information to the informants when they met for an interview. Then they got a new opportunity to withdraw.

A review of 20 research studies over a period of 20 years (1986–2006) showed that the focus was directed toward people with intellectual disabilities and their participation in research (Cleaver et al., 2010). With the exception of three studies that excluded the consent of people with intellectual disabilities, these studies showed that alternative strategies could be useful in obtaining satisfactory consent. The strategies involved obtaining consent only from the person with intellectual disabilities or, if he or she was not competent to give consent, contacting a support person who could provide consent on behalf of the informant. In all cases, this approach involved contacting both the person with intellectual disabilities and a support person to request that they provide consent together. A study by Boxall and Ralph (2009) supported these findings and indicated that service providers or other individuals with whom people with intellectual disabilities may be more familiar may be the most suitable people from whom to acquire satisfactory agreement. To ensure that the content of the research information is understood, it may also be beneficial to repeat the information several times and in different ways (Kittelsaa, 2010; Wennberg and Kjellberg, 2010), as in the author’s study.

The most serious consequences associated with the challenges of obtaining consent from informants with intellectual disabilities primarily pertain to the risk that potential informants will cease participation in studies in which these individuals could likely have contributed a great deal (Ellingsen, 2009). Nevertheless, voluntary informed consent requires that informants understand that participation is voluntary, which requires that information to be understood (Kittelsaa, 2010). In the author’s study (Sigstad, 2013a, 2013b), possibly it had been better if the informants had received an easy-read information sheet by themselves. In that case, we could be less dependent on relying on parents’ ability to explain. Whatever, it is important to ensure that participants have the competence needed to provide consent or to obtain the support that they need from a significant other. Hence, the goal of including as many people with intellectual disabilities as possible in research requires strong planning and arrangement of the consent process. Planning will involve collaboration with possible service providers or with someone close to the informants, the preparation of those who will be participating, facilitation with regard to available information about the research project, expertise on intellectual disabilities, and knowledge of the relevant informants who want to participate. Another way of obtaining consent from informants with intellectual disabilities can be done as a process rather than as an event or two, as was the case in this study. In a process, it is possible that a rather regularly negotiating consent with the participants as the study progresses and the participants see what will happen with their data. In this way, participants can hopefully achieve greater self-determination and control in the research process, and a voluntary informed consent is thus better addressed.

Challenges associated with communication in the interview situation

Previous studies refer to the aspects of research interviews that appear to be particularly influenced by the cognitive and linguistic limitations of informants. These studies appear to show contradictory results. The mentioned challenges may be assessed as sources of error with respect to the quality of the results (Hartley and MacLean, 2006; Heal and Sigelman, 1995; Finlay and Lyons, 2001; Sigelman et al., 1981a), or they may be understood merely as characteristic features that one must consider when conducting research among informants with intellectual disabilities (Finlay and Lyons, 2002; Fujiura and RRTC, 2012; Rapley and Antaki, 1996).

The tendency of informants to acquiesce is one of the challenges that is most frequently reported in this context (Heal and Sigelman, 1995; Finlay and Lyons, 2002; Sigelman et al., 1981a, 1981b). Acquiescence is defined as a disposition to respond “yes” regardless of the question that is asked. Such a response can also be interpreted as a form of compliance in which the informants respond according to what they believe is expected.

In the author’s project, the researcher was prepared for the informants’ tendency to acquiesce (Sigstad, 2013a, 2013b). Therefore, during the interviews, a large amount of time was allotted to ensure that the questions were understood.

Nevertheless, in this project, short “yes” answers were more common than detailed provisions and rich responses. The example below (Sigstad, 2013a) shows a slightly demanding conversation between the interviewer (I) and a participant (student, S), in which the interviewer primarily receives confirmations of questions without answers that provide detailed information:

However, the question is whether the informant’s inadequate responses are related to the interviewer’s difficulties in asking good questions. This call may be an example of questions that were too long and possibly linguistically advanced or questions that were too abstract or executive, so the only answers may be affirmative. As in this case, it is possible that interviewees do not always understand the question or that they lack the skills to respond. They may not remember or may have difficulties finding words. Thus, consent may appear to be the easiest solution.

Finlay and Lyons (2002) reported that yes responses appear to occur more frequently when the answer is unknown, when questions are too long, or when the sentence structure is overly complex. Informants thus risk focusing on only a limited part of the question and losing focus on other significant parts of the question. The following example taken from the same study (Sigstad, 2013a) shows how a participant perceives only part of what is said. The interviewer repeats the informant’s answers and reframes the question to enable the student to understand:

Here, repetition and reformulation seems to be helpful. Otherwise, both abstract and open-ended questions without answers may be difficult to employ (Tøssebro, 1992). Informants may believe that concrete answers are expected, and they may worry about responding incorrectly. In the example below (Sigstad, 2013b), it appears that the informant initially attempts to provide the answer that he thinks is expected. It appears to be easier when the interviewer is more specific and asks the informant to describe an activity:

Heal and Sigelman (1995) argued that the tendency to acquiesce is significantly more pronounced when people with low status are asked questions by interviewers with high status. In other words, an uneven balance of power exists in the relationship between researcher and informant. The example below shows an informant who appears to find it difficult to be critical in front of the interviewer (Sigstad, 2013a). It is difficult for the informant to admit that he does not enjoy school. The interviewer then attempts to provide support by reformulating the informant’s response. The informant again provides a doubtful response, followed by an acknowledgment that he does not enjoy school when the interviewer repeats the same question:

Inconsistency in informants’ statements can pose a challenge (Finlay and Lyons, 2001; Rapley, 1995; Rapley and Antaki, 1996), and the informants may give contradictory answers.

The conversation above displays such an example and shows a call that indicates that interviewer should be particularly careful to understand what is meant by the informant’s responses and whether any uncertainty or misunderstanding caused some type of compliance.

However, Finlay and Lyons (2002) argued that the tendency to provide yes responses is not necessarily related to submissiveness and the tendency to “please.” The grammar and complexity of the questions and the difficulties associated with being evaluated by others also appear to be important. Questions that are formulated to be speculative, hypothetical, future-oriented, or retrospective can be especially difficult.

In the author’s study (Sigstad, 2013a), the interviewer attempted to link questions to the present day and specific tasks that concerned the students:

In this study, it was much easier for the student to talk when the conversation revolved around specific themes. The student showed interest in talking and provided more information.

Another challenge in such interviews is that informants with disabilities may choose to exaggerate reality. Hence, interviewees present a picture of the reality that they prefer (Tøssebro and Lundeby, 2002). In the author’s project, it appears that one of the informants attempts this approach:

In the example above, it may be that the student attempted to introduce a picture of reality that he thought the interviewer would want to hear. It may also be that the student actually did not mind that the teacher decided; this is what he was used to and when he felt safe.

Interviewing adolescents often differs from interviewing adults. When children and adolescents are asked if they can remember any specific information, they can be as accurate in their responses as adults, although they cannot remember as much as adults do (Dockrell, 2004). Specific questions increase the amount of information but the answers are nevertheless less precise. The example below is taken from a study of student participation among students with disabilities (Sigstad, 2013a). This excerpt shows how an informant’s response is extended when the interviewer changes the formulation of the question. By changing the question from a fairly open question to a more hypothetical question, the interviewer gives the participant the ability to be more specific:

As previously mentioned, it appears that there are different ways of understanding related to the inclusion of people with intellectual disabilities in research interviews. Among various methodological challenges, acquiescence has been a particularly important research topic in methodological research on persons with intellectual disabilities. Sigelman et al. (1981b) showed in their studies that informants with intellectual disabilities demonstrated an increased tendency to provide yes responses and that this tendency to acquiesce was negatively correlated with intelligence quotient (IQ). In this respect, Ramirez (2005) conducted an evaluation study assessing the tendency to provide yes responses related to yes–no questions among children with and without disabilities. In contrast to the findings reported by Sigelman et al. (1981a, 1981b), this study did not find significant cognitive group differences with regard to this phenomenon. In addition, there was no significant correlation between IQ scores as a measured tendency toward yes responses. However, these results support the finding that children with intellectual disabilities do not differ from younger children and other peers without intellectual disabilities in terms of yes responses in relation to yes–no questions. These findings are also consistent with the work of Finlay and Lyons (2002).

Rapley and Atanki (1996) noted that the traditional understanding of informants’ submissiveness and willingness to “please” the interviewer is not necessarily an appropriate explanation for this phenomenon. Rather, the authors explain that informants tend to agree because they perceive the research interview as a test situation in which the fear of failure to reply affects the choice to consent. It may be the interview logic that provides such “pseudo-yes-answers” in the face of interviewer reformulations and that leads to informants’ follow-up with several acceptable answers. Dockrell (2004) supported this criticism and showed how an interviewer’s questioning can affect a participant’s ability to admit that he or she does not know the answer to a question. This situation can easily result in the informant attempting to answer yes–no questions even if he or she is unable to answer them. This situation occurs because the interview does not sufficiently account for the social context of the informant. Moreover, according to Rapley and Atanki (1996), the researcher’s interest should be replaced with a more respectful understanding of informants’ interaction skills and cognitive competence that are representative of people with learning disabilities.

My own experience has shown that qualitative interviews among individuals with intellectual disabilities require the researcher to allocate sufficient time, with regard to both informants’ need for processing and any incidental contributions that arise during such interviews (Sigstad, 2013a, 2013b). For example, both incoherent and irrelevant responses are relatively common in such interviews (Finlay and Lyons, 2001). Informants who are concerned about providing a concrete answer in advance appear to have an ongoing need to return to these questions during the interview, possibly because of a lack of control over the ability to disengage from previous themes or because the interviewer moved too rapidly through the interview.

In this respect, Fujiura and RRTC (2012) argued that rather than searching for “mistakes,” interviewers could adopt an inclusive attitude toward informants and seek to understand how things are perceived, processed, and answered. Information should not be passively recalled and directly given to the interviewers; informants should be active participants in the process in which they are supposed to recognize, remember, interpret, and formulate responses. A good qualitative interview presupposes that the interview is adapted to suit the abilities and capacity of the informants rather than the reverse. By improving understanding of how responses are formulated, researchers can strengthen the accuracy of such forms of communication.

Establishment of confidence

If a research interview is to be conducted on the informant’s terms, the interview depends on an understanding of the informant’s social context (Rapley and Antaki, 1996). This understanding requires a good knowledge of the environment of the informants, including home, school, work, or leisure arenas. The research theme forms the basis for the greatest research interest. It is also necessary to have sufficient knowledge of the daily activities of the informants. The aim is to achieve the broadest possible knowledge of the existing arenas of the informants’ lives and to attempt to build the informants’ confidence and knowledge.

Participant observation is often recommended as a replacement or as complementary method to the interview (Sundet, 2010). The purpose of such observation is initially to create trust and communication. It can be difficult to determine the extent to which an interviewer has obtained a common understanding and succeeded in gaining the confidence of the informant. This method can also be challenging because researchers must attempt to maintain distance from their own prejudices, which are often closely related to their own understanding of intellectual disabilities. In the interview study among secondary school students with a mild degree of intellectual disabilities, the method was restricted to interviews (Sigstad, 2013a, 2013b). For this reason, it was especially important to ensure adequate preparation for the interview situation by obtaining the necessary information from teachers, who knew the students, and the support of the students’ parents. One of the students did not want to participate because he perceived the interview as a type of test, which is something that he had previously experienced in a negative way. In a conversation with the parents, the student acquired a new understanding and chose to participate.

Development of the interview guide

An important question in the context of a research interview is related to the extent to which we can assume that the informants’ responses are influenced by the interview situation. Good qualitative answers are provided primarily as a result of thorough preparation of the interview guide. In the planning phase of the study on the teacher–student relationship and student participation (Sigstad, 2013a, 2013b), it was essential to use both simple and concrete questions and to allow both parties room for improvisation (Kittelsaa, 2010).

The selection of questions can be important with respect to the ability of informants to provide adequate responses. The interview structure is relevant in this context. Semi-structured interviews provide an opportunity to focus on predetermined topics and key issues that the researcher wishes to address during the interviews (Dalen, 2011; Kvale and Brinkmann, 2009). At the same time, this approach provides a good opportunity for informants to deepen the discussion or to focus on other themes that they perceive to be significant. The criteria for a good qualitative interview are related to issues that are formulated to promote comprehensive and specific responses from the informants. However, it is not always possible to achieve this scenario when one is interviewing people with intellectual disabilities. In the interview study pertaining to the teacher–student relationship and student participation, this challenge was also an issue (Sigstad, 2013a, 2013b). Compared with ordinary interviews, the interviews in this study had the following characteristics: the students were taciturn; they provided many “I don’t know” responses; responses from the same informant were sometimes inconsistent; and they often chose to eagerly repeat the interview question. The different examples above show how the interviewer reformulates questions and repeats the interviewee’s responses to help the interviewee understand, providing an opportunity for the interviewee to present his or her understanding of the topic area.

Finlay and Lyons (2002) emphasized the need for simple questions worded in short sentences, often associated with specific situations that are familiar to informants. Abstract questions may seem incomprehensible. In a student project in which adults with intellectual disabilities were interviewed about their hopes for the future, the choice of appropriate questions represented a particular challenge (Skipnes and Kambo, 2013). To encourage informants to describe their own dreams, questions were closely related to current topics. All forward-looking questions were related to specific, known venues with important, relevant issues that the informants had previously addressed. The focus on specific subjects appeared to help the informants to more easily describe their desires for the future.

As we have shown, previous methodological research has focused heavily on informants’ yes responses or “acquiescence,” with various explanations of this phenomenon (Finlay and Lyons, 2002; Fujiura and RRTC, 2012; Heal and Sigelman, 1995; Ramirez, 2005; Rapley and Antaki, 1996; Sigelman et al., 1981a). Subsequent studies have indicated that there is not necessarily one simple solution to this phenomenon (Finlay and Lyons, 2002). The tendency toward acquiescence can be reduced by changing the questioning to short questions and simple sentence structure.

Conducting the interview

Thorough preparation with an interview guide is a good starting point and may be an important tool for achieving a successful qualitative interview. However, a prerequisite for achieving satisfactory interaction between the researcher and informant is the participants’ confidence in the interview situation. The selection of an appropriate location for an interview can be important (Kittelsaa, 2010). For some informants, the home will be the place that seems safest, whereas for others, school or the workplace may be perceived as a natural setting for an interview.

In the author’s study (Sigstad, 2013a, 2013b), all interviews were conducted at the students’ schools. Because observation was not utilized in this interview study, a significant amount of time was needed to become acquainted in advance of the actual interview. This time provided an opportunity for conversations with the students, in which they were able to discuss their daily lives both within and outside of school. Thus, they were encouraged to experience more confidence in the actual interview situation.

Despite careful preparation, it is possible that the informants may not understand the questions. They may find it difficult to answer and may resort to simple answers to escape the situation. For a researcher, this situation requires an open and flexible attitude as well as the ability to identify the informant’s challenges, whether they are linguistic or cognitive challenges or are caused by the asymmetry that exists in the power relationship between the individuals.

During the implementation of the author’s interview study, a substantial amount of time was allotted to ensure that the questions were understood (Sigstad, 2013a, 2013b). The interviewer attempted to show interest and curiosity toward the informants by asking alternative questions and sometimes repeating the informants’ answers, thus ensuring that the informants could expand the conversation by discussing their answers further and providing more details (Thorsen and Myrvang, 2008).

Although an interview study is carefully planned, questions may remain regarding the extent to which it is possible to ensure that informants understand the questions and have sufficient time for processing. Because of cognitive and linguistic limitations, it is not always obvious whether the interviewer has grasped the original meaning of the situations that the informants attempt to describe (Guneriussen, 2010). It may be easier for informants to respond to questions that are linked to specific practical tasks. It is possible that the students in this interview study felt unsafe, that they were afraid to give a “wrong” answer, or that they wanted to please the interviewer. For the interviewer, these challenges could be addressed in the following ways: Sufficient time should be allotted to verify whether the students actually understand the questions; the responses of the informants could be repeated; or the interviewer could choose to ask the questions in a slightly different manner.

The interviewer must be prepared for the possibility that open questions may require specific follow-up questions or suggestions that are directly conducive to follow-up discussion. When informants have difficulties expressing themselves, it may be necessary to assist them with suggestions for alternatives or to provide help with completing sentences. Thorsen (2005) emphasized the need to ask leading questions and guide the conversation in such types of interviews, employing types of questions that are not usually used in research. Conducive questions are also employed to verify or disprove an interpretation or to enable the interviewer to obtain more accurate answers.

The use of repetition can be effective as both a question and confirmation of what has been said and can assist in furthering the conversation. The following example is taken from the mentioned interview study (Sigstad, 2013b) and shows a conversation that begins with a yes–no question related to well-being during break time. The interviewer then asks an open question and subsequently follows with questions related to specific activities during recess. By repeating the informant’s response at the end, the interviewer verifies the importance of what the participant wants to communicate:

The affirmative question works here as both a question and a verification. Through the use of repetition, it becomes possible to prompt additional information from the informant while allowing the interviewer to confirm the importance of what has already been said. Thorsen (2005) explained that repetition also helps to create tranquillity in a conversation, which is necessary for participants to have sufficient time for linguistic processing, acquiring new ideas, and finding the appropriate words. In the study by Thorsen and Myrvang (2008), 20 parents were interviewed together with 5 of their adult children with intellectual disabilities. The study focused on life stories and showed obvious contrasts between the parents’ and the children’s narratives. The parents were able to provide rich descriptions of their children’s life stories, whereas the narratives of the developmentally challenged children were more present oriented. With help from the interviewers, the children provided short answers, and, in only a few words, they gave a much more condensed picture of the same stories. Their life narratives were presented either in chronological order or in the context of the life course that they were attempting to describe. This example illustrates the need in qualitative interviews to facilitate understanding of the informants’ context or to employ the assistance of someone close or service providers in the preparation and the implementation of the actual interview. Rather than seeking potential sources of failure that would otherwise be interpreted as threats to the validity of the findings, the interviewer should approach the informants with an inclusive attitude. Such an attitude involves attempting to understand what the informants are attempting to express, observing the ways in which questions are understood and answered, and providing support and good follow-up questions to enable the informants to provide additional information.