Overview ... 145 Functional Assessment ... 145 Interviewing Techniques ... 149 Research Questions ... 152 Method ... 152 Interviews ... 153 Subjective Measures ... 154 Analytical Software ... 155
Coding and Qualitative Analysis ... 155 FSDM Augmental Codes ... 156 [AUG-‐val] Augmental Value ... 156 [AUG-‐con] Augmental Control ... 157 FSDI Codes of Interviewer Statements ... 158 [RuleCrel-‐probe] Self-‐Rule Contextual Relations Probe ... 159 [RuleCfunc-‐probe] Self-‐Rule Contextual Function Probe ... 160 [RuleValid-‐probe] Self-‐Rule Validation Probe ... 161 [S-‐probe] Self-‐ or Other-‐as-‐Story Probe ... 162 [X-‐probe] Self-‐ or Other-‐as-‐Perspective Probe ... 162
Thematic Analysis of Interviews ... 163 Ambivalence about change or direction ... 166 Mindfulness practice as a control strategy ... 172 Describing mindfulness versus practicing mindfulness ... 177 The efficacy of perspective-‐taking ... 179 Values and value oriented self-‐rules ... 183 Generalised responding ... 189
Quantitative Analysis ... 200 FSDM Analysis of the FSDI Interview Transcripts ... 201 Evaluation of the FSDI Interviewer Capabilities ... 212
Discussion ... 216
Overview
While the approach to interviewing used in the previous two studies worked well it led to various questions: What worked in the interview? Which questions yielded rich and accurate data for the different codes? What topographical cues led to different types of questions from the interviewer? This chapter focuses on the development and application of the Functional Self-‐Discrimination Interview (FSDI) as a method for capturing rich data for coding with the Functional Self-‐Discrimination Measure (FSDM).
I first review what a functional assessment is, drawing on the literature in this field to show how I have approached the design of the FSDI. I then provide a qualitative analysis of a series of interviews where I employed the FSDI. In that section I discuss the relationship between interviewer questions and the resulting interviewee
responses suited to coding with the FSDM. This is followed by quantitative analysis of the interviews using the FSDM where I discuss the relationship between code
frequencies and measures of wellbeing. I also provide an independent analysis of my interviewer capabilities from a Motivational Interviewing perspective (Moyers et al. 2014; Moyers et al. 2010). The chapter concludes with a brief discussion of the
results of the study. In the following chapters I present a more complete discussion of the implications of the three studies undertaken in this thesis.
Functional Assessment
The FSDI and the FSDM were designed as a functional assessment (Baer et al. 1968; Chance 1998; Cipani & Schock 2011; Hayes & Brownstein 1986; Nelson & Hayes 1979; Ramnero & Torneke 2008; Vilardaga et al. 2009) of an interviewee’s cognitive and behavioural responses to different types of historically situated events, past, present or future. This approach was predicated on the assumption that reflecting on and recounting a critical incident is the behaviour of verbal operant responding (Hayes et al. 2001d; Leigland 1996, 2005; Shahan 2013). If a question was asked about an event in the person’s history, the interviewee’s response, while guided by the question to some degree, would in the main be a response to the historically situated event. Their response would be in the form of conditioned operant
responses. If the event under question was experienced as aversive, it would have been framed as aversive in recall. Conversely, if the event was appetitive, speaking of the occasion would involve framing the event in positive terms.
There are several key elements of verbal operant responding from an RFT perspective that influenced the design and implementation of the interview
technique described in this chapter. These is a distinction between verbal contextual relations (Crel) and verbal contextual functions (Cfunc); the three-‐term contingencies of
a self-‐rule – antecedent, behaviour and consequence; how augmentals function in a self-‐rule; and, the generalisation of self-‐rules as a behavioural repertoire. I discuss each of these below prior to discussing the interviewing techniques that also informed the FSDI design.
Let’s first consider a simple example of the difference between Crel and Cfunc. Torneke
(2010, p.87) uses the example of “Larry is better than Peter at playing tennis”. In this example, ‘better than’ is the Crel specifying a relation between stimuli Larry and Peter,
while ‘at playing tennis’ (Cfunc) narrows the many possible functions of the relation
down to the particular meaning of ‘playing tennis’. Without the Cfunc, it would not be
clear to the listener in which way Larry was better than Peter. Neither Crel nor Cfunc
should be thought of as particular phrases within a sentence however. Rather they are cues for the overall functions of the sentence. That is, they modify the
‘appropriate’ response associated with the Crel.
To further illustrate the distinction between verbal contextual relations (Crel) and
verbal contextual function (Cfunc) let us consider the value oriented self-‐rule [VOR], "If
I (with the desire to learn and make a difference) study each day this summer, then I will complete my PhD, will likely be a little wiser, and more able to make a
difference." A Crel is the context in which a history of a particular kind of relational
responding is brought to bear on the current situation (Hayes et al. 2001d, p30). In this case, the particular relational responses being brought to bear include
conditional frames (if-‐then), temporal (I-‐will) and other frames such as hierarchy (I with attributes of wisdom and more able to make a difference). The verbal network is complete in this instance in the sense that it describes a context in which a particular
behaviour will result in a particular consequence. A Cfunc is understood to be a
contextual cue that indicates which particular psychological functions associated with an event are modified in accordance with the underlying derived relations (Hayes et al. 2001d, p31). In this instance the psychological function of “daily study” and “completing a PhD” are augmented by virtue of the values I hold of “wanting to learn and make a difference”. These events now matter to me. In this way, just as the “if-‐then” relational response I brought to bear on the events “daily study” and
“complete PhD” is controlled by the context, the specific psychological functions that are transformed – the importance of the events – are also under contextual control (Hayes et al. 2001d, p32).
Technically, this VOR can be deconstructed in terms of a three-‐term contingency (Cipani & Schock 2011; Hayes et al. 1989; Ramnero & Torneke 2008) specifying antecedent [A], behaviour [B], and consequence [C]. "If I (with the desire to learn and make a difference) [A] study each day this summer [B], then I will complete my PhD, will likely be a little wiser, and more able to make a difference [C]." A, B & C are the contextual events within the rule. As the rule provides a proper relational context (a Crel) and fulfilment of a relational response, i.e. the transformation of the stimulus
functions of the related events in the network (Cfunc), the rule is complete.
In a VOR, the Cfunc is defined in part by the values in the verbal context that determine
which functions are transformed. In the VOR above “learning and making a
difference” are the values that function to alter the degree to which daily study and completing the PhD matter as a consequence. This is known as augmenting.
Augmenting is rule-‐governed behaviour due to the relational networks that alter the degree to which events function as consequences (Barnes-‐Holmes et al. 2001d, p109; Leigland 2005). As well as constructing VORs I have learned to construct rules for avoiding unwanted experiences. Often when I sit in front of my computer to write, I want to avoid the anxiety of feeling I don't have anything useful to say. A situation I could actively avoid by following a control oriented self-‐rule [COR] such as, “if I read more then I will know enough to be able to write something coherent tomorrow”. In this COR “feeling I don't have anything useful to say” functions as an augmental as it alters the degree to which reading matters as a strategy for achieving desired
consequences. Although, by following such a self-‐rule I may experience the short-‐ term relief of anxiety, I will also suffer a longer-‐term consequence – the writing will still be there to be done tomorrow and I still may not have anything useful to say.
Such self-‐rules are forms of verbal operants conditioned through a history of learning by consequence that generalise as verbal behavioural repertoires (DeLeon et al. 2013; Hayes et al. 2001b; Hayes et al. 2001d; Neuringer & Jensen 2013). I was able to construct the above rules as I have learned that if I apply myself then in the past I have typically achieved what I set out to achieve. In my history I have learned to verbally track my own behaviour by constructing “if-‐then” self-‐rules, which have allowed me to successfully plan to obtain specified consequences through my actions or potentially avoid unwanted consequences. It was these kinds of verbal operants that I was interested in having interviewees utter (if they were available in their repertoire) in order to do a functional assessment of their verbal behaviour. Further, in addition to classes of self-‐rules, I sought to have interviewees utter characteristic statements about themselves that functioned to derive a sense of esteem, positive or negative. For example “I mustn’t be very smart” or “I am a good writer”; or to take perspective on verbal experience, for example “There’s that not very useful thought again…”
In the FSDI, when probing for verbal operants such as self-‐rules, questions were asked to have interviewees articulate the three-‐term contingencies of self-‐rules and to what extent that class of responding had been generalised. For example, how often they followed the same self-‐rule in different contexts or how typically they attributed the same self-‐categorisations to themselves. The assumption was that when the interviewee turns their attention to the topography of a historically situated event, their routine relational response would be taken to that event, which would be an instance of a generalised operant. The frequencies of patterns of such statements were of interest as a functional assessment of future wellbeing. To confirm that the interviewee’s relational responses were routine, much of my input as the interviewer was in the form of open questions and complex reflections of the interviewee’s
interview techniques employed in the FSDI are outlined next; my approach to questioning is presented in the Coding and Qualitative Analysis section below.
Interviewing Techniques
In addition to considering specific types of questioning, an approach to interviewing known as Motivational Interviewing (MI: Miller & Rollnick 2013) also informed the FSDI design. Originally developed for use in clinical settings, MI is increasingly being adapted into contexts where collaborative conversation styles are employed to promote a person’s motivation for behaviour change (Burke et al. 2003; Lundahl et al. 2010; Rubak et al. 2005). While MI has increasingly been adopted as an
empirically validated counseling style, it lacks a coherent theoretical framework for understanding its processes and efficacy, a deficiency/weakness which a number of scholars have attempted to address (Bricker & Tollison 2011; Markland et al. 2005). For the purposes of this study MI techniques were employed primarily for their utility even though RFT may provide a theoretical framework for understanding and explaining its efficacy. I also employed MI as it provided a method for independently validating the interviewing techniques that I employed when conducting the FSDI (Moyers et al. 2014; Moyers et al. 2010).
MI focuses on eliciting client ‘change talk’ (Miller & Rollnick 2013; Moyers et al. 2014). It is a collaborative conversation style that has been designed to strengthen a person’s own motivation and commitment to change by focusing on addressing ambivalence about change and reinforcing the value for desired change (Hettema et al. 2005; Lundahl et al. 2010; Rubak et al. 2005). MI focuses on invoking that which is intrinsically valued by the person, not installing what is missing. Four key processes underlie the MI approach: partnership, acceptance (a combination of empathy, autonomy support, and affirmation), compassion, and evocation of client ‘change talk’. To successfully engage these processes a competent interviewer practices five key communication skills throughout an interview: asking open questions, affirming, reflecting, summarising, and providing information and advice with permission. I adapted a subset of these processes and communication skills to the FSDI design.
While the FSDI was not designed as an intervention to facilitate behaviour change, it has that potential. The MI approach to evoking client change talk is, in my opinion, essentially an approach to having clients formulate VORs specific to a desired behaviour change (Miller & Rollnick 2013, p164-‐165). When conducting the FSDI, I employed only those MI capabilities designed to cultivate partnership and not those targeting change. This served two purposes. Firstly, it allowed a set of relevant interviewer behaviours to be empirically evaluated (Moyers et al. 2014; Moyers et al. 2010). Second, it provided a segue to future research on interviewing for behaviour change based on a combination of the FSDI and MI. I discuss the implications of this integrated approach in Chapter 6.
The three interviewer behaviours I adapted to the FSDI that were empirically evaluated were: Listening (questions & reflection), Collaboration & Engagement and
Empathy (Moyers et al. 2014; Moyers et al. 2010). The quality of Listening in an interview was identified as a combination of questions and reflection. Questions were open and probed for the speaker’s typical verbal operant responses to the incidents being discussed, examples of which are offered below in the Coding and Qualitative Analysis section. Reflections involved summarising and repeating back to the interviewee something about what they had just said. These reflections were
categorised as Simple or Complex depending on whether they were a literal repeat of what the interviewee had said or they assumed something that was “between the lines” and thus captured deeper meaning. Both open questions and reflections served to elicit the verbal operants “within” the interviewee’s verbal repertoire. Clarity was obtained as a baseline of operant responses was identified, which became evident as recognisably consistent patterns of verbal behaviour across the interview (Leigland 1989, 1996). Questions and reflections also helped qualify how effectively the interviewee was able to take perspective on verbal behaviour as verbal behaviour, that is, their capacity for deictically framing the content of their psychological experience.
Collaboration & engagement, and empathy were considered important qualities for a successful FSDI in order to have the interviewee speak openly about their experience. Collaboration & engagement was evaluated to be occurring if the conversation
appeared to be taking place between two equal partners. Empathy was evaluated as the extent to which I, as the interviewer, understood or made an effort to grasp the interviewee’s perspective and feelings: literally, how much I attempted to “try on” what the interviewee felt or thought. Care was taken not to confuse empathy with warmth, acceptance or genuineness; these qualities were considered independent of empathy, which was specifically about taking the speaker’s view. Reflective listening was an important part of this characteristic, as it showed that I understood the interviewee’s perspective as I conveyed that understanding back to the interviewee without judgement.
The independent evaluation of Listening (questions & reflection), Collaboration & Engagement, and Empathy provided feedback that could be used to improve the FSDI interviewing method. Employing this evaluative approach was intended to be used: 1) as an interviewing integrity measure by MI standards to check the FSDI interviews were not a manipulation for change; and, 2) as a means of providing independent structured, formal feedback about ways to improve interview technique. Full criteria for evaluating these capabilities are presented in Appendix 3: FSDI Interviewer Capability Evaluation. The coding and rating method is discussed in the Quantitative Analysis section below.
Several other interview methods were evaluated for adaptation to the FSDI but none were as applicable as the MI approach. These included: Subject-‐Object interviewing designed to investigate cognitive complexity (Kegan 1994; Kegan et al. 1982; Lahey et al. 1988); interviewing as qualitative research designed to capture retrospective, contemporary and prospective views of an interviewee’s history (Seidman 2013); and, experience-‐based, body-‐anchored qualitative research interviewing designed to capture the first-‐person perspective (Stelter 2010). Each of these approaches
reinforced the importance of the MI processes and skills outlined above.