A Relational Frame Theory approach to Perspective-Taking Developing a Natural Language IRAP for Assessing Perspective-Taking with Regards to Self versus Other Word count: 20094

Lisa Van Raemdonck Student number: 01402215 Supervisor(s): Prof. Dr. Dermot Barnes-Holmes, Dr. Ciara McEnteggart A dissertation submitted to Ghent University in partial fulfilment of the requirements for the degree of Master of Clinical Psychology Academic year: 2018 - 2019

Preface

In the context of completing my masters degree in clinical psychology, I wrote the current thesis.

During the writing process of this thesis there are several people to whom I would like to express

my sincerest gratitude. For which I will take the appropriate time to do so in this ‘preface’.

First, I would like to thank Prof. Dr. Barnes-Holmes from the bottom of my heart, for introducing

me to the relational frame theory and guiding me through the entire writing process. I would also

like to thank Dr. Ciara McEnteggart and Deirdre Kavanagh whom supported me by answering a

multitude of questions and providing valuable advice and feedback on multiple occasions. As a

student novel to the broad field of the relational frame theory and the process of conducting and

reporting on an experimental study, I could not have asked for a more engaged and inspired

research team to be part of.

I would also like to thank my family and friends for supporting me through this challenging

experience and encouraging me during my entire school career. Their support has meant the world

to me and continuously motivated me to work hard and put in the effort.

Abstract

Recently, a Relational Frame Theory (RFT) approach towards perspective-taking,

conceptualising this skill as deictic framing, has become more apparent. There have been several

procedures developed from an RFT-perspective to assess deictic relational responding, such as the

Implicit Relational Assessment Procedure (IRAP). The current study investigates the applicability

of a newly developed version of the IRAP (NL-IRAP) to measure perspective-taking,

implementing natural language statements comprising of psychologically relevant stimuli. To

further investigate the role of diverging interpersonal boundaries, in contrast to previous studies,

the current study used two separate NL-IRAPs to assess whether performances differed

significantly when responding in relation to self versus a different other. On an exploratory basis,

several self-report measures were included, assessing interpersonal boundaries and clinical

concepts often related/ linked to perspective-taking, such as psychotic-like experiences. During the

process of analysing the data a three-way interaction was encountered for a specific trial-type

(Positive event-Negative reaction) that approached significance, suggesting differential responding

relative to self versus other. To pursue the robustness of the obtained trend, data from a related

study was included, also displaying this same pattern of responding. Not only did the three-way

interaction appear to be robust, the influence of the other (different or disliked) may indeed play

an important role. Finally, two correlational clusters were obtained that appeared relatively

consistent with the literature on psychosis and attachment related anxiety/ avoidance, suggesting

higher psychological distress is related to impaired flexibility in deictic relational responding on

the NL-IRAP.

Abstract (Dutch/ NL)

Recent treed een Relationele Frame Theorie (RFT) benadering van perspectiefname steeds

meer op de voorgrond en conceptualiseert dit als deiktische kadering. Vanuit deze benadering zijn

verscheidene methodieken ontwikkelt om deiktische relationele reactiepatronen te meten,

waaronder de ‘Implicit Relational Assessment Procedure’ (IRAP). De huidige studie onderzocht

de functionaliteit van een nieuwe versie van de IRAP (NL-IRAP) om perspectiefname te meten

die gebruik maakt van natuurlijke taalstellingen, evenals psychologisch relevante stimuli. Om de

invloed van divergerende interpersoonlijke grenzen nader te onderzoeken, werd in tegenstelling

tot voorgaande studies, gebruik gemaakt van twee aparte NL-IRAPS om te exploreren of

reactiepatronen significant verschillen met betrekking tot het eigen perspectief versus dit van een

ander. Er werden eveneens zelfrapportage vragenlijsten afgenomen om verbanden te exploreren

tussen prestaties op de NL-IRAP en concepten geassocieerd met interpersoonlijke grenzen en

perspectiefname, zoals psychose. Tijdens de data-analyse werd een marginaal significante drie-

wegsinteractie geobserveerd voor een specifieke testconditie (Positieve gebeurtenis-Negatieve

reactie), die mogelijk duidt op verschillende zelf- en ander-gerelateerde reactiepatronen. Om de

robuustheid van deze trend nader te onderzoeken, werd data van een gerelateerde studie

geïncludeerd waarin ditzelfde patroon werd geobserveerd. De driewegs-interactie bleek niet alleen

robuust, maar ook de specificatie van de ander (verschillend of gehekeld) is hierbij mogelijk van

belang. Tot slot, werden twee correlationele clusters geobserveerd die relatief consistent bleken te

zijn met literatuur over psychose en hechtingsgerelateerde angst/ vermijding. Deze suggereren dat

een hogere mate van psychologisch lijden mogelijk gerelateerd is aan een verminderd vermogen

tot een flexible deiktische relationele wijze van reageren op de NL-IRAP.

Table of Contents �

INTRODUCTION……………………………………………………………………….1

Relational Frame Theory: General Background………………………………….1

Relational Frame Theory and Deictic Relations…………………………………2

Sense of Self: Definition and Clinical Findings………………………………….3

Defining sense of self …………………………………………………….3

Clinical findings on sense of self………………………………………....4

Perspective-Taking: Definition and Clinical Findings…………………………...5

Defining perspective-taking as deictic relational responding…….……....5

Clinical findings on perspective-taking…………………………………..6

The Role of Perspective-Taking in Psychosis Symptomatology…………………7

General background……………………………………………………...7

A relational frame approach………………………………..…………….7

Assessing Perspective-Taking from an RFT Approach………………..…….…..8

Exploring alternative methodologies…………………………………......8

The implicit relational assessment procedure (IRAP)………..………......9

The IRAP and perspective-taking……………………………………….11

The Current Study…………………………………………..…….……….…....12

METHOD……………………………………………………………..….……….…....14

Participants……………………………………………………….…………….14

Design………………………………………………………………………......14

Materials and Apparatus………………………………………………………..15

Natural language-implicit relational assessment procedure

(NL-IRAP)……………………………………………………………...15

Self-focused NL-IRAP……………………………………….......16

Other-focused NL-IRAP…………………………………….......17

The community assessment of psychic experiences (CAPE)….……….19

The psychological flexibility index (PFI)……………………………….19

The experiences in close relationship structures questionnaire – revised

(ECR-RS)……………………………………………………………….20

Experiencing of self scale (EOSS) ……………………………………...21

Inclusion of the other in the self scale (IOS)…………………………….22

Feelings about self thermometer………………………………………...22

Procedure……………………………………………………………………….23

Self-focused NL-IRAP………………………………………………….23

Other-focused NL-IRAP..……………………………………………....25

Self-report questionnaires………………………………………………25

RESULTS…………………………………………………………………………........25

IRAP Data…………………………………………………………………........26

Data preparation and preliminary analyses..……………………………26

Descriptive analyses..……………………………………………...…....27

Statistical analyses.……………………………………………….…….27

Questionnaire Data ...……………………………………………………..……32

Correlations between IRAP Scores and Self-Report Measures………………...33

DISCUSSION………………………………………………………………………......35

Discussion of the Current Findings……………………………………...…......35

Clinical and Theoretical Implications……………………………………..........40

Limitations and Future Directions………………………………………...…....41

CONCLUSION…………………………………………………………………...…....43

REFERENCES…………………………………………………………………...…….44

APPENDIX 1 ………………………………………………………………...……......57

APPENDIX 2 ………………………………………………………………...……......59

APPENDIX 3 ………………………………………………………………...……......61

APPENDIX 4 ………………………………………………………………...……......62

APPENDIX 5 ………………………………………………………………...……......63

APPENDIX 6 ………………………………………………………………...……......64

APPENDIX 7 ………………………………………………………………...……......65

APPENDIX 8 ………………………………………………………………...……......66

1

The historic distinction between a cognitive mediation approach to complex

human behaviour and a functional behavioural account that focuses on learning

principles of readily observable behaviour (often studied in non-humans), has become

less apparent throughout the last century (Stewart, 2016). As evidence for the

applicability and in some cases uniqueness of these general learning principles in more

complex human behaviour emerged (Hughes & Barnes-Holmes, 2016, p.131),

researchers shifted their attention towards an integration of a functional approach and

the study of human language and cognition (see De Houwer, 2011 and De Houwer,

Hughes, & Barnes-Holmes, 2017 for a discussion). Building on Skinner’s account of

human language as operant behaviour, Relational Frame Theory (RFT) has attempted to

conceptualize human language and cognition as arbitrarily applicable relational

responding (AARR) (Stewart, 2016).

Relational Frame Theory: General Background

Recently RFT has been gaining more attention in its attempts to articulate an

overarching framework of language and cognition which aims to connect research and

clinical practice (Barnes-Holmes, Barnes-Holmes, Luciano, & McEnteggart, 2017;

Hughes & Barnes-Holmes, 2016, p.130; O’Connor, Farrell, Munnelly, McHugh, 2017;

see Stewart, 2016, for an overview). RFT conceptualizes language and cognition as

generalized operant behaviours operating under contextual control, that are learned

through a history of multiple exemplars (see Hughes & Barnes-Holmes, 2016 for a

more detailed description). The basic argument is that through extensive verbal

interactions and experiences, an individual learns to derive relations between and

among stimuli in a manner that does not rely only on the physical properties of those

stimuli (Hughes & Barnes-Holmes, 2016). Such responding is referred to as arbitrarily

applicable relational responding (AARR) or relational framing, and this derived effect

was first established by Sidman (1971) during his seminal work on stimulus

equivalence (see Hughes & Barnes-Holmes, 2016, for a more detailed description;

Stewart, 2016). There are multiple ways to relate stimuli arbitrarily, and RFT identifies

particular patterns of such relational responding as relational frames, such as

coordination (see e.g., Alonso-Álvarez & Pérez-Gonzalez, 2017; Lipkens & Hayes,

2009; Ninness et al., 2009; Ruiz & Luciano, 2011), opposition (Alonso-Álvarez &

2

Pérez-Gonzalez, 2017; Lipkens & Hayes, 2009; Ninness et al., 2009), distinction (see

Ming & Stewart, 2017, for a review), hierarchy (Foody, Barnes-Holmes, Barnes-

Holmes, Rai, & Luciano, 2015; Griffee & Dougher, 2002; Slattery, Stewart, & O’Hora,

2011) and perspective-taking or deictic relations (Heagle & Rehfeldt, 2006; McHugh,

Barnes-Holmes, & Barnes-Holmes, 2004; McHugh, Barnes, Holmes, Barnes-Holmes,

& Stewart, 2006; McHugh, Barnes-Holmes, Barnes-Holmes, Stewart, & Dymond,

2007). Detailed descriptions of these frames are unnecessary in the context of the

current research; instead the focus will be on the frames referred to as perspective-

taking or deictic relations (see Hughes & Barnes-Holmes, 2016, for a more detailed

description).

Relational Frame Theory and Deictic Relations

According to RFT, deictic relational responding is presumed to develop later than more

basic or simple forms, such as frames of “same”, “difference” and “opposition”

(Barnes-Holmes, Foody, Barnes-Holmes, & McHugh 2013). Furthermore, the deictic

relations do not appear to have a clear basis in non-arbitrary relational responding, but

instead rely upon a speaker’s perspective functioning as a constant “location” in an

ever-changing environment (Barnes-Holmes et al., 2013; Hughes & Barnes-Holmes,

2016). In other words, as a child learns to use a given language, speaking for the child is

always from his or her perspective. Research in RFT has indicated that there are three

fundamental deictic relations: the interpersonal (I-YOU), the spatial (HERE-THERE)

and the temporal relations (NOW-THEN) (Hughes & Barnes-Holmes, 2016; see

Montoya-Rodríguez, Molina, & McHugh, 2016, for a review). As such, these relations

combine to create the basic relational frame of perspective-taking, which locates an

individual language-user in a time and place relative to others.

Perhaps the earliest attempt to study perspective-taking within RFT involved

developing a verbal protocol that was designed to assess the deictic/perspective-taking

relations (Barnes-Holmes et al., 2013). Although there have been several adaptations

and refinements of this protocol (e.g., Gilroy, Lorah, Dodge, & Fiorello, 2015;

Vilardaga, Estévez, Levin, & Hayes, 2012; McHugh, et al., 2004), the original version

has led to a significant body of research examining and training these deictic frames in

both non-clinical and clinical populations (Kavanagh, Barnes-Holmes, Barnes-Holmes,

3

McEnteggart, & Finn, 2018; see e.g., McHugh et al., 2004; Weil, Hayes, & Capurro,

2011).

The original protocol was developed to target the three key deictic relations in

young children across three different levels of complexity (Barnes-Holmes, Barnes-

Holmes, Roche, & Smeets, 2001; Barnes-Holmes et al., 2013; Hendriks et al., 2016a,

b). The first level considers simple relations, for example, “If [hypothetically speaking]

I (experimenter) have a red brick and YOU (participant) have a blue brick: Which brick

do I have?” (Barnes-Holmes et al., 2013, p.9). The second, level involves reversing the

relation, for example, “If I had a red brick and YOU have a blue brick. If I were YOU

and YOU were me. Which brick would I have? Which brick would you have?” (Barnes-

Holmes et al., 2013, p.11). Increased levels of complexity may involve double reversals

in multiple deictic relations, for example, “I am sitting here on the blue chair and you

are sitting there on the black chair. If I were YOU and YOU were me, and if HERE was

THERE and THERE was HERE: Where would you be sitting? Where would I be

sitting?” (Barnes-Holmes et al., 2013, p. 12). As will be noted later, this type of

protocol was developed specifically for research with children rather than adults and

focused on determining if specific relational skills were present or absent rather than

how fluid or flexible those skills were (Kavanagh et al., 2018).

Sense of Self: Definition and Clinical Findings

Developing a sense of self as distinct from others is vital to each human

individual. The next section will provide a relational frame conceptualisation of self and

its importance for our wellbeing.

Defining sense of self. There are numerous conceptualizations of the self, and

its importance as a concept has been highlighted in many psychological theories.

Reflecting its importance as a psychological construct, there is a large range of self-

report measures that were designed to assess specific conceptualisations of self, for

example, self-esteem or self-schemas (Atkins & Styles, 2016). The current research

focuses on a functional account of the verbal self, which according to RFT is uniquely

human and is perceived to be functionally different from a non-verbal self, which is

associated with pre-verbal children and non-humans (Barnes-Holmes et al., 2001;

McHugh, 2015). According to RFT, learning to use a human language helps to create a

4

verbal sense of self, and thus the verbal self cannot be explained using the types of

learning principles associated with nonhumans (e.g., Pavlovian and direct operant

conditioning; see Atkins & Styles, 2016; Barnes-Holmes et al., 2001).

As noted above, RFT argues that through our verbal interactions with others we

acquire a sense of self as distinct from the perspective of others. As Skinner (1974)

previously described, we learn this through questions and interactions that make us

aware of ourselves and of our inner experiences (e.g., “Are you angry?”, “Are you

playing with a toy?”) (Barnes-Holmes et al., 2001). In addition, RFT suggests that we

thus learn to respond from an I-HERE-NOW perspective relative to YOU-THERE-

THEN (McHugh, 2015); for example, “It’s five o’clock and I am at home cooking”,

“On Monday at ten o’clock you were at school”. As one’s sense of self further

develops, an individual is able to use these deictic frames in a more flexible manner and

in increasingly elaborate verbal interactions. According to RFT, initially these relational

verbal abilities are quite limited such that a young child may initially assume that

everything that they know is also what another person knows (McHugh et al., 2007;

McHugh et al., 2006). As such, the distinction between I-HERE-NOW and YOU-

THERE-THEN is not firmly established, though of course most children seem to

acquire the ability to make this distinction quite readily and without any form of

structured or programmed training (Atkins & Styles, 2016; Barnes-Holmes et al., 2001).

Clinical findings on sense of self. Numerous researchers suggested that there is

a link between human psychological suffering and a dysfunctional sense of self, such as

a negative or disturbed self-concept in depression, personality pathology and psychosis.

(e.g., Cohen, Leibu, Tanis, Ardalan, & Galynker, 2016; Foody, Barnes-Holmes, &

Barnes-Holmes, 2012; Nelson, Thompson & Yung, 2012; Rimes & Watkins, 2005).

Indeed, it has been argued that a sense of self as distinct from others could have

important implications for our psychological functioning. For example, McEnteggart,

Barnes-Holmes, Dillon, Egger and Oliver (2017) recently offered a functional-analytic

account of the relationship between the self, dissociation and voice hearing.

Interestingly, the authors argued that the development of a sense of self and a sense of

other, in terms of the deictic relations, are often impacted upon by a traumatic history.

Specifically, according to RFT the I-HERE-NOW may be strongly coordinated with

OTHERS, and thus the normal boundaries between self and others (as differentiated)

5

are somewhat less distinct. Furthermore, a study by Atkins and Styles (2016) found

exploratory evidence that a “strong” sense of a verbal self predicted well-being. Taken

together, it seems that the role of language in establishing a strong or coherent verbal

sense of self, as distinct from others, is essential in achieving relatively adequate levels

of psychological health and stability.

Perspective-taking: Definition and Clinical Findings

The ability to adopt the perspective of others relative to self appears to be

essential in establishing a stable sense of self. Therefore, a large part of the literature on

the self is devoted to perspective-taking. The next section will outline perspective-

taking as deictic relational responding and its relevance to the clinical domain.

Defining perspective-taking as deictic relational responding. For some time,

the dominant paradigm on perspective-taking has been the theory of mind (ToM), which

explains this as acquiring an understanding of one’s own internal state compared to that

of others (see Baron-Cohen, Tager, Flusberg, & Cohen, 2000, for a more detailed

description of this account). Recently, there has been increasing interest in perspective-

taking from a relational frame perspective (see Montoya-Rodríguez et al., 2016, for a

review). RFT conceptualizes perspective-taking as relational responding based on the

previously described deictic relations (Barnes-Holmes et al., 2013; Hendriks et al.,

2016a). Broadly speaking, the basic idea is that through deictic relational responding, an

individual makes inferences on the mental states and behaviour of others (McHugh &

Stewart, 2012). It is important to note that the specific labels typically used to denote

the three core deictic relations (I, Here, and There) are not the critical stimuli (Barnes-

Holmes et al., 2001; Barnes-Holmes et al., 2013). For example, “I” can also be

represented by a name such as “Louis”, and “HERE” can be represented by a place

such as “library”. In other words, through our verbal interactions with others regarding

our perspective we learn to abstract these relational properties (spatial, temporal and

interpersonal) which serve as a contextual “constant” upon which an ever-changing

environment operates and can be understood (Barnes-Holmes et al., 2001; Hendriks et

al., 2016a). There are several ways in which these deictic frames can interact resulting

in a multitude of possible relational networks (Barnes-Holmes et al., 2001; e.g., YOU-

HERE-NOW, I-HERE-NOW, YOU-THERE-THEN and so on). These elaborate

6

relational networks allow more complex forms of perspective-taking/deictic framing

(Hendriks et al., 2016a). More precisely, because of these interactions it is not only

possible to respond to either our own or another’s behaviour, but we can actually adopt

a multitude of other perspectives (Villatte et al., 2010). We can for example, imagine

how someone must feel in a given situation even though we may not have had a similar

experience. If for example, “Sam is currently feeling sad because he lost his dog”, I can

imagine how he feels by switching from an I-HERE-NOW perspective to an I-THERE-

NOW perspective. In so far as humans are social beings, these perspective-taking skills

seem to be crucial to our social functioning, and most importantly impairments may

therefore negatively impact our interactions with others.

Clinical findings on perspective-taking. Impairments in the ability to take

another’s perspective have frequently been implicated in a large part of the literature on

psychological suffering. A large part of this research has focused specifically on young

children with autism and developmental delays, however perspective-taking most likely

also plays an important role in several forms of psychological suffering in adults. As

mentioned previously, there seems to be a developmental trend in the acquisition of

these deictic relations (Barnes-Holmes et al., 2013; McHugh et al., 2004). Therefore,

the nature of the impairment, as measured using a developmental protocol of the

presence or absence of these relations, may differ in verbally-sophisticated adults

compared to children who have not yet fully developed the skills. Although RFT

research in this domain is still quite limited, there are a few studies examining the role

of perspective-taking in adults with psychological pathologies (see Montoya-Rodríguez

et al., 2016, for a review); for example, in adult populations with pathologies

implicating social impairments (Janssen et al., 2014; Villatte et al., 2008, 2010a, b ).

Overall, there seem to be some important differences between non-clinical and clinical

populations, with clinical samples performing more weakly on tasks that appear to

require more complex deictic relational responding (HERE-THERE and NOW-THEN;

Hendriks et al. 2016a). Hendriks and colleagues (2016a) also suggest diverging

development of perspective-taking in clinical samples. These impairments may be more

specific to certain pathologies than others, and as this line of research has been gaining

more attention, future research may even identify particular patterns of impairment.

7

The Role of Perspective-Taking in Psychosis Symptomatology

One of the clinical domains gaining more attention is the role of deictic

relational responding in schizophrenia and psychosis symptomatology. First, the current

research will look more closely at symptoms of psychosis to then explain how

perspective-taking may be implicated in these phenomena.

General background. Psychosis symptoms indicate an impaired relation with

reality and can, for example, take form as hallucinations or delusions. (Nolen-

Hoeksema, 2014). Psychosis, however, is quite a broad term that has been implicated in

several psychological pathologies such as schizophrenia, dissociative disorder (see

Renard et al., 2017, for a review), borderline disorder (e.g., Niemantsverdriet et al.,

2017), as well as in some non-clinical populations (see Van Os, J., Linscott, R.J., Myin-

Gemeys, L., Delespaul, P., & Krabbendam, L., 2009, for a review). As a continuum-

perspective on these symptoms has been gaining more support (e.g., Shevlin, McElroy,

Bentall, Reininghaus, Murphy, 2017; Unterrassner, Wyss, Wotruba, Haker, & Rössler,

2017), recent reflection on the conceptualization of psychosis symptomatology and its

classification has led to an adjustment of the DSM-IV criteria (see Allardyce, Gaebel,

Zielasek, & Van Os, 2006, for a review; e.g., Demjaha et al., 2009).

The DSM-V outlines five symptom domains of psychosis spectrum disorders:

delusions, hallucinations, disorganized speech, disorganized or catatonic behaviour, and

negative symptoms such as apathy (Nolen-Hoeksema, 2014). A large part of the

research that has focussed on the link between these symptoms and perspective-taking

was undertaken by ToM-researchers, indicating impairments on ToM-tasks in

individuals with psychosis symptoms/disorders (e.g., Achim, Ouellet, Roy, & Jackson,

2011; Bora, Yucel & Pantelis, 2009; Langdon & Coltheart, 2001; Langdon & Ward,

2008; Pijnenborg, Spikman, Jeronimus, & Aleman, 2013). Recently, RFT researchers

have shifted their attention towards this clinical domain and have attempted to further

understand its links with perspective-taking as derived relational responding (Hendriks

et al., 2016b).

A relational frame approach. As stated previously, some RFT researchers have

been examining a potential link between impairments in adopting the perspective of

others with the symptoms of schizophrenia (e.g., Villatte et al., 2010b) and specifically

social anhedonia (e.g., Vilardaga et al., 2012; Villatte, et al., 2008, 2010a). For example,

8

a study by Villatte et al. (2008) indicated weaker performances among participants with

high-anhedonia on ToM tasks and a RFT deictic-relations protocol. That is, the high-

anhedonia group performed significantly weaker on a deictic relations task (I-YOU

reversals) relative to a low-anhedonia group, indicating an impaired flexibility in

perspective-taking. Villatte et al. (2010a) replicated the previous findings with

participants with high social anhedonia producing more errors when making attributions

about another relative to a comparable performance when making attributions to self

(the errors when making attributions to self were comparable to the errors produced by

a control group). In another study, Villatte et al. (2010b) compared non-clinical

participants with a clinical population diagnosed with schizophrenia and found that the

clinical group made significantly more errors on a perspective-taking task, specifically

on I-YOU relations. The clinical group also performed significantly weaker on both

simple and reversed relations relative to the non-clinical group, thus indicating a

diminished ability to correctly make attributions of false-beliefs to others in individuals

with schizophrenia. Overall, these studies suggest an impaired flexibility in deictic

framing involving interpersonal relations for individuals reporting symptoms of

psychosis, with differences emerging between clinical and non-clinical samples, at least

using the RFT-based deictic protocol.

Assessing Perspective-Taking from an RFT Approach

Exploring alternative methodologies. Clearly, there has been a growing

amount of research into perspective-taking from a relational frame theory point of view,

especially in the clinical field (Montoya-Rodríguez et al., 2016). Critically, as noted

earlier, RFT-based measures of perspective-taking were mostly verbally delivered

protocols specifically developed for children and as such were viewed as a skill that was

either present or not. However, a small number of studies have provided evidence that

perspective-taking might be better considered as lying on a continuum ranging from less

to more flexible/evolved (DeBernardis, Hayes, & Fryling, 2014; Kavanagh et al., 2018;

McHugh et al., 2004). Building on this idea, RFT has attempted to develop a measure

that assesses perspective-taking as a verbal ability on which individuals differ in how

fluidly or flexibly they can apply this in their everyday lives (McEnteggart, Barnes-

Holmes, Hussey, & Barnes-Holmes, 2015). Another development was the shift in

9

attention from children towards adults focusing in particular on how a lack of flexibility

might be related to certain forms of psychological suffering in adults (e.g., Villatte et

al., 2010a, b).

Although the original protocol, and its variants, provided some key insights into

perspective-taking and its development, as pointed out by Hussey et al. (2014), it has

also shown to be lacking in certain areas when applied to adults (Kavanagh et al., 2018).

Firstly, the original deictic protocol was designed to test and train the previously

mentioned deictic relations in children and cannot simply be generalized to adults.

Secondly, as previously mentioned, RFT views deictic relational responding as a skill

that lies on a continuum of flexibility, rather than a skill that is simply present versus

absent. A “deficit” approach to perspective-taking does not, therefore, facilitate an

explanation of psychological suffering in terms of flexibility in perspective-taking or the

development of the self generally (Kavanagh et al., 2018; McEnteggart et al., 2017).

Thirdly, although the original protocol aimed to assess the basic skills underpinning

deictic relational responding by targeting the deictic frames, several factors such as IQ

and cognitive functioning seem to impact performances on the protocol (Gore, Barnes-

Holmes, & Murphy, 2010; Hendriks et al., 2016a; McHugh et al., 2004; Vitale, Barnes-

Holmes, Barnes-Holmes, & Campbell, 2008). Finally, because the protocol was

designed to study deictic responding primarily in children with specific relational

deficits (Barnes-Holmes et al. 2013; see e.g., Rehfeldt, Dillen, Ziomek, & Kowalchuk,

2007), it might not be sensitive enough to assess subtle differences in clinical

psychological suffering. Thus, while there are some studies on psychological suffering

which demonstrate deficits, a more sensitive and age-adjusted measure might deliver a

more precise understanding (Kavanagh et al., 2018; Villatte et al., 2008, 2010a, b).

The implicit relational assessment procedure (IRAP). One way in which

flexibility in relational responding, in general, has been studied within RFT involved

developing the implicit Relational assessment procedure (IRAP; Barnes-Holmes,

Barnes-Holmes, Stewart, & Boles, 2010). The IRAP is a computer-based procedure that

aims to assess the relative strength of previously established relational frames, rather

than their mere presence or absence in an individual’s behavioural repertoire (Barnes-

Holmes, Barnes-Holmes, Hussey, & Luciano, 2015; Barnes-Holmes, Barnes-Holmes,

Stewart, & Boles, 2010).

10

An IRAP typically presents participants with both target and label stimuli along

with two response options; for example, a picture of a snake with the word “pleasant”

and the words, ‘True’ and ‘False’ as response options. The participant then confirms

(presses ‘True’) or disconfirms (presses ‘False’) the relationship between the two types

of stimuli by pressing one of two response keys. The presented stimulus-response

mappings can either be consistent or inconsistent with expected response biases. During

a consistent trial, the correct response will likely cohere with the pre-experimentally

established relationship between the two stimuli. For example, if “unpleasant” is shown

with a picture of a snake, the participant should confirm this relationship by pressing the

‘True’ key. If, however the participant chooses the incorrect response (denies that a

snake is unpleasant) a red ‘X’ appears and the participant has to answer correctly in

order to move on. During an inconsistent trial, the correct response is incoherent with

the pre-experimentally established verbal relationship. For example, if the label

“pleasant” is presented with a picture of a bunny, the participant should disconfirm this

by pressing ‘False’. Participants are first presented with blocks of practice trials until

they reach certain criteria before they move on to the test blocks. The combination of

the two categories yields four trial-types (e.g., Bunny-Pleasant-True/False; Bunny-

Unpleasant-True/False; Snake-Pleasant-True/False; Snake-Unpleasant-True/False).

The difference in response latencies between the consistent and inconsistent blocks is

then transformed into DIRAP-scores to reduce contamination by factors related to non-

experimental variables. This DIRAP-score can then be used as an index for the relative

strength of a specific pattern of relational responding.

In most of the previous studies, the IRAP has been used as a measure of implicit

attitudes across a range of domains (see e.g., Roddy, Stewart, & Barnes-Holmes, 2010;

Scanlon, McEnteggart, Barnes-Holmes, & Barnes-Holmes, 2014). As noted above,

however, the IRAP was in fact developed originally as a measure of relational framing,

a point that has been reiterated in recent conceptual and empirical articles (see Barnes-

Holmes et al., 2015 and Hussey, Barnes-Holmes, & Barnes-Holmes, 2015a, for a more

detailed account). Another recent development is the more frequent use of this

procedure to assess more complex relational responses in the clinical domain (Vahey,

Nicholson, Barnes-Holmes, 2015), for example, to assess body image (Parling,

Cernvall, Stewart, Barnes-Holmes, & Ghaderi, 2012), self-esteem (Remue, De Houwer,

11

Barnes-Holmes, Vanderhasselt, & De Raedt, 2013; Vahey, Barnes-Holmes, Barnes-

Holmes, & Stewart, 2009), and of particular interest for the research presented in the

current thesis, perspective-taking (Barbero-Rubio et al., 2016; Kavanagh et al., 2018).

The IRAP and perspective-taking. Barbero-Rubio and colleagues (2016) were

the first researchers to adapt the IRAP as a measure of perspective-taking, along with

more explicit measures of perspective-taking. In their study, they assessed only two

deictic relations: interpersonal (I-YOU) and spatial (HERE-THERE). During the IRAP,

participants were asked to confirm (‘yes’) or disconfirm (‘no’) a relation between a

target and label stimulus. These statements concerned either the participant (self) or the

researcher (other) and participants were required to confirm or disconfirm statements

pertaining to the self versus the other. For example, “Kim” (participants name) was

presented at the top of the screen with a statement underneath pertaining to a property of

either the participant or researcher, such as ‘holding a pen’. If the researcher was

holding a pen, and not the participant (“Kim”), during simple or consistent blocks,

participants had to disconfirm the relation by pressing the ‘no’ key. During reversed

blocks, participants were required to respond in accordance with the instruction, “If I

were YOU and YOU were ME”, and thus ‘yes’ (“Kim is holding a pen”) was the correct

choice. Results indicated differential IRAP performances across simple and reversed

blocks with participants responding more slowly on the reversed trials. The IRAP

performances also correlated with an explicit measure of perspective-taking, with lower

DIRAP-scores being associated with greater accuracy on specific types of reversed trials.

These findings have been partially replicated by Kavanagh et al. (2018), but a

number of differences also emerge. The details of these differences are not critical for

the current research because they may have been related to a range of procedural

differences across the two studies (see Kavanagh, et al., for a detailed discussion). A

more important issue in the context of the current research is that the studies by both

Barbero-Rubio et al. (2016) and Kavanagh, et al. could be criticized on the basis that the

IRAP performances were not driven by perspective-taking per se. Rather, it is possible

that participants simply completed the IRAP by responding in accordance with current

events during simple blocks (i.e., confirming that the researcher was holding and pen)

and then responded in the opposite direction during trials that putatively required

perspective-taking (i.e., simply disconfirming that the researcher was holding a pen). In

12

other words, participants could complete the IRAP without necessarily having to

imagine or think of themselves as literally being the researcher. One way in which it

may be possible to circumvent this interpretive problem with the IRAP is to employ a

version that allows for the presentation of stimuli that require perspective-taking on a

trial-by-trial basis.

The IRAPs employed by Barbero-Rubio et al. (2016) and Kavanagh, et al.

(2018) presented relatively simple stimuli, such as a picture of the researcher holding a

pen, and thus perspective-taking during the actual task may have been limited or

completely absent. If each IRAP trial contained a statement that asked participants to

produce a response that reflected either their own perspective or that of another, with

regard to a specific internal state, then perhaps such a procedure would better capture

perspective-taking during the actual task. Beginning to explore this possibility was the

primary purpose of the research reported in the current thesis.

The Current Study

In aiming to develop an IRAP that may capture perspective-taking more

effectively than previous IRAP studies, it may be wise to employ a new version of the

procedure that allows for the presentation of relatively complex statements. In doing so,

perspective-taking stimuli pertaining to internal states could be employed. Using the

traditional IRAP can accommodate the use of sentences, however, they would still have

to be separated into target and label stimuli which may come across as unnatural

(Kavanagh, Hussey, McEnteggart, Barnes-Holmes, & Barnes-Holmes, 2016). This issue

led to the development of the natural language IRAP (NL-IRAP; see Kavanagh et al.

2016). In a NL-IRAP both the label and target stimuli are simultaneously presented in

the format of a single sentence or statement, similar to everyday language use or verbal

interactions. As Kavanagh, et al. (2016) point out, this may allow researchers to insert

questionnaire-based items that might have otherwise been difficult to assess separately

using the traditional IRAP. On balance, the NL-IRAP is still a very recent development

and thus it seems important to explore its potential as a research tool; this was a

secondary purpose of the current research.

The general aim of the research reported in the current thesis is to determine if

the NL-IRAP can be used to differentiate between taking your own perspective versus

13

taking the perspective of another. This study will thus seek to extend the findings

reported by Kavanagh and colleagues (2018), only using a deictic NL-IRAP. By doing

so this study will attempt to develop an IRAP that assesses perspective-taking

pertaining to the internal states of self versus other. The current study will employ two

separate NL-deictic IRAPs, one which targets self-related responses to positive and

negative events and a second that targets other-related responses to such events. The

research will be largely exploratory, in part because the NL-IRAP is relatively new and

thus it is difficult to predict exactly what it will yield in the domain of perspective-

taking. On balance, a general aim of the study will be to determine if the two types of

NL-IRAP (self versus other perspective) produce any differences in the observed

performances. In particular, if the IRAPs are sensitive to perspective-taking we may

expect to observe generally larger IRAP effects for the self- relative to the other-focused

IRAP. Specifically, one might expect that individuals should be more certain in drawing

conclusions about their own internal states than that of others, and this would be

reflected in stronger IRAP effects for the self-related IRAP. The current study also

involved specifying that the ‘other’ was deemed to be particularly different to the self

and thus there was a reasonable expectation that differences in the two IRAP

performances should emerge.

In addition, the current study also included a range of self-report measures of

psychotic-like experiences and the self, as well as self-report measures of interpersonal

closeness/ boundaries. The main purpose in adding these self-report measures was to

explore whether the ability to take one’s own perspective versus that of another relates

to psychosis symptomology in the general population. Previous studies have

consistently shown an impaired flexibility to take another’s perspective in individuals

with symptoms of psychosis. Thus, if the NL-IRAPs employed in the current research

are tapping in perspective-taking, they should correlate in some manner with such

measures. As noted above, however, this is the first study using a relatively new

procedure (NL-IRAP) and thus no specific hypotheses were formulated.

14

Method

Participants

Thirty-two participants took part in the current study, eight males and twenty-

four females, and ranged from 17 to 32 years old (M = 21,44). The sample size was

decided based upon a meta-analysis on the use of the IRAP in the clinical domain

(Vahey, Nicholson, & Barnes-Holmes, 2015). The experiment took on average an hour

and a half to complete and each participant was paid 15 euro. Participants were selected

through random convenience sampling from the participant pool of the Department of

Experimental, Clinical and Health Psychology of the Ghent University. Participants

were invited to the university research facility to complete the experiment and all

participants signed informed consent prior to commencing the experiment. All

participants consented to take part in all aspects of the study, however the data from two

participants were excluded from the final analysis (see procedure section). The final

sample comprised thirty participants. In addition, during the process of analysing the

data, an interesting three-way interaction effect was observed that approached

significance, so in order to investigate this trend further and to increase statistical

power, the data from a closely related study was combined with the current dataset (i.e.,

see Experiment 6 from Kavanagh et al., in press). In this study, thirty-four participants

participated, four males and thirty females, and ranged from 18 to 50 years old (M =

21.29), and a total of thirty participants were retained after exclusions were applied,

leading to a final combined sample size of sixty participants, 12 males and 48 females,

ranging from 17 to 32 years old (M = 20.75). Both studies were approved by the ethical

committee of the Ghent university. And all procedures were executed conform its

ethical standards and those of the 1964 Helsinki Declaration and its later amendments.

Design

The current experiment was a 2x4 (NL-IRAP x trial-type) within-subjects

design. The order of the NL-IRAPs was counterbalanced across participants and the

four trial-types were presented randomly within each test block. A more detailed

description of the procedure is provided subsequently.

15

Materials and Apparatus

Two computer-based NL-IRAPs and six questionnaires were employed in the

current study. The two NL-IRAPs comprised of a self-focused IRAP and an other-

focused IRAP. The six questionnaires comprised of the Community Assessment of

Psychic Experiences (CAPE), the Psychological Flexibility Indicator (PFI), the

Relationship Structures questionnaire of the Experiences in Close Relationships –

Revised (ECR-RS), the Experiencing of Self Scale (EOSS), the Inclusion of the Other

in the Self Scale (IOS), and the self-warmth thermometer that were all presented by

means of the computer program ‘PsychoPy2’ (v1.83.01, Peirce, 2007, downloaded from

http://www.psychopy.org/index.html). The current study was part of a larger research

project on perspective-taking conducted by the Contextual Behavioural Science (CBS)

research group (see Kavanagh et al., in press). Thus, these specific questionnaires were

employed across all studies in this series in an effort to measure phenomena that are

often considered to be related to perspective-taking and the concept of the self, such as

attachment, relationships, psychological flexibility, and psychosis. All questionnaires

and their instructions (with the exception of the CAPE which has a validated Dutch

version) were translated through a backward-forward method according to the World

Health Organization guidelines (WHO, 2017) by two Dutch speaking colleagues in the

Department of Experimental Clinical and Health Psychology at Ghent University. All

aspects of the study (NL-IRAPs and self-report measures) were conducted on a standard

Dell laptop computer (17.3 inches).

Natural language-implicit relational assessment procedure (NL-IRAP). The

two NL-IRAPs were presented using the Ghent Odysseus GO-IRAP software which is

written in Java (v1.0, downloaded from https://go-rft.com/go-irap/). The statements

used for both IRAPs1 were designed to reflect differential functional properties of

responding from the perspective of the self and from the perspective of another (e.g.,

positive/negative reactions to positive/negative events). A pilot study indicated that

these stimuli produced differential responding across the trial-types and thus were also

employed in the current study (see also, Kavanagh et al., in press). The format used in

the current study is presented in Figure 1.

1 To accommodate the reader, from this point on when referring to the NL-IRAPs used in the current study, we will simply refer to them as ‘IRAPs’.

16

Figure 1. Examples of the four trial-types in the self-focused IRAP. The stimulus valence (e.g., Positive Event) is presented in brackets but was not presented on-screen. Note. The presented format is displayed in English, but all experimental stimuli were presented in Dutch.

Self-focused NL-IRAP. The self-focused IRAP measures the ability to take your

own perspective. The self-focused IRAP presented participants with a sentence

concerning their reaction (positive or negative) in a given situation (positive or

negative). There were four statements for each trial-type resulting in a total of 16

statements (see Table 1). The four trial-types were as follows: Positive event-Positive

reaction (e.g., “When someone says I look good I feel confident.”), Negative event-

Negative reaction (e.g., “When someone I love dies I feel distraught.”), Positive event-

Negative reaction (e.g., “When someone says I look good I feel ugly.”) and Negative

event-Positive reaction (e.g., “When someone I love dies I feel happy.”) that were

presented randomly. Participants were required to respond with either ‘yes’ or ‘no’ by

pressing either ‘d’ or ‘k’. The response options (i.e., ‘yes’ or ‘no’) were displayed at the

bottom left- and right-hand corners of the screen on each trial. Since this is one of the

first studies using this IRAP, there are no psychometric properties available at this

instance.

17

Table 1

Stimuli employed in the self-focused IRAP’s statements for each trial-type

Note. The IRAP statements used in the experiment were presented in Dutch but are presented here in English to accommodate the reader. The statements were presented randomly within each block.

Other-focused NL-IRAP. The other-focused IRAP was similar to the self-

focused IRAP, but required participants to respond to several statements about a

specific other who they were close to, but perceived as very different to them. Once

again, the other-focused IRAP presented participants with a sentence concerning the

others’ reaction (positive or negative) in a given situation (positive or negative) (see

Table 2). Once, again there were four trial-types each comprising of four natural

language statements, leading to a total of 16 statements similar to those of the self-

focused IRAP. The four trial-types were as follows: Positive-Positive2 (e.g., “When

2 To accommodate the reader, from this point on, we will occasionally use a shorter formulation for the trial types (e.g., Positive-Positive). The first part always describes the valence of the event (positive or negative), and the last part describes the valence of the emotional response (positive of negative).

Trial-type Stimuli

Positive Event-Positive Reaction

When someone says I look good I feel confident. If my enemy dies I am relieved.

If I win the lottery I am delighted. Passing my exams makes me feel proud.

Positive Event-Negative Reaction

When someone says I look good I feel ugly. If my enemy dies I am upset.

If I win the lottery I am disappointed. Passing my exams makes me feel frustrated.

Negative Event-Positive Reaction

When someone I love dies I feel happy. Getting a fine makes me feel pleased.

Failing an exam is great. If someone I hate wins the lottery I feel amazing.

Negative Event-Negative Reaction

Getting a fine makes me angry.

When someone I love dies I feel distraught. Failing an exam is disappointing.

When someone I hate wins the lottery I feel bitter.

18

(name) is told they are good looking they feel confident.”), Negative-Negative (e.g.,

“When someone close to (name) dies they feel distraught.”), Positive-Negative (e.g.,

“When (name) is told they are good looking they feel upset.”) and Negative-Positive

(e.g., “When someone close to (name) dies they feel happy.”) that were presented

randomly. Again, participants were required to respond with either ‘yes’ or ‘no’.

Table 2

Stimuli employed in the other-focused IRAP’s statements for each trial-type

Note. The ‘X’ represents the name of the specified other which was defined by each participant individually.

Trial-type Statement

Positive Event-Positive Reaction

When X is told they are good looking they feel confident.

If X’s enemy dies they feel relieved. If X wins the lottery they are delighted.

If X passes an exam they feel proud.

Positive Event-Negative Reaction

When X is told they are good looking they

feel upset. If X’s enemy dies they feel upset.

If X wins the lottery they feel disappointed. If X passes an exam they feel unsatisfied.

Negative Event-Positive Reaction

When someone close to X dies they feel happy.

Getting a fine makes X feel pleased. When X fails an exam they are delighted.

If X’s enemy wins the lottery they feel overjoyed.

Negative Event-Negative Reaction

When someone close to X dies they feel

distraught. Getting a fine makes X angry.

If X fails an exam they feel disappointed. If X’s enemy wins the lottery they feel bitter.

19

The community assessment of psychic experiences (CAPE; Stefanis et al.,

2002). The community assessment of psychic experiences is a self-report questionnaire

adapted from the Peters Delusions Inventory (PDI-21; Peters, Joseph, & Garety, 1999)

in combination with items originating from the SANS (Andreasen, 1989) and SENS

(Selten, Gernaat, Nolen, Wiersma, & Van den Bosch, 1998) scales. The CAPE intends

to measure the life-time occurrence of subclinical psychotic-like experiences in the

general population. The CAPE comprises 42 items across three symptom dimensions: 8

items to measure depressive symptoms (e.g., “Do you ever feel pessimistic about

everything?”); 14 items measuring negative symptoms (e.g., “Do you ever feel that your

feelings lack intensity?”); and 20 items measuring positive symptoms (e.g., “Do you

ever feel that as if you are being followed in some way?”). The CAPE measures both

the frequency and distress of symptoms, each on a four-point Likert scale from 0

(frequency: never; distress: not distressed) to 3 (frequency: nearly always; distress: very

distressed). These can then be summed into separate frequency and distress scores of

psychic experiences, and for each of the three dimensions, that are all weighted to

control for partial non-responses (i.e., dividing the sum score by the total amount of

valid responses). Scores on the CAPE can range from 42 to 168, with lower scores

indicating lower frequency and/ or distress, and higher scores indicating higher

frequency and/ or distress. There are no clinical cut-offs for this measure. All three

dimensions have been shown to be independent, and have shown to have good internal

consistencies (Stefanis et al., 2002, Verdoux, Sorbara, Gindre, Swendsen, & Van Os,

2003, Brenner et al., 2007).The Dutch version has an adequate reliability: depressive

dimension (a = 0.62); negative dimension (a = 0.64); positive dimension (a = 0.63)

(Konings, Bak, Hanssen, Van Os, & Krabbendam, 2006), high internal stability (a =

0.6-0.8), and sufficient discriminant validity with coherent dimensions of interview-

based measures of psychosis (Konings et al. 2006). The current study employed the

Dutch version based on Van Os, Verdoux, and Hanssen (1999).

The psychological flexibility index (PFI; Bond et al., in preparation). The

psychological flexibility index assesses psychological flexibility and of which the

development currently still ongoing (Bond et al., 2017). The PFI is adapted from the

Acceptance and Action Questionnaire-II (AAQ-II) developed by Bernaerts, De Groot

and Kleen (2012). The version used in the current study consists of 82 items, each rated

20

on a 6-point Likert scale, ranging from 1 (disagree strongly) to 6 (agree strongly).

Overall scores on the PFI can range from 82 to 492. Some items used reversed scoring.

Overall scores are calculated by summing scores on all the items. Low scores indicate

lower levels of psychological flexibility, while high scores indicate high psychological

flexibility. There are no psychometrical properties available at present, nor are there

clinical cut-offs. However, both to the English (a = .85; Bond et al., 2011) and Dutch

version (a = .85) (Bernaerts et al., 2012) of the original measure (AAQ-II) on which the

PFI is based, has proven to have good internal consistency. Participants in this study

were presented with a Dutch version of the PFI, which was translated through

backward-forward translation.

The experiences in close relationship structures questionnaire – revised

(ECR-RS; Fraley, Heffernan, Vicary, & Brumbaugh, 2011a). The experiences in

close relationships structures questionnaire (revised version) is a self-report

questionnaire which measures adult attachment across multiple contexts, namely the

relationship with mother, father, romantic partner, and best friend. The ECR-RS had

been adapted from the experiences in close relationships questionnaire (ECR)

developed by Brennan, Clark and Shaver (1998). It consists of 9 items across two

subscales: an anxiety subscale and an avoidance subscale. The anxiety subscale consists

of 3 items assessing the extent to which an individual is concerned the specified other

may reject them (e.g., “I often worry that this person doesn’t really care for me.”), and

the avoidance subscale consists of 6 items assessing how individuals regulate their

attachment related behaviours (e.g., “I usually discuss my problems and concerns with

others.”), each rated on a seven-point Likert scale ranging from 1 (strongly disagree) to

7 (strongly agree). Scores on the anxiety dimension can range from 3 to 21, with higher

scores indicating high attachment related anxiety. Scores on the avoidance dimension

can range from 6 to 42, with higher scores indicating high attachment related avoidance.

A total score for the anxiety and avoidance subscales can be calculated by averaging

scores on each scale separately across the four targets (mother, father, partner and best

friend). A general attachment score can also be calculated by summing the average

score on each scale across targets (mother, father, romantic partner and best friend).

There are no clinical cut-offs for this measure. The internal consistency of both the

anxiety and avoidance subscales has proven to be highly reliable (all as’ = > .80) for

21

each of the four contexts separately, as well as across the four contexts (overall anxiety/

avoidance score) (Fraley et al., 2011a). The test-retest reliability has proven to be

adequate for two domains, the parental (a = .80) and romantic (a = .65) domain, for

each subscale (Fraley, Vicary, Brumbaugh, & Roisman, 2011b). There are no properties

available for the other two domains. As Fraley et al. (2011a) described, both subscales

have shown to be only modestly correlated. Participants in this study were presented

with a Dutch version of the ECR-RS, which was translated through backward-forward

translation.

Experiencing of self scale (EOSS) (Parker, Beitz, & Kohlenberg, 1996). The

experiencing of self scale is a self-report questionnaire that measures the influence of

others on the experience of self. The EOSS scale is comprised of 20 items across four

subscales, each consisting of five items. Participants are presented with self-experience

statements referring to different types of self-experiences (e.g., attitude, feelings), and

are asked to indicate the influence of others along two dimensions that are combined

across: type (casual acquaintance or close relationship) and proximity (present or

absent). These four subscales represent an increasing salience of others on the

experience of self, ranging from least to most salient: casual acquaintances-absent (CA-

A, e.g. “My attitudes are influenced by casual acquaintances when I am alone”), casual

acquaintances-present (CA-P, e.g. “My feelings are influenced by casual acquaintances

when I am with them”), close relationships-absent (CR-A, e.g., “My actions are

influenced by close relationships when I am alone), close relationships-present (CR-P,

e.g. “My wants are influenced by close relationships when I am with them). All items

are rated on a seven-point Likert scale, ranging from 1 (never true) to 7 (always true).

Scores per dimension can range from 5 to 35, with low scores indicating low influence

of others over the experience of self. And high scores indicating a greater influence of

others over the experience of self. There are no clinical cut-offs for this measure. Both

the EOS total score as well as the four subscales have shown to have high internal

consistency (Total a = .91, CA-A a = .87, CA-P a =.83, CR-A a = .93, CR-P a = .88)

(Kanter, Parker, & Kohlenberg, 2001). Participants in this study were presented with a

Dutch version of the ECR-RS, that was translated through backward-forward

translation.

22

Inclusion of the other in the self scale (IOS; Aron, Aron, & Smollan, 1992).

The inclusion of other in the self scale is a pictorial measure of relational

interconnectedness with specified others. It consists of a single item on which

participants are asked to choose one of seven Venn diagrams that depicts two circles

which represent the level of closeness between oneself (diverging overlap) and

“someone who they have a significant relation to, such as their best friend” or “someone

who they have no significant relation to, such as other people generally”. A separate

score is calculated for each subscale (best friend and other people generally) and can

range from 1 (no overlap) to 7 (almost completely overlapping). A low score on score

on this scale indicates a low accordance with the specified other, while a higher score

indicates a high accordance with the specified other. There are no clinical cut-offs for

this measure. The IOS has shown to have adequate alternate-form (a = .93) and test-

retest reliability (r = .83) (Aron et al., 1992). It has also shown to have significant

convergent validity, however, correlations with similar measures were low to moderate

(Aron et al., 1992). Although, a more recent study conducted by Gächter, Starmer and

Tufano (2015) found it to be highly correlated with related measures such as the

Relational Closeness Index across three studies. Participants in this study were

presented with a Dutch version of the ECR-RS, which was translated through

backward-forward translation.

Feelings about self thermometer (Vahey, Barnes-Holmes, Barnes-Holmes,

& Stewart, 2009). The feelings about self thermometer is a subjective measure of self-

esteem which has been adapted from Greenwald and Farnham (2000). The feelings

about self thermometer consists of a single item. Participants are asked to indicate on a

horizontal thermometer how warmly they feel towards themselves, ranging from 0

(cold) to 100 (warm) on a continuous scale with balanced intervals of ten. Lower scores

indicate that one feels generally cold towards themselves, while higher scores on this

measure indicates that one feels warm or positive toward themselves. There are no

clinical cut-offs for this measure, nor a standardized version. Therefore, there are little

to no psychometric properties available. Although, as indicated by Vahey et al. (2009)

and Nelson (2008), research supports their utility, as well as the moderate to high

convergent validity with self-report measures, such as the Rosenberg Self-esteem scale

(r = .74, Greenwald & Farnham, 2000; r = .68, Karpinski, 2004). Participants in this

23

study were presented with a Dutch version of the ECR-RS, which was translated

through backward-forward translation.

Procedure

The experiment was completed on an individual basis in separate, sound-proof

test rooms at the research facility of the Department of Experimental, Clinical, and

Health Psychology at Ghent University. Participants were recruited through random

sampling from the Ghent University participant pool. Participants could not have

participated in experiments using the same/ an associated protocol to eliminate practice

effects. Both studies were conducted by female researchers, and the procedure was

automated to minimize influence caused by the experimenter (the first study was carried

out by L. Van Raemdonck and the second added study by D. Kavanagh). Informed

consent was obtained from all participants. Each participant was presented with the self-

focused and the other-focused IRAPs, of which the order was counterbalanced across

participants (was done manually by the researcher through the use of a counter-

balancing sheet). Finally, participants completed the self-report measures presented in

the following order: CAPE, PFI, ECR-RS, IOS, EOS, and self thermometer.

Self-focused NL-IRAP. The self-focused IRAP comprised a minimum of one

(i.e., one consistent and one inconsistent block) and a maximum of eight pairs of

practice blocks. Followed by three pairs of test blocks (i.e., six test blocks in total). On

each trial, a self-related statement was presented in the middle of the screen (e.g.,

“Getting a fine makes me angry”, see Table 1), with two response options (‘yes’ and

‘no’) displayed at the bottom left- and right-hand corner of the screen. After a brief

introduction of the task (see Appendix 1) in which the importance of responding both

accurate and fast was stressed. The self-focused IRAP commenced with the practice

blocks. Participants were instructed to work out the setup of the task based on the

provided trial feedback. During each trial, participants received feedback on the

computer screen on their relative accuracy and median latency levels as well as

additional feedback by the researcher if necessary (see Appendix 1). If participants did

not meet the set trial-type criteria, which are described subsequently, before completing

a maximum of eight practice blocks, the experiment would be terminated to minimize

practice-effects.

24

Participants were required to respond to each statement with either ‘yes’ or ‘no’

by using either the “d” key corresponding with the response option presented at the

bottom left of the screen or the “k” key corresponding with the response option

presented at the bottom right. The position of the response options (i.e., ‘yes’ and ‘no’)

alternated across trials in a quasi-random order to ensure participants were responding

to the actual content rather than simply matching response options and their orientation.

Hence, the response options were never presented in the same position (i.e., ‘yes’

presented on the left) for more than three consecutive trials. Between selecting a correct

response and presenting the next trial within each block, an inter-trial interval of 400 ms

occurred. When participants answered incorrectly a red ‘X’ would appear on the screen

along with the statements. In order to move on to the next trial, participants were

required to select the correct response. The pattern of trial presentation followed this

model for all 32 trials comprising a block, along with corrective feedback on each trial

(i.e., a red ‘X’ when answering incorrectly). Trials within a block were presented in a

quasi-random order, and each statement was presented twice in each block. Participants

always commenced with a consistent block which required a response coherent with the

experimentally established history/ provided feedback, more precisely positive events

produce positive reactions and negative events produce negative reactions (i.e.,

Positive-Positive/yes, Negative-Negative/yes, Positive-Negative/no, Negative-

Positive/no). After which an inconsistent block followed requiring an opposing

response which was indicated by a standardized instruction (“The previously correct

and incorrect answers have now been reversed”) presented at the beginning of each

inconsistent block. More precisely, positive events produce negative reactions and

negative events produce positive reactions (i.e., Positive-Positive/no, Negative-

Negative/no, Positive-Negative/yes, Negative-Positive/yes). After each block

participants were presented with feedback on their accuracy and speed during that

specific block presented on the computer screen.

After each block of trials, the IRAP program provided feedback on participants’

accuracy and median latency during that block. In contrast to many previous IRAP

studies the current study set the required criteria on trial-type level rather than block

level. Their accuracy level reflected the percentage of correctly emitted first responses

(required no correction). Participants needed to achieve an average accuracy of

25

minimum 75% on each trial-type (i.e., a minimum of 2 errors were allowed per trial

type). The average latency was calculated by averaging the interval between presenting

the statement and the first correctly emitted response across all 32 trials within a block.

Participants also needed to achieve a maximum median latency of maximum 5000 ms

on each trial-type. These criteria were based on previous studies using this procedure

(Kavanagh et al., in press). After participants met the set trial-type criteria (latency of

<= 5,000 and accuracy level of >= 75%) on either pair of practice blocks, they

immediately moved on to the first pair of test blocks. Though, there were no accuracy or

latency requirements in order to proceed to the following test blocks. The IRAP

program continuously provided participants with feedback on their performance at the

end of each block, to maintain accurate and fast responding.

Other-focused NL-IRAP. Before completing this IRAP, participants were first

asked to name someone they consider being close to yet who they perceive to have a

personality that is very different from theirs. This person’s first name was then inputted

in the NL-IRAP program by the researcher. Again, followed by a brief instruction (see

Appendix 1) to stress responding both accurate and fast. All aspects of the format of the

other-focused IRAP were identical to the self-focused IRAP, but with statements

regarding the specified, different other, rather than statements regarding self (e.g.,

“Getting a fine makes Lisa angry.”).

Self-report questionnaires. After completing the IRAPs, participants were

asked to complete the six self-report measures in the following order: CAPE, PFI, ECR-

RS, EOSS, IOS, and the feelings about self thermometer.

Results

The Results section is divided into three sections. The first section presents the

data and analyses from the two NL-IRAPs. The initial set of analyses that were

conducted revealed an interaction effect that approached significance, so in order to

determine if the interaction effect was relatively robust, additional data that had been

collected by another researcher, but using the same IRAPs, were added to the current

data set (details presented below). The remaining analyses were then conducted with the

combined data sets. The second section presents the data from the self-report

26

questionnaires; and the third section presents a set of correlational analyses between the

NL-IRAPs and the questionnaires.

IRAP Data

Data preparation and preliminary analyses. Thirty-two participants

completed the study, but one participant was excluded after failing to meet the trial-type

criteria (see below), and one participant was excluded due to software failure during the

Other-focused NL-IRAP. The final sample consisted of 22 females and 8 males (N =

30) and ages ranged from 17 to 32 years (M = 21,3). All aspects of data processing for

the IRAPs adhered to standard conventions (e.g., Nicholson & Barnes-Holmes, 2012),

with one modification. Specifically, all participants were required to meet pre-

determined performance criteria, which involved maintaining a median latency of

<=5,000 and an accuracy level of >= 75% correct across each of the trial types of the

test blocks. These stricter performance criteria are consistent with a general strategy that

was being employed within the research group within which the current study was

conducted.

The primary focus of the analyses conducted on the IRAP data involved

comparing within and between the two IRAPs. As such, effects for each of the four

trial-types for each of the two IRAPs were calculated, the results of which are presented

in Figure 2. Specifically, the IRAP latency data were converted into D-scores using a

standardized D algorithm for transforming the difference in response latencies between

consistent and inconsistent test blocks (see Appendix 2 for all steps in this procedure).

We used the D algorithm as formulated by Greenwald Nosek and Banaji (2003), which

as described by Hussey, Thompson, McEnteggart, Barnes-Holmes, Barnes-Holmes and

(2015b) is used in most IRAP studies (e.g., Vahey et al., 2015). Following this data

transformation, positive scores indicated that during history-consistent blocks

participants responded ‘yes’ more quickly than ‘no’ during Positive-Positive and

Negative-Negative trial-types; and responded ‘no’ more quickly than ‘yes’ during

Positive-Negative and Negative-Positive trial-types. Negative DIRAP-scores indicated the

opposite pattern (e.g. responding ‘no’ more quickly than ‘yes’ during Positive-Positive

or Negative-Negative trial-types).

27

Descriptive analyses. As illustrated in Figure 2, both IRAP’s produced positive

DIRAP-scores across each of the four trial-types, although the effects for the Negative-

Positive trial-type were close to zero. The size of the effects was similar across the two

IRAPs for three of the trial-types but differed dramatically for the Positive-Negative

trial-type, with a relatively large effect for the self-focused IRAP and a relatively weak

effect for the other-focused IRAP. In general, therefore, both IRAPs yielded effects that

were consistent with what might be predicted based on conventional assumptions,

except for the very weak effects observed for the Negative-Positive trial-type. The

relatively weak effect for the Positive-Negative trial-type for the other-focused IRAP

could also be seen as somewhat unexpected.

Figure 2. Mean DIRAP-scores on the self-focused NL-IRAP trial types and the other-focused trial types for the single study (N=30). Positive DIRAP-scores indicate a tendency towards history consistent responding and negative DIRAP-scores indicate scores indicate a tendency towards history-inconsistent responding. *Specifies mean DIRAP-scores that differ significantly from zero.

Statistical analyses. The data were subjected to a repeated measures 2 (self

versus other IRAP) x 2 (Positive versus Negative Event) x 2 (Positive versus Negative

Reaction) analysis of variance (ANOVA). The critical three-way interaction approached

significance [F(1,29) = 3.658, p = .066, !"#= 0.11], which appears to be driven largely

by the dramatic difference between the two IRAPs for the Positive event-Negative

reaction trial-type. Eight one-sample t-tests (see Table 3) indicated that three of the four

-0.05

0

0.05

0.1

0.15

0.2

0.25

0.3

0.35

Positive Event -Positive Reaction

Positive Event-Negative Reaction

Negative Event-Positive Reaction

Negative Event-Negative Reaction

Mea

n D

IRAP

-sco

res

Trial type

Self-focused

Other-focused

* * * *

*

28

trial-types for the self-focused IRAP (Positive-Positive: M = .16, SD =.24, t(29) =

3.696, p <.001; Positive-Negative: M = .17, SD = .31, t(29) = 3.026, p < .05; Negative-

Negative: M = .19, SD = .27, t(29) = 3.806, p < .001) differed significantly from zero

(all ps < .05), whereas this was the case for only two of the trial-types for the other-

focused IRAP (ps < .01) (Positive-Positive: M = .19, SD = .26, t(29) = 3.968, p < .001;

Negative-Negative: M = .23, SD = .37, t(29) = 3.370, p < .05); remaining ps > .61. The

trend towards significance for the three-way interaction, combined with the pattern of

significant and non-significant one-sample t-tests, suggests that the self- and other-

focused IRAPs were differentially sensitive for only one particular trial-type (Positive

event- Negative reaction).

Table 3

One sample t-tests for the single study (self versus different other)

Trial-type Mean P

Self Positive Event-Positive Reaction .160 < .001*

Self Positive Event- Negative Reaction .173 <.05*

Self Negative Event- Positive Reaction .018 .678

Self Negative Event-Negative Reaction .188 < .001*

Other Positive Event-Positive Reaction .187 < .001*

Other Positive Event- Negative Reaction .029 .610

Other Negative Event- Positive Reaction .021 .662

Other Negative Event-Negative Reaction .227 <.05*

* Specifies significant p values (p < .05).

29

On balance, it is important to recognize that the three-way interaction only

approached significance and therefore the effect should be interpreted with caution. In

order to determine if the effect was relatively robust it was decided to add data from a

closely related study that was conducted within the research group. The additional data

were taken from a study that was similar in all respects to the current study, except the

other-focused IRAP targeted an individual the participants disliked (rather than

someone who was simply different; see Appendix 1 for the instructions for this IRAP).

A total of 30 participants had completed this study and all of these data were simply

added to the existing data set and then the analyses were re-run once only, thereby

avoiding potential “p-hacking” issues.

The data were subjected to a mixed four-way 2x2x2x2 ANOVA with the three

same within-participant variables as specified previously, but with a single between-

participant variable (Liked versus Disliked other) added to the analysis. The critical

three-way interaction was now significant, [F(1,58) = 5.224, p = .026, !"#= .08];

however, the four-way interaction, with Liked versus Disliked other was not [F(1,58) =

.002, p = .963, !"#= 3.74-05]. Thus, the original three-way interaction appeared to be

robust, with little if any suggestion that it was moderated by whether or not participants

liked or disliked the individual targeted by the other-focused IRAP. To further examine

which effect drove this three-way interaction, we conducted four separate paired t-tests

(see Table 4). Results were in line with our previous findings indicating the three-way

interaction was driven by the difference between Self (M = .16, SD = .32) and Other (M

= .03, SD = .33) on the Positive event-Negative reaction trial-type (t(59) = 2.195, p <

.05); none of the other paired t-tests were significant (ps >.12) . Eight one-sample t-tests

(see Table 5) yielded the same pattern of significant versus non-significant effects for

both IRAPs, as was observed for the original data set. Again, three of the four trial-

types for the self-focused IRAP (Positive-Positive: M = .18, SD =.30, t(59) = 4.713, p

<.001; Positive-Negative: M = .16, SD = .32, t(59) = 3.924, p < .001; Negative-

Negative: M = .18, SD = .27, t(59) = 5.233, p < .001) differed significantly from zero

(all ps < .05), whereas this was the case for only two of the trial-types for the other-

focused IRAP (ps < .01) (Positive-Positive: M = .16, SD =.30, t(59) = 4.144, p <.001;

Negative-Negative: M = .20 SD = .34, t(59) = 4.538, p < .001); remaining ps > .20. A

graphical representation of the three-way interaction from the combined sample of 60

30

participants is presented in Figure 3; the pattern of effects is broadly similar to those

obtained from the original sample of 30 participants, with the most substantive

difference between the two IRAPs being observed for the Positive-Negative trial-type.

Given that a clearly significant interaction effect was only obtained from the combined

data set, all subsequent analyses of the questionnaire data, and the correlations, were

conducted using this final combined data set.

Figure 3. Mean DIRAP-scores on the self-focused NL-IRAP trial-types and specified, different other-focused trial-types for the combined sample (N=60). Positive DIRAP-scores indicate a tendency towards history consistent responding and negative DIRAP-scores indicate scores indicate a tendency towards history-inconsistent responding. *Specifies mean DIRAP-scores that differ significantly from zero.

-0.15

-0.1

-0.05

0

0.05

0.1

0.15

0.2

0.25

0.3

Positive Event -Positive Reaction

Positive Event-Negative Reaction

Negative Event-Positive Reaction

Negative Event-Negative Reaction

Mea

n D

IRAP

-sco

res

Trial type

Self-focused

Other-focused

* * * *

*

31

Table 4

Paired t-tests for the combined sample

Trial-type Mean Difference P

Self Positive Event-Positive Reaction/ Other Positive Event–Positive Reaction

.019 .717

Self Positive Event- Negative Reaction/ Other Positive Event-Negative Reaction .132 <.05*

Self Negative Event- Positive Reaction/ Other Negative Event-Positive Reaction .080 .119

Self Negative Event-Negative Reaction/ Other Negative Event- Negative Reaction -.019 .726

* Specifies significant p values.

Table 5

One sample t-tests for the combined study (self versus different and disliked other)

Trial-type Mean P

Self Positive Event-Positive Reaction .179 < .001*

Self Positive Event- Negative Reaction .164 < .001*

Self Negative Event- Positive Reaction .029 .396

Self Negative Event-Negative Reaction .181 < .001*

Other Positive Event-Positive Reaction .160 < .001*

Other Positive Event- Negative Reaction .032 .458

Other Negative Event- Positive Reaction -.051 .197

Other Negative Event-Negative Reaction .199 < .001*

* Specifies significant p values (p < .05).

32

Questionnaire Data

A summary of the means and standard deviations of each questionnaire and

relevant subscales is provided in table 6. Given that there are no clinical cut-offs for

most questionnaires we will be interpreting them in terms of how high or low the scores

are relative to the mid-range/mean. The weighted overall and subscale scores on the

CAPE, ranging from 1.54 to 2.60, were relatively low compared to a maximum

weighted score of 4.00, thus indicating a low to average presence of psychotic-like

symptoms in the current study sample.

The mean overall PFI score in this sample was relatively high at 360.62 (/492),

indicating high psychological flexibility.

The ECR-RS scores for both the avoidance and anxiety subscales ranged from

11.55 to 23.55 (/42) for the former, and from 5.47 to 9.77 (/21) for the latter. These data

indicate participants’ scores were low to average for both avoidant and anxious

attachment-related behaviours.

The mean overall EOSS score of 75.77 was at the mid-way point (/140),

indicating average control by others over the experience of self. The EOSS subscale

scores ranged from 11.85 to 25.2 (/35), indicating low to average control by others over

the experience of self. Scores for casual acquaintances and close relationships that were

present were higher than for relations that were absent (casual acquaintances: 17.75 vs

25.2; close relationships: 11.85 vs 20.97). Thus, others that were present seemed to have

greater impact on participants’ feelings, wants, opinions, attitudes, and actions than

absent others.

Finally, the IOS scores for best friend at 4.77 were around mid-way point (/7),

while the scores for other people (2.92) were at the lower end. The higher scores for

best friend imply a closer relationship with respect to other people.

In line with previous study samples (see Vahey et al., 2009), the mean score on

the Feelings toward Self Thermometer was situated in the mid-range (48.27/100),

indicating that on average participants did not feel particularly cold, nor warm towards

themselves.

Overall, therefore, and consistent with the use of a non-clinical sample, no

unexpected or atypical findings emerged from the questionnaires.

33

Table 6

Descriptive statistics for the questionnaires for the combined sample (N = 60)

Questionnaire

M SD

CAPE (weighted scores) Overall Frequency 1.81 .36 Frequency of Positive Symptoms 1.54 .36 Frequency of Negative Symptoms 1.98 .49 Frequency of Depressive Symptoms 2.18 .54 Overall Distress 2.19 .46 Distress associated with Positive Symptoms 1.77 .51 Distress associated with Negative Symptoms 2.07 .55

Distress associated with Depressive Symptoms

2.60 .69

Psychological Flexibility Index (PFI) ECR-RS

360.62 24.42

Attachment-related avoidance (Mother) 16.98 8.41 Attachment-related anxiety (Mother) 5.47 3.37 Attachment-related avoidance (Father) 23.55 9.93 Attachment-related anxiety (Father) 7.23 5.11 Attachment-related avoidance (Partner) 11.55 4.91 Attachment-related anxiety (Partner) 9.77 4.98 Attachment-related avoidance (Best Friend) 12.92 5.59 Attachment-related anxiety (Best Friend) 6.92 4.18 EOSS Overall EOSS 75.77 14.91 Casual Acquaintances-absent 17.75 5.89 Casual Acquaintances-present 25.2 3.91 Close relationships-absent 11.85 6.04 Close relationships – present 20.97 6.27 IOS Best Friend 4.77 1.36 Other people Feelings about Self Thermometer

2.92 48.27

1.15 26.45

Note. None of the questionnaires and their subscales have fixed clinical cut-offs. All CAPE subscales have a maximum weighted score of 4.00. The PFI has a maximum score of 492. All of the ECR-RS attachment related avoidance subscales have a maximum score of 42. The maximum score for all the Attachment related anxiety subscales is 21. The EOSS has a maximum score of 140 for the overall score, and a maximum score of 35 for each subscale. Both IOS scales have a maximum score of 7. Finally, the Feelings about Self Thermometer has a maximum score of 100.

Correlations between IRAP Scores and Self-Report Measures

A range of correlational matrices were calculated to explore the relationships

between performance on the self and other IRAPs and the self-report measures. The

34

results of these correlational analyses are presented across Appendix 3 to 8, with the

correlational analyses divided according to the six different self-report measures. The

correlational analyses involved five correlations for each IRAP, one correlation for the

overall score and four correlations for each of the trial-type scores for each IRAP (i.e., a

total of 10 correlations across the two IRAPs (self and other). A total of 21 correlations

were calculated across the six self-report measures for each of the IRAP performance

measures. In presenting the results of the correlational analyses, the r value for each

correlational is presented in the Tables. Correlations that were significant at p < .05

(without correction) and at p < .002 (with Bonferroni correction) are presented. The

correction involved dividing .05 by 21 because each IRAP performance score was

correlated with 21 different self-report measures.

The general strategy that was adopted in reporting on and discussing these

correlational analyses involved focusing on those correlations that were significant at p

< .002 and/or clustered together for a particular IRAP trial-type and specific self-report

measure. The clearest example of such clustering (i.e. five out of the eight correlations

were significant at the .05 level) may be observed in Table 6 for the correlations

between the CAPE and Negative-Negative trial-type for the Other-focused IRAP (see

Appendix 3). Interestingly, no correlations between this trial-type and the Cape scores

were recorded for the Self-focused IRAP. The correlational analyses thus suggest that

increased levels of psychological suffering, as a result of psychotic-like symptoms, was

associated with a reduction in confirming that others react negatively to negative events

in their lives.

The only other evidence of a clustering of significant correlations involving a

highly significant effect was with the ECR-RS and the Positive-Negative trial-types and

the Self-Focused IRAP (see Appendix 5). Specifically, the results suggest that increased

levels of psychological suffering in attachment-related self-reports were associated with

a reduction in denying that positive events produce negative reactions (with regard to

self).

Although a number of other significant correlations were recorded across the six

Tables, these were all at the .05 level with no evidence of clustering on a particular trial-

type for those measures that involved multiple subscales.

35

Discussion

The general aim of the research reported in the current thesis was to explore the

extent to which a NL-IRAP may be sensitive to perspective-taking in terms of how self

versus others react to positive versus negative events. Another aim was to examine

whether IRAP performances related to self-report measures assessing clinical concepts

often linked to perspective-taking in general, such as self-warmth and psychotic

symptoms. A potentially interesting interaction effect emerged when analysing the

original data set that was collected by the author of the current research. When the data

were combined with a data set that was collected by a different researcher, the original

pattern of effects remained largely unchanged, and thus the analyses and subsequent

discussion will focus on the combined data set.

Discussion of the Current Findings

The main finding arising from a comparison of the two IRAPs (self versus other)

was the difference on the positive-negative trial-type. Specifically, participants showed

a relatively strong bias towards denying that the self reacted negatively to positive

events, but this bias was absent for others. This difference indicates that the two IRAPs,

as expected based on the findings of Kavanagh et al. (2018), were tapping into separate

reactions to self versus others, at least with regard to one trial-type. At the current time

it remains unclear why this trial-type yielded a clear difference, but one possible

interpretation is as follows. Denying that you react negatively to positive events may be

relatively easy with regard to self (because you are very familiar with your own

reactions in this regard). With respect to others, however, participants may be less

certain. For example, some individuals, particularly those who are deemed to be

different or disliked relative to self, might be seen as generally negative and “never

happy,” even when good things happen to them. Or to put it another way, if you

consider yourself to be a generally positive and happy individual, others who are

deemed to be different/disliked might be seen as more negative relative to you (e.g.,

Chen et al., 2014; Suls, Lemos, & Stewart, 2002; Trafimow, Armendariz, & Madson,

2004; Zuckerman & O’Loughlin, 2006).

If the foregoing interpretation is correct how might we explain the absence of

any clear differences for the other three trial-types? For the positive-positive and

36

negative-negative trial-types, perhaps these trial-types failed to evoke relatively strong

reactions to self versus other because they required responses that either confirmed or

disconfirmed ‘common-sense’ reactions. Or in other words, the IRAPs simply asked the

participants to relate positive to positive and negative to negative, and the psychological

impact of self versus other was greatly diminished for these trial-types. But how might

we interpret the relatively weak effects obtained for both the self and other IRAPs on

the Negative-Positive trial-type? Similar to the Positive-Negative trial-type, it may have

been more difficult to respond in terms of basic common-sense, and thus it produced far

weaker effects. On balance, the trial-type could be interpreted as asking how resilient

you and others are to negative events. Or in other words, how well do you and others

cope when negative events occur in your lives. Given that reacting positively to

negative events may be seen as relatively difficult for everyone (i.e., self and others),

any difference between self and other IRAPs might be expected. On balance, it remains

unclear why the effect was relatively weak for both IRAPs (why did participants find it

difficult to simply deny that self and others tend to react positively to negative events?).

At this point, the foregoing interpretations could be seen as excessively

speculative and post-hoc, and thus it may be wise to consider an alternative perhaps

simpler interpretation of the IRAP effects. One possibility is to use a recently proposed

model of IRAP performances that might be relevant. The model was developed to

account for differential arbitrarily applicable relational responding effects (DAARRE)

that are often observed in IRAP performances. In contrast to an earlier model of IRAP

effects (see Barnes-Holmes et al., 2010), the DAARRE model focuses on both the

relationship between each element within a trial-type (e.g., between the event and the

reaction in the current research), and also on the psychological functions of the

individual elements, including the response options (i.e., ‘yes’ and ‘no’ in the current

study). Within the DAARRE model, the relationship between the elements (event-

reaction) is referred to as a “Crel” property and the psychological functions of the

individual elements is referred to as a “Cfunc” property. The reader should note that the

response options (yes and no) also are seen as possessing “Cfunc” properties; in the

current case ‘yes’ would have a generally positive “Cfunc” property and ‘no’ would

have a generally negative “Cfunc” property. A detailed description of the model and

how it has been used to interpret a range of IRAP effects that have been reported in the

37

recent literature is not necessary here (see, Finn, Barnes-Holmes & McEnteggart, 2018,

for more detail), but the basic model as applied to the current findings is described

subsequently.

According to the DAARRE model, when participants complete an IRAP their

responses will be influenced to a greater or lesser degree by the relations that are

presented within each trial-type and by the extent to which each of the elements within

the trial-type cohere with each other (e.g., in the current study, self with positive, self

with negative, yes with positive, no with positive and so on). Thus, for example, when

responding on the positive-positive or the negative-negative trial-types in the current

study, the relationship between the events and the reactions would be highly coherent

(i.e., positive events usually produce positive reactions and negative events usually

produce negative reactions). In this case, therefore, the relationship between the event

and the reaction would be highly dominant in determining participants’ responses for

both self and other IRAPs. The patterns observed in the current research accord with

this interpretation in that they produced relatively strong tendencies to confirm highly

coherent relations (i.e., positive-positive and negative-negative).

Figure 4. Application of the basic DAARRE model. Implementing two types of stimulus sets: the labels ‘Events’ (top) and the target stimuli ‘Reactions’ (bottom). Displayed at the bottom there are two response options (‘Yes’ and ‘No’). The negative and positive labels indicate the valence of each function (Cfunc): Crel (relational properties) and Cfunc (functional properties).

38

In turning to the Positive-Negative and Negative-Positive trial-types the relative

dominance of the relationship between the event and the reaction may be much reduced

(because they do not cohere). Given that the IRAPs targeted self or a different/disliked

other, the coherence between the positive or negative reaction (of self or other) could be

a dominating influence. Thus, when responding on the Positive-Negative trial-type,

denying negativity with regard to self would be relatively strong (at least for a

normative sample of participants); this response bias would be considerably weaker

with regard to others, particularly if they are defined as different/disliked. Again, this is

the pattern observed in the current research (i.e., participants showed a significant bias

for responding ‘no’ more quickly than ‘yes’ on the self-focused IRAP, but not on the

other-focused IRAP). In turning to the Negative-Positive trial-type, the reaction is

specified as positive and thus the bias for responding ‘no’ with regard to self would be

much reduced, and indeed this is the pattern observed on this trial-type. In fact, this is

the only trial-type that fails to produce a significant bias for both self and other.

Of course, the foregoing interpretation of the current findings remain post-hoc,

and somewhat speculative, but at least they are consistent with a formal model of IRAP

performances, and thus there is some basis for future research to formulate specific

hypotheses that may be tested empirically. For example, if the two IRAPs targeted liked

versus disliked others, perhaps the difference observed on the Positive-Negative trial-

type would be much reduced or absent for the liked other IRAP (see Kavanagh, et al., in

press, for indirect evidence consistent with this prediction).

39

Figure 5. Application of the DAARRE model to the current research design using two separate deictic NL-IRAPs. Implementing two types of stimulus sets: the labels ‘Events’ (top) and the target stimuli ‘Reactions’ (bottom). Displayed at the bottom there are two response options (‘Yes’ and ‘No’). The negative and positive labels indicate the valence of each function (Cfunc): Crel (relational properties) and Cfunc (functional properties). The dotted arrow indicates the pull produced by possible contamination of the response option.

In turning to the correlations obtained between the IRAP performances and the

self-report measures, two key clusters were identified. The clearest of these was

recorded between the Negative-Negative trial-type and the CAPE, but only for the

other-focused IRAP. In accordance with our expectations there appeared to be several,

though few, significant relations between performance on the deictic IRAPs and the

occurrence of psychosis symptomatology in our non-clinical study sample. The

correlations were negative, and thus a reduced bias towards confirming that others react

negatively to negative events was associated with increased levels of self-reported

psychotic-like symptoms. In simple terms, it appears that as level of distress (in terms of

psychotic-like experiences) increases it becomes more difficult to judge the negative

reactions of others to negative events. This could be seen as largely consistent with the

literature on psychosis and its links with perspective-taking. That is, symptoms of

psychosis have been associated with impairments in the ability/ flexibility of

perspective-taking or deictic framing, such as social anhedonia (e.g. Vilardaga et al.,

2012; Villatte, et al., 2008,2010a). On balance, it remains unclear why performances on

the other trial-types (e.g. Positive-Negative) did not correlate with the CAPE. One

possibility is that the relative dominance of the control by Crel versus Cfunc properties

(according to the DAARRE model) differed across individuals who were high versus

low in distress. That is, perhaps those high in distress were dominated more by Cfunc

properties and as such found it more difficult to confirm that others were negative. Or to

40

put it more simply, “I am damaged, but others are not.” To further explore these

correlations, it may be useful to consider employing a clinical sample to determine if

they produce different trial-type patterns on IRAPs similar to those employed in the

current study.

Another, less apparent cluster appeared for the Positive-Negative trial-type and

the ECR-RS questionnaire assessing attachment related behaviours, but only for the

self-focused IRAP. The correlations were again negative, and thus a reduced bias

towards denying that others react negatively to positive events was associated with

increased levels of self-reported attachment related avoidant/anxious behaviours. In

simple terms, it appears that as psychological suffering in attachment-related self-

reports increases, individuals have more difficulty denying that the self reacts in a

negative manner to positive things. If the DAARRE model offered previously is correct,

then this interpretation could be simplified as problems in attachment are associated

with self-negative bias (e.g., Dewitte, De Houwer, Buysse, 2008; Passanisi, Gervasi,

Madonia, Guzzo, & Greco, 2015; Wearden, Peters, Berry, Barrowclough, & Liversidge,

2008). Once again, it may be useful for future research to pursue this tentative

interpretation with a clinical sample.

Clinical and Theoretical Implications

In reflecting upon the findings reported in the current thesis and those reported

by Kavanagh, et al. (in press), it is interesting to note that in Experiments 1-4 there was

no evidence of a clear difference between the self and other IRAP performances for the

positive-negative trial-type. The critical point here is that in those first four

Experiments, the other was either unspecified or specified as liked. It appears therefore

that only when the specified other is defined as dissimilar or disliked do you obtain

clear evidence of different response patterns across the self-focused and other-focused

IRAPs. Although this finding will need to be replicated in future studies it does indicate

that the IRAPs employed here, and in similar studies, may be sensitive to the diverging

boundaries between self and other, and may thus prove useful in clinical research

targeting perspective-taking in general.

At a more abstract or theoretical level, the current research may have important

implications for the on-going development of the DAARRE model. For example, in

41

attempting to explain some of the current findings the model was used in a post-hoc

fashion to interpret the idiosyncratic (trial-type) IRAP effects, and the correlations

between those effects and responses on the employed self-report measures. In future

studies it may prove useful to use the model to predict specific response biases on

particular IRAPs (e.g., people low on self-esteem may show a bias towards responding

‘yes’ on a self-focused Positive-Negative trial type and a bias towards responding ‘no’

on a self-focused Negative-Positive trial type). Ideally, the DAARRE model could be

used to predict or even guide development of effective interventions in treatment

studies, focused on ameliorating various forms of psychological suffering. For example,

would an intervention designed to increase self-worth or self-esteem reduce a bias

towards responding ‘no’ on a self-focused Negative-Positive trial-type? At the present

time there are very few studies that have attempted to use the IRAP to train specific

biases in this manner, but preliminary work in this area shows some promise (see

Murphy, Lyons, Kelly, Barnes-Holmes, & Barnes-Holmes, 2019, who used the IRAP to

train relational responding with regards to coordination and distinction relations in

children with autism spectrum disorder; see Murphy & Barnes-Holmes, 2017 for a

review).

Limitations and Future Directions

Once again, it is important to emphasize the exploratory nature of the current

research, and thus more direct tests of the NL-IRAP in the domain of perspective taking

are needed. Indeed, a number of issues would seem to require direct attention in moving

forward. First, a clear distinction between a different (liked) versus disliked other may

be one that was not particularly salient in the current study (i.e., there was little

evidence that this variable impacted IRAP performances differentially across the two

data sets that were analysed in the current thesis). Perhaps future research could attempt

to increase the salience of this variable by asking participants to engage actively with

thinking about the distinction between individuals who are deemed to be different but

liked versus those who are different but disliked. These more fine-grained ways of

processing this distinction may prove fruitful in explaining perspective-taking as a

dimensional concept, thus contrasting with some of the currently available categorical

views, which view perspective taking abilities as either present or absent.

42

Second, the procedures employed in the current study provide limited insight

into the potential strategies employed during exposure to the IRAP. As illustrated by the

post-hoc theorizing presented above in terms of the DAARRE model, the performance

on an IRAP may be interpreted in multiple ways. For example, for some participants it

may be that Crel control dominants but for other Cfunc properties dominate, and to

complicate matters even more the relative dominance of these properties may differ

across trial-types. It will be important in future studies, therefore, that specific

predictions are made concerning what particular properties are likely to dominate for

particular types of participants and perhaps for particular trial-types.

Third, an observation made by the researcher conducting the current study was

that some participants remarked that some of the statements presented in the IRAP were

somewhat extreme (e.g., “If my enemy dies I am relieved.”), and thus perhaps hindered

relatively rapid responding compared to less extreme statements (e.g., “Failing an exam

is disappointing.”). Future studies using statements similar to those employed in the

current study should perhaps ensure that all statements are not deemed particularly

extreme and should be related to everyday experience.

Finally, the current study sample was a random convenience sample mainly

comprising college students of Ghent University. Although some researchers have

pointed out that this age group is most vulnerable to the emergence of psychosis

symptoms (see e.g., McGrath et al., 2016; Vilardaga et al. 2012), the current research

employed a non-clinical sample and therefore does not allow any decisive conclusions

to be made in terms of the clinical relevance of the findings. As previously mentioned,

therefore, future studies could consider employing deictic IRAPs in a sample of

individuals with a diagnosis of a psychosis spectrum disorder.

43

Conclusion

In closing, the current study provides new insights into IRAP research on

perspective-taking in adults, although definitive conclusions arising from the findings

remain limited. In contrast to other studies, the current research involved implementing

two separate NL-IRAPs (self and other), which presented statements targeting

“genuine” perspective-taking, rather than simple deictic relating (e.g., Barbero-Rubio et

al., 2016 and Kavanagh, et al., 2018). The current research was also unique in

manipulating interpersonal boundaries in terms of examining perspective-taking with

regard to whether the “other” was simply different or disliked. Although clear

differences for this variable did not emerge, it may be important for future studies to

attempt to increase the salience of this variable to determine if it could in fact impact

upon IRAP performances. Overall, therefore, the current study should be seen as largely

exploratory in terms of laying the groundwork upon which future studies using the NL-

IRAP may build, as a potential measure of perspective-taking in both non-clinical and

clinical populations.

44

References

Achim, A. M., Ouellet, R., Roy, M. A., & Jackson, P. L. (2011). Assessment of

empathy in first-episode psychosis and meta-analytic comparison with previous

studies in schizophrenia. Psychiatry Research, 190(1), 3–8. doi:

10.1016/j.psychres.2010.10.030

Allardyce, J., Gaebel, W., Zielasek, J., & Van Os, J. (2007). Deconstructing Psychosis

Conference February 2006: The Validity of Schizophrenia and Alternative

Approaches to the Classification of Psychosis. Schizophrenia Bulletin, 33(4), 863-

867. doi: 10.1093/schbul/sbm051

Alonso-Álvarez, B. & Pérez-González, L. A. (2017). Contextual control over

equivalence and nonequivalence explains apparent arbitrary applicable relational

responding in accordance with sameness and opposition. Learning and Behavior,

45(3), 228–242. doi: 10.3758/s13420-017-0258-1

Andreasen, N. C. (1989). The Scale for the Assessment of Negative Symptoms (SANS):

conceptual and theoretical foundations, The British Journal of Psychiatry,

155(Suppl. 7), 49-58. doi: 10.1192/S0007125000291496

Aron, A., Aron, E. N., & Smollan, D. (1992). Inclusion of Other in the Self Scale and

the Structure of Interpersonal Closeness. Journal of Personality and Social

Psychology, 63(4), 596-612. doi: 10.1037/0022-3514.63.4.596

Atkins, P. W. B. & Styles, R. G. (2016). Measuring self and rules in what people say:

exploring whether self-discrimination predicts long-term wellbeing. Journal of

Contextual Behavioral Science, 5(2), 71-79. doi: 10.1016/j.jcbs.2016.05.001

Barbero-Rubio, A., López- López, J. C., Luciano, C., & Eisenbeck, N. (2016).

Perspective-taking measured by Implicit Relational Assessment Procedure

(IRAP). Psychological Record, 66(2), 243-252. doi: 10.1007/s40732-016-0166-

3

Barnes-Holmes, D., Barnes-Holmes, Y., Hussey, I., & Luciano, C. (2015). Relational

Frame Theory: Finding its historical and intellectual roots and reflecting upon its

future development. In R. D. Zettle, S. C. Hayes, D. Barnes-Holmes, & A.

Biglan (Eds.), The Wiley Handbook of Contextual Behavioral Science (pp. 115-

128). John Wiley & Sons, Ltd.

Barnes-Holmes, D., Barnes-Holmes, Y., Luciano, C., & McEnteggart, C. (2017). From

45

the IRAP and REC model to a multi-dimensional multi-level framework for

analyzing the dynamics of arbitrarily applicable relational responding. Journal

of Contextual Behavioral Science, 6(4), 434–445. doi:

10.1016/j.jcbs.2017.08.001

Barnes-Holmes, Y., Barnes-Holmes, D., & McHugh, L. (2004). Teaching derived

relational responding to young children. Journal of Early and Intensive Behavior

Intervention, 1(1), 3-12. doi: 10.1037/h0100275

Barnes-Holmes, Y., Barnes-Holmes, D., Roche, B., & Smeets, P. M. (2001). The

Development of self and perspective-taking: A Relational Frame analysis.

Behavioral Development Bulletin, 10(1), 42-45. doi: 10.1037/h0100482

Barnes-Holmes, D., Barnes-Holmes, Y., Stewart, I., & Boles, S. (2010). A Sketch of the

Implicit Relational Assessment Procedure (IRAP) and the Relational Elaboration

and Coherence (REC) Model. The Psychological Record, 60(3), 527-542. doi:

10.1007/BF03395726

Barnes-Holmes, Y., Foody, M., Barnes-Holmes, D., & McHugh, L. (2013). Advances

in research on deictic relations and perspective-taking. In S. Dymond & B.

Roche (Eds.), Advances in relational frame theory: Research and

application (pp. 127-148). Oakland, CA, US: Context Press/New Harbinger

Publications.

Baron-Cohen, S., Tager-Fluchberg, H., & Cohen, D. (2000). Understanding other

minds: Perspectives from developmental cognitive neuroscience (2nd Ed.).

Oxford: Oxford University Press.

Bernaerts, I., De Groot, F., & Kleen, M. (2012). De AAQ-II, een maat voor

experiëntiële vermijding: normering bij jongeren. Gedragstherapie, 4, 389-399.

Retrieved from

https://www.tijdschriftgedragstherapie.nl/scripts/shared/artikel_pdf.php?id=TG-

2012-4-3

Bond, F., Hayes, S., Baer, R., Carpenter, K., Guenole, N., Orcutt, H., … Zettle, R.

(2011). Preliminary psychometric properties of the Acceptance and Action

Questionnaire-II: A revised measure of psychological inflexibility and

experiential avoidance. Behavior Therapy, 42, 676-88. doi:

10.1016/j.beth.2011.03.007

46

Bond, F. W., Lloyd, J., Barnes-Holmes, Y., Torneke, N., Luciano, L., Barnes-Holmes,

D., & Guenole, N. (2017). A new measure of psychological flexibility based on

RFT. Symposium at the Assocation for Contextual Behavioural Science World

Conference 15, 22-25 June 2017, Seville, Spain.

Bora, E., Yucel, M., & Pantelis, C. (2009). Theory of mind impairment in

schizophrenia: Meta-analysis. Schizophrenia Research, 109(1–3), 1–9. doi:

10.1016/j.schres.2008.12.020

Brenner, K., Schmitz, N., Pawliuk, N., Fathalli, F., Joober, R., Ciampi, A., & King, S.

(2007). Validation of the English and French versions of the Community

Assessment of Psychic Experiences (CAPE) with a Montreal community

sample. Schizophrenia Research, 95(1–3), 86–95. doi:

10.1016/j.schres.2007.06.017

Chen, Y., Zhong, Y., Zhou, H., Zhang, S., Tan, Q., & Fan, W. (2014). Evidence for

implicit self-positivity bias: an event-related brain potential study. Experimental

brain research, 232(3), 985-994. doi: 10.1007/s00221-013-3810-z

Cohen, L., Leibu, O., Tanis, T., Ardalan, F., & Galynker, I. (2016). Disturbed self

concept mediates the relationship between childhood maltreatment and adult

personality pathology. Comprehensive Psychiatry, 68, 186-192. doi:

10.1016/j.comppsych.2016.04.020

DeBernardis, G. M., Hayes, L. J., & Fryling, M. J. (2014). Perspective taking as a

continuum. The Psychological Record, 64(1), 123-131. doi: 10.1007/s40732-

014-0008-0

De Houwer, J. (2011). Why the cognitive approach in psychology would profit from a

functional approach and vice versa. Perspectives on Psychological Science, 6(2),

202–209. doi: 10.1177/1745691611400238

De Houwer, J., Hughes, S., & Barnes-Holmes, D. (2017). Bridging the Divide Between

Functional and Cognitive Psychology. Journal of Applied Research in Memory

and Cognition, 6(1), 47–50. doi: 10.1016/j.jarmac.2017.01.001

Demjaha, A., Morgan, K., Morgan, C., Landau, S., Dean, K., Reichenberg, A., …

Dazzan, P. (2009). Combining dimensional and categorical representation of

psychosis: The way forward for DSM-V and ICD-11? Psychological Medicine,

39(12), 1943–1955. doi: 10.1017/S0033291709990651

47

Dewitte, M., De Houwer, J., & Buysse, A. (2008). On the Role of the Implicit Self-

Concept in Adult Attachment. European Journal of Psychological Assessment,

24(4), 282-289. doi: 10.1027/1015-5759.24.4.282

Fraley, R. C., Vicary, A. M., Brumbaugh, C. C., & Roisman, G. I. (2011b). Patterns of

Stability in Adult Attachment: An Empirical Test of Two Models of Continuity

and Change. Journal of Personality and Social Psychology, 101(5), 974-992.

doi: 10.1037/a0024150

Fraley, R. C., Heffernan, M. E., Vicary, A. M., & Brumbaugh, C. C. (2011a). The

Experiences in Close Relationships – Relationship Structures Questionnaire: A

method for assessing attachment orientations across relationships. Psychological

Assessment, 23(3), 615-625. doi: 10.1037/a0022898

Finn, M., Barnes-Holmes, D., & McEnteggart, C. (2018). Exploring the single-trial-

type-dominance-effect in the IRAP: Developing a Differential Arbitrarily

Applicable Relational Responding Effects (DAARRE) Model. The

Psychological Record, 68(1), 11-25. doi: 10.1007/s40732-017-0262-z

Foody, M., Barnes-Holmes, Y., & Barnes-Holmes, D. (2012). The role of Self in

Acceptance and Commitment Therapy (ACT). In McHugh, L. & Stewart, I.

(Eds.). The self and perspective taking: Contributions and Applications from

Modern Behavioral Science (pp. 125-142). New York: New Harbinger

Publications.

Foody, M., Barnes-Holmes, Y., Barnes-Holmes, D., Rai, L., & Luciano, C. (2015). An

empirical investigation of the role of self, hierarchy, and distinction in a

common act exercise. Psychological Record, 65(2), 231–243. doi:

10.1007/s40732-014-0103-2

Gächter, S., Starmer, C., & Tufano, F. (2015). Measuring the Closeness of

Relationships: A Comprehensive Evaluation of the ‘Inclusion of the Other in the

Self’ scale. PLOS ONE, 10(6): e0129478. doi: 10.1371/journal.pone.0129478

Gilroy, S. P., Lorah, E. R., Dodge, J., & Fiorello, C. (2015). Establishing deictic

repertoires in autism. Research in Autism Spectrum Disorders, 19, 82–92. doi:

10.1016/j.rasd.2015.04.004

Gore, N. G., Barnes-Holmes, Y., & Murphy, G. (2010). The relationship between

intellectual functioning and relational perspective-taking. International Journal

48

of Psychology and Psychological Therapy, 10(1), 1-17.

Greenwald, A. G. & Farnham, S. D. (2000). Using the Implicit Association Test to

measure self-esteem and self-concept. Journal of Personality and Social Psychology,

79(6), 1022–1038. doi: 10.1037//0022-3514.79.6.1022

Griffee, K. & Dougher, M. J. (2002). Contextual control of stimulus generalization and

stimulus equivalence in hierarchical categorization. Journal of the Experimental

Analysis of Behavior, 78(3), 433–447. doi: 10.1901/jeab.2002.78-433

Heagle, A. & Rehfeldt, R. A. (2006). Teaching perspective-taking skills to typically

developing children through derived relational responding. Journal of Early and

Intensive Behavior Intervention, 3(1), 1-34. doi: 10.1037/h0100321

Hendriks, A. L., Barnes-Holmes, Y., McEnteggart, C., De Mey, H. R. A., Janssen, G. T.

L., & Egger, J. I. M. (2016a). Understanding and remediating social-cognitive

dysfunctions in patients with serious mental illness using Relational Frame

Theory. Frontiers in Psychology, 7, 1-7. doi: 10.3389/fpsyg.2016.00143

Hendriks, A. L., Barnes-Holmes, Y., McEnteggart, C., De Mey, H. R. A., Witteman, C.

L. M., Janssen, G. T. L., & Egger, J. I. M. (2016b). The relationship between

theory of mind and relational frame theory: Convergence of perspective-taking

measures. Clinical Neuropsychiatry, 13(2), 17-23.

Hughes, S. & Barnes-Holmes, D. (2016). Relational Frame Theory: The Basic Account.

In R. D., Zettle, S. C., Hayes, S. C., D., Barnes-Holmes, & A., Biglan, (Red.).

The Wiley Handbook of Contextual Behavioral Science (1th Ed., pp.129-169).

New York: John Wiley & Sons.

Hussey, I., Barnes-Holmes, D., & Barnes-Holmes, Y. (2015). From Relational Frame

Theory to implicit attitudes and back again: Clarifying the link between RFT and

IRAP research. Current Opinion in Psychology, 2, 11–15. doi:

10.1016/j.copsyc.2014.12.009

Hussey, I., McEnteggart, C., Barnes-Holmes, Y., Kavanagh, D., Barnes-Holmes, D.,

Parling, T., & Lundgren, T. Flexible perspective-taking: New concepts and a

new behavioural measure. ACT CBS Conference. Dublin (December, 2014).

Hussey, I., Thompson, M., McEnteggart, C., Barnes-Holmes, D., & Barnes-Holmes, Y.

(2015). Interpreting and inverting with less cursing: A guide to interpreting

IRAP data. Journal of Contextual Behavioral Science, 4(3), 157-162. doi:

49

10.1016/j.jcbs.2015.05.001

Janssen, G., De Mey, H., Hendriks, A., Koppers, A., Kaarsemaker, M, Witteman, C., &

Egger, J. (2014). Assessing deictic relational responding in individuals with

social anxiety disorder: Evidence of perspective-taking difficulties. The

Psychological Record, 64(1), 21-29. doi: 10.1007/s40732-014-012-3

Kanter, J. W., Parker, C. R., & Kohlenberg, R. J. (2001). Finding the self: A behavioral

measure and its clinical implications. Psychotherapy: Theory, Research,

Practice, Training, 38(2), 198-211. doi: 10.1037//0033-3204.38.2.198

Karpinski, A. (2004). Measuring self-esteem using the Implicit Association

Test: The role of the other. Personality and Social Psychology Bulletin, 30(1),

22–34. doi: 10.1177/014616720325883

Kavanagh, D., Barnes-Holmes, Y., Barnes-Holmes, D., McEnteggart, C., & Finn, M.

(2018). Exploring differential trial-type effects and the impact of a read-aloud

procedure on deictic relational responding on the IRAP. The Psychological

Record. Advance online publication. doi: 10.1007/s40732-018-0276-1

Kavanagh, D., Hussey, I., McEnteggart, C., Barnes-Holmes, Y., & Barnes-Holmes, D.

(2016). Using the IRAP to explore natural language statements. Journal of

Contextual Behavioral Science, 5(4), 247–251. doi: 10.1016/j.jcbs.2016.10.001

Kavanagh, D., Roelandt, A., Van Raemdonck, L., Barnes-Holmes, Y., Barnes-Holmes,

D., & McEnteggart, C. (in press). The On-going Search for Perspective-taking

IRAPs: Exploring the Potential of the Natural Language IRAP. The

Psychological Record. Retrieved from https://go-rft.com/wp-

content/uploads/2019/02/Accepted-Deictic-NL-manuscript.pdf

Konings, M., Bak, M., Hanssen, M., Van Os, J., & Krabbendam, L. (2006). Validity

and reliability of the CAPE: A self-report instrument for the measurement of

psychotic experiences in the general population. Acta Psychiatrica

Scandinavica, 114(1), 55-61. doi: 0.1111/j.1600-0447.2005.00741.x

Langdon, R. & Coltheart, M. (2001). Visual perspective-taking and schizotypy:

Evidence for a simulation-based account of mentalizing in normal adults.

Cognition, 82(1), 1–26. doi: 10.1016/S0010-0277(01)00139-1

Langdon, R. & Ward, P. (2008). Taking the perspective of the other contributes to

awareness of illness in schizophrenia. Schizophrenia Bulletin, 35(5), 1003-1011.

50

doi: 10.1093/schbul/sbn039

Lipkens, R. & Hayes, S. C. (2009). Producing and recognizing analogical relations.

Journal of the Experimental Analysis of Behavior, 91(1), 105–126. doi:

10.1901/jeab.2009.91-105

McEnteggart, C., Barnes-Holmes, Y., Dillon, J., Egger, J., & Oliver, J. E. (2017).

Hearing voices, dissociation, and the self: A functional-analytic perspective.

Journal of Trauma & Dissociation, 18(4), 575-594. doi:

10.1080/15299732.2016.1241851

McEnteggart, C., Barnes-Holmes, Y., Hussey, I., & Barnes-Holmes, D. (2015). The ties

between a basic science of language and cognition and clinical applications.

Current Opinion in Psychology, 2, 56-59. doi: 10.1016/j.copsyc.2014.11.017

McGrath, J. J., Saha, S., Al-Hamzwadi, A. O., Alonso, J., Andrade, L., Borges, G., …

Kesseler, R. R. (2016). Age of Onset and Lifetime Projected Risk of Psychotic

Experiences: Cross-National Data From the World Mental Health Survey.

Schizophrenia Bulletin, 42(4), 933-941. doi: 10.1093/schbul/sbw011

McHugh, L. (2015). A Contextual Behavioural Science approach to the self and

perspective taking. Current Opinion in Psychology, 2, 6-10. doi:

10.1016/j.copsyc.2014.12.030

McHugh, L., Barnes-Holmes, Y., & Barnes-Holmes, D. (2004). Perspective-taking as

relational responding: A developmental profile. The Psychological Record,

54(1), 115-144. doi: 10.1007/BF03395465

McHugh, L., Barnes-Holmes, Y., Barnes-Holmes, D., & Stewart, I. (2006).

Understanding False Belief as Generalized Operant Behavior. The Psychological

Record, 56(3), 341-364. doi: 10.1007/BF03395554

McHugh, L., Barnes-Holmes, Y., Barnes-Holmes, D., Stewart, I., & Dymond, S.

(2007). Deictic Relational Complexity and the Development of Deception. The

Psychological Record, 57(4), 517-531. doi: 10.1007/BF03395592

McHugh, L. & Stewart, I. (2012). The self and perspective taking: Contributions and

applications from Modern Behavioral Science. Dublin: New Harbinger

Publications.

51

Ming, S. & Stewart, I. (2017). When things are not the same: A review of research into

relations of difference. Journal of Applied Behavior Analysis, 50(2), 429–455.

doi: 10.1002/jaba.367

Montoya-Rodríguez, M. M., Molina, F. J., & McHugh, L. (2016). A review of

Relational Frame Theory research into deictic relational responding. The

Psychological Record, 67(4), 569-579. doi: 10.1007/s40732-016-0216-x

Murphy, C. & Barnes-Holmes, D. (2017). Teaching Important Relational Skills for

Children with Autism Spectrum Disorder and Intellectual Disability Using

Freely Available (GO-IRAP) Software. Austin Journal of Autism & Related

Disabilities, 3(2), 1041. Retrieved from https://go-rft.com/wp-

content/uploads/2018/01/murphy-and-barnes-holmes.pdf

Murphy, C., Lyons, K., Kelly, M., Barnes-Holmes, Y., & Barnes-Holmes, D. (2019).

Using the Teacher IRAP (T-ITAP) interactive computerized programme to teach

complex flexible relational responding with children with diagnosed autism

spectrum disorder. Behavior Analysis in Practice, 12(1), 52-65. doi:

10.1007/s40617-018-00302-9

Nicholson, E. & Barnes-Holmes, D. (2012). The Implicit Relational Assessment

Procedure (IRAP) as a measure of spider fear. The psychological record, 62(2),

263-277. doi: 10.1007/BF03395801

Nelson, S. (2008). Feeling thermometer. In P. J. Lavrakas (Ed.). Encyclopedia of survey

research methods (pp. 275-277). Thousand Oaks, CA: SAGE Publications, Inc.

doi: 10.4135/9781412963947.n183

Nelson, B., Thompson, A., & Yung, A. R. (2012). Basic self-disturbance predicts

psychosis onset in the ultra high risk for psychosis “prodromal” population.

Schizophrenia Bulletin, 38(6), 1277–1287. doi: 10.1093/schbul/sbs007

Niemantsverdriet, M. B. A., Slotema, C. W., Blom, J. D., Franken, I. H., Hoek, H. W.,

Sommer, I. E. C., & Van Der Gaag, M. (2017). Hallucinations in borderline

personality disorder: Prevalence, characteristics and associations with comorbid

symptoms and disorders. Scientific Reports, 7, 1–8. doi: 10.1038/s41598-017-

13108-6

Ninness, C., Dixon, M., Barnes-Holmes, D., Rehfeldt, R. A., Rumph, R., McCuller, G.,

… McGinty, J. (2009). Constructing and deriving reciprocal trigonometric

52

relations: A functional analytic approach. Journal of Applied Behavior Analysis,

42(2), 191–208. doi: 10.1901/jaba.2009.42-191

Nolen-Hoeksema, S. (2014). Schizophrenia Spectrum and Other Psychotic Disorders

Along the Continuum (6th Ed.). Abnormal Psychology (pp. 216-249). New York:

McGraw-Hill Education.

O’Connor, M., Farrell, L., Munnelly, A., & McHugh, L. (2017). Citation analysis of

relational frame theory: 2009–2016. Journal of Contextual Behavioral Science,

6(2), 152–158. doi: 10.1016/j.jcbs.2017.04.009

Parling, T., Cernvall, M., Stewart, I., Barnes-Holmes, D., & Ghaderi, A. (2012). Using

the implicit relational assessment procedure to compare implicit pro-thin/anti-fat

attitudes of patients with anorexia nervosa and non-clinical controls. Eating

Disorders, 20(2), 127-143. doi: 10.1080/10640266.2012.654056

Passanisi, A., Gervasi. A. M., Madonia, C., Guzzo, G., & Greco, D. (2015).

Attachment, Self-Esteem and Shame in Emerging Adulthood. Procedia-Social

and Behavioral Sciences, 191, 342-346. doi: 10.1016/j.sbspro.2015.04.552

Peirce, J. W. (2007). PsychoPy - Psychophysics software in Python. Journal of

Neuroscience Methods, 162(1-2),8-13 doi: 10.1016/j.jneumeth.2006.11.017

Peters, E. R., Joseph, S. A., & Garety, P. A. (1999). Measurement of delusional ideation

in the normal population: introducing the PDI (Peters et al. Delusions

Inventory). Schizophrenia Bulletin, 25(3), 553-576. doi:

10.1093/oxfordjournals.schbul.a033401

Pijnenborg, G. H. M., Spikman, J. M., Jeronimus, B. F., & Aleman, A. (2013). Insight

in schizophrenia: Associations with empathy. European Archives of Psychiatry

and Clinical Neuroscience, 263(4), 299–307. doi: 10.1007/s00406-012-0373-0

Rehfledt, R. A., Dillen, J. E., Ziomek, M. M., & Kowalchuk, R. K. (2007). Assessing

relational learning deficits in perspective-taking in children with high-

functioning autism spectrum disorder. The Psychological Record, 57(1), 23-47.

doi: 10.1007/BF03395563

Remue, J., De Houwer, J., Barnes-Holmes, D., Vanderhasselt, M.-A., & De Raedt, R.

(2013). Self-esteem revisited: Performance on the implicit relational assessment

procedure as a measure of self- versus ideal self-related cognitions in dysphoria.

Cognition & Emotion, 27(8), 1441-1449. doi: 10.1080/02699931.2013.786681

53

Renard, S. B., Huntjens, R. J. C., Lysaker, P. H., Moskowitz, A., Aleman, A., &

Pijnenborg, G. H. M. (2017). Unique and overlapping symptoms in

schizophrenia spectrum and dissociative disorders in relation to models of

psychopathology: A systematic review. Schizophrenia Bulletin, 43(1), 108–121.

doi: 10.1093/schbul/sbw063

Rimes, K. & Watkins, E. (2005). The effects of self-focused rumination on global

negative self-judgements in depression, Behavioral Research Therapy, 43, 1673-

1681. doi: 10.1016/j.brat.2004.12.002

Roddy, S., Stewart, I., & Barnes-Holmes, D. (2010). Anti-fat, pro-slim, or both?: Using

two reaction-time based measures to assess implicit attitudes to the slim and

overweight. Journal of Health Psychology, 15(3), 416-425. doi:

10.1177/1359105309350232

Ruiz, F. J. & Luciano, C. (2011). Cross-domain analogies as relating derived relations

among two separate relational networks. Journal of the Experimental Analysis of

Behavior, 95(3), 369–385. doi: 10.1901/jeab.2011.95-369

Scanlon, G., McEnteggart, C., Barnes-Holmes, Y., & Barnes-Holmes, D. (2014). Using

the implicit relational assessment procedure (IRAP) to assess implicit gender

bias and self-esteem in typically-developing children and children with ADHD

and with dyslexia. Behavioral Development Bulletin, 19(2), 48-59. doi:

10.1037/h0100577

Selten, J. P., Gernaat, H. B., Nolen, W. A., Wiersma, D., & Van den Bosch, R. J.

(1998). Experience of negative symptoms: Comparison of schizophrenic patients

to patients with a depressive disorder and to normal subject. American Journal

of Psychiatry, 155(3), 350-354. doi: 0.1176/ajp.155.3.350

Shevlin, M., McElroy, E., Bentall, R. P., Reininghaus, U., & Murphy, J. (2017). The

psychosis continuum: Testing a bifactor model of psychosis in a general

population sample. Schizophrenia Bulletin, 43(1), 133–141. doi:

10.1093/schbul/sbw067

Sidman, M. (1971). Reading and auditory-visual equivalences. Journal of Speech and

Hearing Research, 14(1), 5-13. doi: 10.1044/jshr.1401.05

Skinner, B.F. (1974). About behaviorism. New-York: Knopf.

Slattery, B., Stewart, I., & O’Hora, D. (2011). Testing for transitive class containment

54

as a feature of hierarchical classification. Journal of the Experimental Analysis

of Behavior, 96(2), 243–260. doi: 10.1901/jeab.2011.96-243

Stefanis, N. C., Hanssen, M., Smirnis, N. K., Avramopoulos, D. A., Evdokimidis, I. K.,

Stefanis, C. N., …, Van Os, J. (2002). Evidence that three dimensions of

psychosis have a distribution in the general population. Psychological Medicine

32(2), 347–358. doi: 10.1017/S0033291701005141

Stewart, I. (2016). The fruits of a functional approach for psychological science.

International Journal of Psychology, 51(1), 15-27. doi: 10.1002/ijop.12184

Suls, K., Lemos, K., & Stewart, H. L. (2002). Self-Esteem, Construal, and Comparisons

With the Self, Friends, and Peers. Journal of Personality and Social Psychology,

82(2), 252-261. doi: 10.1037/0022-3514.82.2.252

Unterrassner, L., Wyss, T. A., Wotruba, D., Haker, H., & Rössler, W. (2017). The

intricate relationship between psychotic-like experiences and associated

subclinical symptoms in healthy individuals. Frontiers in Psychology, 8, 1-15.

doi: 10.3389/fpsyg.2017.01537

Vahey, N. A., Barnes-Holmes, D., Barnes-Holmes, Y., & Stewart, I. (2009). A first test

of the Implicit Relational Assessment Procedure as a measure of self-esteem:

Irish prisoner groups and university students. The Psychological Record, 59(3),

371-387. doi: 10.1007/BF03395670

Vahey, N. A., Nicholson, E., & Barnes-Holmes, D. (2015). A meta-analysis of criterion

effects for the Implicit Relational Assessment Procedure (IRAP) in the clinical

domain. Journal of Behavior Therapy and Experimental Psychiatry, 48, 59-65.

doi: 10.1016/j.jbtep.2015.01.004

Van Os, J., Linscott, R. J., Myin-Gemeys, L., Delespaul, P., & Krabbendam, L. (2009).

A systematic review and meta-analysis of the psychosis continuum: evidence for

a psychosis proneness-persistence-impairment model of psychotic disorder.

Psychological Medicine, 39(2), 179-195. doi: 10.1017/S0033291708003814

Van Os, J. & Reininghaus, U. (2016). Psychosis as a transdiagnostic and extended

phenotype in the general population. World Psychiatry, 15(2), 118–124. doi:

10.1002/wps.20310

Van Os, J., Verdoux, H., & Hanssen, M. (1999). Dutch version of the CAPE

(Community Assessment of Psychic Experiences). Retrieved from

55

http://cape42.homestead.com/index.html

Verdoux, H., Sorbara, F., Gindre, C., Swendsen, J. D., & Van Os, J. (2003). Cannabis

use and dimensions of psychosis in a nonclinical population of female subjects.

Schizophrenia Research 59(1), 77–84. doi: 10.1016/S0920-9964(01)00401-7

Vilardaga, R., Estévez, A., Levin, M. E., & Hayes, S. C. (2012). Deictic relational

responding, empathy and experiential avoidance as predictors of social

anhedonia: Further contributions from relational frame theory. The

Psychological Record, 62(3), 409-432. doi: 10.1007/BF03395811

Villatte, M., Monestès, J.-L., McHugh, L., Freixa, I Baqué, E., & Loas, G. (2008).

Assessing deictic relational responding in social anhedonia: A functional

approach to the development of Theory of Mind impairments. International

Journal of Behavioral Consultation and Therapy, 4(4), 360-373. doi:

10.1037/h0100867

Villatte, M., Monestès, J.-L., McHugh, L., Freixa, I Baqué, E., & Loas, G. (2010a).

Assessing perspective taking in schizophrenia using Relational Frame Theory.

The Psychological Record, 60(3), 413-436. doi: 10.1007/BF03395719

Villatte, M., Monestès, J.-L., McHugh, L., Freixa, I Baqué, E., & Loas, G. (2010b).

Adopting the perspective of another in belief attribution: Contribution of

Relational Frame Theory to the understanding of impairments in schizophrenia.

Journal of Behavior Therapy and Experimental Psychiatry, 41(2), 125-134. doi:

10.1016/j.jbtep.2009.11.004

Vitale, A., Barnes-Holmes, Y., Barnes-Holmes, D., & Campbell, C. (2008). Facilitating

responding in accordance with the relational frame of comparison: Systematic

empirical analyses. The Psychological Record, 58(3), 365-390. doi:

10.1007/BF03395624

Wearden, A., Peters, I., Berry, K., Barrowclough, C., & Liversidge, T. (2008). Adult

attachment, parenting experiences, and core beliefs about self and others.

Personality and Individual Differences, 44(5), 1246-1257.doi:

10.1016/j.paid.2007.11.019

Weil, T. M., Hayes, S. C., & Capurro, P. (2011). Establishing a deictic relational

repertoire in young children. The Psychological Record, 61(3), 371–390. doi:

10.1007/BF03395767

56

Zuckerman, M. & O’Loughlin, R. R. (2006). Self-Enhancement by social comparison:

A Prospective Analysis. Personality and Social Psychology Bulletin, 32(6), 751-

760. doi: 10.1177/0146167205286111

57

Appendix 1

Instructions Self- and Other-NL-IRAP.

Self-NL-IRAP. Other-NL-IRAP.

“In the following task you will be

required to respond to some statements,

these statements will be about you. Even

though, these statements concern you, the

computer will tell you what is correct and

what is not. You will have to either

conform (‘yes’) or disconfirm (‘no’) the

statement by pressing ‘d’ or ‘k’, be aware

that the position of ‘yes’ and ‘no’ may

change. Try to be as accurate and fast as

you can. First you will receive a few

practice trials to make you familiar with

the task and then you will move on to the

test trials.”

Inconsistent blocks (only for the first

inconsistent block): “Now correct and

incorrect are reversed. What was correct

before is now incorrect, and what was

incorrect before is now correct.”

“If you could think about someone you

now well but consider to be completely

different from you. As if your

personalities are from different planets.

Then you can write down their name.”

“In the following task you will be

required to respond to some statements,

these statements will be about ‘X’. Even

though, these statements concern ‘X’, the

computer will tell you what is correct and

what is not. You will have to either

conform (‘yes’) or disconfirm (‘no’) the

statement by pressing ‘d’ or ‘k’, be aware

that the position of ‘yes’ and ‘no’ may

change. Try to be as accurate and fast as

you can. First you will receive a few

practice trials to make you familiar with

the task and then you will move on to the

test trials.”

Inconsistent blocks (only for the first

inconsistent block): “Now correct and

incorrect are reversed. What was correct

before is now incorrect, and what was

incorrect before is now correct.”

58

In case their performance on either IRAP required additional feedback provided by

the researcher (below the test criteria or marginal performance).

Because they made too many mistakes: “If you start making too many mistakes it is

sometimes better to slow down a bit, the speed will come as you practice.”

Because they were too slow: “Try to respond as accurate and fast as you can.”

Note. ‘X’ indicates the specified other as defined by each participant individually.

59

Appendix 2

Explanation of the steps involved in calculating the DIRAP-scores for the current study design, following the description of Hussy, Thompson, McEnteggart, Barnes-Holmes and Barnes-Holmes (2015, p. 9)

Scoring

1. First there is a need to define the measurement unit. In accordance to previous

IRAP-studies, reaction times are usually defined as the time from stimulus-onset

until the first correctly emitted response.

2. Only data from the test-blocks is used.

3. To correct for ‘fast’ responding, all IRAP data for a participant were deleted if

10% of participants' response latencies on the test blocks were less than 300 ms.

4. To remove outliers, response latencies of more than 10,000 ms were deleted

for each trial-type.

5. DIRAP-scores were calculated for each trial-type (four trial-types) in each

separate block pair (six test blocks leading to a total of three block pairings),

using the following formula: D = (MB – MA)/ SDAB

Where;

MA = mean latencies block A,

MB = mean latencies block B,

SDAB = standard deviation of latencies in both blocks A and B

This included:

Calculating 24 mean latencies for each trial-type in each block.

Calculating 12 standard deviations for each trial-type across the different

pairs of test blocks (i.e., six test blocks leading to three pairs of test

blocks).

Calculating a difference score for each trial-type for each pairing of test

blocks (mean latency of a consistent versus an inconsistent block).

Processing

60

6. Exclude all data for each participant that did not meet the set accuracy

(>=75%) and latency criteria (<= 5000 ms) during the test blocks. These mastery

criteria are set across each trial-type within the test blocks. Stating that when a

participant fails to meet these set criteria for one of the trial-types within a test

block all data for this participant is deleted.

7. DIRAP- scores for each trial-type (four trial-types) are then averaged across

block pairs.

e.g., Dpositive-positive = (Dpair1 + Dpair2 + Dpair3)/ 3

Note. The IRAP program used for this study (GO-IRAP v1.0) calculated D-scores automatically performing steps 3 and 4. However, steps 1, 5, 6, and 7 were done manually. 1Blocks A and B are defined by the responding contingencies within them (e.g., in our example test block A = consistent block 1 vs. block B = inconsistent block 1) and do not refer to the order of presentation of the blocks.

61

Appendix 3

Correlations for the CAPE (Weighted scores)

Self Other

Overall Positive-

Positive

Positive-

Negative

Negative-

Positive

Negative-

Negative

Overall Positive-

Positive

Positive-

Negative

Negative-

Positive

Negative-

Negative

Overall frequency -.127 .122 -.209 -.012 -.025 -.159 -.083 .036 -.178 -.251*

Overall distress -.236 -.061 -.135 -.173 .006 -.129 .147 .227 -.093 -.419**

Positive frequency -.075 .227 -.062 .092 .051 -.099 -.011 -.031 -.115 -.163

Depressive frequency -.149 -.078 -.176 -.078 -.087 -.211 -.106 .056 -.171 -.319*

Negative frequency -.111 .081 -.290* -.072 -.056 -.118 -.109 .078 -.169 -.191

Positive distress -.315* -.068 -.113 -.222 -.052 -.041 .200 .195 -.116 -.177

Depressive distress -.220 -.050 .054 -.265* .047 -.157 .090 .146 -.017 -.365*

Negative distress -.072 -.047 -.206 -.123 .041 -.051 .062 .116 -.041 -.386**

* Indicates significant p values (p < .05).

** Indicates significant p values after correcting for multiple testing (Bonferroni correction) (p < .002).

62

Appendix 4

Correlations for the PFI

Self

Other

Overall Positive-

Positive

Positive-

Negative

Negative-

Positive

Negative-

Negative

Overall Positive-

Positive

Positive-

Negative

Negative-

Positive

Negative-

Negative

Overall PFI -.124 .105 .032 -.096 -.142 -.289* -.201 -.190 -.004 -.309*

* Indicates significant p values (p < .05).

** Indicates significant p values after correcting for multiple testing (Bonferroni correction) (p < .002).

63

Appendix 5

Correlations for the ECR-RS

Self Other

Overall Positive-

Positive

Positive-

Negative

Negative-

Positive

Negative-

Negative

Overall Positive-

Positive

Positive-

Negative

Negative-

Positive

Negative-

Negative

Attachment-related anxiety

(Mother)

.036 .197 -.125 -.188 -.029 -.062 -.152 -.112 -.126 .101

Attachment-related

avoidance (Mother)

-.120 .221 -.410** .036 -.160 -.094 -.058 -.125 -.161 -.152

Attachment-related anxiety

(Father)

.010 .015 -.064 -.160 .103 -.122 -.087 -.112 -.104 -.034

Attachment-related

avoidance (Father)

-.046 -.007 -.016 .112 .044 -.120 .083 .025 -.197 -.179

Attachment-related anxiety

(Partner)

.124 .148 -.043 -.131 .311* -.059 -.039 -.090 .075 -.004

Attachment-related

avoidance (Partner)

-.191 -.218 -.325* -.262* -.068 -.059 -.117 -.116 -.038 -.036

Attachment-related anxiety

(Best -friend)

.065 .183 -.330* -.021 .016 -.039 -.101 -.064 .063 -.100

Attachment-related

avoidance (Best friend)

-.078 -.197 -.171 -.015 -.118 -.044 -.047 -.036 .132 -.313*

* Indicates significant p values (p < .05).

** Indicates significant p values after correcting for multiple testing (Bonferroni correction) (p < .002).

64

Appendix 6

Correlations for the EOSS

Self

Other

Overall Positive-

Positive

Positive-

Negative

Negative-

Positive

Negative-

Negative

Overall Positive-

Positive

Positive-

Negative

Negative-

Positive

Negative-

Negative

Overall EOSS -.198 .036 -.359* -.150 -.085 -.201 -.134 -.201 -.144 -.136

Casual

Acquaintances-

absent

-.106 .086 -.212 -.122 .160 -.143 -.170 -.249 .002 .052

Casual

Acquaintances-

present

-.082 .188 -.211 -.003 .012 -.053 .073 -.011 -.085 -.031

Close

relationships-

absent

-.170 -.043 -.226 -.145 -.091 -.111 -.079 -.133 -.047 -.147

Close

relationships –

present

-.156 -.070 -.304* -.100 -.272* -.204 -.130 -.109 -.247 -.211

* Indicates significant p values (p < .05).

** Indicates significant p values after correcting for multiple testing (Bonferroni correction) (p < .002).

65

Appendix 7

Correlations for the IOS

Self

Other

Overall Positive-

Positive

Positive-

Negative

Negative-

Positive

Negative-

Negative

Overall Positive-

Positive

Positive-

Negative

Negative-

Positive

Negative-

Negative

Best friend -.014 .024 .008 .244 -.140 -.060 -.152 -.113 -.239 .063

Other people .094 -.002 .221 .310* -.107 -.103 -.071 -.131 -.068 -.119

* Indicates significant p values (p < .05).

** Indicates significant p values after correcting for multiple testing (Bonferroni correction) (p < .002).

66

Appendix 8

Correlations for the Feelings about Self Thermometer

Self

Other

Overall Positive-

Positive

Positive-

Negative

Negative-

Positive

Negative-

Negative

Overall Positive-

Positive

Positive-

Negative

Negative-

Positive

Negative-

Negative

Overall -.018 .035 -.146 .023 -.229 -.047 -.020 -.102 -.254* -.112

* Indicates significant p values (p < .05).

** Indicates significant p values after correcting for multiple testing (Bonferroni correction) (p < .002).