Are we missing the family context? An exploratory, practice-based analysis of YP-CORE sensitivity to family expressiveness

by Frankie Brown and John McLeod (Supervisor)

When we work with young people in individual therapy, it can throw up a unique set of challenges compared to working with adults, one of which being that the young person often lives in a family context which might be contributing to, or perpetuating, the symptoms. In the UK and Ireland, these symptoms are often captured in therapy through the use of outcome and problem scales, with one of the most widely-used measures being the Clinical Outcome in Routine Evaluation (CORE), designed to assess general psychological problems such as anxiety and depression (Barkham et al., 2001; Evans et al., 2002). Whilst the Clinical Outcome in Routine Evaluation for Young People (YP-CORE) has been found to be broadly equivalent to the adult versions in sensitivity to changes in psychological well-being (Twigg et al., 2009), the YP-CORE has no family interaction domain, despite a substantial body of research documenting that emotional difficulties experienced by young people are closely linked to attachment styles and unhelpful patterns of interaction within the family (Connell, 2009; Gatta et al., 2015).

Family interactions
We recognise that the way the main caregiver(s) interacts with their child provides that child with their first sense of how to be in this world – internalising signals and messages from their parents (Flores, 2010; Immordino-Yang & Damasio, 2007) – as well as replicating behaviours (Perry & Szalavitz, 2006), which can drastically shape the growth and future development of the child (Lewis, 2009). Attachment theory (Bowlby, 1969; 1988) further defines this interaction describing it as a “lasting psychological connectedness between human beings” (Bowlby, 1969: 194) which can shape the attachment style of the child – both in childhood and in later life. This was an important discovery, which has seen a renewed interest in recent years, because attachment styles are associated with emotional regulation; children that have their emotional expressions acknowledged and responded to in a supportive and consistent manner, are more likely to become secure, develop openness to seeking support, express and experience both positive and negative emotions, whilst also being able to regulate their own emotions (Wallin, 2015). This contrasts with individuals whose emotional, and sometimes physical needs, were not supported or were met with inconsistent support. These individuals may become anxious, evidenced through ‘dwelling on’ or ‘exaggerating’ their problems, or may develop avoidant styles of interacting and might suppress their emotions as a result of being rebuffed (Jones, Brett, Ehrlich, Lejuez & Cassidy, 2014).

Through past research, we know that emotional regulation in children is fostered through positive Family Expressiveness (FE), particularly by mothers who could regulate their own emotions (Are & Schaffer, 2016). As therapists, parents, or as any professional working with minors, having this knowledge is useful although it is only one part of the jigsaw. Preliminary research into Family Expressiveness and the role it plays throughout the lifespan of an individual was conducted through a systematic literature review of 48 papers, identifying that FE is positively correlated to parental stress or adversity (McCue-Horwitz et al., 2007), as well as to the social, behavioural, psychological and physical functioning of the parent/ caregiver (for example Beek et al., 2015; Loon et al., 2014; Noguchi & Ollendick, 2010).

A small selection of results seemed to dispute that parental stress or adversity impacted Family Expressiveness (Dinning and Berk, 1989) or indeed that children were affected by poor FE (Loon et al., 2014). Possible explanations for these results could include good inter- parental relationships (Marakovitz et al., 2011) which moderated the level of stress felt by the parent(s), or the presence of another source of positive FE (Westerink & Giarratano, 1999). Additionally, it is important that we take note that Family Expressiveness was seen to manifest differently in different cultures (Beek et al., 2015; Suveg et al., 2014).

In summary, good communication, including Family Expressiveness, is a protective factor against adversity, whilst poorer communication, including FE, is related to negative outcomes (Diamond, Diamond & Levy, 2014; Manzeske et al., 2009) (Fig. 1). Family Expressiveness seems to be both impacted by, and impacting on, poor health and social problems whereas, at times, other influencing factors are present (Barry & Fleming, 1990; Schroeder & Kelley, 2009).

Figure 1. The Role of Family Expressiveness in an Attachment-Based Theory of Psychological Health (adapted from Diamond, Diamond & Levy, 2014: 51)

Enhancing responsiveness to family dynamics in pluralistic therapy with adolescent clients
When we engage a pluralistic style of therapy, it provides a framework for integration and builds on existing traditions of integrative therapy by actively promoting client-therapist collaboration and shared decision-making. In a pluralistic approach, we can enhance the capacity of the therapist and client to work productively together by agreeing tasks and goals, utilising feedback measures as conversational tools, and drawing on the strengths and resources of the client (Cooper and McLeod, 2012; McLeod, 2018). It is important to note that when working with adolescents, even in individual therapy, we may also need to extend these pluralistic skills to other stakeholders, including the parents/carers of the child.

Understanding the importance of Family Expressiveness, and concerned that this was not routinely captured through useful measures of outcome, I conducted a small-scale exploratory study to determine whether the use of a family-oriented outcome and feedback measure, the Family Expressiveness Questionnaire (FEQ), might be valuable in pluralistic counselling with adolescent clients. Specifically, the aim was to explore the possibility that, in counselling adolescent clients, the FEQ scale might yield information, that was not captured within YP-CORE, but that was relevant to the adolescents’ presenting issues.

Method
Twelve adolescents were invited to participate in the study at The Village Counselling Service in Tallaght, a low-cost service with over 160 counsellors. Clients were aged between 15 and 17 (m=15.75; male 5, female 7). Potential participants, and their parents, were offered a brief explanation of the study at an appropriate point during the routine intake at The Village Counselling Service. If both the young person and their parent or carer indicated an interest in taking part, they were invited to read an information sheet, which was tailored to provide information in age appropriate language, and invited to ask questions before completing a consent (adult) and assent (minor) form if they were willing to proceed.

Many of the routine Village Counselling Service assessment practices were upheld – including contracting around confidentiality and what to expect from counselling – with the parent/carer present. Parents were then invited to wait outside whilst the young person completed a YP-CORE and a semi-structured assessment interview, designed to gather information on the background of the client, as well as the presenting problems. For this study an additional measure of Family Expressiveness (FEQ) was introduced prior to the semi-structured interview, which participants were asked to complete. Upon completion of both the measures and assessment, the client and parent were debriefed by the researcher and each were provided with a debriefing sheet in age appropriate language.

YP-CORE Measure
The YP-CORE is designed to be an easily understood questionnaire, for eleven to sixteen-year- olds, to measure psychological distress. It is adapted from the CORE-OM (CORE- Outcome Measure), proposed to be a generic pan-theoretical and pan-diagnostic measurement (Barkham et al., 2001). Both these measures comprise of the four domains considered to be the most important aspects of psychological health: Risk; Functioning; Wellbeing; Symptoms (Twigg et al., 2009). Totalling the scores provides the therapist, or researcher, with an indication as to the level of psychological difficulty – the higher the score, the more distress.

Family Expressiveness Questionnaire Measure
The FEQ is a 40-item measure that we can use to assess the individual’s perception of patterns of emotional expressiveness within their family (Halberstadt, 1986). The domains of the FEQ fall under four subscales: Positive-Dominant (PD) highlighting aggressive forms of “pride, admiration, or support” (Cotar-Konrad, 2016); Positive-Submissive (PS) highlighting the softer emotions of “sympathy, readiness to help or to do a favour” (Cotar-Konrad, 2016); Negative-Dominant (ND) identifying contempt, criticism, blame and anger; Negative-Submissive (NS) identifying crying, sadness, sorrow and embarrassment (Halberstadt, 1986). An important feature of the FEQ is that we can use it to estimate the balance within a family between a more negative or a more positive expressive environment. We obtain the overall emotional balance by subtracting the totalled positive subscales from the totalled negative subscales.

In the present study, I revised the language in some items to suit an Irish adolescent (e.g., “argue” rather than “quarrel”), creating the Family Expressiveness Questionnaire (Revised Irish Version).

Furthermore, for the purposes of the study, two additional questions were inserted at the end of the FEQ with the option to indicate on a five-point Likert scale whether they ‘strongly disagreed’, ‘disagreed’, ‘neither agreed nor disagreed’, ‘agreed’ or ‘strongly agreed’ that:

1. The statements in this questionnaire made sense to me.
2. Some of the situations described in this questionnaire are relevant to the problems that I want to talk about in counselling.

Ethical issues
The ethical considerations of this study were thoroughly reviewed to ensure best practice and well-being of participants and parents, including confidentiality, in adherence with data protection (GDPR, 2016) and Children First Guidelines (The Department of Children and Youth Affairs, 2017). I considered that, by asking the adolescent to give details about their family, the study might cause unease or low-level distress to the adolescent or parent therefore, in order to promote non-malfeasance, it was deemed extremely important to provide a brief explanation into relationships and dynamics at the preliminary stage, to reassure that the responsibility of Family Expressiveness does not fall on any single person (including blame or judgement). Additionally, both parent and adolescent were debriefed, and debriefing counsellors were offered. The study received ethical approval from the Ethics Committee, Institute for Integrative Counselling and Psychotherapy (IICP).

Analysis
Using the Statistical Package for the Social Sciences database (SPSS) a series of descriptive analyses (mean, standard deviation) were conducted for the FEQ (Revised Irish Version) and YP-CORE questions, subscales and total scores, as well as the FEQ balance score. I then ran Pearson Correlations between all FEQ (Revised Irish Version) and YP-CORE questions, subscales and total scores as well as the FEQ balance score to determine whether the FEQ (Revised Irish Version) is providing information that is distinct from YP-CORE data. Both the descriptive analysis and Pearson correlation tests were further split into those who found the FEQ (Revised Irish Version) relevant to what they wanted to talk about in therapy (n=7) and those who did not find it relevant (n=5). The FEQ (Revised Irish Version) will further be referred to as FEQ.

Results
In the interest of clarity, I include a summary of key statistical findings below. More detailed analysis is available upon request.

Descriptive Analysis
The mean (average) YP-CORE score for these clients, collected at intake, was 20.17 (standard deviation 6.235). This is similar to YP-CORE scores recorded in other studies (O’Keeffe et al., 2015; Twigg et al., 2009) who found their adolescent mean scores to be 19.44 (SD=7.43) and 19.0 (SD = 7.5) respectively. The mean FEQ subscales recorded were also similar to mean scores reported in other studies (Kao, Nagata & Peterson, 1997; Hofer, Eisenberg & Reiser, 2010) as demonstrated in Table 1.

Table 1. Comparison of mean FEQ subscales with other studies

In terms of the perceived relevance of the FEQ, 91.6% of the participants either ‘strongly agreed’ (n=4) or ‘agreed’ (n=7) that the statements in the FEQ made sense to them whilst 8.3% (n=1) stated they ‘neither agreed nor disagreed’ with this statement. Seven participants (58.3%) ‘agreed’ (n=5) or ‘strongly agreed’ (n=2) that some of the situations described in the FEQ were relevant to problems that they would like to talk about in counselling. Three participants (25%) ‘disagreed’ (n=2) or ‘strongly disagreed’ (n=1) that it was relevant to their problems they want to discuss and two (16.6%) ‘neither agreed nor disagreed’.

Inferential Analysis FEQ Subscales and Balance and YP-CORE Total
Overall, there was a strong positive correlation between the Family Expressiveness Questionnaire (FEQ) Balance (Negative Affect) and the YP-CORE Total Score (r=0.765, p=0.004) (valued at p=0.01), indicating that both measures were sensitive to general levels of distress being experienced by participants.

Participants were split by those who either ‘agreed’ or ‘strongly agreed’ (n=7) that the FEQ was relevant to their therapy and those who did not (n=5). In those who found the FEQ relevant, there was a significant positive correlation between the Negative Dominant subscale of the FEQ and the YP-CORE Total (r=0.918, p=0.003) (p valued at 0.05). It also documented that the FE Balance had a significant negative correlation with Positive Dominant (r=-0.904, p=0.005), Positive Submissive (r=-0.839, p=0.018) subscales and a positive correlation with Negative Dominant (r=0.852, p=0.015) subscale. Other positive and negative correlations emerged, such as between FE Balance and the YP-CORE Total (r=0.752, p=0.051) however, these were not statistically significant.

In contrast, those who felt that the FEQ was not relevant to their therapy recorded statistically significant correlations between the YP-CORE and the FE Balance (r=0.950, p=0.013) and the Positive Submissive Subscale of the FEQ (r=-0.943, p=0.016). The FE Balance also had a positive correlation to the Positive Submissive subscale (r=0.980, p=0.003).

Inferential Analysis FEQ and YP-CORE Questions
A number of potentially meaningful findings emerged when YP-CORE questions were compared with FEQ Balance scores. Nine of the ten YP-CORE questions had only a weak correspondence to overall FEQ totals. The single YP-CORE question that was strongly linked to the FEQ score was “It’s been hard to go to sleep or stay asleep (r=0.670, p=0.017) (where p is valued at 0.05).

When I analysed the FEQ questions in terms of their level of correlation with specific YP- CORE statements both positive and negative correlations were found, as identified in Table 2 and Table 3 below:

Table 2: YP-CORE and FEQ questions with positive correlations.

Ten FEQ items showed low similarity with either specific YP-CORE statements, or the total YP-CORE score as shown below in Table 4.

Table 4: Low correlation with all YP-CORE questions and the YP-CORE total score

 

 

Discussion
This exploratory pilot study offers evidence that the YP-CORE and FEQ are in broad agreement: the more severe the mental health of the adolescent (high YP-CORE score) the more negative the Family Expressiveness (higher FE balance). Specifically, the YP-CORE total does correspond with some FEQ items indicating, for example, that the worse the well-being of the young person, the less the family seems to tell each other how happy they are, the more they argue, say they are not happy with someone else’s behaviour and show dislike for someone.

However, I found that some family interactions, such as blaming each other for family problems or showing deep affection or love for someone, had no correlation with the YP- CORE. Additionally, for those clients who experienced difficulties in Family Expressiveness, and who indicated an interest in discussing these issues in therapy (58%), the YP-CORE alone does not appear to be sensitive enough to discern all the domains and nuances that the young person experiences.

Taken as a whole, these findings suggest that it could be valuable for those of us working with young people in therapy to ask our clients if they would like to talk about their Family Expressiveness. It also highlights the importance of the counsellor’s understanding of, and ability to work with, family dynamics if deemed relevant by the young person. In addition, the FEQ may represent a useful way to enable clients to identify troubling family interaction patterns, providing the young person with an opportunity to identify tasks and goals that they wish to explore with their therapist and to track the effectiveness of therapy in addressing such problems.

For participants in the present study who found the Family Expressiveness Questionnaire relevant, I discovered that levels of distress in the young person were linked to the Negative-Dominance there was at home (contempt, criticism, blame and anger), whereas those who did not find the FEQ relevant to their therapy were more likely to record high scores in the areas of Positive- Submissiveness (sympathy, offering to do a favour). This may suggest that it is conceivably easier for an adolescent to manage the effect of less Positive-Submissiveness than it is to manage more Negative-Dominance. During adolescence the reliance on parental relationships for emotional regulation is lessening and moving towards peer relationships (Crittenden, 2002; Potard et al., 2017). It is possible therefore that the young people in this study were able to receive the sympathy and supportive affect missing from home through their peer relationships.

This paper demonstrates that Family Expressiveness has an important role to play in the life of an adolescent and, as such, it is essential that we do not see the young person as existing in a vacuum but, instead, as being impacted by both their family and the culture of the society in which they reside. At the same time, it is crucial that we keep in mind that what works for individuals differs greatly from one to another, and at various points in therapy (McLeod & Cooper, 2015). Examples of how working with the client’s parents and the family dynamic can be incorporated into therapy can be found in many approaches including psychodynamic (Diamond et al., 2010), CBT and humanistic (Haine-Schlagel et al., 2012), integrative (Martinez et al., 2015), and group-based interventions (Pereira et al., 2015). A pluralistic framework for practice provides a basis for integrating family-informed ideas and methods into therapy in a respectful, responsive and collaborative manner (McLeod, 2018).

Limitations of this study
This present study is based on a small sample of clients in one service and the results require validation in further studies. Many participants asked for further clarification or examples, despite the majority answering that the FEQ made sense to them (83.3%). Indeed, Halberstadt (1986) would say that there is ‘noise’ in the data when examining FE, as to ask people to compare, or accurately remember, their family environment can be difficult. This may be increasingly heightened for those with a learning difficulty such as being on the autistic spectrum who may not be able to read social cues (National Institute of Mental Health, 2011) or dyslexia, cited as the reason for declining participation by two adolescents. Consequently, learning difficulties, as well as the challenging concept of FE, are deemed important considerations prior to future use and research. On the other hand, participants appeared to be genuinely interested in the study, and seemed to complete the measures in a careful and authentic manner.

Finally, the results of my study are not intended to show causation, only a relationship; FE has not been shown to cause poor adolescent well-being, just as adolescent well-being has not been shown to affect the FE. Either may be true, both are likely.