Investigating the Effects of Challenging Behaviour on the Sibling Relationship: Influenced by Behaviour Topography and Shaped by Attributions of Controllability?
Dublin Core
Title
Investigating the Effects of Challenging Behaviour on the Sibling Relationship: Influenced by Behaviour Topography and Shaped by Attributions of Controllability?
Creator
Lauren Laverick-Brown
Date
2018
Description
Challenging behaviour (CB) displayed by individuals with an intellectual disability (ID) is consistently identified as a stressor on the relationship that they have with their typically developing (TD) sibling. Given the potentially damaging effects of CB on the quality of the sibling relationship and the wellbeing of the TD sibling, understanding the cognitions that underpin TD siblings’ emotional and behavioural responses to CB is essential to direct sibling-targeted psychoeducational interventions. This study considered whether siblings’ responses to CB vary according to behavioural topography. Further, the study considered whether any effects detected were shaped by attributions made by TD individuals regarding the controllability of their siblings’ CB. Thirty-eight siblings of individuals with ID, and 36 participants with a nondisabled sibling, completed a web-based questionnaire measuring participants’ positive and negative affect towards their sibling, the nature of their sibling’s CB, and controllability perceptions regarding their sibling’s CB. The results of this study reiterate that CB is a stressor on the sibling relationship, with externally directed CB (i.e. aggression, destruction) eliciting greater negative affect in siblings compared to internally directed behaviours (i.e. self-injury). However, it could not be concluded with an appropriate level of significance (i.e. p<.05) that this was due to participants perceiving their siblings as more in control of their externally directed behaviours. These findings may have resulted from the diverse nature of the participant group. Further research is required to examine specific differences in the emotional impact of each type of challenging behaviour (and then subsequently, whether any differences detected arise due to contrasting perceptions of behaviour controllability).
Source
Participants
Seventy-four TD individuals who had a sibling completed this study. Participants were allocated to one of two conditions according to whether they had a sibling with an ID, or did not (i.e. their sibling was TD). There were 38 participants who had a sibling with ID (82% female, Mage=27.32, SD=9.65) and 36 participants with a TD sibling (92% female, Mage=28.61, SD=10.81); ranging between 13 to 60 years old. Siblings diagnoses are reported in Table 1.
Table 1: Diagnoses of participants' siblings
Participants were recruited on a voluntary basis through social media advertisements posted by the researcher and by disability organisations (who also sent emails to their followers), and through word of mouth. The researcher developed a digital research flyer summarising study’s purpose and procedure, distributed as described above. To incentivise participation, participants had the option to enter themselves in a prize draw for a £20 Amazon voucher upon completion of the study.
A minimum participation age was determined after inputting the text of each questionnaire included in the study into Coh-Metrix (Version 3.0; Graesser, McNamara, Louwerse & Cai, 2004): a web-based software tool assessing the cohesion and coherence of a text. Coh Metrix provides an index of readability by generating the reading age of a piece of text, and the reading age determined of the questionnaires was “grade six”; indicating that 10/11-year-olds should have the ability to comprehend and respond to questions. Thus, it was decided that the questionnaires were suitable for TD individuals aged 12 or above.
Consent was gained from all those over 16 years of age, and parental consent for and assent from those aged 12-15 years of age (see “Ethical Considerations” below for further information).
Design
The study was of a correlational design, investigating the relationship between the following continuous variables: the quality of the sibling relationship, CB displayed by the sibling with ID, attributions of controllability made by participants in respect to their sibling’s behaviour, and participants’ general relational abilities.
As part of further analysis, the intention was to examine whether there were effects of having a brother/sister with a disability, gender, and birth order (i.e. whether participants were older/younger than their sibling) (all between-subjects factors) on the sibling relationship.
Materials
During this study, four self-report questionnaires were administered to all participants: the Positive and Negative Affect Scale (PANAS) (Watson, Clark & Tellegen, 1988) (Appendix A), the Behavioral Problems Inventory (Short Form) (BPI-S) (Rojahn, Matson, Lott, Esbensen, & Smalls, 2001) (Appendix B), the Controllability Beliefs Scale (CBS) (Dagnan, Grant & McDonnell, 2004) (Appendix C), and Social Competence and Close Friendship subscales taken from the Harter Self-Perception Profile for Adolescents (Harter, 2012) (Appendix D). The development and presentation of the questionnaires was done using online Qualtrics software (Qualtrics, Provo, UT).
The Harter Self-Perception Profile for Adolescents (Harter, 2012) is a multidimensional measure of how young people evaluate their scholastic, social, athletic, and job competencies, as well as physical appearance, romantic appeal, behavioural conduct, and close friendship. However, for the purposes of this study, only the subscales regarding social competence and close friendship were included to detect an individual’s general ability in forming and maintaining relationships with others, which might be a confounding influence on detecting the quality of the sibling relationship. Furthermore, the phrasing of the questionnaire was deemed suitable for both adult and adolescent participants.
The questions are presented as two clauses (e.g. "Some people know how to make others like them, but…”, and “Some individuals do not know how to make others like them”). Participants are able to select whether each clause is “really true for me” or “less true for me”, though are required to make the one selection out of four options across both clauses that is most self-descriptive. These responses are coded into a 4-point scale, with “1” representing poorer social/friendship abilities, and some items are negatively coded. Sufficient levels of validity and reliability of the Profile have been reported within a range of population groups (e.g. Donnellan, Trzesniewski & Robins, 2015; Rose, Hands, & Larkin, 2012).
A modified version of the PANAS (Watson et al. 1988) was used to assess participants’ feelings towards their brother/sister with a disability, which were then used to infer the quality of the sibling relationship i.e. greater positive affect would indicate a positive and fulfilling sibling relationship, whilst greater negative affect would indicate poor sibling relationship quality. The PANAS is a self-report questionnaire that consists of two separate scales containing emotion-based items that encapsulate positive and negative affect. Participants were asked to think about their sibling and whether they had felt each emotion towards them, rating this on a 5-point scale to specify how often they feel that emotion, ranging from 1 (very slightly or not at all) to 5 (extremely often). Higher total scores on each scale indicated greater positive/negative affect. “Total negative affect” and “total positive affect” scores were obtained for each participant; whereby higher scores pertain to greater affect.
The PANAS has been widely utilised to measure variation in affect, and previous research investigating its psychometric properties concludes it to have high reliability and validity across many populations (e.g. Merz, Malcarne, Roesch, Ko, Emerson, et al., 2013; Bakhshipour & Dezhkam, 2006). In this study, certain items of the PANAS were adapted to ensure that they were recognisable to younger participants; for example, “hostile” and “strong” were changed to “angry” and “happy”, respectively. The items “jittery”, “active” and “determined” were excluded as the researcher did not view them as relevant to the sibling relationship. Nevertheless, statistical analysis revealed that internal consistency remained, with the positive and negative affect scales showing high reliability in the current sample, Cronbach’s αnegative=.87; Cronbach’s αpositive=.93.
The BPI-S (Rojahn et al. 2001) is a psychometrically sound behaviour rating instrument (Rojahn, Rowe, Sharber, Hastings, Matson, et al., 2012; Mascitelli, Rojahn Nicolaides, Moore, Hastings et al., 2015) constituting a series of items referencing examples of CB. When completing the BPI-S, respondents consider whether a specific individual (in this study, participants’ sibling) engages in a behaviour, and then rate its frequency on a 1-to-6-point scale; corresponding to responses ranging from “never” to “daily”. The original BPI-S also contains a severity-rating subscale; however, this was excluded from the study, as rating the severity of behaviour was deemed to be too complicated for younger participants to judge.
The BPI-S contains questions relating to three types of problem behaviours: self-injurious, stereotypic, and aggressive/destructive behaviours. For the purposes of this study, the behavioural items of the BPI-S were grouped and scored according to whether they constituted IDB (i.e. self-injury) or EDBs (i.e. aggression and destruction). Items referencing stereotyped behaviour were excluded, as it was not possible to neatly categorise them into IDB or EDB. As an addition to the questions of the BPI-S, an opportunity for “free text” was included immediately after, whereby participants could describe any behaviours of concern that were not specified by the questionnaire and rate their frequency. Total scores for the BPI-S were obtained, as well as separate total scores for IDB and EDB frequency, whereby higher scores represent a greater number of incidences of CBs.
Lastly, the CBS (Dagnan et al., 2004) is a 15-item measure designed to capture participants’ perceptions regarding an individual’s (in this case, their siblings’) control over their CB. Responses are scored on a 1-to 5-point scale, corresponding to ‘disagree strongly’, ‘disagree slightly’, ‘unsure’, ‘agree slightly’ and ‘agree strongly’. Ten items are worded such that agreement reflects participants attributing high control over behaviour (e.g. ‘They are trying to wind me up’). In contrast, five items are phrased whereby agreement indicates participants attributing low control over behaviour (e.g. ‘They don’t mean to upset people’); thus, these items are reversed scored. A “total CBS” score was calculated for each participant, with higher scores pertaining to perceptions of greater control over behaviour. Moreover, Dagnan et al. (2004) report good internal reliability, with a Cronbach's alpha of 0.89.
Demographic information relating to participants’ age, gender, birth order (i.e. were they older/younger than their sibling) and the diagnosis of their disabled sibling (if their brother/sister was disabled) was collected prior to participants completing the questionnaires.
Procedure
After receiving expressions of interest from prospective participants and confirming they had a sibling (with or without an ID), the researcher issued a participant information sheet detailing the nature and aims of the study. Both groups of siblings followed the same study procedure but received participant information sheets that were relevant to their role in the study. The researcher also provided a weblink to the online consent form hosted by Qualtrics. Once participants completed the consent form, they answered demographic questions and generated a participant code to ensure anonymity of responses. Participants were informed prior to the study commencing that they could withdraw at any time, either by closing the webpage or by contacting the researcher and asking to be removed from the dataset.
Initially, participants responded to items of the Close Friendship and Social Competence subscales of the Harter Self Perception Profile. Following completion of these questions, participants then completed the PANAS, BPI-S and CBS (in that order). Upon finishing the CBS, participants who had a sibling with ID proceeded to a debrief form that outlined the study in detail and provided contact information for support organisations (if needed following discussion of their encounters with CB). Control participants received a debrief form detailing their role in determining the baseline/typical sibling relationship.
The procedure differed slightly for participants aged under 16 years old. With one exception, who contacted the researcher directly (but ultimately could not participate due to lack of parental consent), this group expressed their desire to participate through their parents contacting the researcher. In response, the researcher sent a consent form for a parent/guardian to complete, giving their permission for their child to participate in the study. Two participant information sheets were also provided; one for parents and another simplified version of the adult participant information sheet for individuals under 16 years old. Once the researcher had received the completed consent form, the weblink to the study was emailed. It was stressed to parents that, though they may wish to support their child in understanding the questions of the study, they should refrain from guiding their child’s answers.
After clicking the weblink, younger participants completed an assent form and were informed about the participation withdrawal procedures, if required. The presentation of the questionnaires was the same as for those aged 16 years old and above. However, the debrief form was simplified in its language and content to ensure it was accessible to younger participants. Contact information for organisations who could support this group of participants specifically was also provided. Additionally, younger participants with a non-disabled sibling received a simplified version of the adult participant debrief form relevant to their role in the study. After reading the debrief sheet, all participants were given the opportunity to enter into a prize draw for a £20 Amazon voucher. The study lasted roughly 15-20 minutes.
All participant information sheets, consent forms and debrief sheets are listed in Appendices E – S.
Ethical Considerations
This study was reviewed and approved by the Psychology Department Research Ethics Committee at Lancaster University.
The topic of this study revolved around participants’ experiences of CB, which could involve reflection upon sensitive experiences (including those of violence and destructive behaviour) that elicit negative psychological reactions (such emotional upset, worry, stress, and shame). Furthermore, the minimum age specified for participants is 12 years old, so some participants recruited would be minors (i.e. a vulnerable participant group).
In case the discussion of CB experiences elicited negative psychological reactions in participants, contact information for sources of wellbeing support was given as part of the study debrief for both young and adult participants (e.g., talking to a trusted family member or a teacher; information and contact details for free services such as Childline, the Samaritans, The CB Foundation etc.). Offering access to support services was particularly important to younger participants, who may not feel able to speak to their parents about any issues they have.
Furthermore, consent was required from all participants over the age of 16. If a participant indicated being under the age of 16, consent was sought from a parent/guardian, whilst assent was obtained from all 12-to-15-year-old participants. Consent and assent were monitored throughout the study. All participants were given sufficient opportunity to understand the nature, aims and expected outcomes of research participation. Complex technical information was suitably adapted so that participants aged under 16 years old could give consent to the extent that their capabilities allowed.
Seventy-four TD individuals who had a sibling completed this study. Participants were allocated to one of two conditions according to whether they had a sibling with an ID, or did not (i.e. their sibling was TD). There were 38 participants who had a sibling with ID (82% female, Mage=27.32, SD=9.65) and 36 participants with a TD sibling (92% female, Mage=28.61, SD=10.81); ranging between 13 to 60 years old. Siblings diagnoses are reported in Table 1.
Table 1: Diagnoses of participants' siblings
Participants were recruited on a voluntary basis through social media advertisements posted by the researcher and by disability organisations (who also sent emails to their followers), and through word of mouth. The researcher developed a digital research flyer summarising study’s purpose and procedure, distributed as described above. To incentivise participation, participants had the option to enter themselves in a prize draw for a £20 Amazon voucher upon completion of the study.
A minimum participation age was determined after inputting the text of each questionnaire included in the study into Coh-Metrix (Version 3.0; Graesser, McNamara, Louwerse & Cai, 2004): a web-based software tool assessing the cohesion and coherence of a text. Coh Metrix provides an index of readability by generating the reading age of a piece of text, and the reading age determined of the questionnaires was “grade six”; indicating that 10/11-year-olds should have the ability to comprehend and respond to questions. Thus, it was decided that the questionnaires were suitable for TD individuals aged 12 or above.
Consent was gained from all those over 16 years of age, and parental consent for and assent from those aged 12-15 years of age (see “Ethical Considerations” below for further information).
Design
The study was of a correlational design, investigating the relationship between the following continuous variables: the quality of the sibling relationship, CB displayed by the sibling with ID, attributions of controllability made by participants in respect to their sibling’s behaviour, and participants’ general relational abilities.
As part of further analysis, the intention was to examine whether there were effects of having a brother/sister with a disability, gender, and birth order (i.e. whether participants were older/younger than their sibling) (all between-subjects factors) on the sibling relationship.
Materials
During this study, four self-report questionnaires were administered to all participants: the Positive and Negative Affect Scale (PANAS) (Watson, Clark & Tellegen, 1988) (Appendix A), the Behavioral Problems Inventory (Short Form) (BPI-S) (Rojahn, Matson, Lott, Esbensen, & Smalls, 2001) (Appendix B), the Controllability Beliefs Scale (CBS) (Dagnan, Grant & McDonnell, 2004) (Appendix C), and Social Competence and Close Friendship subscales taken from the Harter Self-Perception Profile for Adolescents (Harter, 2012) (Appendix D). The development and presentation of the questionnaires was done using online Qualtrics software (Qualtrics, Provo, UT).
The Harter Self-Perception Profile for Adolescents (Harter, 2012) is a multidimensional measure of how young people evaluate their scholastic, social, athletic, and job competencies, as well as physical appearance, romantic appeal, behavioural conduct, and close friendship. However, for the purposes of this study, only the subscales regarding social competence and close friendship were included to detect an individual’s general ability in forming and maintaining relationships with others, which might be a confounding influence on detecting the quality of the sibling relationship. Furthermore, the phrasing of the questionnaire was deemed suitable for both adult and adolescent participants.
The questions are presented as two clauses (e.g. "Some people know how to make others like them, but…”, and “Some individuals do not know how to make others like them”). Participants are able to select whether each clause is “really true for me” or “less true for me”, though are required to make the one selection out of four options across both clauses that is most self-descriptive. These responses are coded into a 4-point scale, with “1” representing poorer social/friendship abilities, and some items are negatively coded. Sufficient levels of validity and reliability of the Profile have been reported within a range of population groups (e.g. Donnellan, Trzesniewski & Robins, 2015; Rose, Hands, & Larkin, 2012).
A modified version of the PANAS (Watson et al. 1988) was used to assess participants’ feelings towards their brother/sister with a disability, which were then used to infer the quality of the sibling relationship i.e. greater positive affect would indicate a positive and fulfilling sibling relationship, whilst greater negative affect would indicate poor sibling relationship quality. The PANAS is a self-report questionnaire that consists of two separate scales containing emotion-based items that encapsulate positive and negative affect. Participants were asked to think about their sibling and whether they had felt each emotion towards them, rating this on a 5-point scale to specify how often they feel that emotion, ranging from 1 (very slightly or not at all) to 5 (extremely often). Higher total scores on each scale indicated greater positive/negative affect. “Total negative affect” and “total positive affect” scores were obtained for each participant; whereby higher scores pertain to greater affect.
The PANAS has been widely utilised to measure variation in affect, and previous research investigating its psychometric properties concludes it to have high reliability and validity across many populations (e.g. Merz, Malcarne, Roesch, Ko, Emerson, et al., 2013; Bakhshipour & Dezhkam, 2006). In this study, certain items of the PANAS were adapted to ensure that they were recognisable to younger participants; for example, “hostile” and “strong” were changed to “angry” and “happy”, respectively. The items “jittery”, “active” and “determined” were excluded as the researcher did not view them as relevant to the sibling relationship. Nevertheless, statistical analysis revealed that internal consistency remained, with the positive and negative affect scales showing high reliability in the current sample, Cronbach’s αnegative=.87; Cronbach’s αpositive=.93.
The BPI-S (Rojahn et al. 2001) is a psychometrically sound behaviour rating instrument (Rojahn, Rowe, Sharber, Hastings, Matson, et al., 2012; Mascitelli, Rojahn Nicolaides, Moore, Hastings et al., 2015) constituting a series of items referencing examples of CB. When completing the BPI-S, respondents consider whether a specific individual (in this study, participants’ sibling) engages in a behaviour, and then rate its frequency on a 1-to-6-point scale; corresponding to responses ranging from “never” to “daily”. The original BPI-S also contains a severity-rating subscale; however, this was excluded from the study, as rating the severity of behaviour was deemed to be too complicated for younger participants to judge.
The BPI-S contains questions relating to three types of problem behaviours: self-injurious, stereotypic, and aggressive/destructive behaviours. For the purposes of this study, the behavioural items of the BPI-S were grouped and scored according to whether they constituted IDB (i.e. self-injury) or EDBs (i.e. aggression and destruction). Items referencing stereotyped behaviour were excluded, as it was not possible to neatly categorise them into IDB or EDB. As an addition to the questions of the BPI-S, an opportunity for “free text” was included immediately after, whereby participants could describe any behaviours of concern that were not specified by the questionnaire and rate their frequency. Total scores for the BPI-S were obtained, as well as separate total scores for IDB and EDB frequency, whereby higher scores represent a greater number of incidences of CBs.
Lastly, the CBS (Dagnan et al., 2004) is a 15-item measure designed to capture participants’ perceptions regarding an individual’s (in this case, their siblings’) control over their CB. Responses are scored on a 1-to 5-point scale, corresponding to ‘disagree strongly’, ‘disagree slightly’, ‘unsure’, ‘agree slightly’ and ‘agree strongly’. Ten items are worded such that agreement reflects participants attributing high control over behaviour (e.g. ‘They are trying to wind me up’). In contrast, five items are phrased whereby agreement indicates participants attributing low control over behaviour (e.g. ‘They don’t mean to upset people’); thus, these items are reversed scored. A “total CBS” score was calculated for each participant, with higher scores pertaining to perceptions of greater control over behaviour. Moreover, Dagnan et al. (2004) report good internal reliability, with a Cronbach's alpha of 0.89.
Demographic information relating to participants’ age, gender, birth order (i.e. were they older/younger than their sibling) and the diagnosis of their disabled sibling (if their brother/sister was disabled) was collected prior to participants completing the questionnaires.
Procedure
After receiving expressions of interest from prospective participants and confirming they had a sibling (with or without an ID), the researcher issued a participant information sheet detailing the nature and aims of the study. Both groups of siblings followed the same study procedure but received participant information sheets that were relevant to their role in the study. The researcher also provided a weblink to the online consent form hosted by Qualtrics. Once participants completed the consent form, they answered demographic questions and generated a participant code to ensure anonymity of responses. Participants were informed prior to the study commencing that they could withdraw at any time, either by closing the webpage or by contacting the researcher and asking to be removed from the dataset.
Initially, participants responded to items of the Close Friendship and Social Competence subscales of the Harter Self Perception Profile. Following completion of these questions, participants then completed the PANAS, BPI-S and CBS (in that order). Upon finishing the CBS, participants who had a sibling with ID proceeded to a debrief form that outlined the study in detail and provided contact information for support organisations (if needed following discussion of their encounters with CB). Control participants received a debrief form detailing their role in determining the baseline/typical sibling relationship.
The procedure differed slightly for participants aged under 16 years old. With one exception, who contacted the researcher directly (but ultimately could not participate due to lack of parental consent), this group expressed their desire to participate through their parents contacting the researcher. In response, the researcher sent a consent form for a parent/guardian to complete, giving their permission for their child to participate in the study. Two participant information sheets were also provided; one for parents and another simplified version of the adult participant information sheet for individuals under 16 years old. Once the researcher had received the completed consent form, the weblink to the study was emailed. It was stressed to parents that, though they may wish to support their child in understanding the questions of the study, they should refrain from guiding their child’s answers.
After clicking the weblink, younger participants completed an assent form and were informed about the participation withdrawal procedures, if required. The presentation of the questionnaires was the same as for those aged 16 years old and above. However, the debrief form was simplified in its language and content to ensure it was accessible to younger participants. Contact information for organisations who could support this group of participants specifically was also provided. Additionally, younger participants with a non-disabled sibling received a simplified version of the adult participant debrief form relevant to their role in the study. After reading the debrief sheet, all participants were given the opportunity to enter into a prize draw for a £20 Amazon voucher. The study lasted roughly 15-20 minutes.
All participant information sheets, consent forms and debrief sheets are listed in Appendices E – S.
Ethical Considerations
This study was reviewed and approved by the Psychology Department Research Ethics Committee at Lancaster University.
The topic of this study revolved around participants’ experiences of CB, which could involve reflection upon sensitive experiences (including those of violence and destructive behaviour) that elicit negative psychological reactions (such emotional upset, worry, stress, and shame). Furthermore, the minimum age specified for participants is 12 years old, so some participants recruited would be minors (i.e. a vulnerable participant group).
In case the discussion of CB experiences elicited negative psychological reactions in participants, contact information for sources of wellbeing support was given as part of the study debrief for both young and adult participants (e.g., talking to a trusted family member or a teacher; information and contact details for free services such as Childline, the Samaritans, The CB Foundation etc.). Offering access to support services was particularly important to younger participants, who may not feel able to speak to their parents about any issues they have.
Furthermore, consent was required from all participants over the age of 16. If a participant indicated being under the age of 16, consent was sought from a parent/guardian, whilst assent was obtained from all 12-to-15-year-old participants. Consent and assent were monitored throughout the study. All participants were given sufficient opportunity to understand the nature, aims and expected outcomes of research participation. Complex technical information was suitably adapted so that participants aged under 16 years old could give consent to the extent that their capabilities allowed.
Publisher
Lancaster University
Format
Data/SPSS.sav
Identifier
LaverickBrown2018
Contributor
Rebecca James
Rights
Open
Relation
None
Language
English
Type
Data
Coverage
LA1 4YF
LUSTRE
Supervisor
Chris Walton
Project Level
MSc
Topic
Clinical
Sample Size
Seventy-four typically developing individuals
Statistical Analysis Type
None
Files
Collection
Citation
Lauren Laverick-Brown, “Investigating the Effects of Challenging Behaviour on the Sibling Relationship: Influenced by Behaviour Topography and Shaped by Attributions of Controllability?,” LUSTRE, accessed April 24, 2024, https://www.johnntowse.com/LUSTRE/items/show/79.