Spanish Family Quality of Life Scales: Under and over 18 years old

Abstract Background Researchers, professionals, and families have shown increasing concern with the family quality of life (FQoL) of people with intellectual disability (ID) and their families. The goals of this research were (a) to explore how Spanish families understand FQoL by developing 2 different measurement tools for families with a member with ID under and over 18 years old, and (b) to provide 2 diagnostic instruments that will be useful for designing action plans. Method The study comprised 4 stages: (a) focus groups, (b) expert assessment, (c) pilot study, and (d) normalisation and standardisation. The data were collected in 5 regions in Spain, and 1,205 families with a member with ID took part in the normalisation and standardisation of the scales. Results Both FQoL scales were consistent and have valid psychometric characteristics. Conclusions The scales have a diagnostic purpose for use in designing action plans aimed at producing significant changes in families’ lives.


Introduction
In recent decades, researchers, professionals and families themselves have shown growing concern with the quality of life (QoL) of persons with intellectual disability (ID) and their families. Although research on the QoL of persons with ID has a long history and has generated a large corpus of results with consequences for both theory and practical applications for improving the lives of these people (e.g., Brown, 1997;Cummins, 1996;Cummins, 1997a;Keith & Schalock, 2000;Schalock, 1996Schalock, , 1997Schalock, Gardner, & Bradley, 2007;Schalock & Verdugo, 2002), the same has not happened with regard to family quality of life (FQoL).
In comparison with research on the QoL of persons with ID, research on FQoL is still limited (Summers et al., 2005) and recent. Nevertheless, there is widespread agreement on the importance of this new construct, both for improving families ' abilities to cope with a child with ID and for assessing the possible impact of the services and supports these families receive, which represent a signifi cant investment in terms of fi nancial and professional resources. To address these issues, various research teams have put considerable effort into conceptualising, measuring, and improving the quality of the families ' lives (Aznar & Casta ñ ó n, 2005;Brown et al., 2006;Hoffman, Marquis, Poston, Summers, & Turnbull, 2006;Hu, Summers, Turnbull, & Zuna, 2011;Isaacs et al., 2007). This work has led to the development of three scales for measuring FQoL, including different domains: (a) the Beach Center Family Quality of Life Scale (Beach Center FQOL Scale; Hoffman et al., 2006); (b) the Family Quality of Life Survey (FQOLS-2006;Brown et al., 2006;Isaacs et al., 2007); and (c) the Latin American Family Quality of Life Scale (Latin American FQoL Scale; Aznar & Casta ñ ó n, 2005).
The research presented here is in line with these initiatives and seeks to provide Spanish professionals and families with instruments based on what they understand by FQoL, their values and beliefs, and is sensitive to Spain ' s social, economic, and cultural reality. As pointed out by Gallimore, Weisner, Kaufman, and Bernheimer (1989), families build and organise their daily routines according to their beliefs, values, objectives, and the circumstances in which they live.
A QoL scale ought to be able to capture as faithfully as possible the situation of families at the macro-, meso-, and microsystem levels in the country where it is used. Spain has a number of particularities in different fi elds that justify the need for a special scale. For example, the rights of people with ID and their families are recognised by law, but are often ignored in practice; administrative procedures for accessing services can be a trying barrier for families; labour law does not favour reconciliation between work and parental attention to children with ID; around 80% of adults with ID live with their relatives despite the cost involved for parents; and there is a lack of respite care services (Gin é et al., 2011). In the fi eld of education, the transition from school to adult life is not well supported. From the social and economic point of view, the government has recently cut down on some aids to people with disability; unemployment now exceeds 20% of the population (Instituto Nacional de Estad í stica, 2012).
That is why, in spite of their great value, adapting one of the existing QoL scales does not seem the best way of measuring FQoL in Spain. Consequently, it is justifi ed to explore what Spanish families understand by FQoL through grounded theory and identify the most important areas that would serve as the basis for drawing up the fi rst version of the scale for the pilot study, as will be explained. Another feature of this study is that it presents two scales: one for families of children under 18 years old and another for families with children over 18 years old. Although it is true that many items are relevant throughout the whole life span, it is also true that the needs of persons with ID change substantially over time and, in consequence, so do the needs of the families and the impact on their perception of QoL. The wish to enhance the sensitivity to such changes in people ' s lives and their possible impact on FQoL, often associated with the services received, made it advisable to have two partially different instruments. Cummins (1997b) opted for two instruments to measure individual QoL, one for children and another for adults. The Supports Intensity Scale (Thompson et al., 2004) also has two instruments to measure support needs, one for adults and another that is currently being developed for children.
In addition, it was considered necessary to have instruments that would not only measure FQoL, but would also be useful diagnostic and intervention tools (Zuna, Summers, Turnbull, Hu, & Xu, 2010).
In this study, we set out to explore how Spanish families understand QoL by developing two different measurement tools in keeping with Spain ' s social and cultural context: one for families with children up to the age of 18 years, the other for families with adults over 18. The second objective of the study was to produce an instrument suited to the characteristics and needs of the population at which it is aimed, and one that would serve as a useful diagnostic tool for designing intervention plans capable of bringing about signifi cant improvements in families ' QoL.

Method
This section presents the procedure employed in developing the scales and the characteristics of the participants in the pilot study and fi eldwork.

Procedure
The study involved a qualitative inquiry in order to conceptualise and organise the different areas of FQoL, and a quantitative process that included trialling an initial pilot version of the two scales and the fi nal version. The study therefore had fi ve stages.

From conceptualisation of FQoL to the development of the Spanish Family Quality of Life Scales (CdVF-E).
Seventeen focus groups, each comprising between eight and 10 members of families with persons with ID, were held to identify and defi ne the areas of how FQoL is conceived of by Spanish families.
Four researchers taking on different roles (one coordinator, one co-coordinator and two observers) participated in each of the focus groups, of which audio and video recordings were made. The 17 sessions held in the different regions were then transcribed and the transcriptions analysed using ATLAS.ti (Version 5.1; Muhr, 2006) discourseanalysis software. The procedure employed for analysing the focus group data was as follows: (a) each researcher identifi ed the most signifi cant text segments or quotations, assigned them a code, and defi ned different categories; (b) the two researchers who had analysed the same transcriptions shared their analyses, compared their codes and categories, and established an initial version of codes, categories, and defi nitions; (c) two discussion meetings in which all the researchers from each team took part were then held to agree on a single proposal regarding codes, areas ( Emotional Wellbeing, Family Interaction, Health, Financial Wellbeing, Parents ' Organisation and Skills, Family Accommodation, and Social Inclusion and Participation), and defi nitions.

Development of the primary survey items.
On the basis of the domains and the quotations transcribed from the focus groups, the different teams produced items for each of the areas, which were subsequently revised in accordance with the criterion that the items (statements) should refer to the key aspects of FQoL.
At this stage, each of the FQoL areas had between 10 and 15 items consisting of statements, and the families were asked to rate each item in terms of both importance and satisfaction. This process concluded with two proposed FQoL scales, one for over-18year-olds and one for under-18-year-olds, each with 105 items. Both scales had a 5-point Likert format and were self-administered.
Experts ' assessment and revision. Experts from different fi elds (research, professional, and public administration) reviewed the scales. The under-18-year-old scale was reviewed by 21 experts and the over-18year-old scale by 22 experts. The experts were requested to give their view on (a) the instructions given to the families on how to respond to the scale statements; (b) the clarity, suitability, and importance of the items; (c) the type of response (in terms of importance and satisfaction); and (d) the length of the instrument. All the experts ' comments were discussed by the research teams until a consensus was reached. The fi nal wording of the items on each scale was drawn up, taking into account the experts ' comments. A total of 101 items remained on each scale.
Pilot test and revision. As a result of the experts ' contributions and refl ection by the researcher group, as will be explained, the characteristics of the two scales used for the pilot study were as follows: (a) each scale contained statements from the seven areas identifi ed through the focus groups, (b) the items were not grouped by domains within the instrument, (c) each item required a response on a 5-point Likert scale, (d) these statements were to be answered in terms of importance and frequency, (e) items could be scored as not applicable , and (f) the two scales were self-administered.
Participants were recruited from associations of parents of children with ID in the Spanish regions of Andalusia, the Basque Country, the Canary Islands, Catalonia, and Madrid. The lead researcher in each region contacted the coordinators of the associations by telephone and described the details of the study and what was required for participation. Whenever an association decided to participate, its coordinator contacted the families about the study. Families who agreed to participate were given a research package that included an informed consent form (in accordance with the ethical guidelines of the Spanish Psychological Society), an information sheet, a form on which to record the family ' s relevant demographic details, a copy of the pilot FQoL scale and an envelope in which to enclose, seal, and return the scale and demographic details form to the association once they had been fi lled in. Families were asked to complete the scale from the viewpoint of the family as a whole, irrespective of whether the mother or the father fi lled it in. The associations undertook to post the envelopes with the scales and forms on to the respective research groups. The families were aware that they could raise any queries they might have with the lead researcher at any time during the process.
An analysis of the item-total correlation matrix was used to identify low-discrimination items ( r Ͻ .15) and these were eliminated from the fi nal version of the scale.
Field test. Stratifi ed sampling was used to select families with persons with ID from associations of parents of persons with ID in the fi ve regions of Spain mentioned earlier. The procedure employed for sending out and collecting the scales was the same as in the pilot study. The selection criteria for participation in the sample were the degree of disability (mild, moderate, and severe) and age (under/ over 18 years old) of the person with ID, and the family ' s place of residence (urban/rural area).
Altogether, 1,679 scales for families with a child with ID under 18 years old and 1,487 scales for families with a child with ID over 18 years old were sent out. A total of 607 (36.15%) under-18 scales and 598 (40.21%) over-18 scales were returned, although four of the latter had to be discarded as they were incomplete.
A total of 116 families participated in the pilot study. Of these, 61 families took part in the trial of the scale for families with children under 18 years old, and 55 in the trial of the scale for families with children over 18 years old.
A total of 1,205 people, all different from those in the pilot study, took part in the fi eld test. Of these, 598 completed the over-18 scale and 607 the under-18 scale. The demographic characteristics of the participants in the fi eld test process are shown in Table 1.

Results
This section describes the fi nal characteristics of the two scales that were produced and their psychometric properties.

Final characteristics of the two scales
The characteristics of the fi nal versions of the two scales were as follows: (a) the items of both scales are representative of the seven FQoL areas identifi ed (Emotional Wellbeing, Family Interaction, Health, Financial Wellbeing, Parents ' Organisation and Skills, Family Accommodation, and Social Inclusion and Participation); (b) the items relating to the different areas were arranged in random order; (c) each of the areas had between six and 12 items consisting of statements; (d) respondents indicated their extent of agreement with each statement on a 5-point Likert scale; (e) there was a not applicable option; (f) there were 61 items in the under-18-year-old scale and 67 in the over-18-year-old scale; and (g) the two scales were self-administered.
Appendix 1 shows a small sample of the items in the fi nal version of each scale.

Psychometric characteristics of the scales
From the returned responses, an item correlation matrix was made for both the over-18 and under-18 scales. Items with a correlation value of r Ͻ .15 were discarded, leaving the under-18 scale with 61 items and the over-18 scale with 67 items. The reliability according to the Cronbach ' s alpha coeffi cient for both scales was .96.
An intercorrelation matrix showed that the seven domains helped to measure FQoL effectively and effi ciently ( Table 2 shows the mean and standard deviation score by domain for each scale).
However, a factor analysis showed that a main factor accounted for 30% of variance in the over-18 scale and 28% of the variance in the under-18 scale. Both scales measured a single factor (overall FQoL) and, as Samuel, Rillotta, and Brown (2012) stated, the domains could be combined in order to describe FQoL as a whole. The previously identifi ed FQoL  (Brown et al., 2006;Isaacs et al., 2007), although some aspects are occasionally included in different dimensions. Unlike the Beach Center FQOL Scale (Hoffman et al., 2006) and the FQOLS-2006 (Brown et al., 2006;Isaacs et al., 2007), the scale presented here (CdVF-E) does not include any dimension regarding the support received by the families or the persons with disability. The reason for this is that the scale seeks to measure the effects of the various types of support on FQoL rather than whether or not families receive such support. Supports should be regarded more as an input than as an output. Certainly, the construct of supports is mediated by multiple inputs (i.e., fi nance, community, services), but we felt it appropriate not to include in the scales the supports that come directly from services as there are other tools specifi cally designed to capture the support needs of people with ID and their families.
Following the recommendations by Cummins (2002) and the Beach Center researchers (Hoffman et al., 2006), the CdVF-E does not use " importance " as a family response variable. Cummins said, " However, I am forming the view that importance may not be a useful construct to measure " (Cummins, 2002, p. 3), and that " importance " becomes a " weak and ambiguous concept " (Cummins, 2002, p. 3). For their part, Samuel et al. (2012) recall that the scores families give to " importance " are usually high on most items, and so it is not useful for measuring FQoL.
How respondents perceive satisfaction within the domains of FQol has been an important issue from the earliest systematic assessment of FQoL (Barnes & Olson, 1982). Certainly, in the more recent scales -the Beach Center FQOL Scale (Hoffman et al., 2006) and the FQOL-2006(Brown et al., 2006Isaacs et al., 2007) -levels of satisfaction are considered to be a good measure to describe family perceptions of QoL. Nevertheless, as Rillotta, Kirby, and Shearer (2010) and Zuna et al. (2009) state, one of the results generally found when these scales have been used is that the degree of satisfaction expressed tends to be high and therefore has little discriminating power. Additionally, when attempts were made to supplement the responses of the families with an interview, the results were contradictory (Rillotta, Kirby, & Shearer, 2010;Zuna et al., 2009).
We think that the results we obtained using the CdVF-E scales may help us to fi nd other ways to complement the measurement of FQoL and to develop action plans. We can also supplement the results obtained from the scales with other measures, such as interviews, to explore the satisfaction levels of the families. areas should therefore be interpreted as trends and used only as a guide for more clinical work.

Discussion
The fi rst goal of this study was to produce two scales that would enable measurement of FQoL of Spanish families with children with ID throughout their life span. In this regard, it can be said that two valid and reliable scales for measuring the FQoL of families with children and adults with ID were constructed.
The areas identifi ed in relation to Spanish families are generally consistent with the two most important FQoL scales (Samuel et al., 2012). This supports the existence of common areas affecting FQoL, such as health, economic and material welfare, and family interactions. Some of the areas identifi ed with regard to Spanish families, such as emotional wellbeing and parenting, coincide most clearly with the dimensions of the Beach Center FQOL Scale (Hoffman et al., 2006). However, the family accommodation area, which refl ects aspects to do with acceptance of, and adaptation to, their child ' s disability by the different family members, includes, among other things, some aspects that in the FQOLS-2006 (Brown et al., 2006;Isaacs et al., 2007) are included in Careers and Preparing for Careers, and also other matters related to Infl uence of Values. The participation and social inclusion area in the Spanish scale (CdVF-E) refers to the social relations of the family and the person with ID, and includes aspects that are covered by the Leisure and Recreation and Community Interaction dimensions in the FQOLS-2006 (Brown et al., 2006;Isaacs et al., 2007). In summary, the content of the areas of the CdVF-E coincide to a large extent with the Beach Center FQOL Scale (Hoffman et al., measurement to application; that is, translating concepts into practice in order to produce observable improvement in the life of persons and families (Brown, Schalock, & Brown, 2009;Zuna et al., 2009). Professionals in Spain now have available to them instruments validated in that country enabling them to move from measurement to action. The scales provide them with tools to identify families ' strengths and weaknesses and QoL trends that may be useful in setting goals, planning, and implementing family-centred intervention plans. This is the objective that gives meaning to all our work: helping, through our research, to promote changes in the lives of families in terms of a better QoL. That is why we have given both scales to FEAPS (Spanish Confederation of Organisations in Favour of Persons with Intellectual Disability) and have developed a user ' s manual together with FEAPS so the scales can be used as intervention tools once professionals have received the appropriate training, which we are currently engaged in providing. The CdVF-E can be used in two ways: the scales can be self-administered like other scales, such as the Beach Center FQOL Scale (Hoffman et al., 2006), and they can also be used for clinical purposes by means of an interview. The CdVF-E scales are a modest contribution to cross-cultural studies on FQoL and to the construction of FQoL theory, as called for by Zuna et al. (2009Zuna et al. ( , 2010.

Limitations of the study
One limitation concerns the composition of the participant sample. Recruitment of families from parent associations may mean that the participants in this study are not representative of the wider population of families with a member with ID (Hoffman et al., 2006;Summers et al., 2005).
A second limitation relates to the fact that although the people who answered the questionnaire were asked to do so from the family perspective, there is no way of being sure this instruction was always followed.
A fi nal limitation has to do with the scope of the population at which the Spanish FQoL scales are aimed. They were constructed to provide Spanish professionals and families with a person with ID with instruments attuned to their social and cultural characteristics. Although such contextualisation was one of the objectives, it may also be a limitation in regard to the possible use of these scales in other countries. If they are used elsewhere, care must be taken to ensure that the social, cultural, and economic conditions, and the legislation and access to services are similar.
The decision to produce two scales (one for children up to 18 years old and the other for adults over 18) has a theoretical basis in the fact that the different scales respond better to families ' changing needs throughout the life span. This decision was also supported by the comments of families in the focus groups that revealed different concerns among younger and older parents, which seemed necessary to capture. Different versions of scales are common practice in the fi eld of individual QoL assessment. For example, Cummins (1997b) developed one scale for adults and another for children, whereas Schalock and Keith ' s (1993) Quality of Life Questionnaire is aimed at schoolchildren. In addition, both the Latin American FQoL Scale (Aznar & Casta ñ ó n, 2005) and the Beach Center FQOL Scale (Hoffman et al., 2006) are fundamentally designed for young families (Samuel et al., 2012). Consequently, two scales have been constructed that are sensitive to the needs of Spanish families with young and adult children.
Lastly, in addition to measuring FQoL, our scales also have a diagnostic purpose for use in designing action plans aimed at producing signifi cant changes in families ' lives. This facilitates the work of professionals by enabling them to link the measurement of FQoL with the goal of fostering real improvements in the lives of all family members, including the person with disability (Zuna et al., 2010).
In summary, both scales -the under-18 and over-18 CdVF-E -represent a step forward in investigating the QoL of families with persons with ID in relation to both research and especially professional practice. The scales appear to be useful instruments for working with families throughout the life span and for drawing up action plans that will help to improve the QoL of the families and persons with ID.

Conclusions
As was proposed at the outset, Spanish professionals and families now have two scales available to them -one for children and young people up to 18 years old and another for persons over 18 years oldconstructed on the basis of what families in Spain understand by family QoL, which are sensitive to their political, social, and cultural reality, and are standardised and validated. Although a large number of items are common to both instruments, having two scales makes it possible to better capture situations that change in the course of families ' life cycles and potentially have an impact on their QoL.
According to Samuel et al. (2012), interest in the QoL and FQoL fi eld has recently shifted from The family member with ID has healthy eating habits. 1 2 3 4 5 0 My family has enough fi nancial stability to face the future without any major concerns. The family member with ID has a group of friends. 1 2 3 4 5 0