Research in Developmental Disabilities, Vol. 16, No. 1, pp.

51-74, 1995
Pergamon 1995 Elsevier Science Ltd
Printed in the USA. All rights reserved
0891-4222/95 $9.50 + .00

0891-4222(94)00028-X

Quality of Life:
Its Definition and Measurement

David Felce and Jonathan Perry

Welsh Centre for Learning Disabilities Applied Research Unit, University of Wales
College of Medicine

A model of quality of life is proposed that integrates objective and subjective

indicators, a broad range o f life domains, and individual values. It takes
account of concerns that externally derived norms should not be applied with-
out reference to individual differences. It also allows for objective comparisons
to be made between the situations of particular groups and what is normative.
Considerable agreement exists that quality o f life is multidimensional.
Coverage may be categorised within five dimensions: physical weUbeing, mate-
rial wellbeing, social wellbeing, emotional wellbeing, and development and
activity. A research agenda is discussed as are the particular problems caused
by difficulties in understanding and communicating.

Quality of life is an elusive concept approachable at varying levels of gen-

erality from the assessment of societal or community wellbeing to the spe-
cific evaluation of the situations of individuals or groups. Conceptualisation
has reflected such variation. Broad social indicators have been used to
chart the wellbeing of populations at the aggregate level (Flax, 1972; Liu,
1976; Schneider, 1976). Social and psychological indicators have been
developed to reflect individual welfare (Bigelow, McFarland, & Olson,
1991; Bradburn, 1969; Campbell, Converse, & Rodgers, 1976; Heal &

Requests for reprints should be sent to David Felce, Meridian Court, North Road, Cardiff, CF4
3BL, Wales.
This research was conducted under a grant from the Welsh Office/Department of Health. The
authors are grateful to Sharon Landesman-Ramey, Robert Cummins, Roy Brown, John Jacobson,
and Jim Mansell for comments on an earlier draft.
51
52 D. Felce and J. Perry

Chadsey-Rusch, 1985). Operational definitions of quality of life are

diverse, with variability fuelled not only by use of societal or individualis-
tic perspectives but also by the range of applicable theoretical models or
academic orientations.
Many writing on the subject of quality of life have commented on this
diversity. Liu (1976) stated that there were as many quality of life defini-
tions as people, emphasising the axiom that individuals differ in what they
find important. Baker and Intagliata (1982) point to there being as many
definitions as the number of people studying the phenomenon, a comment
that throws the spotlight on the lack of agreement between those attempting
to operationalise the concept. In their view, writers had done little up to that
point to achieve definitional consistency. A decade later, Cummins,
McCabe, Gullone, & Romeo (in press) has echoed this position in observ-
ing that not 1 of over 80 quality of life scales identified has achieved a
level of acceptance against which other scales can be validated.
Many of these scales, together with other measures of process and out-
come, have been developed to evaluate fundamental changes in services
that play a major role in the day-to-day lives of people in need of support,
such as those with mental retardation, chronic psychiatric morbidity,
physical disabilities, or infirmity due to old age. In the field of mental
retardation alone, the last three decades have seen the introduction of
major policy movements: deinstitutionalisation, investment in early inter-
vention and ongoing family support, integration in schooling, and reform
of adult day services. Although policy formulation has followed from our
best efforts to pursue fundamental values in the light of knowledge or
assumptions about human behaviour, and although evaluation has accom-
panied the process of change, the impact of these policies on quality of
life cannot be wholly substantiated in the absence of an agreed assess-
ment methodology.
The inadequacy of such a situation was recognised by Landesman (1986),
who called for a concerted effort to explore the meaning of two concepts,
quality of life and life satisfaction, and establish criteria to assess outcome
in such terms. In response, a monograph on the subject (Schalock, 1990),
together with a number of other relevant books, articles, and chapters have
been published in the last 5 years (Borthwick-Duffy, 1992; Brown, 1988;
Brown, Bayer, & MacFarlane, 1989; Goode, 1988; Parmenter, 1988). The
Schalock (1990) monograph provides a starting point for reviewing the sta-
tus of research in this field. However, as in many edited texts, synthesis was
not achieved across the diverse contributions. The purposes of this article
are, therefore, (a) to review the literature concerned with the definition of
quality of life, (b) to identify common definitional themes, (c) to construct a
model of quality of life based on common ground, and (d) to suggest
research that needs to be done to move the field forward.
 Quality of Life 53

THE EVOLVING SCOPE OF EVALUATION:

TREATMENT OUTCOME TO QUALITY OF LIFE

The widening scope of evaluation in mental retardation is well illustrated

by the evolution of research on residential services. Early research of altema-
fives to institutional care manifested a traditional treatment effectiveness
approach, with effectiveness being equated with developmental gain (Tizard,
1964). Though moves to broaden the lens of evaluative effort were made
almost immediately (King, Raynes, & Tizard, 1971), measurement of devel-
opmental progress has consistently featured in research on residential ser-
vices up to the present (Aninger & Bolinsky, 1977; Felce, de Kock, Thomas,
& Saxby, 1986; Hemming, Lavender, & Pill, 1981; Lowe, de Paiva, & Felce,
1993; Schroeder & Henes, 1978).
Emerson (1985) criticized the marked rehance researchers have placed on
adaptive behaviour gain as an outcome measure and the limited attention
given to other indicators, such as user satisfaction (Conroy & Bradley, 1985;
Scheerenberger & Felsenthal, 1977), social relationships (de Kock, Saxby,
Thomas, & Felce, 1988; Romer & Heller, 1983), or activity patterns
(Mansell, Jenkins, Felce, & de Kock, 1984; Sackett & Landesman-Dwyer,
1977). T'mard had reeognised that intelligence quotient data had not captured
the changes in the quality of regime estabhshed in the Brooklands experi-
ment (Tizard, 1964). His collaboration with King and Raynes in the analysis
of staff--resident management practices (King et al., 1971) was important in
conceptualising quality of care as an independent entity potentially related to
but not definitionally the same as service or treatment outcome. Moreover,
their attempt to encapsulate quality of regime was paralleled by other emerg-
ing environmental assessments reflecting developmental psychology
(McLain, Silverstein, Hubbell, & Brownlee, 1977), social psychology
(Moos, 1974), behavioural psychology (Risley & Cataldo, 1973),
behavioural ecology (Sackett & Landesman-Dwyer, 1977), and normalisa-
tion (Wolfensberger & Glenn, 1975).
In keeping with this broadening attention to the situations, social process-
es, activities, and outcomes experienced by people with disabilities, service
agencies increasingly incorporated a range of target achievements within
their mission statements (DHSS, 1971; Welsh Office, 1983). Maximising
developmental growth was frequently cited, but so too were normal patterns
of living, maintenance of family ties, development of social relationships,
participation in community and civic life, exercise of personal choice, and
other characteristics of a typical community lifestyle. Research responded to
this emerging demand for multidimensional measurement. For example,
several evaluations comparing community and institutional settings
(Burchard, Hasazi, Gordon, & Yoe, 1991; Conroy & Bradley, 1985; Felce,
1989; Felce, Kushlick, & Smith, 1980; Lowe & de Paiva, 1991) have
54 D. Felce and J. Perry

addressed the majority of the following: environmental quality, staffing,

staff turnover, staff performance, the use of professional services, develop-
mental progress, resident activity patterns, family and friendship contact,
community activities, and family and consumer satisfaction.
However, although evaluation practice has undoubtedly shifted to
address issues of quality of life, a comprehensive assessment approach
encapsulating its full complexities has not emerged. Although measures
that reflect use of the community, social networks, the pursuit of construe-
five activities, and choice and variety in activity bear on and are important
facets of quality of life, they do not constitute a full operationalisation of
the concept. In his review, Emerson (1985) concluded that evaluation of
personal satisfaction and psychological wellbeing had been seriously
neglected and pointed to emerging models in other fields (Andrews &
Withey, 1976; Baker & Intagliata, 1982; Campbell et al., 1976).

CONCEPTUAL MODELS OF QUALITY OF LIFE

Borthwick-Duffy (1992) has presented three perspectives on quality of
life: (a) quality of life defined as the quality of one's life conditions, (b)
quality of life defined as one's satisfaction with life conditions, and (c)
quality of life defined as a combination of both life conditions and satis-
faction. These and a fourth perspective are illustrated in Figure 1. The lat-
ter, which we will argue has advantages over the first three, similarly
depicts quality of life as a combination of life conditions and satisfaction
but emphasises the need to take account of personal values, aspirations,
and expectations.
Landesman's appeal for definition of concepts (Landesman, 1986) sug-
gested that there were two separate phenomena, quality of life and satisfac-
tion with life. One possible model is that depicted in Figure la. Quality of
life is the sum of a range of objectively measurable life conditions experi-
enced by an individual. These may include physical health, personal cir-
cumstances (wealth, living conditions, etc.), social relationships, functional
activities and pursuits, and wider societal and economic influences.
Subjective response to such conditions is the domain of personal satisfac-
tion with life. One interpretation of this model would follow from the argu-
ment that no citizen has the right to satisfaction with life but only the right
to life and equality of opportunity. Life conditions may well affect personal
satisfaction, but neither this potential relationship nor the subjective
appraisal itself is germane to quality of life assessment. General indicators
of quality of life may, therefore, be established by the assessment of a vari-
ety of life conditions across the population. The quality of life of a particu-
lar individual or subgroup would be located by comparing their position to
the total population distribution.
 Quality of Life 55

(a) Quality of life is defined in terms of Life Conditions

(b) Quality of Life is defined in terms of Satisfaction with life

(c) Quality of Life defined as a combination of Life Conditions awl Satisfaction

~ p

ev
r a
s I

n e
a s

(d) Quality of Life defined as a combination of Life Conditions and Satisfaction

weighted by Scale of Importance

FIGURE 1. Conceptualisations of quality of life.

For somewhat different reasons, Edgerton (1990) has also equated qual-
ity of life with the conditions under which life is lived. He cautions
against there being objective standards by which one can define a decent
or reasonable quality of life. Satisfaction with life is described as a sepa-
rate and ultimately more important criterion of individual welfare. The
56 D. Felce and J. Perry

individual's personal autonomy to maintain or change their quality of life is

a paramount consideration. Such a position is also reflected in the model in
Figure lb.
The formulation in Figure la is at odds with the tradition of treating
quality of life as the ultimate criterion by which to judge personal welfare.
In defining quality of life in terms of life conditions without any subjective
interpretation of how the individual perceives and reacts to such conditions,
room is left to advance such subjective assessments as personal satisfaction
with life as more important. From this perspective, the model in Figure lb
would appear preferable in defining quality of life as synonymous with per-
sonal satisfaction. Overall satisfaction with life may reflect satisfaction in a
number of life domains: material comforts, health, work, recreation, learn-
ing, and creative expression (Flanagan, 1978); living situation, relations
with family, social relations, leisure, work, finances, safety, health, and reli-
gion (Lehman, 1988); and, marriage, family life, friendships, standard of
living, work, neighbourhood, city or town of residence, the state of the
nation, housing, education, health, and the self (Campbell, 1981). It may be
hypothesised that satisfaction is influenced by external and objectively
quantifiable life conditions, and such a model has been tested (Allen,
Benfler, & Gutek, 1985).
However, equating quality of life solely with personal satisfaction and
treating the latter as a commentary on the acceptability of one's life condi-
tions is not to be accepted uncritically. Edgerton (1990) comments on the
possible independence between life conditions or events and subjective
reports of wellbeing. He cites his and others' research (Costa, McCrae, &
Zonderman, 1987; Keogh & Puliis, 1980; Lazarus & Lannier, 1979) to sus-
tain the argument that reports of welibeing owe more to internal disposi-
tions than to external conditions. Significant life events may induce tempo-
rary changes in reports of wellbeing, but regression to a characteristic level
will occur in time. Hence, the longitudinal pattern is relatively stable with
the best predictor of current satisfaction being its past rating. If this argu-
ment is supported, reports of wellbeing may only be sensitive to gross and
immediate changes in life conditions and not provide the subtle indicator of
quality of life that may be assumed.
Further, the basis for reaching a judgement on satisfaction is compara-
tive; it is impossible to divorce expressions of satisfaction from their con-
text. Thus, a person may be satisfied with his or her wages until he or she
discovers that a colleague of equivalent seniority, competence, and respon-
sibilities at work earns significantly more. Such contextual factors may be
specific, as in the preceding example, or general, as in the typical circum-
stances a person has experienced during his or her lifetime and seen experi-
enced by peers. It is tempting to equate high satisfaction with maximised
welfare as if it were the expression of preference under the conditions of
 Quality of Life 57

free choice, unconstrained opportunity, equality of expectation, and a stan-

dard comparative frame of reference. This is clearly not the case; satisfac-
tion is a personal assessment, the frame of reference is personal and affect-
ed by experience and the judgement of what is possible and typical for a
person in one's situation. Expectations and referents may be set low or high
and may themselves change over the life course and in response to personal
circumstance. In particular, expressions of satisfaction may adjust to rather
than reflect circumstance. Quality of life defined as synonymous with per-
sonal satisfaction, without regard to widely different life conditions, is a
less appealing formulation when reasonable independence and autonomy
of action cannot be assumed.
Clearly, there are some members of society whose independence and
autonomy is limited to a great or even atypical degree. Most people with
mental retardation lack independence skills, a deficit that constrains their
autonomy and frequently results in them inhabiting worlds of other people's
construction. The autonomy to maintain or change life conditions in line
with subjective appraisal that underpins the primacy of personal satisfaction
in the argument advanced by Edgerton earfier may be missing. Moreover, if
satisfaction is a measure of comparison, people whose circumstances and
options to date may make them particularly prone to having low expecta-
tions may be the most likely to report satisfaction in situations that the
majority would find intolerable. Edgerton, Bollinger, and Herr (1984), Close
and Halpern (1988), and Flyun (1989) have shown that individuals remain
philosophical or satisfied about the present and remarkably optimistic about
the future, despite the adverse conditions under which they live (including
poverty, poor housing, threats to health, threats to safety, victimisation,
social isolation, experience of loss, failure to gain or retain employment).
Holland (1990) found that expressions of satisfaction failed to differentiate
among a range of living environments, from impoverished traditional hospi-
tal wards to typical community housing, settings that differed markedly on
many objective characteristics. A definition of quality of life that ignores
objective assessment of life conditions may, therefore, not provide an ade-
quate safeguard for the best interests of vulnerable and disadvantaged peo-
ple. Expressions of satisfaction may simply reflect the intractability of con-
ditions commonly experienced by those with limited skills, autonomy, and
attachment to the mainstream society and its economy.
Nevertheless, expressions of satisfaction remain expressions of satisfac-
tion, a datum that most commentators agree lies somewhere near the heart
of quality of life. Personal appraisals of wellbeing have a validity for which
there is no substitute ff one person's values are not to be imposed on anoth-
er. Their accepted importance, tempered by some appreciation of the diffi-
culties inherent in their interpretation, may account for the widespread
view that the concept of quality of life must combine both objective and
58 D. Felce and J. Perry

subjective components (Brown et al., 1989), a view that produces a third

formulation (Fig. lc). Life domains may be assessed objectively by biolog-
ical, material, social, behavioural, or psychological indicators. Subjective
feelings about each area of life may also be reflected in reports of satisfac-
tion and wellbeing. For example, Bigelow et al. (1991) view quality of life
as an equation that balances the fulfilment of needs with the meeting of
demands, and that involves the subjective assessment of wellbeing together
with objective components, such as social functioning. Measures of quality
of life in the field of mental retardation have also been constructed to take
account of both objective and subjective facets (Schalock, Keith, &
Hoffman, 1990). This formulation is sensitive to the individual's own view
of his or her life circumstances, while maintaining an independent perspec-
tive on those circumstances.
Quality of life has also been defined "as the satisfaction of an individu-
al's values, goals and needs through the actualisation of their abilities or
lifestyle" (Emerson, 1985, p. 282). This definition is consistent with the
concepmalisation that satisfaction and wellbeing stem from the degree of
fit between an individual's perception of their objective situation and their
needs or aspirations (Andrews & Wlthey, 1976; French, Rogers, & Cobb,
1974). The conceptual model put forward by Campbell et al. (1976) exem-
plifies this approach. Individuals judge their objective situation in each of
various life domains according to standards of comparison based on aspira-
tions, expectations, feelings of what would be just, reference group com-
parisons, personal needs, and personal values. Their resulting evaluation is
their satisfaction with that domain. Domain satisfactions combine to pro-
duce a general sense of weUbeing. Inwaindividual mediators between expe-
rience of life and satisfaction with life are also included in models address-
ing mental health or disability issues. Baker and Intagliata (1982) refer to
need levels, beliefs, and attitudes as significant components of the bio-psy-
cho system element of their model. Parmenter (1988) lists beliefs, goals,
values, and aspirations within the self element of his model.
Cummins (1992a) has suggested that quality of life should comprise both
objective and subjective assessment and that the manner of combining sub-
jective assessments across separate life domains to produce an overall
appraisal should take account of the importance the individual places on
the particular aspect being considered. Such a principle could also be
applied to objective assessments: The significance attached to objective life
conditions might also take account of the individual's scale of values. This
formulation gives rise to a three-factor model in which personal values as
well as life conditions and life satisfaction interact to determine quality of
life (Fig. ld). The significance of either the objective or subjective assess-
ment of a particular life domain is interpretable only in relation to the
importance the individual places on it. For example, size of income (the
 Quality of Life 59

objective measure) may contribute little to quality of life for a person

whose values are nonmaterialist, although satisfaction with income (e.g.,
enough to meet personal needs) might still carry a high weight.
Capacity to put an individual weighting on how subjective and objective
assessments are combined across life domains is a strength of this model.
How to combine objective or subjective appraisals across issues of rele-
vance is not specified in the model depicted in Figure lc. Measures lacking
a basis for ranking the importance that different individuals give to various
aspects of their overall situation frequently combine them with equal
weighting (Schalock et al., 1990). Alternatively, assumptions about the rel-
ative importance of different concerns may be used to generate a weighting
structure (Wolfensberger & Glenn, 1975). In both cases, the method of
combining items in the overall assessment can be criticized as arbitrary.
The use of personal assessments of importance to weight the significance
of objective and subjective indices also deals well with the concerns
expressed by Edgerton (1990) that only individuals can decide the trade-off
between competing aspects of their own personal welfare. Similarly, family
goals are central to the model of family functioning put forward by
Landesman, Jaccard, and Gunderson (1991). Moreover, as long as a similar
model is applied to the assessment of quality of life among the general pop-
ulation, the position of individuals or groups can be judged by comparison
to a total population distribution.
Indeed, judging whether a reasonable level of quality of life exists is a
separate issue to its measurement. Life conditions and satisfaction with life
will inevitably vary, and neither ideal conditions nor perfect satisfaction
can be arranged for or achieved by every member of a society or societal
subgroup. Aggregated data for a defined group of interest may be com-
pared to those for the total population to establish whether aspects of quali-
ty of life are similarly distributed or narrowly clustered either in their
favour or to their disadvantage. In other words, cultural norms and ranges
may provide a standard of reference. A view that an acceptable quality of
life had been achieved could require that both expressed satisfaction with
various aspects of life and objective descriptors of those aspects were in
keeping with, or at least not inferior to, the typical for the society as a
whole. (This argument should not be mistaken as indicating that particular
interested parties, such as family members, program professionals, or advo-
cates, with or without disabilities, constitute a sufficient or more significant
external reference group for making quality of life judgements.)

Agreement on Relevant Life Domains

Despite disagreement on the definition of quality of life, there is consid-
erable overlap among researchers on relevant domains for assessment.
60 D. Felce and Z Perry

Fifteen key literature sources that describe conceptual models or opera-

tionalisations of quality of life have been reviewed (Fig. 2 footnote). The
majority of aspects mentioned in these sources can be grouped under five
domain headings (Fig. 2). The figures in parentheses in the following para-
graph give the number of the 15 sources that include consideration of the
aspect discussed, either in the precise terms adopted in Figure 2 (e.g., mate-
rial wellbeing, interpersonal relationships) or via a synonym (e.g., standard
of living, social relations).
Physical wellbeing is a main heading in the classification put forward by
Bhmden (1988) and subsumes health (8), fitness (3), and physical safety
(4). Material weUbeing is again taken from Blunden (1988) and subsumes
finance or income (9), quality of the living environment (12), and privacy,
possessions, meals or food, transport, neighbourhood, security, and stability
or tenure (variously 3-7). Social wellbeing includes two major dimensions.
A strong concern for the quality and breadth of interpersonal relationships
is reflected in the relationships a person has within their family or house-
hold life (9) and those with relatives in the extended family or with more
general friends and acquaintances (14). Community activities (12) and the
level of community acceptance or support (4) together reflect a similarly
strong concern for community involvement. Development and activity is
concerned with the possession and use of skills in relation to both self-
determination - - competence or independence (10) and choice or control
(7) - - and the pursuit of functional activities - - work (10), leisure (10),
housework (4), education (4), and productivity or contribution (5).
Emotional wellbeing, affect or mood, satisfaction, or fulfilment was men-
tioned in 12 of the 15 sources with self-esteem, status/respect, and religious
faith also being cited in between 2 and 6.
Although the categorization of domains in Figure 2 is based on a sub-
stantial overlap between sources, it is important to emphasize that the num-
ber and nature of quality of life domains depicted have not been empirical-
ly derived, but pragmatically specified by us, as a means of classifying the
content encompassed by the literature. There seems to be considerable
agreement that quality of life is a multidimensional concept, but, beyond
that, the merits of particular formulations are not known.

An Overall Model of Quality of Life

Figure 3 brings together our categorisation of quality of life domain
areas with the three-element model of quality of life discussed earlier and
depicted in Figure ld. Quality of life is defined as an overall general well-
being that comprises objective descriptors and subjective evaluations of
physical, material, social, and emotional wellbeing together with the extent
of personal development and purposeful activity, all weighted by a personal
 FIGURE 2. Domains relevant to quality of life. Based on: Andrews & Withey (1976); Baker & lntagUata (1982); Bigelow et ai.
(1991); Blunden (1988); Borthwtck-Duffy et al. (1992); Brown & Bayer (1992); Campbell et al. (1976); Cnmmins (1992a); Flanagan
(1978); Franklin, Simmons, Soiovitz, Ciemons, & Miller (1986); Heal & Chadsey-Rusch (1985); O'Brien (1987); Parmenter (1988);
Schalock et al. (1990); Stark and Goldsbury (1990).
62 D. Felce and J. Perry

Quality of Life

I ] Objective LifeConditions
Physical Wellbeing
Material Wellbeing
Social Wellbeing
r Developmentand Activity
Emotional WeUbeing
x
t
e ,/
t/ 1 "'.'~
r
n /
t
s
SubjectiveFeelingofWeHbeing
a i
ct Satisfaction with:
1 / Physical Wellbeing
/_ . . . . . . . . Material Wellbeing
I Social Wellbeing
n Developmentand Activity
f Emotional Wellbeing
1
n
e
n
c
PersonalValuesandAspirations
e Importance of:
s Physical Wellbeing
Material Wellbeing
Social Wellbeing
Developmentand Activity
Emotional Wellbeing

OverallAssessmentofweilbeing
(Quality of Life)

FIGURE 3. A model of quality of fife.

set o f values. The three elements are shown in dynamic interaction with
each other. Changes in some objective facet of life may change satisfaction
or one's personal values or both. Similarly, changes in values may change
satisfaction and precipitate change in some objective circumstance. In the
same way, a change in a sense of satisfaction may lead to reappraisal of
values and lifestyle.
As well as affecting each other, the three elements are capable of chang-
ing independently as a result of external influences. Such external influ-
 Quality of Life 63

ences might include genetic, social, and material inheritance, age and matu-
ration, developmental history, employment, peer influences and reference
points, and other social, economic, and political variables. As the three ele-
ments that define quality of life are all open to external influence, assess-
ment of all three is necessary to any measurement system purporting to
capture quality of life. Knowledge of one set cannot predict another
because the relationships between them may not remain constant.

A Research Agenda
Reaching a valid definition and measurable operationalisation of the qual-
ity of life construct is a central goal. A sustained research effort in three
related areas is required before the field can establish the validity of such a
multidimensional construct. These involve: (a) the refinement of a measure-
ment technology for the assessment of objective aspects of wellbeing, per-
sonal development, and activity; (b) the refinement of a methodology for
gaining subjective appraisals of a broad range of personal satisfaction con-
ceres and of the importance an individual places on various lifestyle issues;
and (e) research on how to establish whether individuals with conspicuous
cognitive limitations and communication deficits consequent on their mental
retardation have sufficient receptive and expressive language to discriminate
what is being asked and to give answers that reflect their feelings accurately.
Allied to this latter effort, for people with mental retardation who clearly
cannot speak for themselves, is the need to investigate whether a valid rep-
resentation of individual concerns and values can be given by anyone other
than the subject, or whether only a restricted assessment of quality of life in
terms of objective indicators is possible.

Refining objective indicators. Many measures relevant to the five domain

areas exist. Borthwick-Duffy (1992) has suggested that multiple methods
and informants should be used to determine what relationships exist among
variables and whether they provide the basis for a meaningful and valid
construct. Our own recent research has followed from a similar idea.
Because existing evaluative research has largely restricted the number of
measures used to a single approach per aspect evaluated, there is remark-
ably little information on agreement across different measurement meth-
ods. We are, therefore, currently engaged in a project that is using several
alternative methods for measuring each of a number of quality aspects in
staffed housing services (Perry & Felce, in press). Ratings are concerned
with material and social wellbeing, personal development, and activity.
Analysis of agreement and disagreement between ostensibly similar mea-
sures may help to refine measures and lead to general acceptance of the
utility and properties of the resulting methods.
64 D. Felce and J. Perry

Establishing the social validity of a taxonomy based on the proposed five

life domains is an equally important contribution to refining an agreed
measurement technology. The adequacy of the scope of the items covered
in Figure 2 needs to be examined by research that establishes which indica-
tors of quality of life are frequently mentioned by the general public. The
framework may also be checked against the particular concerns and opin-
ions of those people with mental retardation able to speak for themselves.
Again, relevant work has already been accomplished. For example, the
Audit Commission and the Social Services Inspectorate Wales (in prepara-
tion) has identified 42 imperatives distilled from a variety of consultation
exercises with users in Wales. These cover many of the life domain areas
described earlier: issues of self-determination, choice, and control (11
items); access to community activities and a wider social life (6 items);
opportunities for productivity, contribution, and recognition (9 items); and
issues concerning social equity, civic fights, and respectful treatment, espe-
ciaily by staff (12 items). (The remaining four items referred to being given
help and information.)

Assessment of satisfaction and personal priorities. Initially, focus for the

measurement of satisfaction was derived from a relatively restricted assign-
ment to gain the views of service users about changes to residential ser-
vices (Scheerenberger & Felsenthal, 1977; Seltzer, 1981). More recently,
scales have been developed to tap subjects' satisfaction with their lifestyles
more generally. However, current examples still only cover a proportion of
the domain issues highlighted in Figure 2. The Lifestyle Satisfaction Scale
(Heal & Chadsey-Rusch, 1985) contains 29 yes-no items and addresses sat-
isfaction with residence, living arrangements, neighbourhood, friends, the
use of leisure time, and generic or professional services. The satisfaction
section of the Quality of Life Questionnaire (Schaiock et al., 1990) has 10
items rated on a three-point scale: an overall view on life, how much enjoy-
ment respondents derive from it, how well off they are compared to others,
whether most events or activities are rewarding or not, their satisfaction
with their living arrangements, how well they are treated by neighbours,
whether their education prepared them for what they are currently doing,
the extent of their problems, whether they feel lonely, and whether they
feel out of place in social situations. Subjective appraisal of such concerns
as health, emotional wellbeing, material circumstances, productivity, con-
tribution, and self-determination are relatively poorly represented.
Further research is also required on such methodological issues as ques-
tionnaire length, questioning style, and response format. Presumably for
the purposes of keeping scales short, issues that are addressed in the exam-
pies mentioned here are done so by reference to one or at most a few items.
However, brevity produces generality if the complexity of such major life
 Quality of Life 65

domains as homelife, social life, or work is not to be oversimplified. This

makes greater demands on the language abilities of respondents, requiring
them both to respond to abstract terms, such as the extent of their problems,
and to collate experience relevant to the issue in question and distil a sum-
mary position. Such short scales appear, therefore, to ignore the advice of
Wyngaarden (1981) to adopt a strategy of avoiding abstract questions.
Although individuals may answer abstract questions, their answers may not
carry a full understanding or appraisal of what is germane to the abstrac-
tion. It would be useful to conduct research in which the answers to a short
list of general questions were compared to those answers individuals might
make to a more detailed list of more specific and tangible concerns.
Scales also typically employ simple, structured answer formats to facili-
tate ease of response. Levine (1985) found that people with mild mental
retardation had difficulty in responding to a four-point Likert scale; they
tended to confuse the two intermediate responses. He therefore simplified
the scale to one with only three points. Whereas simple yes-no or three-
point structured answers may promote the ability of people with communi-
cation deficits to respond at all, there are disadvantages to such designs in
that respondents may have a tendency to acquiesce (Sigelman, Budd,
Spanhel, & Sehoenrock, 1981) or to choose the last of a possible range of
responses offered (Sigelman, Budd, Winer, Schoenrock, & Martin, 1982).
Shaw and Budd (1982) have indicated that acquiescence may be related to
cognitive disability, and this, linked with perceived social desirability, may
produce inflated reporting of satisfaction. The use of pictures or simple
drawings to denote responses does not reduce the acquiescence problem.
Moreover, the adoption of simple response formats underlines the doubts
expressed earlier about complex general or abstract questions. The two
seem inconsistent with each other.
Qualitative interviewing approaches (Atkinson & Ward, 1987; Flynn,
1989), guided by topic lists but otherwise allowing an open-ended dialogue,
have possible advantages in being less prescriptive of the manner of
response and by avoiding providing "model" answers. Prescriptions for
good interviewing practice in relation to people with mental retardation
(Flynn, 1986) show much in common with guidelines for such qualitative
research in general (Walker, 1985): emphasising clear, simple questions,
patience, and sensitivity on the part of the interviewer; allowance of ade-
quate time for the respondent to answer; and the desirability of checking the
veracity of responses against other sources of data. Research is needed on
how to construct quantitative assessments of satisfaction from information
gained from qualitative interviewing, or at least, on how to take account of
more detailed qualitative accounts alongside the use of quantitative scales.
However, caution about the veracity of answers given in open-ended dia-
logue is also indicated by the research of Sigelman and her colleagues
66 D. Felce and J. Perry

(Sigelman, Schoenrock et al., 1981). While recognising that closed ques-

tioning creates certain validity problems, they concluded that open-ended
questions appear to reduce responsiveness and create validity problems of
their own (see later). Moreover, methods for controlling interviewer bias or
influence in qualitative interviewing are also required, particularly among
people susceptible to acquiescence.
Holland's (1990) investigation of service user views confirmed the need
to know more about the effects of interview format and style. Given a
structured interview with a simple response format, her sample reported
almost uniformly high satisfaction irrespective of issue addressed or the
considerable differences in residential situation, from small house to large
traditional facility. However, a more detailed follow-up of a subsample
using qualitative interviewing over several meetings showed that residents
could be dissatisfied about major aspects of their situations, such as not liking
the other people with whom they lived. Such views were not necessarily
expressed in the first interchange; concerns tended to emerge over succes-
sive interviews as the relationship between researcher and respondent was
said to grow. A closer investigation of the interplay between elicited infor-
mation and the nature of the relationship established between the inter-
viewer and interviewee is required. In addition, whether and how specific
instances of dissatisfaction are reflected in global assessments of satisfac-
tion requires greater understanding.
Ascertaining individuals' personal values, that is, reaching a view of how
important a particular issue is to a person within the wider scheme of
things, may call for even greater cognitive abilities than the expression of
satisfaction, because the former involves ranking of relative values and the
latter merely the assignment of a value to a single issue. How difficult a
task this is and the methodological development required are subjects for
further investigation. Cummins (1992a) has constructed a scale that asks
respondents not only to indicate their satisfaction in a number of life
domains but also to rank the degree to which the life domain issue is
important to them. He uses visual analogue Likert scales for both purposes
and has embedded practical tests within the assessment protocol to check
that the respondent's discriminative ability is on a par with the complexity
of the scales used.

From whom is it possible to gain personal views? There is a tremendous

range in cognitive and communicative competency among people labelled
as having mental retardation. At one end of the spectrum are people who
have virtually no understanding of spoken or signed language and no
means of expression. At the other are people who have language approach-
ing that of people not so labelled. Much is known about the interaction
between degree of mental retardation and the level of a person's language
 Quality of Life 67

ability from research on receptive and expressive language and the devel-
opment of adaptive behavior scale norms. What this research implies for
the content and format of questioning will enable the field to determine
what proportion of people can register what complexity of opinion under
what conditions.
Detailed qualitative accounts have been obtained from people with mental
retardation, most of whom were able enough to live largely independently
(Edgerton et al., 1984; Flynn, 1989). Information is also available on respon-
dents in quantitative studies. The abilities of participants in the study by Heal
and Chadsey-Rusch (1985) were described in terms of intelligence quotients
taken from people's files, one group ranging from 22 to 72 with a mean of 47
and the other from 47 to 97 with a mean of 64. These data are unsatisfactory,
but, if taken at face value, they do illustrate the tendency for respondents in
these studies to have moderate or mild mental retardation, or even intelli-
gence outside the range for inclusion. Even so, acquiescence bias, which was
deliberately accounted for within their study, was significantly above zero. It
is not possible to see the extent to which acquiescence was attributable to
those with moderate and severe disabilities within the sample.
The series of studies summarised in Sigelman, Schoenroek et al. (1981)
addressed four questions: (a) to what extent can people with mental retar-
dation respond to questions, regardless of truth or accuracy; (b) how con-
sistent are such answers over time; (c) how valid and free of systematic
bias are such answers; and (d) what types of question optimise responsive-
ness, reliability, and validity? Questions used a variety of yes-no, picture
choice, either-or, verbal multiple choice, and open-ended formats. They
found that some people with severe mental retardation and the great major-
ity of people with profound mental retardation were unable to respond at
all. The proportion of questions that children and adults with severe, mod-
erate, or mild mental retardation answered were, respectively: 46.1% and
58.5%, 84.8% and 78.0%, and 90.9% and 73.5%. Unpredictability of
responsiveness increased with severity of disability. However, higher rates
of responding were obtained among children with severe mental retardation
in a later study (78.7%) by using repetition to facilitate understanding and
by restricting question types to those most likely to be answered (picture
choice and yes-no formats).
Obtaining an answer is a necessary but not sufficient condition of sound
interviewing. The Sigelman group's second and third questions (Sigelman,
Schoertrock et al., 1981) related to consistency and validity of answers.
They found that, for those who could respond, 85.2% of answers were con-
sistent across similar repeated questions requiring yes-no responses, and
71.9% were consistent across repeated three- or four-answer multiple
choice formats. Only 46.2% consistency was found for a four-choice pic-
ture format (smiting or frowning faces) to indicate how respondents felt
68 D. Felce and J. Perry

about aspects of residential life. Eight open-ended questions were also

investigated. Roughly two-thirds responded to four factual questions about
themselves (e.g., name, birthday) but achieved only 63.3%, 42.9%, 63.6%,
and 61.9% consistency across repeated answers to each. The remaining
four were factual but required descriptive answers (e.g., variety of social or
leisure activities undertaken). Consistency was checked in terms of
whether possible categories of activity were or were not mentioned in suc-
cessive interviews. High consistency (83.9%) was found overall, but this
was mainly attributable to possibilities not being mentioned. Consistency
about what did occur was found to be much lower, for example, only
18.1% of respondents consistently mentioned an activity undertaken both
times. The researchers concluded that open-ended questions calling for
descriptive listing of activities or experiences yield little information, because
many people with mental retardation are unable to respond at all, and those
who do say very little.
Validity of answers was examined by comparing responses across alter-
natively worded questions designed to elicit similar information and by
gaining reports from third parties, neither of which are completely satisfac-
tory approaches to checking veracity. As previously noted, their research
revealed that the tendency to acquiesce is a significant problem. Across a
variety of questions, about half of the samples studied answered yes where
the answer no was indicated either from the answer to an oppositely word-
ed question or from known fact. Problems of acquiescence were found to
be significantly inversely associated with intelligence, indicating that the
majority of people with severe mental retardation will provide affirmative
answers irrespective of question content. Agreement between respondents
and care staff was found to be mixed, with some good agreement on yes-no
questions about commonly occurring activities (e.g., watching television)
but poor agreement on yes-no questions relating to less commonly per-
formed activities or on questions using multiple-choice formats. Agreement
was rare between respondent and staff answers to open-ended questions.
The f'mdings noted here illustrate the difficulties that confront those wish-
ing to gain valid and reliable information from people with mental retarda-
tion about complex issues. Research, whether qualitative or quantitative,
must not only report results but also the checks incorporated within it to
check for consistency and validity. Cummins (1992a) has constructed practi-
cal tests to ascertain whether respondents can make size or order discrimina-
tions that match the level of complexity of the Likert scales used. For exam-
pie, for a five-point scale, he ascertains whether an individual can order five
different sized blocks by length. Failure to do so would bring into question
the person's response to a pictorial scale on the basis that the more concrete
task is easier than one that involves symbolic interpretation. He then pro-
ceeds to test for use of an abstract scale against known preferences. This
 Quality of Life 69

seems a line of methodological rigour well worth further development.

Cummins (1992b) reports preliminary indications from a study specifically
addressing the level of intellectual functioning required to use his scale that
the dividing line between those capable of responding from those not
appears to be about the middle of the moderate level of mental retardation.
The lack of language ability among a substantial percentage of people
with mental retardation means that it will not be possible to get accurate
views from them on their satisfaction with life or on the importance they
assign to various life domain issues. It may prove that other people can
accurately reflect an individual's opinions and feelings, or it may be the
case that quality of life assessment in such situations must be restricted to
objectively measurable phenomena, interpreted via norms and ranges relat-
ing to the general population. In the absence of being able to gain informa-
tion directly from the person with mental retardation, the perspectives of oth-
ers have been taken, principally of the most immediate family or other caters
(Schalock et al., 1990). However, the existing literature on how well third
parties represent the views of individuals is not encouraging (Cummins,
1992a). On the assumption that some third parties might do so better than
others, Cummins (1992a) has suggested that research could address the
question of whether there are certain characteristics of third parties that
might discriminate whether they were a good source of testimony. In rela-
tion to people with mental retardation who can speak for themselves reli-
ably, such research could measure the personal characteristics of third
parties and establish which predicted a good match between third party
views and those of the individual themselves. Characteristics explored
could include relationship to the person (e.g., relative, friend, service work-
er), role in relation to the person (e.g., cater, peer, advocate), gender (same,
different), age (e.g., same age, 10 years different, 20 years different), length
of association with the person (e.g., all individual's life through to recent
meeting), frequency of current contact (e.g., frequent, moderate, occasion-
al), and extent of current shared experience (e.g., with or without mental
retardation, user of same services or not, similar activity pursuits or not).

CONCLUSION
In conclusion, a model of quality of life is proposed that integrates objec-
tive and subjective indicators, collectively reflecting a broad range of life
domains, through an individual ranking of the relative importance of each
domain. This model accommodates both concerns that objective data
should not be interpreted without reference to personal autonomy and pref-
erences and concerns that expressions of satisfaction are themselves rela-
tive to the individual's temperament and the circumstances and experiences
that have shaped their frame of reference. Although in the proposed model
70 D. Felce and J. Perry

individuals may weight certain life conditions as less important than others,
the objective assessment of those life conditions is still a part of the overall
assessment and may be compared to those typically experienced. Consider-
able agreement among writers on this topic in the disability field exists on
bow to categorise life domains relevant to the conceptualisation of qual-
ity of life.
The conceptualisation offered is based on literature concerning the mea-
surement of quality of life in the general population as well as for disability
groups. Cummins (1992b) has argued that one of the characteristics of a
good quality of life measure is that it will enable comparison between
groups of people irrespective of their precise characteristics. In particular,
measures of quality of life for people without disabilities should be equally
applicable to people with disabilities, and vice versa, thus allowing com-
parison of the quality of life of people with disabilities to that of the gener-
al population. The five domains used in this article to categorise the content
areas relevant to quality of life are all of a general nature, and this is desir-
able from the perspective of developing a quality of life measurement sys-
tem with broad utility. However, it would be wise not to assume that such
broad utility follows simply from the seeming general applicability of
items to a wide range of people. Rather, the development of quality of life
measures for people with disabilities should go hand in hand with and par-
allel measures of quality of life for the general population (see Cummins,
1992a, 1992c).
The development of a common approach to quality of life applicable
across societal groups and to the total population is vital if information
regarding one section of society is to be interpreted with confidence.
Problems in inferring what alternative quality of life might be reasonable in
the individual case are compounded by the fact that life conditions and sat-
isfaction with life inevitably vary across individuals in all groups within
society. However, aggregated data for a defined group of interest, com-
pared to the statistical distribution for the population as a whole, can be
used to reflect whether life conditions and satisfaction in various domains
are typical of the general pattern or have a significantly different profile.
Social policy may then respond to conspicuous inequality.
Finally, we have outlined some of the research steps that will contribute
to the proposed multielement approach to the quality of life construct.
Although we agree that continued conceptual and methodological develop-
ment is important, we also recognise that such is the extent of social
inequality, inferior opportunities, and disadvantaged circumstances experi-
enced by people with mental retardation that existing research tools may
very well adequately assess whether success accompanies our efforts to
bring about beneficial change. Perhaps the most important implication of
the current state of conceptua/isation of quality of life is that it sets a broad
 Quality of Life 71

agenda. Research and service development targeted on a narrowly defined

range of outcomes will inevitably fail to improve other issues of consider-
able significance in people's lives. The strength of that significance is a
matter of individual weighting, a fact that directs researchers and service
developers toward an even closer alliance with people with disabilities in
order to respond to their interests, preferences, and concerns. However,
cognitive and language limitations are real. A rational approach needs to be
taken to the gaining of user views and, where this is not possible, to the
investigation of the validity of third party opinions. The interpretation of
objective data on quality of life against societal norms may provide a sound,
if limited, approach in many cases.

REFERENCES
Allen, H. M., Bender, P. M., & Gntek, B. A. (1985). Probing theories of individual wellbeing:
A comparison of quality-of-life models assessing neighbourhood satisfaction. Basic and
Applied Social Psychology, 6, 181-203.
Andrews, E M., & Withey, S. B. (1976). Social indicators ofweUbeing: Americans' perceptions
of life quality. New York: Penum Press.
Aninger, M., & Bolinsky, K. (1977). Levels of independent functioning of retarded adults in
apartments. Mental Retardation, 15, 12-13.
Atkin~n, D., & Ward, L. (1987). Friends and neighbours: Relationships and opportunities in the
community for people with a mental handicap. In N. Malin (Ed.), Reassessing community care
(pp. 232-248). London: Croom Helm.
Audit Commission and Social Services Inspectorate Wales. (in prep). Learning difficulties, quaff-
ty and the All Wales Strategy. Cardiff: Welsh Office.
Baker, E, & Intagliata, J. (1982). Quality of life in the evaluation of community support systems.
Evaluation and Program Planning, 5, 69-79.
Bigelow, D. A., McFarland, B. H., & Olson, M. M. (1991). Quality of life of community mental
health program clients: Validating a measure. Community Mental Health Journal, 27, 43-55.
Blunden, R. (1988). Quality of life in persons with disabilities: Issues in the development of services.
In R. I. Brown (Ed.), Quality of lifefor handicappedpeople (pp. 37-55). London: Croom Helm.
Borthwick-Duffy, S. A. (1992). Quality of life and quality of care in mental retardation.
In L. Rowitz (Ed.), Mental retardation in the year 2000 (pp. 52-66). Berlin: Springcr-Verlag.
Borthwick-Duffy, S. A., Widaman, K. E, Little, T. D., & Eyman, R. K. (1992). Foster family care
for persons with mental retardation. Washington DC: American Association on Mental
Retardation.
Bradbum, N. M. (1969). The structure of psychological wellbeing. Chicago: Aldine.
Brown, R. I. (1988). Quality of life for handicapped people. London: Croom Helm.
Brown, R. I., & Bayer, M. B. (1992). Rehabilitation questionnaire manual: A personal guide to
the individual's quality of life. Ontario: Captns Press.
Brown, R. I., Bayer, M. B., & MacFarlane, C. M. (1989). Rehabilitation programmes: The per-
formance and quality of life of adults with developmental handicaps. Toronto: Lugus
Productions Ltd.
Burchard, S. N., Hasazi, J. S., Gordon, L. R., & Yoe, J. (1991). An examination of lifestyles and
adjustment in three community residential alternatives. Research in Developmental Disabilities,
12, 127-142.
Campbell, A. (1981). The sense ofwellbeing in America. New York: McGraw Hill.
Campbell, A., Converse, P. E., & Rodgers, W. L. (1976). The quality of American life:
Perceptions, evaluation and satisfactions. New York: Russell Sage Foundation.
72 D. Felce and J. Perry

Close, D. W., & Halpem, A. S. (1988). Transitions to supported living. In M. P. Janicki, M. W.

Krauss, & M. M. Seltzer (Eds.), Community residences for persons with developmental dis-
abilities: Here to stay (pp. 159-171). Baltimore: Paul H. Brookes Publishing Co.
Conroy, J. W., & Bradley, V. J. (1985). The Pennhurst Longitudinal Study: A report offive years
research and analysis. Philadelphia: Temple University Developmental Disabilities Center.
Costa, R T., Jr., McCrae, R. R., & Zonderman, A. B. (1987). Environmental and dispositional
influences on wellbeing: Longitudinal follow-up of an American national sample. British
Journal of Psychology, 78, 299-306.
Cummins, R. A. (1992a). Comprehensive quality of life scale -- intellectual disability (3rd ed.).
Melbourne: Psychology Research Centre.
Cnmmins, R. A. (1992b). Personal communication.
Cummins, R. A. (1992c). Comprehensive quality of life scale (3rd ed.). Melbourne: Psychology
Research Centre.
Cummins, R. A., McCabe, M. e., Gullone, E., & Romeo, Y. (1994). The comprehensive quality of
life scale: Instrument development and psychometric evaluation on college staff and students.
Educational and Psychological Measurement, 54, 372-382.
de Kock U., Saxby, H., Thomas, M., & Felce, D. (1988). Community and family contact: An eval-
uation of small community homes for adults. Mental Handicap Research, 1, 127-140.
Department of Health and Social Security (DHSS). (1971). Better services for the mentally handi-
capped. London: HMSO.
Edgerton, R. B. (1990). Quality of life from a longitudinal research perspective. In R. L. Schalock
(Ed.), Quality of life: Perspectives and issues (pp. 149-160). Washington DC: American
Association on Mental Retardation.
Edgerton, R. B., Bollinger, M., & Herr, B. (1984). The cloak of competence: After two decades.
American Journal of Mental Deficiency, 88, 345-351.
Emerson, E. B. (1985). Evaluating the impact of deinstitutionalisation on the lives of mentally
retarded people. American Journal of Mental Deficiency, 90, 277-288.
Felce, D. (1989). Staffed housing for adults with severe and profound mental handicaps:
The Andover Project. Kidderminster: BIMH Publications.
Felce, D., de KocL U., Thomas, M., & Saxby, H. (1986). Change in adaptive behaviour of severe-
ly and profoundly mentally handiapped adults in different residential settings. British Journal
of Psychology, 77, 489-501.
Felce, D., Kushlick, A., & Smith, J. (1980). An overview of the research on alternative residential
facilities for the severely mentally handicapped in Wessex. Advance in Behaviour Research
and Therapy, 3, 1-4.
F!anagan, J. C. (1978). A research approach to improving our quality of life. American Psychologist,
33, 138-147.
Flax, M. J. (1972). A study in comparative urban indicators: Conditions in 18 large metropolitan
areas. Washington DC: The Urban Institute.
Flynn, M. (1986). Adults who are mentally handicapped as consumers: Issues and guidelines for
interviewing. Journal of Mental Deficiency Research, 30, 369-377.
Fly[m, M. (1989). Independent living for adults with a mental handicap: A place of my own.
London: Cassel.
Franklin, J. L., Simmons, J., Solovitz, B., Clemens, J. R., & Miller, G. E (1986). Assessing quality
of life of the mentally ill: A three-dimensioual model. Evaluation and the Health Professions, 9,
376-388.
French, J. R. P., Rogers, W., & Cobb, S. (1974). Adjustment as person-environment fit. In G. V.
Coelho, D. A. Hamburg, & J. E. Adams (Eds.), Coping and adaption (pp. 316-333). New
York: Basic Books.
Goode, D. (1988). Quality of life: A review and synthesis of the literature. Valhalla, NY:
The Mental Retardation Institute.
Heal, L. W., & Chadsey-Rusch, J. (1985). The Lifestyle Satisfaction Scale (LSS): Assessing indi-
viduals' satisfaction with residence, community setting and associated services. Applied
Research in Mental Retardation, 6, 475-490.
 Quality o f Life 73

Hemming, H., Lavender, T., & Pill, R. (1981). Quality of life of mentally retarded adults trans-
ferred from large institutions to new small units. American Journal of Mental Deficiency, 86,
157-169.
Holland, A. (1990). People with mental retardation living in community homes: Their views and
the quality of the service. Unpublished doctoral thesis. University of Loughborough.
Keogh, B. K., & Pullis, M. E. (1980). Temperament influences on the development of exceptional
children. In B. K. Keogh (Ed.), Advances in special education (Vol 1, pp. 112-121).
Greenwich, CT: JAI Press.
King, R., Raynes, N., & Tizard, J. (1971). Patterns of residential care. London: Routledge and
Kegan-Paul.
Landesman, S. (1986). Quality of life and personal life satisfaction: Definition and measurement
issues. Mental Retardation, 24, 141-143.
Landesman, S., Jaccard, J., & Gunderson, V. (1991). The family environment: The combined
influence of family behaviour, goals, strategies, resources and individual experiences.
In M. Lewis & S. Feinman (F_As.),Social influences and socialization in infancy (pp. 63-96).
New York: Plenum.
Lazarus, R., & Lannier, R. (1979). Stress related transactions between person and environment. In
L. Pervin, & M. Lewis (Eds.), Perspectives in international psychology (pp. 142-156). New
York: Plenum.
Lehman, A. E (1988). A quality of life interview for the chronically mentally ill. Evaluation and
Program Planning, 11, 51-62.
Levine, H. G. (1985). Situational anxiety and everyday life experiences of mildly mentally retard-
ed adults. American Journal of Mental Deficiency, 90, 27-33.
Liu, B. C. (1976). Quality of life indicators in U.S. metropolitan areas: A statistical analysis. New
York: Praeger Publishers.
I~we, K., & de Paiva, S. (1991). NIMROD: An overview. London: HMSO.
Lowe, K., de Paiva, S., & Felce, D. (1993). Effects of a community-based service on adaptive and
maladaptive behaviours: A longitudinal study. Journal oflnteUectaal Disability Research, 37,
3-22.
Mansell, J., Jenkins, J., Felce, D., & de Kock, U. (1984). Measuring the activity of severely and
profoundly mentally handicapped adults in ordinary housing. Behaviour Research and
Therapy, 22, 23-29.
McLain, R. E., Silverstein, A. B., Hubbell, M., & Browulee, L. (1977). Comparison of the resi-
dential environment of a state hospital for retarded clients with those of various types of corn-
munity facilities. Journal of Coramanity Psychology, 5, 282-289.
Moos, R. (1974). The Social Climate Scales: An overview. Palo Alto: Community Psychologists
Press.
O'Brien, J. (1987). A guide to life-style planning. In B. Wilcox & G. T. Bellamy (Eds.), The
Activities Catalog: An alternative curriculum for youth and adults with severe disabilities
(pp. 175-189). Baltimore: Paul H. Brookes Publishing Co.
Parmenter, T. R. (1988). An analysis of the dimensions of quality of life for people with physical
disabilities. In R. I. Brown (Ed.), Quality of life for handicapped people (pp. 7-36). London:
Croom Helm.
Perry, J., & Felce, D. (in press). Objective assessments of quality of life, how much do they agree
with each other.'?Journal of Community and Applied Social Psychology.
Risley, T. R., & Cataldo, M. E (1973). Planned Activity Check: Materials for training observers.
Lawrence, KS: Center for Applied Behavior Analysis.
Romer, D., & HeUer, T. (1983). Social adaptation of mentally retarded adults in community set-
tings: A social-ecological approach. Applied Research in Mental Retardation, 4, 303-314.
Sackett, G. P., & Landesman-Dwyer, S. (1977). Toward an ethology of mental retardation.
In P. Mittler (Ed.), Research to practice in mental retardation: Vol II. Education and training
(pp. 27-37). Baltimore: University Park Press.
Schalock, R. L. (1990). Quality of life: Perspectives and issues. Washington DC: American
Association on Mental Retardation.
74 D. Felce and J. Perry

Schalock, R. L., Keith, K. D., & Hoffman, K. (1990). Quality of life questionnaire:
Standardization manual. Hastings NE: Mid-Nebraska Mental Retardation Services.
Scheerenberger, R. C., & Felsenthal, D. (1977). Community settings for MR persons: Satisfaction
and activities. Mental Retardation, 15, 3-7.
Schneider, M. (1976). The "quality of life" and social indicators research. Public Administration
Rev/ew, 36, 297-305.
Schroeder, S., & Henes, C. (1978). Assessment of progress of institutionalized and deinstitutional-
ized retarded adults: A matched-control comparison. Mental Retardation, 16, 147-148.
Seltzer, G. B. (1981). Community residential adjustment: The relationship among environment,
performance and satisfaction. American Journal of Mental Deficiency, 85, 624-630.
Shaw, J. A., & Budd, E. C. (1982). Determinants of acquiescence and nay saying of mentally
retarded persons. American Journal of Mental Deficiency, 87, 108-110.
Sigehnan, C. K., Budd, E. C., Spanhel, C. L., & Schoenrock, C. J. (1981a). When in doubt day
Yes: Acquiescence in interviews with mentally retarded persons. Mental Retardation, 19,
53-58.
Sigelman, C. K., Budd, E. C. Winer, J. L., Schoenrock, C. J., & Martin, P. W. (1982). Evaluating
alternative techniques of questioning mentally retarded persons. American Journal of Mental
Defic/ency, 85, 511-518.
Sigeiman, C. K., Schoenrock, C. J., Wmer, J. L., Spanhel, C. L., Hromas, S. G., Martin, E W., Budd,
C., & Bensberg, G. J. (1981b). Issues in interviewing mentally retarded persons: An empirical
study. In R. H. Brunmink~ C. E. Meyers, B. B. Sigford, & IC C. ~ (Eds.), Deinst/tut/ona//za-
tion and community adjustment of mentally retardedpersons (pp. 114-129). Washington, DC:
American Association on Mental Deficiency.
Stark, J. A., & Goldsbury, T. (1990). Quality of life from childhood to adulthood. In R. L.
Schalock (Ed.), Quality of life: Perspectives and issues (pp. 71-83). Washington DC:
American Association on Mental Retardation.
T'h~ard, J. (1964). Community services for the mentally handicapped. London: Oxford University
Press.
Walker, R. (1985). Applied qualitative research. Aldershot: Gower.
Welsh Office. (1983). All Wales strategy for the development of services for mentally handi-
capped people. Cardiff: Welsh Office.
Wolfensberger, W., & Glenn, L. (1975). Program analysis of service systems: Handbook and
manua/(3rd ed.). Toronto: National Institute on Mental Retardation.
Wyngaarden, M. (1981). Interviewing mentally retarded persons: Issues and strategies. In R. H.
Bmnininks, C. E. Meyers, B. B. Sigford, & K. C. Lakin (Eds.), Deinstitutionalization and
community adjustment of mentally retarded people (pp. 107-113). Washington, DC: American
Association on Mental Deficiency.