Issues in the Conceptualization and Measurement of Participation: An Overview

S5

SPECIAL COMMUNICATION

Issues in the Conceptualization and Measurement of Participation: An Overview Marcel P. Dijkers, PhD ABSTRACT. Dijkers MP. Issues in the conceptualization and measurement of participation: an overview. Arch Phys Med Rehabil 2010;91(9 Suppl 1):S5-16. While participation is increasingly defined as the key outcome of rehabilitation, disagreements on and shortcomings in the definition, operationalization, and measurement of this concept abound and interfere with the progress of clinical services and research. This article explores a number of the major issues related to the quantification of participation and makes suggestions for new directions, using the following orienting questions: What is the definition of participation? Where is the border between Participation and Activity? Is there more to participation than performance? What domains should be included in a participation measure? What are the appropriate metrics in quantifying participation? How do we define adequate participation? How should participation be operationalized? What is the proper measurement model for participation instruments? How should we collect data on participation? How do we evaluate the quality of a participation instrument? Key Words: Data collection; Environment; Interpersonal relations; Quality of life; Questionnaires; Rehabilitation; Reproducibility of results; Role; Social adjustment; Social environment; Social isolation; Social support; Validation studies as topic. © 2010 by the American Congress of Rehabilitation Medicine “. . . the ultimate aim of rehabilitation is to maximize a person’s participation in society.”1(p970) ARTICIPATION IS A KEY outcome of rehabilitation P and of other medical and social service programs supporting persons who, because of impairments resulting from injury, birth defect, disorder, or aging, are involved in family, household, community, and society to a lesser degree than they, their service providers, or society may desire. Everyone involved as a clinician or researcher in medical rehabilitation (and other professionals such as those employed in psychiatric rehabilitation, substance abuse rehabilitation, geriatrics, and

developmental disabilities services and research) has some idea what participation is: it is the domain of functioning that is beyond impairment and performance of activities. It is more or less what we mean with instrumental ADLs (also called advanced/extended ADLs), community re-entry or participation, social or role or social role participation, social or societal integration, community living or (re)integration, independent living, normalization, psychosocial functioning or integration, handicap, social health or inclusion or adjustment or disability or disablement, social role valorization, and a number of other terms, varying by professional domain and by whether we want to emphasize a process of normalizing or the resulting status.2,3 Unfortunately, we do not have a standard definition of participation or of any of these other partly overlapping concepts. That is no problem in itself—there probably is limited consensus on many other concepts that are commonly used in rehabilitation. However, the problem is more significant with respect to participation than in relation to other terms that are key to rehabilitation. In addition, participation appears to be a part of the social model of disability, not the medical model, and issues such as the proper relationship of individual to society, biological and social standards for normality, and so forth, play a role in defining and operationalizing the concept. Participation at first blush appears to be a simple concept to measure, but each attempt to construct an instrument needs to address issues in conceptualization and operationalization that get at the core of science epistemology and methodology (eg, value-free measurement) and of metrologic theory and practice (eg, CTT vs clinimetrics). Because of the norming of participation by social roles and cultural values, the potential for developing a single measure that is appropriate across age groups, sexes, socioeconomic classes, and cultures is debatable. The purpose of this article, which is introductory to a special issue devoted to the conceptualization, operationalization, and measurement of participation, is to lay out the major issues involved with these 3 steps leading to quantification of participation. A number of articles published in recent years have pointed out various predicaments in these areas and made

List of Abbreviations From the Department of Rehabilitation Medicine, Mount Sinai School of Medicine, New York, NY. Presented to the American Congress of Rehabilitation Medicine, October 14 –19, 2008, Toronto, ON, Canada. Supported by an honorarium by the Rehabilitation Institute of Chicago using funds received from the National Institute on Disability and Rehabilitation Research (grant no. H133B040032); the National Institute on Disability and Rehabilitation Research (grant nos. H133B040033 and H133N060027), Office of Special Education Services, U.S. Department of Education, to Mount Sinai School of Medicine. No commercial party having a direct financial interest in the results of the research supporting this article has or will confer a benefit on the author or on any organization with which the author is associated. Correspondence to Marcel Dijkers, PhD, Dept of Rehabilitation Medicine, Box 1240, Mount Sinai School of Medicine, One Gustave Levy Place, New York, NY 10029-6574, e-mail: [email protected]. Reprints are not available from the author. 0003-9993/10/9109S-00348$36.00/0 doi:10.1016/j.apmr.2009.10.036

ADLs CHART CIQ CTT DCP EMA ICF IRT NA PDA SCI WHO

activities of daily living Craig Hospital Assessment and Reporting Technique Community Integration Questionnaire classical test theory Disability Creation Process ecological momentary assessment International Classification of Functioning, Disability and Health item response theory not applicable personal digital assistant spinal cord injury World Health Organization

Arch Phys Med Rehabil Vol 91, Suppl 1, September 2010

S6

CONCEPTUALIZATION OF PARTICIPATION, Dijkers

suggestions for getting us out of them.2,4-8 A few articles focused on systematically reviewing participation measures also have useful suggestions for improved measurement.9-14 This article aims to show that the problems at various stages of the process of measurement are connected and offers some suggestions for improved measurement of participation. However, its primary objective is to point out where questions exist and where existing answers should be questioned. Most measures and examples used in the following reflect the issues encountered in conceptualizing and operationalizing the participation of adults; however, these problems all are equally applicable to the case of children. In measuring the participation of children, some additional issues come into play, which are discussed by others.7,12,15 Also, the current review focuses largely on issues in conceptualization and measurement as a topic in medical (physical) rehabilitation. Flynn and Aubry3 offer an informative review of “Integration of persons with developmental or psychiatric disabilities: conceptualization and measurement” that is up to date with developments in those 2 areas of scholarship through 1999. Within medical rehabilitation, and in other fields, the definition and taxonomy of participation of the WHO’s ICF16 are increasingly used in the development of participation measures. Where in the following sections Participation (Participation Restrictions) and Activity (Activity Limitations) are capitalized, the concepts as defined and operationalized in the ICF are referred to. Where participation and activity are used without capitalization, other or broader conceptualizations or descriptions are denoted.

The concept of Participation as delineated in the ICF gains more clarity when considered next to the ICF concept of Activity and its definition. However, concepts deserve their own definition and should not be defined by default.5 Consequently, the field is still in need of a consensus definition of participation. Even with the ex cathedra definition of Participation by WHO, there still is no consensus on the definition of participation, or community integration, the term used most popularly.20 Appendix 1 provides a sample of definitions, collected from various sources. The collection may serve as a starting point for discussion, clarification of the concept, and possible consensus. (It should be noted that these definitions are largely limited to participation quantifiable as performance, as discussed in the section “Is there more to participation than performance?”) The definitions in appendix 1 were developed by researchers or other professionals. Another approach is to use focus groups or other qualitative research methodologies to get lay people’s definition of participation.28,33,34 This may lead to surprises. For instance, Mars et al35 suggest that in the eyes of older adults with a chronic physical illness, only positive experiences count as social participation. McColl et al20 found differences in the definitions offered by respondents with positive versus those with negative self-evaluations. One concern with these approaches to conceptualization is that frequently respondents do not know how to describe participation.34,35 If the interviewer needs to give examples or descriptions, there is reason to doubt that delegation of authority is a fruitful approach to the definition of scientific constructs.

WHAT IS THE DEFINITION OF PARTICIPATION? “Participation” has its roots in the Latin “pars” and “capere”—“part” and “to take.” The meanings of the word in English dictionaries often include the following: partaking in something, association with others in a relationship, taking part with others in an activity, and social interaction in a group. However, nowadays most researchers who are interested in measuring participation go straight to the ICF and read its definition of Participation: “involvement in a life situation.”16(p10) This is not a very useful definition: being born, and dying, and everything we do in between involve being in a life situation.9,12 The obverse of Participation that the WHO offers, Participation Restrictions, has its own definition: “problems an individual may experience in involvement in life situations,” which also does not help constrain the scope of Participation.16(p10) If a store clerk is not attentive, or FedEx does not show up in time to pick up my latest grant application, I experience a problem in a life situation, but these situations hardly seem to be issues resulting from the intersection of health problems and the environment—a key claim in both the ICF and the DCP17 about the emergence of Participation Restrictions. The explanations in the ICF manual16(p15,p213) are not helpful. For instance, in specifying that the qualifier “performance” can be applied to the concepts Participation and Activity in order to operationalize them, it is noted that performance can also be understood as “the lived experience” of people in the actual context in which they live.16(p15) For those who perform qualitative research, “lived experience” has a meaning that is diametrically opposed to what the ICF intends: not a system of categories to classify aspects of an objectively existing reality, but the subjective experience of people who are going through a particular life event or process such as relearning self-care lost as a result of injury or illness18 or (not) being excluded from activities with nondisabled peers.19

WHERE IS THE BORDER BETWEEN PARTICIPATION AND ACTIVITY—OR DO WE NOT NEED ONE? While the ICF’s definition of Participation is fuzzy at best, the ICF definitions of Activity (and its obverse, Activity Limitation) are much clearer and in line with previous theory and research.36 Activity is defined as “the execution of a task or action by an individual” and Activity Limitations as “difficulties an individual may have in executing activities.”16(p14) The decision to define Activity (Limitations) and Participation (Restrictions) as separate constructs, but to offer only a single taxonomy that does not differentiate the 2, has resulted in confusion in spite of the fact that an appendix to the ICF offers 3 options to separate Activities from Participation in the taxonomic listing. Distinguishing between concrete, person-level, immediately observable Activities and more abstract, community-society– level Participation is an old tradition in rehabilitation and has been the basis of fruitful research and efficient administration of services. The refusal of WHO to make a split in the taxonomy parallel to the one in the ICF conceptual framework creates problems, at least for those who insist that the taxonomy should follow the ICF’s theory. Pragmatists may be happy to select one of the ICF appendix options for distinguishing Activity and Participation and go to work. An alternative to basing a distinction on theoretical grounds is to use empirical, data-driven techniques to draw a dividing line. One study found evidence that items in a functional assessment instrument could be sorted into groups that paralleled the Activity-Participation distinction.37 However, in a second study, the same authors found evidence that led them to reverse their position.38 Post et al39 too, in a factor analysis of either IMPACT-S items or subscores, found no separation between Activities and Participations, but instead factors that cut across the divide, in spite of different answer stems in items

Arch Phys Med Rehabil Vol 91, Suppl 1, September 2010

CONCEPTUALIZATION OF PARTICIPATION, Dijkers

for Activity Limitations (“limitations”) and Participation Restrictions (“problems”). It is unlikely that brute empiricism will resolve this dilemma. Everyone agrees that there is a causal relationship between the 2 entities, in that, ceteris paribus, particular Activity Limitations are likely to give rise to specific Participation Restrictions. This relationship will result in correlations between selected Activity Limitations and selected Participation Restrictions (eg, between transferring oneself and neighborhood mobility) that may be stronger than the correlation between one Activity Limitation and the next, such as transferring oneself and writing messages, or than the association between one Participation Restriction and the next, such as neighborhood mobility and maintaining formal relationships. (For instance, Jette et al40 found that the correlation between a “physical movement activity”40(p 2092) summary activity score and a community participation summary score reflecting mobility, functioning in work, and other daily activities was stronger than the correlation between the latter and a summary score of “applied cognitive activity.”40(p 2092) On the other hand, the latter was a better predictor of “social and home participation,”40(p 2090) which summarized communication, social relationships, and home management roles.) In this case, factor analysis or similar techniques cannot make distinctions that parallel our theories. In addition, the proper measurement of Activity Limitations and of Participation Restrictions likely requires method differences, such as a metric of difficulty for the former, and one of frequency for the latter, that almost guarantee a separation between the 2 in factor analyses and similar multivariate approaches.38 We should be loath to accept method variance as evidence that supports theory. Disablement may be compared with a stone thrown into a quiet pond—the successive waves moving outward from the point of entry cause one another and almost imperceptibly turn into one another. There is so little differentiation within and between impairments, functional limitations,36 activity limitations, and participation restrictions that it is nigh impossible to create a simple taxonomy in which the first branching is a differentiation among the 4. We should remember that the first requirement of a taxonomy is that it makes distinctions that are useful for the task at hand; picking 1 of the 3 WHO-suggested ways of differentiating Activity and Participation in the ICF taxonomy might suffice. Of course, there should be correspondence between our criteria for distinguishing Activity from Participation and the definition of Participation (and the definition of Activity); working back from the criteria to the definition may be a good way of sharpening our definition of the concept Participation. IS THERE MORE TO PARTICIPATION THAN PERFORMANCE? The familiar measures of participation and related constructs41-43 focus on performance: the degree to which a person fulfills roles and has relationships, displays community presence, and otherwise performs a portfolio of actions that can be witnessed by an observer. It is this aspect of participation that society and rehabilitation service providers have a primary interest in: the degree to which individuals with disability take part in household, community, and society activities. However, over the last decade, a number of authors have made claims that just getting the objective facts does not give us the full story.20,33,34,44 They have started to develop measures that incorporate alternative aspects of participation they have declared important.45,46 Appendix 2 provides a sampling of the concepts mentioned in the literature and reflected in various measures.

S7

Participation as performance is most commonly quantified by determining someone’s status (married or not) and by a count of the frequency with which actions that are core to roles are performed or the hours that are dedicated to role performance. Because “separate but equal” is not seen as ever equal when it comes to people with disability, the location of one’s residence and the site of performance of major life roles is a common indicator of participation, as is the nature of the other people in whose company life unfolds—for example, professional supervisors, peers with stigma, or self-selected disabled or nondisabled friends and colleagues. Difficulty is used as a metric in some participation performance measures, although it may be too subjective to be used by outside observers. While various aspects of participation qua performance have received attention in the literature, there has not been a systematic review of the aspects of participation that are subjective and reflect the desires and satisfactions of the people involved. The items in appendix 2 are preliminary and derived primarily from published (ad hoc) measures. Further study of the literature and further research will yield additional factors and combinations. The last 2 items in appendix 2 are of special interest. Satisfaction can be attached to various aspects of participation as performance, as well as to the other qualitative aspects of participation. For instance, marriage and family scholars have spent years defining and measuring numerous aspects of marital satisfaction—with the sexual relationship, with the domestic division of labor, with agreement on child rearing principles, and so forth; the literature on job or employment satisfaction is hardly smaller. Importance is a potential qualifier of performance scores, used in weighting various domains that are part of a participation measure by their importance to the individual in question or to society as a whole. The same qualifier can be applied to almost all aspects of participation listed in the subjective subgroup. For instance, if one tries to quantify participation satisfaction as an average of satisfaction in various domains (employment, parenthood, and so forth), the relative importance of those domains needs to be taken into account, as is done in the Participation Objective, Participation Subjective45 and the Participation Assessment with Recombined Tools—Subjective component.46 Last, direct rating of importance is a possible way of studying response shift phenomena, which affect most of the subjective aspects of participation.66 It is impossible to quantify, in the typical study of rehabilitation outcomes, all the subjective aspects of Participation listed in appendix 2. Further research needs to determine which ones can be measured reliably, validly, and easily. Then, more research should determine whether these best conceptually different aspects are divergent enough that separate quantification is justified. Low intercorrelations as well as different patterns of correlation with other relevant variables such as activity level, depression, and participation as performance may be used in investigating this.67 For instance, it is quite possible that the correlation between a desire to do more, less, or the same of a particular activity is associated with satisfaction with performance, so strongly that routine application can omit one or the other.48 WHAT DOMAINS SHOULD BE INCLUDED IN A PARTICIPATION MEASURE? Participation qua performance involves discharging normative roles in community and society. In postindustrial societies, that is a broad area, ranging from involvement in religious activities to employment, from discharging of civic duties to an economic role as a consumer. Most creators of participation Arch Phys Med Rehabil Vol 91, Suppl 1, September 2010

S8

CONCEPTUALIZATION OF PARTICIPATION, Dijkers

measures also include roles in a nuclear family and household. Measures of participation and those of related concepts probably differ most from one another in the area of domain selection. Beyond a few basics (work/school, running a household, relations with family and kin), there is no agreement about what should be included. This discrepancy may be dictated in part by the characteristics of the target population (children vs adults; religious United States vs agnostic western Europe) or of the desired length of the instrument, but the differences between measures targeting the same groups with the same approximate number of items also are substantial. Both the ICF16 and the DCP68 offer lists of activities and participation elements. Some of these entries are not useful in deciding on the content of measures because they occur so infrequently (eg, getting a mortgage—ICF code d865: complex economic transactions) as to be useless in differentiating well integrated from less integrated individuals. However, by including the more generic parent category (d86: economic life), sufficiently frequent occurrences may be captured. Whether it is useful to include in an instrument both categories that are used frequently and their more global parent categories53 is still to be determined. It would be a valuable exercise to use a Delphi or similar process with a group of experts (researchers, clinicians, consumers) to develop a core list of domains to be covered in instruments—possibly differentiated by age group, other characteristics, and short versus long forms. WHAT ARE THE APPROPRIATE METRICS IN QUANTIFYING PARTICIPATION? Mentioned previously were a number of metrics (frequency of actions, hours dedicated to roles, etc.) that have commonly been used in quantifying participation qua performance. The definitions of participation in appendix 1 imply or sometimes explicitly state that participation is a macro phenomenon in the sense that it is reflected in the occupying of statuses or the discharge of roles. Such a definition suggests that a measure of participation could ask simple questions such as “Are you married?” or “Are you a member of a religious community?” However, this approach has 2 disadvantages. One is that statuses often are slow to change, especially after sudden-onset disorders that affect role fulfillment. Most young people would be considered a student even if they had not attended school in months. Marriages sometimes are not formally dissolved until years after one partner leaves the other because of an inability to deal with a disability. Inquiring into the fulfillment of roles (“Do you currently fulfill the husband role?”) rather than into the having of statuses is preferable.2 However, this approach would result in uncomplicated dichotomies, and in many instances the definition of the role is not simple or lacks consensus among cultural or socioeconomic groups. (What is the “husband role”?) In order to obtain categories of finer grain than dichotomies, many investigators ask about hours dedicated to major, clearly defined roles— hours working or running a household. For less clearly defined roles, they may ask about the frequency of actions and activities that constitute a core component of the role, such as number of times a month the person goes shopping or does volunteer work. This results in some instruments using a mishmash of metrics,41,42,69 although others manage to do with only 1.70 If a Delphi exercise is performed, it would be useful to ask about metrics to be used with each indicator of participation that is short-listed. The issue of appropriate metrics presumably is less of a problem with the subjective aspects of participation. Whether satisfaction or self-efficacy is considered, a metric suggests itself effortlessly, and it would seem that this metric is applicable across the domains of participation that might be inArch Phys Med Rehabil Vol 91, Suppl 1, September 2010

cluded in the measure—for instance, “How satisfied are you with your employment; your civic activities?” and so forth. However, consensus between respondents and investigator on the scope of each of these domains is still required to obtain total scores that are comparable from 1 individual to the next. Appendix 2 on each line may list several terms that are presumed to be equivalent, such as having difficulty, being restricted, having limitations, or having a problem; however, they may not be equivalent in the eyes of the individuals who complete participation instruments.71 HOW DO WE DEFINE ADEQUATE PARTICIPATION? For some applications of a participation measure, the issue of adequate or normal participation as performance is irrelevant— change over time after onset of injury and differences between clinical groups can be measured without referring to what is normal or minimally acceptable from a societal or clinical perspective. However, in many clinical and research applications, it is important to have an indication that subjects are involved in community and society at the level of their nondisabled peers.2 A number of methods have been used to do so. To date, no investigator has developed a traditional normreferenced test of participation, using the study of a large and diverse group of randomly selected nondisabled individuals to determine norms for participation, overall or by subgroups defined by sex, age, social class, or other relevant characteristics. The closest that researchers have come to this is the use of a small sample of people without a disability to develop a mean and SD for their participation instrument, which subsequently has been used by others as a basis of comparison for clinical samples.69 The CHART incorporates norm-referencing into its scoring algorithm in the sense that the maximum score on each of its 6 subscales is 100 —the score attained by “most” of a sample of 88 nondisabled subjects.42 Thus, anyone with a disability scoring 100 on a subscale is not considered to be handicapped (to use the adjective corresponding to the concept CHART attempts to operationalize.) However, this algorithm fails to reflect that in each society within the group of nondisabled people as well as within the group of individuals with a disability, there are differences in participatory performance, with many scoring above the able-bodied modal category. In addition, CHART subscale scores will depart from a normal distribution, creating statistical analysis problems most investigators do not appear to appreciate properly. Quite another approach to norm-referencing was taken by van Brakel et al,54 whose Participation Scale asks subjects to compare themselves to their peers with respect to performance in various domains, such as, “Do you contribute to the household economically in a similar way to your peers?” and “Do you take as much part in casual recreational/social activities as do your peers? (sports, chat, meetings).”54(p202) Peers are defined as “those who are similar to the respondent in all respects (sociocultural, economic and demographic) except for the disease or disability.”54(p196) In these kinds of instruments, participation restrictions may be underreported because respondents select peers with a disability rather than nondisabled peers as comparators.72,73 To the degree that respondents know what is normal for nondisabled persons and the range of performance for these peers is narrow, the van Brakel technique shades into a criterion-referenced approach, a road previously taken in the Perceived Handicap Questionnaire.74 In the Perceived Handicap Questionnaire, participants are instructed to compare their sit-

CONCEPTUALIZATION OF PARTICIPATION, Dijkers

uations with those of peers, first nondisabled ones and then those with a similar disability. This approach eliminates part of the problem alluded to in the previous paragraph; however, the comparison basis selected by respondents is still unknown: do they choose individuals of the same age group, life cycle stage, sex, social class, and so forth, or others with widely variant demographic characteristics, with whose lifestyle they may be familiar? Because there is little agreement on the content and extent of social roles and social participation in postindustrial societies,2 it is unlikely that a truly norm-referenced instrument to measure participation can be developed, unless for narrowly defined subgroups such as school-age children in a single culturally homogeneous society. There is a great risk of imposing the values of the instrument creator in selecting domains to include in a participation measure, weighting these domains to reflect societal or community views, and developing algorithms to score various levels of performance within each domain.2 While extensive feedback from panels of consumers and other stakeholders may serve to reduce the impact of the values of “western white middle class overachiever intellectuals” on the selection of domains33 or the weighting of items,75 using the same democratic methods in working out the details of a scoring algorithm is much harder. Value-free measurement may be impossible—if for no other reason than that the consumer panels one might use in developing each detail of a measure themselves are steeped in a particular culture and subculture. HOW SHOULD PARTICIPATION BE OPERATIONALIZED? Operationalization is the translation of the theoretical definition of a construct into measurement operations and frequently is not a difficult or controversial process. However, in the case of participation, it appears there are a number of issues suggesting that our theoretical definitions are not as clear as we would like to think, and that operationalization brings a number of difficulties. One issue is that of domains and subscores. Participation Assessment with Recombined Tools—Objective component investigators considered both a total score and subscores for various domains of life included in this participation instrument.46 However, the number of domains defined by these experts and the specific items proposed for each were different. In addition, when the discussion turned to the development of norms based on parallel data for a nondisabled population sample, additional disagreements became evident. Some thought that individuals could not be characterized as having normal participation unless they had scores reflecting adequate participation in all 3 or 4 domains to be distinguished; others thought that deficits in one domain could be compensated for by high scores in another, and that only the total score should be considered an indication of the level of participation, whatever the utility of the domain subscales. Clearly, these differences in opinion have implications for the scoring of a measure, a key part of operationalization. Apparently, most of the individuals who wrote the definitions in appendix 1 did not ask themselves these types of questions. When in the development of the CHART large groups of subjects were asked to rate the relative importance of the various items included in this measure to the operationalization of the concept handicap, there was tremendous variation from one person to the next. Surprisingly, the mean ratings produced by individuals with SCI were very similar to the mean of the ratings proposed by professionals, suggesting sharing of the values underlying these ratings but an uncertain translation of these principles into scoring algorithms.75

S9

Another major issue in the operationalization of participation is how answers of “NA” are to be scored. If given the option (as in the Participation Scale,) people may select this answer because an item truly is NA to them—such as child care for childless adults or lawn care by apartment dwellers. Many instrument developers provide a checkbox for this purpose. If no such option is provided, subjects may leave an item blank, or (the most problematic choice) select the “never do” option. However, people with a disability may also answer “NA” when, because of persistent insurmountable barriers, they have become accustomed to not being able to occupy a status and are not even aware that they might want to play a particular role. For instance, a young adult who after an SCI has not even been able to apply for employment because work places are not accessible, or has been turned down repeatedly because of prejudicial hiring practices, may dismiss the possibility of employment and direct her energies elsewhere. From a participation-as-performance point of view, “not working” may be the correct answer, rather than “NA,” which might be appropriate for an 80-year-old.8,76 The fact that in these instances the cause of low participation lies in the environment rather than the person is irrelevant—as the ICF correctly conceptualizes the situation, Participation Restrictions result from the confluence of Impairments, Activity Limitations, and Environmental factors. The purpose of a participation instrument is to quantify participation, not its causes. A logical consequence of this view may be that in scoring participation qua performance all “NA” answers should be equated with the “never do” category or with the performance category that reflects or includes “no performance.” The individuals who mistakenly select NA will receive the proper credit, and even for those who have the excuse of age or sex or severe disability for not being involved in a particular life domain, “never do” is not an unreasonable selection. Being a parent or a wage worker is one way of participating in community and society, and they have chosen, or have been forced, not to take part in that particular way; consequently, their participation is less than maximal. Given limits in time, resources, and energy, everyone is compelled to make a choice about which roles to play in life, and everyone’s participation is less than the theoretical maximum. Alternatively, the modal nondisabled participation score can be attained through the combination of high scores in many domains, and, for instance, not having a garden to look after would make available time and energy to devote to other roles. The section on measurement models takes up this argument. WHAT IS THE PROPER MEASUREMENT MODEL FOR PARTICIPATION INSTRUMENTS? Developers of participation instruments, apparently without exception, have used the traditional psychometric model to score their measures and to assess their reliability. The reasoning is that participation is a latent trait that shows its impact through a series of indicators such as employment, parenting, and churchgoing. These indicators, because they are all caused by the same trait, ought to be correlated. Finding correlations between them is evidence of dependable measurement, and calculating a score by summing items provides a reliable result because random errors in the measurement of individual items will tend to cancel one another out. The total score reflects the strength of the latent trait in the individuals measured. Whether the researchers use CTT or IRT, in the end these are the underlying assumptions of participation measure developers. Clinimetrics is a method of measuring constructs, developed in clinical medicine, which in basic assumptions differs from CTT and IRT. Feinstein77 coined the term clinimetrics to Arch Phys Med Rehabil Vol 91, Suppl 1, September 2010

S10

CONCEPTUALIZATION OF PARTICIPATION, Dijkers

describe a method by which a limited number of key indicators reflecting all aspects of the concept of interest are selected, scored (reflecting the importance of each, as necessary, by means of the item category values), and combined into a total score. The archetypical example of a clinimetric measure is the Apgar78 score, which quantifies the survival potential of a newborn based on 6 vital signs that are not correlated. Feinstein77,79 listed 6 characteristics which he saw as central to the clinimetric approach: 1. Clinical expertise is used to select items to be included in a measurement instrument. While sometimes that expertise comes from the instrument developer only, more commonly multiple clinicians provide input, consulting the literature to identify items. In the last 2 decades, interviewing patients about what symptoms and problems they experience, and which ones of these are most bothersome, has also been used as a means of generating items. 2. Items are not weighted in calculating an index total score. The simplicity of scoring at the bedside trumps other considerations. This simplicity is also limiting, and nowadays, in the era of widespread computer availability, many clinimetric measures use patient-rated importance in calculating total scores. 3. Items are chosen to be heterogeneous, to reflect all significant factors affecting the concept of interest. Intercorrelations between items are not a selection criterion; in fact,

high intercorrelations may be a reason to throw out 1 item in a pair as redundant. This heterogeneity is almost certain to result in multidimensionality of the scale: “A clinician wants to combine different elements, not different expressions for essentially the same thing.”79(p126) 4. Ease of use is a prime consideration in clinimetric measures. It must be possible to obtain information on each constituent item and to calculate total scores quickly in order to have the results available to make clinical decisions during the medical encounter. Since Feinstein conceived of these ideas, desktop computers and PDAs have become available that make score calculations more feasible, even when scoring algorithms are complex. Increasing use of clinimetric measures in research (rather than clinical care) has further decreased the emphasis on ease of use. 5. Face validity is a primary criterion in selecting items and evaluating the measure as a whole. If the items individually and the scoring algorithm used to combine them do not make clinical sense, the index is likely to be rejected. Face validity for Feinstein also takes on overtones of simpleness, and of comprehensiveness to cover all relevant aspects of a concept. 6. The patients’ feelings and beliefs must be solicited and incorporated into those instruments that aim to operationalize any construct incorporating subjective elements. Eliciting input from the patient should be done using simple

Fig 1. The relationship between indicators and constructs in psychometric and clinimetric measurement of abstract concepts.

Arch Phys Med Rehabil Vol 91, Suppl 1, September 2010

CONCEPTUALIZATION OF PARTICIPATION, Dijkers

means, and the dictates of preordained scoring systems (read: CTT) and similar methodology-driven issues should not stand in the way. The difference between psychometrics and clinimetrics is made clear in figure 1. In psychometrics, the various items used to operationalize a construct are assumed to be the effect of the latent trait of interest. In clinimetrics, the indicator items reflect 2 or more behaviors or other characteristics that together define the latent concept of interest. Fayers et al80 named them causal indicators to make clear that together they create the construct, but defining is a better term for how the items relate to the construct. The term psychometrics has elicited animosity,81-83 especially from those who claim that there is nothing new here, and that psychometricians are aware that some constructs have been measured using items that were not developed with the methods of psychometrics and their emphasis on internal consistency.84 Regardless of whether we believe that participation is a latent trait similar to intelligence or authoritarianism or is more like one of the constructs operationalized by clinimetricians such as survival potential in the Apgar score, the elements that go into the measurement of participation qua performance have many of the characteristics of the latter. One can be participatory doing one thing or another but even people who score high on whatever measure of participation a researcher may develop are unlikely to be active in all domains that may go into the total: household, work, volunteering, socializing, churchgoing, leisure activities, and so forth. The clearest example is the triad of going to school, household activities, and work: if these 3 key activities are quantified in terms of hours spent a day, the number of hours in a day sets a limit on what is humanly possible. No single person is likely to score the item maximum on 2 of these, let alone all 3.8 The same is true for indicators of participation in which a choice is not forced by lack of time or energy but by preferences or circumstances. Some people spend their leisure time making music, and others like to engage in sports. People with no nearby family may compensate by spending much time socializing with friends. The items involved in these duos and triads are likely to have very low or even negative correlations, which under psychometric rules is sufficient reason to throw them out of the set of indicators of the concept of interest. Common sense argues against creating a participation scale without work, school, and household responsibilities, however. It is feasible to develop a scoring algorithm that takes low or even negative associations into account. For instance, the productive activity subscore of the CIQ41,69 allows a maximum score if one is a full-time student or worker, or by a combination of hours in work or school or volunteering. The score on each dimension of the CHART42,85 is computed by totaling just a few rather broadly defined items, and the maximum is capped such that one often can obtain the ceiling score with a 0 score on 1 or more of the items. These scoring algorithms suggest that the developers of these 2 measures let common sense lead them to embrace clinimetrics. The fact that the items on any 1 participation measure are causes (definers) of participation rather than effects of a latent trait participatoriness does create a problem when one attempts to develop an IRT-based participation measure. In any sample of adults, even adults with a disability, many more will be working than going to school. IRT interprets the popularity of items as an expression of their difficulty, and ends up slotting working among the easier items, and school among the more difficult ones. The same phenomenon occurs with all items which differ in their popularity rather than difficulty, whatever

S11

that might mean with respect to participation. Churchgoing is less difficult than political/civic participation; going to the movies is less difficult than going to a classical music concert, and so forth. The IRT fit statistics will reflect that the items were shoehorned into a scale and should be removed to improve the measure. However, it clearly would not do to claim that an index of participation that lacks school attendance, for instance, is a valid reflection of participation. One possible solution is to combine items, as was done in the CIQ. Rather than using an item “hours in school” and an item “hours at work,” one could create an item “hours in school and/or at work.” However, the metrics with which the item information was collected may not allow such simple summation. In addition, it is possible that even combined items misfit, and one ends up with 1 or a few super-items: “frequency of A and/or B and/or C and/or . . . .” Clearly, IRT is an inappropriate technique here.8 The fact that it sometimes works42,53,86 and sometimes can be made to work may be explained by the fact that participation levels and profiles among individuals with a disability are determined not just by sex, age, and lifestyle preferences. As is made clear in the ICF conceptual model, participation is also affected by impairments and activity limitations in interaction with environmental and personal factors. If in one’s sample there is an adequate variety of impairment levels, there may be positive correlations between all participation items in an instrument, and these correlations may be sufficiently strong to make IRT analysis– based scoring formulas feasible. However, a valid measure of participation should be reliable even if one controls for impairment (which may be a major cause of participation but does not equal participation), for instance by performing separate analyses for low-disability, medium-disability, and high-disability groups. This suggestion flies in the face of the claim that IRT produces measures that are item-independent and sample-independent. This claim may be true for traits (such as reading comprehension and functional status) measured with effect indicators, which indeed can be substituted for one another without affecting the resulting scores. However, when one is trying to measure a clinimetric concept, items are not freely substitutable. The sometime success of IRT when applied to the items of a participation measure should not detract from the basic fact that participation is not a concept that fits the assumptions of IRT. If there is good reason to abandon CTT and IRT and rely on the strictures of clinimetrics to develop a participation measure, it is also possible to bring in explicit weighting of the indicators by their importance, as determined by society or by the individual subject. Weighting of indicators has no official standing in the methods of either CTT or IRT, and has been shown, following the “It don’t make no nevermind” principle of Wainer,87 to have an effect on total scores only under special circumstances, such as an instrument with a small number of items. However, developers of participation measures often have a strong preference for using weighting items or item categories to reflect the relative importance of indicators.41,42,45 Otherwise, being productive in terms of hours worked would be equally valuable as being productive in terms of hours spent chatting on the telephone with friends, a situation that goes against the understanding of participation most people have. HOW SHOULD WE COLLECT THE DATA ON PARTICIPATION? Almost without exception, self-report methods have been used in medical rehabilitation research to collect participation information. However, it may be useful for rehabilitation reArch Phys Med Rehabil Vol 91, Suppl 1, September 2010

S12

CONCEPTUALIZATION OF PARTICIPATION, Dijkers

searchers to consider other data collection methodologies, including systematic observation, experience sampling, instrumental recording, and qualitative methods. Self-reports of behaviors over a previous period, unless part of a fixed daily routine, are notoriously unreliable. The CHART question “With how many strangers have you initiated a conversation in the last month (for example, to ask information or place an order)?” is likely to produce only a crude approximation of the frequency (0 vs 1–2 vs 3–5 vs 6⫹ times) of this behavior. EMA refers to a set of techniques to bring the reporting of the information needed by a researcher closer to the time point at which the behavior occurs and to have the report produced within the relevant setting rather than in a research office or over the telephone.88 At one extreme, EMA methods require the subject to complete a diary at the end of each day in which the activities or moods of the day are recorded—the end-of-day diary. Because even a delay of less than 24 hours may allow random and systematic biases (memory, telescoping, latency effects, and so forth), a more typical design is that the subject is alerted at multiple preset or random times a day using a pager, special preprogrammed wristwatch, or PDA, and then is required to report key pieces of information using a small carry-along diary, the PDA, or another type of palm-top computer. Because times are sampled, the method is also referred to as the experience sampling method, especially if the focus is on subjective experiences.89 In the arena of participation assessment, EMA could be used to address such questions as the following: “In what activity are you currently engaged?” “How often in the past hour have you done activity X?” “With which people, if any, are you currently interacting?” By means of appropriate methods of aggregating these discrete reports over a day and individual days over a longer period,89-91 the researcher can obtain a reliable overview of the person’s participation.92 To date, EMA methods have been used in participation research on a very limited basis.93,94 Multiple single-axial or multiaxial accelerometers can be used to determine time spent in different postures (sitting, lying, standing) and various physical activities (walking, cycling).95 Further development of software algorithms may make it possible to make even more differentiation within the broad category of activity.96 A data logger attached to the spokes of a wheelchair makes it possible to determine the distance the user covers each day, and using the Global Positioning System what areas the user frequents.4,97 The studies that have used this technology to date have focused more on activities than on participation.98,99 Qualitative research methods have been used by several investigators in the development of participation measures, mostly to assess what consumers consider to be important domains of participation33,100 or to develop a definition of participation.28 Cognitive interviewing to determine the target population’s understanding of the questions and answer categories in participation instruments is an increasingly applied qualitative methodology.48,101 While there are many examples of qualitative research used to answer substantive research questions,102-104 this still is a method that is underused. Investigators who are interested in both participation as performance and the subjective issues related to it or in the relationship between participation and barriers/facilitators8 might do well exploring what various qualitative research methodologies have to offer. HOW DO WE EVALUATE THE QUALITY OF A PARTICIPATION INSTRUMENT? Collection of data relevant to the reliability and validity of participation measures is often done, it would appear, on auArch Phys Med Rehabil Vol 91, Suppl 1, September 2010

topilot. Investigators calculate the Cronbach alpha and other indicators of metrologic quality without paying attention to the nature of their instrument. The major issue is the use of parameters that are appropriate to a psychometric concept with a construct that is measured, intentionally or de facto, along clinimetric lines. If the items are not considered to emanate from the latent trait participatoriness, then they need not be correlated, and the Cronbach alpha and related quality indicators are irrelevant. Similarly, IRT fit parameters are not a suitable indicator for determining whether an item belongs in a measure, and factor analytic approaches also are inappropriate, unless for each domain we have multiple indicators and we hypothesize that they cluster by domain.53 In instruments that assess, for example, satisfaction with various participation domains, the correlations between items should not be strong. Even if a large trait element is reflected, there is no reason to assume that satisfaction with domain A is the same as satisfaction with domain B. For instance, people with a disability tend to be satisfied with their interactions with friends and family members but dissatisfied with their employment and financial status.105,106 The same reasoning applies to autonomy, respect and dignity experienced in interactions, and all personal perspective elements listed in appendix 2. Regretting floor and ceiling effects (and adding items to an instrument to get rid of them) makes no sense if one aims to quantify deviation from normal. In countless clinical populations, many individuals maintain participation levels on a par with nondisabled people. If this level is the cut-off for normal performance, there is no need to undo artificially the basic premise of the instrument. The fact that distributions with strong cellar or ceiling effects are disadvantageous when it comes to statistical analysis is another issue; with the availability of powerful nonparametric multivariate statistical analysis techniques, this problem is becoming a historic curiosity. This holds true for participation qua performance as well as the subjective elements of participation. For instance, the fact that on the Keele Assessment of Participation 53% of elderly respondents were at ceiling59 (suggesting that they had adequate autonomy in their participation when it comes to evaluating activities) is not problematic per se. With respect to validity of a participation instrument, correlations with other measures of participation qua performance should not be expected to be high. The differences between measures in terms of domains included, metrics used, and algorithms applied to develop a total score suggest that correlations high enough to be encouraging about the validity of a new instrument compared with an established one will be the exception rather than the rule. We should not give up efforts to evaluate the quality of our instruments, initially and for every new population and new application. To the contrary, we probably should do more of it. However, researchers must do a better job of thinking through what evidence is applicable given their conceptualization and operationalization of participation, and then only produce, use, and present that evidence. CONCLUSIONS While medical rehabilitation has a large number of measures of activity and activity limitations, most clinicians and researchers use the FIM, which after its development in the 1980s has grown into an industry standard. The situation with respect to the measurement of participation and participation restrictions is radically different. While there are a few dominant measures, they tend to be specific to use with particular diagnostic groups (CHART in SCI; CIQ in traumatic brain injury, and so forth), age groups (Frenchay Activities Index in

CONCEPTUALIZATION OF PARTICIPATION, Dijkers

geriatrics), or geographic areas (Life-H in Quebec). Since the FIM was launched, rehabilitation researchers have learned much about measurement107 and about the benefit of having an industry standard for quantifying key characteristics of our clientele. The time is ripe to develop a measure of participation qua performance that can function as the industry standard for medical rehabilitation and other areas serving individuals with functional deficits. Consensus on development of a minimal set of measures that adequately tap into the patients’ individual perspectives on participation (satisfaction, autonomy, belonging, and so forth) will take a longer time. This article has sketched a number of significant issues that must be dealt with in the conceptualization, operationalization, and measurement of participation. Defining and operationalizing participation requires developers of measures to consider these choices as a package. Given the complexity of participation and its manifestations in various groups defined by age, culture, and accessibility of the physical and sociocultural environment, it will be far from easy to create a measure that is simple yet reflects participation adequately based on the conceptualization of most stakeholders. However, improvements over the currently available instruments are feasible and needed. APPENDIX 1: DEFINITIONS OF PARTICIPATION AND RELATED TERMS/CONSTRUCTS Community integration can be viewed as the return of individuals to their age, sex, and culturally appropriate role functions. . . . People are considered to have healthy participation if they take part in all life areas or situations in which they wish to participate, in a manner or to the extent that is expected of an individual without restrictions in that culture or society.21(p992) Community integration includes (a) assimilation— being able to fit in with other people, knowing your way around, and being accepted; (b) social support— being part of a network of family, friends, and acquaintances; (c) occupation— having things to do for fun, and to do meaningful and productive activity during the main part of the day; and (d) independent living—independence in everyday tasks and in making everyday decisions and life choices.22(p8) . . . community integration implies “normalization” (Wolfensberger and Tullman, 198223 ), that is, the circumstance in which individuals with disabilities live, work, play and lead their daily lives without distinction from and with the same opportunities as individuals without disabilities.24(p570) Community Integration: Presence, participation and interaction in the consumer’s natural environments.25 Fundamental to the definition of community integration is the expectation that people with psychiatric disabilities can have the same opportunity to live in the community as everyone else. To live in the community requires the ability to seek employment, housing, education and to pursue leisure and civic activities of one’s own choice.26 Participation is involvement in a life situation.16(p10) Participation restrictions are problems an individual may experience in involvement in life situations.16(p10) . . . social participation, as proposed by the conceptual model of the Disability Creation Process (DCP), is more explicitly defined as “Optimal accomplishment of daily activities and social roles valued by the person or socio-economic environment which ensure survival and development in society throughout life.”27(p1232) Social integration [is] a process, unfolding over time, through which individuals who have been . . . disabled increasingly develop and exercise their capacities for connect-

S13

edness (construction and successful maintenance of interpersonal relationships) and citizenship (rights and privileges, responsibilities).28(p471) Community integration is the opportunity to live in the community, and be valued for one’s uniqueness and abilities, like everyone else.29(p4) Participation is the involvement in life situations, which includes being autonomous to some extent or being able to control [one’s] own life, even if one is not actually doing things themselves.30(p578) Social participation occurs in the conduct of actions in which individuals share (a part of) their resources with others. . . . Three types of social participation can be distinguished on the basis of the resources that are shared. . . . Collective social participation activities [are] defined as the common acting of group members, whereby the intention is directed toward the group itself and not toward reaching an outside goal. The main resource that is shared among the group members is time. . . . Productive social participation [is] defined as the rendering of services, goods, and benefits for others. Political participation involves acts of decision making about social groups and the allocation of resources.31(p510-1) Social integration refers to the satisfaction experienced by the individual in their life situation, their sense of community affiliation and their self-image.32(p535) APPENDIX 2: ASPECTS OF PARTICIPATION Quantitative/objective/externally observable: performance ● ● ● ● ● ● ● ● ● ● ●

●

Status (marital, occupational, educational, and so forth) Frequency of activities Hours of role functioning, of activities in the community Location (normative or not) Company (caregivers, others with devalued identity, nondisabled friends, family, neighbors, and so forth) Difficulty,47-51 restriction,52 limitation,53 problemness54,55 Hierarchical position Social network (size, variety, interaction frequency) Reciprocity of interactions (initiation, giving and receiving assistance and social support, and so forth) Reliance on aides/aids50,56,57 Use of (generic) community resources (public transportation, social and health services, religious, recreational facilities, banks, shops and other economic institutions, and so forth) Etc. Qualitative/subjective/evaluative: individual perspectives

● ● ● ● ● ●

Autonomy (choice, control, independence, self-determination)1,33,51,57-59 Self-efficacy, sense of competence Difficulty,47-51 restriction,52 limitation,53 problemness54,55 Sense of belonging/membership/being accepted33,60 Opportunity and access33 Value and meaning33,44,58 X X

● ● ● ● ● ●

Social, cultural, personal relevance Preference, interest

Productivity Respect and dignity33 Desire for more/less/same/different45,61,62; fit with ideal self63 Closeness, affection, social support received, emotional attachment61 Responsibility33 Influence33 Arch Phys Med Rehabil Vol 91, Suppl 1, September 2010

S14 ● ● ● ●

Social support given33 (Dis)agreement, unity Challenge, mastery (balance of routine and challenge) Satisfaction with X X X X X

●

Participation overall44,48,51,53,62,64,65 Frequency Choice57 Contacts/interaction partners Etc.

15.

16.

17.

Importance X X X X X

●

CONCEPTUALIZATION OF PARTICIPATION, Dijkers

Globally46 For self-concept (role salience63 and so forth) For sustaining self or others For maximizing enjoyment in life45 Etc.

18. 19.

20.

Etc. 1. 2.

3.

4.

5.

6.

7.

8.

9.

10.

11.

12.

13.

14.

References Cardol M, De Jong BA, Ward CD. On autonomy and participation in rehabilitation. Disabil Rehabil 2002;24:970-4. Dijkers MP, Whiteneck G, El-Jaroudi R. Measures of social outcomes in disability research. Arch Phys Med Rehabil 2000; 81:S63-80. Flynn RJ, Aubry T. Integration of persons with developmental or psychiatric disabilities: Conceptualization and measurement. In: Flynn RJ, Lemay RA, editors. A quarter-century of normalization and social role valorization: evolution and impact. Ottawa: University of Ottawa Pr; 1999. p 271-303. Harris F. Conceptual issues in the measurement of participation among wheeled mobility device users. Disabil Rehabil Assist Technol 2007;2:137-48. Schuntermann MF. The implementation of the International Classification of Functioning, Disability and Health in Germany: experiences and problems. Int J Rehabil Res 2005;28:93-102. O’Halloran R, Larkins B. The ICF activities and participation related to speech-language pathology. Intern J Speech Lang Pathol 2008;10:18-26. McConachie H, Colver AF, Forsyth RJ, Jarvis SN, Parkinson KN. Participation of disabled children: how should it be characterised and measured? Disabil Rehabil 2006;28:1157-64. Whiteneck GG, Dijkers MP. Measuring difficult to measure constructs: participation and environmental factors. Arch Phys Med Rehabil 2009;90(11 Suppl):S22-35. Dalemans R, de Witte LP, Lemmens J, van den Heuvel WJ, Wade DT. Measures for rating social participation in people with aphasia: a systematic review. Clin Rehabil 2008;22:542-55. Eadie TL, Yorkston KM, Klasner ER, et al. Measuring communicative participation: a review of self-report instruments in speech-language pathology. Am J Speech Lang Pathol 2006;15: 307-20. Mortenson WB, Miller WC, Auger C. Issues for the selection of wheelchair-specific activity and participation outcome measures: a review. Arch Phys Med Rehabil 2008;89:1177-86. Bedell G, Coster W. Measuring participation of school-aged children with traumatic brain injuries: considerations and approaches. J Head Trauma Rehabil 2008;23:220-9. Wong YL, Solomon P. Community integration of persons with psychiatric disabilities in supportive independent housing: a conceptual model and methodological considerations. Ment Health Serv Res 2002;4:13-28. Magasi SR, Heinemann AW, Whiteneck GG, Quality of Life/ Participation Committee. Participation following traumatic spi-

Arch Phys Med Rehabil Vol 91, Suppl 1, September 2010

21.

22.

23. 24.

25. 26.

27. 28.

29.

30.

31.

32.

33.

34.

35.

nal cord injury: an evidence-based review for research. J Spinal Cord Med 2008;31:145-56. Coster W, Khetani MA. Measuring participation of children with disabilities: issues and challenges. Disabil Rehabil 2008;30:63948. World Health Organization. International classification of functioning, disability and health: ICF. Geneva: World Health Organization; 2001. Levasseur M, Desrosiers J, St-Cyr TD. Comparing the Disability Creation Process and International Classification of Functioning, Disability and Health models. Can J Occup Ther 2007;74 Spec No.:233-42. Guidetti S, Asaba E, Tham K. The lived experience of recapturing self-care. Am J Occup Ther 2007;61:303-10. Heah T, Case T, McGuire B, Law M. Successful participation: the lived experience among children with disabilities. Can J Occup Ther 2007;74:38-47. McColl MA, Carlson P, Johnston J, et al. The definition of community integration: perspectives of people with brain injuries. Brain Inj 1998;12:15-30. Resnik LJ, Allen SM. Using International Classification of Functioning, Disability and Health to understand challenges in community reintegration of injured veterans. J Rehabil Res Dev 2007;44:991-1006. Winkler D, Unsworth C, Sloan S. Factors that lead to successful community integration following severe traumatic brain injury. J Head Trauma Rehabil 2006;21:8-21. Wolfensberger W, Tullman S. A brief outline of the principle of normalization. Rehabil Psychol 1982;27:131-45. Bond GR, Salyers MP, Rollins AL, Rapp CA, Zipple AM. How evidence-based practices contribute to community integration. Community Ment Health J 2004;40:569-88. Golden Gate Regional Center. List of terms. Available at: http:// www.ggrc.org/fi_terms.htm. Accessed September 8, 2008. UPenn Collaborative on Community Integration—Issues. Available at: http://www.upennrrtc.org/issues/view.php?id⫽24. Accessed September 8, 2008. Rochette A, Korner-Bitensky N, Levasseur M. “Optimal” participation: a reflective look. Disabil Rehabil 2006;28:1231-5. Ware NC, Hopper K, Tugenberg T, Dickey B, Fisher D. Connectedness and citizenship: redefining social integration. Psychiatr Serv 2007;58:469-74. Salzer MS. Introduction. In: Salzer MS, editor. Psychiatric rehabilitation skills in practice: a CPRP Preparation and Skills Workbook. Columbia: United States Psychiatric Rehabilitation Association; 2006. Perenboom RJ, Chorus AM. Measuring participation according to the International Classification of Functioning, Disability and Health (ICF). Disabil Rehabil 2003;25:577-87. Bukov A, Maas I, Lampert T. Social participation in very old age: cross-sectional and longitudinal findings from BASE. Berlin Aging Study. J Gerontol B Psychol Sci Soc Sci 2002;57:P510-7. Lefebvre H, Cloutier G, Josee Levert M. Perspectives of survivors of traumatic brain injury and their caregivers on long-term social integration. Brain Inj 2008;22:535-43. Hammel J, Magasi S, Heinemann A, Whiteneck G, Bogner J, Rodriguez E. What does participation mean? an insider perspective from people with disabilities. Disabil Rehabil 2008; 30:1445-60. Haggstrom A, Lund ML. The complexity of participation in daily life: a qualitative study of the experiences of persons with acquired brain injury. J Rehabil Med 2008;40:89-95. Mars GM, Kempen GI, Mesters I, Proot IM, van Eijk JT. Characteristics of social participation as defined by older adults with a chronic physical illness. Disabil Rehabil 2008; 30:1298-308.

CONCEPTUALIZATION OF PARTICIPATION, Dijkers

36. Marino RJ. Domains of outcomes in spinal cord injury for clinical trials to improve neurological function. J Rehabil Res Dev 2007;44:113-22. 37. Jette AM, Haley SM, Kooyoomjian JT. Are the ICF activity and participation dimensions distinct? J Rehabil Med 2003;35:145-9. 38. Jette AM, Tao W, Haley SM. Blending activity and participation sub-domains of the ICF. Disabil Rehabil 2007;29:1742-50. 39. Post MW, de Witte LP, Reichrath E, Verdonschot MM, Wijlhuizen GJ, Perenboom RJ. Development and validation of IMPACT-S, an ICF-based questionnaire to measure activities and participation. J Rehabil Med 2008;40:620-7. 40. Jette AM, Keysor J, Coster W, Ni P, Haley S. Beyond function: predicting participation in a rehabilitation cohort. Arch Phys Med Rehabil 2005;86:2087-94. 41. Willer B, Rosenthal M, Kreutzer JS, Gordon WA, Rempel R. Assessment of community integration following rehabilitation for traumatic brain injury. J Head Trauma Rehabil 1993;8:75-87. 42. Whiteneck GG, Charlifue SW, Gerhart KA, Overholser JD, Richardson GN. Quantifying handicap: a new measure of longterm rehabilitation outcomes. Arch Phys Med Rehabil 1992;73: 519-26. 43. Holbrook M, Skilbeck CE. An activities index for use with stroke patients. Age Ageing 1983;12:166-70. 44. Ueda S, Okawa Y. The subjective dimension of functioning and disability: what is it and what is it for? Disabil Rehabil 2003;25: 596-601. 45. Brown M, Dijkers M, Gordon WA, Ashman T, Charatz H. Participation Objective, Participation Subjective: a measure of participation combining outsider and insider perspectives. J Head Trauma Rehabil 2004;19:459-81. 46. Dijkers M, Cicerone K, Heinemann A, Brown M, Whiteneck G. PART-S: a new measure of satisfaction with participation. Arch Phys Med Rehabil 2009;90:e39. 47. Tate RL, Pfaff A, Veerabangsa A, Hodgkinson AE. Measuring psychosocial recovery after brain injury: change versus competency. Arch Phys Med Rehabil 2004;85:538-45. 48. Yorkston KM, Baylor CR, Dietz J, et al. Developing a scale of communicative participation: a cognitive interviewing study. Disabil Rehabil 2008;30:425-33. 49. Washington LA, Wilson S, Engel JM, Jensen MP. Development and preliminary evaluation of a pediatric measure of community integration: the Pediatric Community Participation Questionnaire (PCPQ). Rehabil Psychol 2007;52:241-5. 50. Noreau L, Fougeyrollas P, Vincent C. The LIFE-H: assessment of the quality of social participation. Technol Disabil 2002;14: 113-8. 51. Trigg R, Wood VA. The Subjective Index of Physical and Social Outcome (SIPSO): a new measure for use with stroke patients. Clin Rehabil 2000;14:288-99. 52. Bedell GM. Developing a follow-up survey focused on participation of children and youth with acquired brain injuries after discharge from inpatient rehabilitation. NeuroRehabilitation 2004;19:191-205. 53. Gandek B, Sinclair SJ, Jette AM, Ware JE Jr. Development and initial psychometric evaluation of the Participation Measure for Post-Acute Care (PM-PAC). Am J Phys Med Rehabil 2007;86: 57-71. 54. van Brakel WH, Anderson AM, Mutatkar RK, et al. The Participation Scale: measuring a key concept in public health. Disabil Rehabil 2006;28:193-203. 55. Franchignoni F, Ferriero G, Giordano A, Guglielmi V, Picco D. Rasch psychometric validation of the Impact on Participation and Autonomy questionnaire in people with Parkinson’s disease. Eura Medicophys 2007;43:451-61.

S15

56. Hwang JL, Davies PL, Taylor MP, Gavin WJ. Validation of school function assessment with elementary school children. OTJR Occup Participation Health 2002;22:48-58. 57. Gray DB, Hollingsworth HH, Stark SL, Morgan KA. Participation Survey/Mobility: psychometric properties of a measure of participation for people with mobility impairments and limitations. Arch Phys Med Rehabil 2006;87:189-97. 58. Hemmingsson H, Jonsson H. An occupational perspective on the concept of participation in the International Classification of Functioning, Disability and Health—some critical remarks. Am J Occup Ther 2005;59:569-76. 59. Wilkie R, Peat G, Thomas E, Hooper H, Croft PR. The Keele assessment of participation: a new instrument to measure participation restriction in population studies: combined qualitative and quantitative examination of its psychometric properties. Qual Life Res 2005;14:1889-99. 60. McColl MA, Davies D, Carlson P, Johnston J, Minnes P. The Community Integration Measure: development and preliminary validation. Arch Phys Med Rehabil 2001;82:429-34. 61. Eklund M, Bengtsson-Tops A, Lindstedt H. Construct and discriminant validity and dimensionality of the Interview Schedule for Social Interaction (ISSI) in three psychiatric samples. Nord J Psychiatry 2007;61:182-8. 62. Sandstrom M, Lundin-Olsson L. Development and evaluation of a new questionnaire for rating perceived participation. Clin Rehabil 2007;21:833-45. 63. Plach SK, Heidrich SM. Women’s perceptions of their social roles after heart surgery and coronary angioplasty. Heart Lung 2001;30:117-27. 64. Levasseur M, Desrosiers J, St-Cyr Tribble D. Do quality of life, participation and environment of older adults differ according to level of activity? Health Qual Life Outcomes 2008;6:30. 65. King GA, Law M, King S, et al. Measuring children’s participation in recreation and leisure activities: construct validation of the CAPE and PAC. Child Care Health Dev 2007;33:28-39. 66. Schwartz CE. Applications of response shift theory and methods to participation measurement: a brief history of a young field. Arch Phys Med Rehabil 2010;91(9 Suppl 1):S38-43. 67. Yorkston KM, Kuehn CM, Johnson KL, Ehde DM, Jensen MP, Amtmann D. Measuring participation in people living with multiple sclerosis: a comparison of self-reported frequency, importance and self-efficacy. Disabil Rehabil 2008;30:88-97. 68. Fougeyrollas P, Cloutier R, Bergeron H, Coté J, Coté M, St Michel G. Revision of the Quebec Classification: Handicap Creation Process. Quebec: International Network on the Disability Creation Process; 1997. 69. Willer B, Ottenbacher KJ, Coad ML. The Community Integration Questionnaire: a comparative examination. Am J Phys Med Rehabil 1994;73:103-11. 70. Ostir GV, Granger CV, Black T, et al. Preliminary results for the PAR-PRO: a measure of home and community participation. Arch Phys Med Rehabil 2006;87:1043-51. 71. Bode R. Measuring participation: the patient-reported outcomes measurement information system experience. Arch Phys Med Rehabil 2010;91(9 Suppl 1):S60-5. 72. Dijkers M. Comments on van Brakel et al.’s Participation Scale. Disabil Rehabil 2006;28:1360-2; author reply 1363-4. 73. van Brakel WH. Reply to Professor Dijkers. Disabil Rehabil 2006;28:1363-4. 74. Tate D, Forchheimer M, Maynard F, Dijkers M. Predicting depression and psychological distress in persons with spinal cord injury based on indicators of handicap. Am J Phys Med Rehabil 1994;73:175-83. 75. Dijkers M, CHART study group. Scoring CHART: survey and sensitivity analysis. J Spinal Cord Med 1991. Arch Phys Med Rehabil Vol 91, Suppl 1, September 2010

S16

CONCEPTUALIZATION OF PARTICIPATION, Dijkers

76. Dijkers MP, Yavuzer G, Ergin S, Weitzenkamp D, Whiteneck GG. A tale of two countries: environmental impacts on social participation after spinal cord injury. Spinal Cord 2002;40:351-62. 77. Feinstein AR. Clinimetrics. New Haven: Yale Univ Pr; 1987. 78. Apgar V. A proposal for a new method of evaluation of the newborn infant. Curr Res Anesth Analg 1953;32:260-7. 79. Feinstein AR. Multi-item “instruments” vs Virginia Apgar’s principles of clinimetrics. Arch Intern Med 1999;159:125-8. 80. Fayers PM, Hand DJ, Bjordal K, Groenvold M. Causal indicators in quality of life research. Qual Life Res 1997;6:393-406. 81. Streiner DL. Clinimetrics vs. psychometrics: an unnecessary distinction. J Clin Epidemiol 2003;56:1142-5; discussion 1146-9. 82. de Vet HC, Terwee CB, Bouter LM. Clinimetrics and psychometrics: two sides of the same coin. J Clin Epidemiol 2003;56: 1146-7. 83. Zyzanski SJ, Perloff E. Clinimetrics and psychometrics work hand in hand. Arch Intern Med 1999;159:1816-7. 84. Streiner DL. Being inconsistent about consistency: when coefficient alpha does and doesn’t matter. J Pers Assess 2003; 80:217-22. 85. Mellick D, Walker N, Brooks CA, Whiteneck G. Incorporating the cognitive independence domain into CHART. J Rehabil Outcomes Meas 1999;3:12-21. 86. Dumont C, Bertrand R, Fougeyrollas P, Gervais M. Rasch modeling and the measurement of social participation. J Appl Meas 2003;4:309-25. 87. Wainer H. Estimating coefficients in linear models: it don’t make no nevermind. Psychol Bull 1976;83:213-7. 88. Stone A, Shiffman S, Atienza A, Nebeling L, editors. The science of real-time data capture: self-reports in health research. New York: Oxford Univ Pr; 2007. 89. Stone AA, Turkkan JS, Bachrach CA, Jobe JB, Kurtzman HS, Cain VS, editors. The science of self-report: implications for research and practice. Mahwah: Lawrence Erlbaum Associates; 2000. 90. Reis HT, Gable SL. Event-sampling and other methods for studying everyday experience. In: Judd CM, Reis HT, editors. Handbook of research methods in social and personality psychology. New York: Cambridge Univ Pr; 2000. p 190-222. 91. Schwartz JE, Stone AA. Strategies for analyzing ecological momentary assessment data. Health Psychol 1998;17:6-16. 92. Stone AA, Shiffman S. Capturing momentary, self-report data: a proposal for reporting guidelines. Ann Behav Med 2002;24:236-43. 93. Seekins T, Ipsen C, Arnold NL. Using ecological momentary assessment to measure participation: a preliminary study. Rehabil Psychol 2007;52:319-30.

Arch Phys Med Rehabil Vol 91, Suppl 1, September 2010

94. Glueckauf RL, Quittner AL. Assertiveness training for disabled adults in wheelchairs: self-report, role-play, and activity pattern outcomes. J Consult Clin Psychol 1992;60:419-25. 95. Tulen JH, Stronks DL, Bussmann JB, Pepplinkhuizen L, Passchier J. Towards an objective quantitative assessment of daily functioning in migraine: a feasibility study. Pain 2000;86: 139-49. 96. Schasfoort FC, Bussmann JB, Stam HJ. Correlation between a novel upper limb activity monitor and four other instruments to determine functioning in upper limb complex regional pain syndrome type I. J Rehabil Med 2005;37:108-14. 97. Cooper RA, Thorman T, Cooper R, et al. Driving characteristics of electric-powered wheelchair users: how far, fast, and often do people drive? Arch Phys Med Rehabil 2002;83:250-5. 98. Tolerico ML, Ding D, Cooper RA, et al. Assessing mobility characteristics and activity levels of manual wheelchair users. J Rehabil Res Dev 2007;44:561-72. 99. Dudek NL, Khan OD, Lemaire ED, Marks MB, Saville L. Ambulation monitoring of transtibial amputation subjects with patient activity monitor versus pedometer. J Rehabil Res Dev 2008;45:577-85. 100. Castel LD, Williams KA, Bosworth HB, et al. Content validity in the PROMIS social-health domain: a qualitative analysis of focus-group data. Qual Life Res 2008;17:737-49. 101. Dijkers M, Brown M, Waterston L, Oldfield B. Cognitive interviewing techniques to improve a participation questionnaire used with persons with traumatic brain injuries. Arch Phys Med Rehabil 2006;87:e3. 102. Lawlor K, Mihaylov S, Welsh B, Jarvis S, Colver A. A qualitative study of the physical, social and attitudinal environments influencing the participation of children with cerebral palsy in northeast England. Pediatr Rehabil 2006;9:219-28. 103. Naraian S. “I didn’t think I was going to like working with him, but now I really do!”: examining peer narratives of significant disability. Intellect Dev Disabil 2008;46:106-19. 104. Ryan S. “I used to worry about what other people thought but now I just think . . . well I don’t care”: shifting accounts of learning difficulties in public places. Health Place 2008;14: 730-9. 105. Lund ML, Nordlund A, Bernspang B, Lexell J. Perceived participation and problems in participation are determinants of life satisfaction in people with spinal cord injury. Disabil Rehabil 2007;29:1417-22. 106. Anke AG, Fugl-Meyer AR. Life satisfaction several years after severe multiple trauma—a retrospective investigation. Clin Rehabil 2003;17:431-42. 107. Johnston MV, Keith RA, Hinderer SR. Measurement standards for interdisciplinary medical rehabilitation. Arch Phys Med Rehabil 1992;73:S3-23.