Concept of Participation
By linking the meaningful concepts identified in the participation instruments, it was possible to determine which ICF categories the instruments include. In this study an instrument was considered to assess the concept of participation and included if its domains cover a minimum of three chapters (domains) between d3 Communication and d9 Community, social and civic life in the ICF component activities and participation. This broad definition of participation was used since there is no consensus regarding how activity is differentiated from participation [2, 5, 11–14] and selecting chapter headings provided objective criteria. In considering which activities and participation domains the instruments cover, an even broader definition of participation was used by also including d1 Learning and applying knowledge and d2 General tasks and demands since these domains may have been considered relevant to the concept of participation by the instrument developers. Perenboom and Chorus , however, considered a question to be assessing participation if it asks about "actual or perceived participation (involvement, autonomy, social role)" (page 578) and so different results would be obtained using this definition.
Content of the Participation Instruments
Although all the instruments cover six to eight of the nine activities and participation domains, there are differences in the actual content. All of the instruments include content from domains d6 Domestic life, d7 Interpersonal interactions and relationships, d8 Major life areas and d9 Community, social and civic life. However, there are differences in whether the domains d3 Communication, d5 Self-care and certain aspects of d4 Mobility are considered aspects of participation.
Four instruments (PM-PAC, P-Scale, ROPP, WHODAS II) intend to assess d3 Communication based on the materials describing their development and ICF categories from d3 Communication were noted for all these instruments. Meaningful concepts linked to categories in d3 Communication were also identified in the KAP and POPS which is likely not the major focus, as the questions have meaningful concepts linked to multiple ICF domains. For example, in the POPS the question 'How many times do you speak with your neighbour?' includes the meaningful concept 'conversation' which was coded as d350 Conversation but it is only a minor meaningful concept and the major meaningful concept is 'relationship with neighbour(s)', coded as d7501 Informal relationships with neighbours. In some instruments, such as the PM-PAC, assessing communication is a major focus ('How much are you limited in watching or listening to the television or radio?'). Empirical findings suggest that it is difficult to demonstrate discriminant validity among participation domains [15, 17] and this may be a result of overlapping content. In future studies it may be beneficial to identify and code the major and minor meaningful concepts, since this could assist with developing a priori hypotheses regarding expected correlations between instrument domains.
All of the instruments contain meaningful concepts linked to categories in d5 Self-care with the exception of the POPS. When the POPS was developed self-care was not included since participation was operationalized as "engagement in activities that are intrinsically social, that are part of household or other occupational role functioning, or that are recreational activities occurring in community settings" (page 463) and self-care did not qualify . The PM-PAC does not intend to assess self-care  but there were two meaningful concepts linked to d5 Self-care. One question in the PM-PAC asks about 'exercising' which was coded as d5701Managing diet and fitness and the other question asks about 'providing self-care to yourself', which was coded as d5 Self-care. In terms of mobility, all of the instruments contain meaningful concepts linked to categories in d4 Mobility and all the instruments intend to include content from this domain. Three instruments (IPA, PARTS/M, WHODAS II) operationalize moving in the home using specific phrases such as 'getting out of bed', 'getting out of a chair' (PARTS/M) or 'getting up and going to bed' (IPA). In the other instruments, mobility includes broader statements such as 'moving or getting around the home' (KAP, PM-PAC, P-Scale, ROPP) and in the POPS mobility includes only using transportation.
Two instruments, the P-Scale and WHODAS II, were considered to have content not related to the concept of participation, which was defined broadly as ICF categories in the activities and participation domains d1 Learning and applying knowledge to d9 Community, social and civic life. The P-Scale has one question which only asks about the observable attitudes of others ('In your home, are the eating utensils you use kept with those used by the rest of the household?'). The WHODAS II contains five questions which ask about content related to body functions (e.g. 'remembering' which was linked to b144 Memory functions) or were not covered/not defined (e.g. 'barriers or hindrances in the world around you'). By linking the meaningful concepts to the ICF classification it was evident that not all questions appear to assess participation as defined in the ICF. This information may assist users in understanding what the questions assess and aid in selecting an instrument depending on his or her purpose, since this may or may not be an issue.
Linking the Meaningful Concepts to the ICF
The methodology published by Cieza et al.  was used to identify and link meaningful concepts to the ICF. Our results for the activities and participation codes for the WHODAS II can be compared to a study by Cieza and Stucki , which also linked the WHODAS II to the ICF. It is difficult to compare the results from these two studies directly since Cieza and Stucki  used an older version of the linking rules  and we modified the linking rules by including 'health condition' as a meaningful concept if it was included in the instructions. Cieza and Stucki  identified 38 meaningful concepts and in our study we had 45 not including coding 'health condition', however, we did not include the five questions in the WHODAS II on general health and it appears that Cieza and Stucki  did. Both studies had the same number of meaningful concepts linked to body functions (n = 3), environmental factors (n = 1) and 'not defined' (n = 2). There were some differences. We linked 38 meaningful concepts to categories from activities and participation and Cieza and Stucki  linked 30 meaningful concepts and we linked one meaningful concept to 'not covered' whereas these authors linked two meaningful concepts.
The implications of not reliably determining if the meaningful concepts can be linked to the ICF classification or differences in the ICF categories and codes selected can impact the results and how the questions in the instruments are interpreted. It has been recognized that there are a number of challenges with using the linking rules (e.g. establishing the meaningful concepts contained in the assessment items) . Offering on-line training on how to use the ICF linking rules and presenting difficult coding examples are types of initiatives that could help improve the standardization of this methodology.
Participation and Other ICF Categories and Codes
Meaningful concepts included in the instructions as well as within each question were examined to determine the context in which aspects of participation are assessed. The ICF states that disability is a dynamic process which results from the interaction of the ICF components (body structures, body functions, activities and participation) and the contextual factors (environment, personal factors) . It is helpful to identify what is asked in relation to participation; for example, for every participation topic area (e.g. dressing, working inside the home) included in the PARTS/M, a question is asked if participation is impacted by pain and/or fatigue. Clinically it is useful to determine the impact of factors such as pain and fatigue, since similar to environmental factors they can be potentially modified in order to enhance participation.
As stated by Nordenfelt  and others , activity and participation must occur in an environment. In the ICF there is reference to a 'standard environment' versus 'usual environment' and this distinction is one way activity is differentiated from participation . It is interesting how environmental factors asking about assistance or equipment are included in some instruments (IPA, KAP, PARTS/M, PM-PAC, POPS, P-Scale) but not in other instruments (ROPP, WHODAS II). The PARTS/M specifically assesses the use of assistance and the frequency which accommodations, adaptations or special equipment is used. Asking about the use of equipment and assistance is important clinically since a person's environment can often be modified to enhance their participation. Further qualitative and quantitative studies will determine if respondents inherently consider their environment when answering the questions.
Similar to environmental factors, there is variation in whether a participation restriction is attributed to a health condition. In the WHODAS II and IPA the instructions state that the respondent should consider their health condition or disability. In the PARTS/M there are specific questions which ask if the person's participation is limited by their illness or physical impairment. Dubuc et al.  demonstrated the importance of specifying whether the participation restriction is a result of a health condition or not, especially for areas which are highly influenced by environmental factors. By asking if the participation restriction is a result of a health condition, it underestimated the influence of the environment since subjects focussed on the implications due their health and did not often consider the restrictions in the physical and social environment . More research should determine the best way to assess these influencing factors. The PARTS/M offers the advantage of asking specific questions with and without the influence of health and the environment which may help determine the causes of the participation restrictions and also provide potentially 'pure measures' of participation. None of the instruments have meaningful concepts coded as personal factors, which is not surprising since this data is often collected separately (e.g. age, gender) in research studies. Further studies should compare questions that either attribute or do not attribute participation to factors such as the environment or health conditions to determine if these phrases influence a person's response.
There are several limitations to this study which need to be considered when interpreting the results. In this study only instruments which were developed using the ICF were included and the meaningful concepts were linked to the ICF classification, which limits the findings to how participation is conceptualized in the ICF. In addition, the criteria assume it is desirable to have an instrument cover the majority of areas within a multidimensional concept such as participation and so it may not be suitable for instruments which focus on selected areas such as employment. By linking the meaningful concepts in the questions to the ICF classification it provided an objective evaluation, however, it is possible that we did not capture the correct meaning of the questions. Since very few studies have linked the instruments used in this study to the ICF classification, the results from this study should be confirmed in other studies. Interpreting the questions and determining the meaningful concepts can be influenced by culture and the experience of the coders and enhancements to the ICF linking rules will help improve the assessment of content validity in these types of studies.