The Action Project Method Applied in Nursing Home Settings

Abstract

The Action-Project Method (A-PM) is a qualitative research approach used to understand the actions and experiences co-constructed by individuals. We applied the A-PM in a nursing home (NH) setting with the aim to explore how NH residents and the people closest to them describe their priorities for care and act on these priorities. Due to the health of residents, the demands on staff and family members, and general issues around scheduling, applying the A-PM in the setting required adaptations. The core focus of this article is on the necessary adaptations to apply the A-PM in a NH setting. (1) The A-PM is typically longitudinal with multiple data collection cycles; however, given the circumstances surrounding the residents’ health and the highly structured institutional setting, we opted for a single round of data collection. (2) During recruitment, three residents asked to participate alone, which we accommodated to acknowledge their experiences. (3) The setting posed challenges to data collection such as ensuring privacy and avoiding interruptions, but these challenges often reinforced participants’ experiences. With the necessary adaptations the A-PM in NHs gives voice to participants while contextualizing that within their core relationships.

Introduction

The Action-Project Method (A-PM) is a consensus-based method for conducting qualitative research. It has previously been used in counselling psychology as well as family and health studies to explore the actions and experiences that are co-constructed between individuals (Young et al., 2002, 2011). The method is specifically concerned with how people engage together to move towards joint and individual goals over time (Young et al., 2002). Our aim is to share our experiences and lessons learned when applying the A-PM in a nursing home (NH) setting for a project designed to gain insight into the experiences of residents. The resident experience of quality of care has primarily been studied with interview-based approaches (Coughlan & Ward, 2007; Helgesen et al., 2020), which can lack authentic insight into the resident experience and reduce the residents’ role in NH to care recipients. The A-PM has a unique ability to change the focus on residents from self-contained care recipients to agents making goal-oriented actions within interpersonal processes. This level of resident agency is rare in NH research and it acknowledges the different ways the resident experience is inextricable from contributions from others as well as NH culture. We used the A-PM, which positions residents and their care partners as active participants within the limits of institutional care, to explore the goal-directed actions of residents and the people closest to them. Unlike other qualitative approaches that rely on one-on-one interviews or focus groups, the A-PM emphasizes relational dynamics. The data collected with A-PM in our study reflect the everyday actions and language of the residents as they engage with their care partners (e.g., family members, friends, neighbours, companions, or paid care staff). The results of this work are reported elsewhere (Gruneir et al., 2021).

The A-PM has been adapted and used in many settings, such as learning institutions (Marshall et al., 2020; Marshall et al.,2011; Marshall et al. 2014; Wall et al., 2019; Young et al., 1999; Zaidman-Zait et al., 2014), community meetings rooms (Young et al., 2001), private homes (Marshall et al., 2018) and counselling offices (Marshall et al., 2008; Young et al., 2011, 2021). In these studies, researchers used the method to explore how socially connected partners act separately and together to navigate day-to-day life. NHs provide specific challenges to applying this method, especially pertaining to data collection in the institutional setting and the cognitive and physical limitations of the residents. The advantage of the A-PM in this setting is that it considers the relationships among residents, families, and care staff in ways that many other methods do not. Social relationships are essential to quality of care in NHs (Bergland & Kirkevold, 2005; Bowers et al., 2001; Haugan, 2013) and the resident experience is intrinsically linked to their relationships with staff (Bowers & Roberts, 2015; Heliker & Nguyen, 2010), peers (Achterberg et al., 2003; Bowers & Roberts, 2015) and family members (Bauer et al., 2013; Baumbusch & Phinney, 2013). Understanding the different types and dynamics of relationships is therefore paramount when exploring the resident experience of the care they receive.

In this article we present how contextual characteristics of the NH setting required us to adapt our application of the A-PM. We begin by outlining the background of the study as well as our team, then account for the theoretical underpinnings of the A-PM and its typical protocol. We then present our considerations of the barriers for applying the A-PM in a NH setting and the adaptations we made to the method as well as the utility of the protocol in the context of institutional care.

Study Context

The Overall Project

This article is part of a larger research study exploring priorities for care in NHs by residents, family members, staff, and health system decision-makers (Gruneir et al., 2021). The goal of the research is to develop a core set of measures to report on burdensome symptoms and potentially inappropriate care practices that would be useable by different stakeholder groups. To collect data directly from NH residents and those close to them on their priorities for care, we applied the A-PM, which, to the best of our knowledge, has not been used in this setting before. This study was approved by The University of Alberta Research Ethics Board and by each of the participating NHs.

The research team included two co-leads (one epidemiologist and one nursing/health services researchers) as well as one Research Associate with prior A-PM experience. The team also included one postdoctoral fellow with training in gerontology, and a Research Associate with experience in qualitative analysis. Additionally, two research assistants helped conduct the data collection (one with training in sociology and one with training in nursing) and two other research assistants helped with the data analysis (one with training in public health and one with training in social work). All data collection and analysis were overseen and guided by an experienced researcher with expertise in the A-PM.

Study Setting

Data for this study were collected from three NHs in Edmonton, Alberta, Canada. The Canadian NH system is publicly funded and admission to NH is based on a standardized assessment for level of need for care. The day-to-day care in NHs is primarily provided by care aides who assist in tasks such as transferring residents in and out of bed, dressing and personal hygiene. In Alberta, NHs are “(…) designed specifically for individuals with complex, unpredictable medical needs who require 24-hr onsite Registered Nurses assessment and/or treatment” (Alberta Health, 2018). Nursing home residents are primarily older adults, 80% of whom are 75 years of age or older, with substantial day-to-day care needs including a high prevalence of dementia and other conditions that impact cognitive and physical function (CIHI, 2019; Hoben et al., 2019).

Study Sample

Given that our focus was on the resident experience in NH, we chose to recruit the residents first and then ask them to identify a potential care partner (either family or staff member) with whom they would like to participate, to explore the co-construction of action between residents and people with meaningful relation to their care. We recruited 17 residents from three major not-for-profit NHs in Edmonton, Alberta, 11 participants were female and five were male. We did not collect specific ages from the participants. We contracted a licensed practical nurse (LPN) from each NH to assist with recruitment. Eligible residents were approached and if they were interested in participating, the LPN would obtain written consent for contact by the research team. Residents had to be 60 years of age or older and be able to carry out a 10–15-min conversation in English. The age criteria were to keep the focus on the relevant population of NH residents. People under 60 make up a very small group in NH (approx. 6%) (Alberta Health, 2018) and typically have very different needs such as high levels of physical disability, which makes their experience potentially very different and outside the scope of this study.

Five residents identified as their care partner a family member (two adult children, three partners) and seven identified a staff member (one paid companion, one recreation therapist, and five care aides working for the NH). Three residents chose to participate without a care partner, and we opted to include their perspective to explore as diverse a range of experiences as possible. One resident died before data collection could be completed and another was excluded because of recording failure during data collection. The final sample included 15 residents.

Study Method

We chose the A-PM because it provides insight into the ways that residents’ lives are co-constructed within the context of institutional care. NH care is relational, directly influencing the experience of care recipients and providers (Baumbusch & Phinney, 2013; Heliker & Nguyen, 2010). In addition to care staff, residents also rely profoundly on family or close others as their care partners for support and care, even when living in a NH setting (Drageset, 2004; Kiely & Flacker, 2003). The A-PM data were collected without researchers present in the room, aiming at minimizing the power imbalance between researcher and participants. However, even researchers as an implied audience might influence how the participants (residents, as well as care partners) respond. We will discuss these concerns in more detail throughout the article.

Contextual Action Theory

We adopted contextual action theory (Young et al., 2002) to frame the research. Contextual Action Theory (CAT) was originally developed in career and counselling psychology, and from this theoretical approach, individuals’ behaviours are seen as embedded in the social context and actions as goal-oriented and intentional, although not necessarily rational (Young et al., 2011). Actions are assumed to reflect socially constructed meanings. The intent of CAT is therefore not to ask ‘why’ people act in certain ways but instead to ask ‘what’ people are doing to understand the socially constructed and underlying meanings that reflect the goals of behaviors (Young et al., 2021).

Current NH research is concerned with framing behavior in older adults around strategies for communication and including them in resident-centered care approaches (Ellis & Astell, 2017; Haberstroh et al., 2011; Sion et al., 2020; Wong et al., 2020). NH residents, with varying degrees of cognitive impairment, are similarly viewed in the light of CAT as goal-directed individuals who are active in co-constructing their experiences and actions with others. The core aspect of these approaches is the volitional perspective on the role of residents in NH care.

Based on CAT, individual action is considered from the perspectives of manifestation. This includes perspectives of (1) the manifest behavior, (2) the internal cognitive and emotional processes related to behavior and (3) the social meanings people construct around their behaviors (ibid). Action is also considered on three different levels, including as (1) action elements, such as verbal and nonverbal behavior (for example, strings of words put together forming a question); (2) functional steps that serve to reach a goal or future state through linked sequences of actions; (3) goals representing the meaning of action processes and functional steps towards these goals (ibid). Finally, action is considered in systems. Systems of action are (1) individual or joint, both of which are characterized as relatively short-term and anchored cognitively, socially, and environmentally in people’s daily lives; (2) projects, which indicate a series of actions constructed around a common goal, when occurring over a middle-term length of time; (3) careers, which represent the construction and organization of projects in people’s lives with high significance, when occurring over a long-term period (Young et al., 2011). These different considerations of action inform the A-PM and how it is used for qualitative research.

Action-Project Method

In the A-PM, action is the unit of analysis and which might stem from either one individual or socially related individuals working towards goals over time (Marshall et al., 2012). As actions are considered socially constructed, they are often assessed in dyads of individuals who are jointly engaged in some meaningful way. This allows researchers to explore perspectives, levels, and systems of action on both the individual and supra individual level.

CAT informs the A-PM, which typically is applied over multiple rounds of data collection. Within a data collection cycle (Figure 1), researchers collect data in two rounds. Round one includes collection of three types of data after gathering consent: (1) the introductory conversation; (2) the dyadic conversation; (3) the video-recall interview. In round two reseachers present of the results of the initial data analysis of the data from round one in the form of narratives to participants to collect their feedback as data. Figure 1 illustrates the sequence of an A-PM project data collection cycle.

Figure 1.

Sequence of A-PM data collection cycle.

Data Collection: Round 1

A-PM data collection begins with the participating dyad (resident and their chosen care partner) sitting with 2–3 members of the research team for the orientation interview in empty common areas, meeting rooms or private resident rooms. Here, researchers ask the dyad members general questions about themselves, and issues related to the research topic. The next step of the data collection process is a joint conversation between the dyad members. The researchers will ask the dyad to discuss their experience in NH and what they value in terms of the care provided. The researchers leave the room after prompting the conversation, which is audio- and video-recorded. The researchers re-join the dyad after 5–10 min, stop the recording, and then meet with each of the dyad members for the video-recall. The participants sit in separate rooms with a member of the research team. The researcher plays back the video of the joint conversation between the dyad members in approximately one-minute segments, and the participants are asked questions about their internal processes such as “What were you thinking during that part of the conversation?” (Marshall et al., 2012). The video-recall is also audio-recorded.

Analysis: Round 1

After the first round of data collection, all recordings are transcribed. Researchers pair up and code action elements in the transcriptions. Dyad conversations are coded based on actions and not topic of conversations between participants. As action is the unit of analysis, the coding targets actions taken by participants such as “expressing [type of] emotion”, “describing past event”, or “acknowledging dyad partner”.

The video-recall interviews are used to provide additional context and understanding of the participants’ intentions and cognitions. Once the full transcript is coded, researchers begin to describe participants’ goals and how their actions lead to those goals. Goals are then linked to identify longer-term projects. Since A-PM is focused on co-construction between individuals, projects for each participant as well as joint projects shared between the participants are identified. A tentative narrative is written for the dyad and the individuals based on the analysis of goals and projects.

Data Collection: Round 2

The second round of data collection starts with the presentation of individual narratives. Researchers present the tentative narratives written about the actions and projects that emerged during the first round of data collection to each of the participants. The purpose is to elicit feedback from the dyad members regarding the accuracy of the initial analysis of data. The individual narrative presentation is audio-recorded and transcribed. After each participant has had a chance to give feedback on the accuracy of the analysis, the researchers bring the dyad members together in a room and present the joint narrative for the dyad as a whole. This narrative presentation is again audio-recorded and transcribed.

Analysis: Round 2

The narrative feedback in the second round of analysis is used to alter any misconceptions the researcher might have made or add perspectives or nuance that might have been overlooked during the first round of analysis. After researchers make the appropriate changes to the narrative of the individuals and the dyad, a cross-case analysis is conducted. This part of the analysis allows research teams to explore cases that are similar or divergent in themes, patterns, or projects. Typically, cross-case analysis is conducted in iterations. The cases are organized into groups of three or four to facilitate comparison, either through random selection and/or relevant participant characteristics. This is typically done at least twice to ensure that cases are compared against as many other cases as possible. The team works towards consensus, comparing projects and goals across cases to ensure findings accurately represent the research participants’ actions.

Longitudinal Procedure

An A-PM study is usually longitudinal and conducted over a period of between 6 and 10 months (Marshall et al., 2008, 2014, 2018, Young et al., 2001a, 2001b). At the end of the data collection cycle illustrated in Figure 1, a monitoring process begins. Typically, researchers have periodic follow-up interviews (by phone or videoconference) with participants to discuss recent actions and progress on their projects during the monitoring period. At the end of the monitoring period, the interview-based data collection cycle is repeated to gather a longitudinal perspective on development of experiences of the participants. Our A-PM study was cross-sectional rather than longitudinal, something which we will go into more detail about in the discussion. We will be using the rest of this article to illustrate and discuss our considerations surrounding the specific characteristics of the NH setting for the application of an A-PM approach.

NH-Specific Characteristics That Influenced Application of A-PM

Practical Considerations

Some of our primary considerations for applying the A-PM in the context of NH, which to the best of our knowledge no previous A-PM studies have dealt with, related to the institutional structure, schedule, and protocols of NHs. These types of contextual characteristics often intersect with the previously mentioned characteristics of NH residents (dementia, changed physical function and other cognitive decline) (Hoben et al., 2019) and can act as barriers to data collection in a NHs setting.

Institutional Care as a Context

We found that going into residents’ homes presented valuable insight into their experience by creating more opportunities for residents to lead the conversations and show us, and their care partner, their world. For instance, some residents would ask to show the researchers around the NH and their room before conducting the initiation interview. Residents raised issues such as privacy, personal space, or autonomy, all of which they could illustrate by pointing to their current surroundings during the data collection. While the accounts of the residents might have been sufficient in depicting these experiences, the conceptual link between the actions and goals of residents as they navigated their experience within the social context of a NH became even more clear as it also surrounded the researchers and research study itself.

Residents’ Health Conditions

Nursing home residents often have extensive health care needs, including multiple chronic conditions (with a high prevalence of dementia) and physical impairments. To accomodate these conditions, we made several decisions about the frequency and timing of data collection. We chose to eliminate the longitudinal aspect of data collection and instead collect data in one cycle in which the video-recall was conducted immediately after the dyad interview and the narrative review was conducted within 6 weeks of the first round of data collection. In addition to reducing the potential burden on residents, this minimized the scheduling burden on facilities and staff. Cognitive impairment is highly prevalent in the population so our condensed timeline for data collection enabled us to maximize engagement with residents who may have had more difficulty with a longitudinal design. The fact that one resident died before we had a chance to do the narrative review illustrates how this adaptation to a cross-sectional study was necessary.

To accommodate residents’ cognitive and sensory impairments, we also adapted the process of the video-recall interviews. Typically, during the video-recall, parts of the joint conversation video-recording would be played back to each participant individually and a member of the research team would ask questions to get more information about the intentions, goals and meanings behind that part of the joint conversation. However, some of the residents had hearing problems or cognitive impairments, which required the researcher to repeat what was said in the video-recorded conversation, sometimes also requiring further condensing or paraphrasing to enhance the resident’s understanding.

With our modifications of the timeline and the process of the video-recalls, we found that participants were very willing to share their experiences and used each point in the data collection cycle to do so, ultimately resulting in rich data. The adaptations we made to accommodate residents’ needs and their willingness to participate increased the time needed to collect data. While the general timeline for the study was reduced by eliminating the longitudinal aspect, we spent more time in the first round of data collection to accommodate the needs of the residents and the facilities. We realized after our initial interviews that we had to space the interviews out due to residents requiring more time and assistance with the process, as well as the occasional showing researchers around the facility and providing further introduction into their daily lives.

Scheduling

Scheduling data collection required coordination with residents and their care partners. For residents, fitting a research interview into the care schedule was challenging. Interviews needed to be scheduled around other fixed activities such as meals, showers, medication administration, physical therapies, and recreational activities. There was therefore only a small window each day that worked for the residents, which then had to be coordinated with the availability of the care partner. Interviews involving staff took place during their shift, which required flexibility during the scheduling of the interviews. Staff members were busy and did not always have predictable schedules. If the care partner was a family member, that required scheduling a time when they could be at the NH. In all cases, coordination with staff was necessary as we needed to ensure that the resident would be ready and available at the scheduled time. Because we chose to conduct a cross-sectional version of the A-PM rather than a longitudinal, we were able to schedule the joint conversation and the video-recall back-to-back. Also, because staff were already on-site, there was no additional travel required on their part to participate. The second round of data collection was scheduled for 6 weeks after to accommodate fluctuations in family and staff members’ schedules as well as residents’ health concerns.

Privacy

Ideally, completely private spaces would be available for the two types of interviews (joint conversations and video-recalls) embedded in the data collection process. However, many NH residents live in shared rooms with limited seating and space for visitors. Most NHs have limited meeting spaces outside the residents’ rooms. This presented some challenges for data collection. Sometimes there were private areas available for our team to conduct the joint conversations or the video-recall interviews, but many times data collection would take place in the resident’s room or when dyad members had to be split up, in another shared (but unoccupied) space. Our data collection was challenged as we had to spend more time looking for private spaces to conduct the interviews to keep the conversations confidential and avoid the institutional lack of privacy influencing what the dyad members chose to talk about. While our use of secluded common spaces is not ideal, it does illustrate the lack of respect for the residents’ privacy within the NH setting.

Meaningful Relationships

Relationships are at the core of the A-PM, and we carefully considered the types of relationships that might be meaningful for residents in exploring their experience of and priorities for care. In the A-PM, researchers can determine in advance which types of relationships can be included as dyads in a study. It can be beneficial to have an idea of what types of relationships might be relevant to one’s study, but it is important to let the participants define the dyads for them to be truly meaningful to their experience.

Dyads

We asked residents to choose a dyad partner who they had a close relationship with and who was involved with their care and/or decision-making. This resulted in dyads composed of residents and care aides (n = 5), other staff (one recreational therapist, one paid companion) (n = 2), spouses or partners (n = 3) and adult children (n = 2).

We gave residents the option to choose staff members as their care partners because the relationships that can form between staff and resident can become deeply intimate (Cook & Brown-Wilson, 2010). However, sometimes residents seemed to choose staff to avoid burdening family or to avoid having to discuss challenging topics with people close to them (e.g., medical, or emotional challenges).

We found that the interactions in different types of relationships varied, especially between staff and family member dyads in general. Dyads with family members often included actions centering on navigating changes in long-term relationships, physical separation and adaptation to new roles in the relationship, such as children or romantic partners/spouses becoming caregivers. In dyads with staff members, the staff member tended to take on an interviewer role rather than conversation-partner role, reflecting the task-like nature of the jobs as well as their professionalized approach to the role as dyad partner. The fact that data collection occurred during staff members’ shifts and in their workplace may have contributed to the observation that staff often treated residents as someone whose needs they were hired to meet during data collection. This asymmetrical relationship between residents and staff kept power imbalances relevant for these dyads. Staff members have control over residents’ daily routines, which could impact whether residents were willing to speak out against a staff member if they disagreed or objected to the conversation. Additionally, the use of the A-PM let us directly observe the impact of staffing-related issues to the experience of residents. When staff members participated, our data collection was under time constraint as the staff would need to return to their workday. While we were able to work around this with scheduling, it did allow us to observe the way the pace of the staff’s work would impact the residents and their interactions directly.

Single Participants

As previously outlined, the unit of analysis in A-PM is action, particularly joint action, explored through dyadic conversations. Although we did not initially expect to be working with single participants, some residents expressed interest in participating in the study but did not want to involve a family or staff member. We considered their perspectives important and opted to include them. Some past A-PM studies have included single participants as well as dyads (Marshall et al., 2018). The single participants had different reasons for wanting to participate without a care partner, either because they found the topic too personal or because they did not want to burden others by asking them to participate. We decided that residents’ engagement with members of the research team would emulate the data of the dyad participants, or at least provide data with divergent action patterns to enrich the analysis. While a care partner would have been ideal for the study, the single participants still engaged in interpersonal processes and co-constructed their actions and experiences with others, in these cases with the researchers. Furthermore, excluding single residents would have meant excluding people without close relationships involved in their care, who make up a significant part of the NH population.

Discussion

In our primary research, our aim was to understand residents’ experience of quality NH care within the context of the relationships that construct the experience. To accommodate the specific needs of residents and the setting, we made several adaptations to the A-PM data collection protocol that we have described in this article, which aims to highlight the adaptability of the A-PM while maintaining its emphasis on nuance in individual and joint action research. The most notable adaptations include the shortened data collection timeframe and the flexible approach to care partner recruitment. In recruitment, we asked residents to choose anybody who they considered important to their care and/or assisted in decision-making. This resulted in a mix of family members and staff members as care partners; it also resulted in residents opting to participate alone.

Nursing homes are well-recognized as a challenging setting within which to collect data for research, regardless of the study methods. Several of the challenges that we grappled with in our research have been raised by others, including the need for an intermediary in the NH to facilitate recruitment, complicated scheduling, staff workload and its impact on their ability to participate, and the limited quiet spaces for data collection (Marshall et al., 2020; Maas et al., 2002). One of the primary challenges, of course, is the advanced degree of illness and the high frequency of dementia in the NH population. This, in part, led to our decision to shorten the data collection timeframe. This meant that we no longer benefited from the longitudinal aspect typical to A-PM applications, however, it also reduced the potential for high loss to follow-up and the challenges posed by repeated scheduling. We cannot fully know how this impacted our study. The benefit of a longitudinal A-PM study is the insight into how actions and experiences change over time, which we lack. However, the A-PM has previously been used with a similar reduced data collection timeframe and yielded nuanced findings (Marshall et al. 2018, 2020), and the richness of data we accumulated by accommodating the needs of the setting and the population mitigates the potential loss of rigor.

In our study, we also restricted inclusion to residents who could hold a conversation for approximately 10 min. This meant that we excluded anyone with moderate to severe cognitive impairment and impairment of speech. While we worked with the recruiting nurse to make sure they understood the scope of the study and the type of participants we were looking for, we had no control over or insight into how they approached the NH residents. Because we were removed from the original recruitment, we don’t know how the participants were approached, how many might have been approached but declined to participate, who was overlooked and what the personal biases of the recruitment nurse might have meant for the residents that were deemed eligible for participation. Strategies such as integrating augmentative and alternative communication supports (Light et al., 2019) as well as attention to non-verbal communication (Ellis & Astell, 2017; Zaidman-Zait & Young, 2008) may be useful in expanding inclusion to residents with advanced cognitive impairment or other communication impairments. Although we did not use these strategies, they could be integrated within the A-PM data collection and analysis protocols to enable inclusion of a larger segment of the NH population.

Our study showed that the A-PM can be successfully implemented in NHs with the described adaptations (and likely others) and can add to the existing research by tapping into both the relational aspects of care and the action-oriented nature of those relationships. For example, by allowing the resident to choose either a family member or staff member to join them, we were able to observe both expected and unexpected differences in their interactions. While family members often led the conversation with the resident, their interaction patterns were generally conversational in nature, whereas staff members more often took on an interviewer-like role reflecting the task-oriented nature of their work. In our study and others, residents have reported the centrality of relationships to their NH experience and the A-PM enables those relationships, and the ways that they are enacted, to be the focus of interrogation.

However, power imbalances remain relevant. Two main types of power imbalances were key to our application of the method in the NH setting: (1) the power imbalance between researchers and residents and (2) the power imbalance between care partners and residents. It was our experience that the presence of the researchers did not seem to impede residents’ openness to talk as they were eager to share their experiences and correct researchers when needed. If anything, the researchers introduced an artificial presence, which did not seem to significantly impact the openness of participants. The main power imbalance remained between residents’ and their chosen care partners particularly staff members. It is plausible that residents could be concerned about the consequences for speaking up against staff members, which could translate to them being reluctant to express their true goals and take authentic action in front of staff who could view it as criticism. This did not come up in any of our data collection, but it is important to distinguish between the artificial power of the researchers and the actual power of the dyad members. It is our experience that no qualitative method can claim to be completely free of power imbalances and that the A-PM remains useful in its attempt at enhancing the voice of the residents within interpersonal processes with a limited artificial power of a researcher present.

Regardless of the dyad type, by focusing on relational dynamics, we captured well-meaning but paternalistic interactions and the tendency for care partners to focus on positive topics and steer residents away from more serious concerns. For those residents who chose to participate alone, the method is flexible enough to include them while keeping the focus on their actions and goal-oriented behaviours. In our case, including residents who wanted to participate alone was itself important to capturing different ways in which their relationships, or limited relationships, to both family and staff shaped their experiences within the NH.

We have shown that the A-PM can be successfully adapted for use in the NH setting. By doing so, we were able to collect rich data on the resident experience embedded within relationships to key family and staff members. Other NH researchers should consider the A-PM as a useful addition to the cadre of research methods already in use but also look to further adaptations for its use with a wider population. The successful use of the A-PM in NHs should also prompt others to consider its application in other congregate living settings.

Strengths and Limitations

There are limitations to consider. First, when we worked with single participants the researchers were not just the implied audience but a direct audience. Arguably, this could have influenced the residents to align their actions to the prompts provided by the researcher. We used an unstructured interview approach to keep the data collection with single participants conversational in nature, however, we did still miss out on the nuance of the shared experiences of residents and their care partners that the dyad interviews reflected. At the same time, we found that single participants were very open to share their experience and engage in co-construction of actions with the interviewers in a way that offered rich insight and projects that reflected similar themes as to what we observed with the dyads.

Second, we included some participants who required support in the conversations, resulting in researchers needing to paraphrase during the video-recalls. We felt that it was better to give voice to residents with moderate cognitive impairments since their voices are often not heard in research even though this strategy deviated from the typical A-PM protocol. This did mean that nuance might have been lost or the researchers’ interpretation of the residents’ own words might have influenced the participants to agree with the paraphrasing. Third, dyads often had a hard time getting the conversation started, which consequently might have meant that the research team ended the conversation too early.