In 2011, the Canadian Journal on Aging/La Revue canadienne du vieillissement (CJA/RCV) celebrated its 30th anniversary. At that time, the incoming editor-in-chief wrote an editorial in which she pointed out that the journal needed to build on past successes, identify challenges, and reflect on gerontological research trends as it moved forward as a “mature” journal (Penning, Reference Penning2011). Penning also invited input from scholars on how to ensure the journal’s continued success. Systematic, in-depth analyses of Canada’s only gerontology-focused academic journal can contribute to meeting the goals laid out by Penning by providing detailed examinations of its publishing history. Knowing exactly where one has come from can help with deciding where one needs to go. With that in mind, this case study analysis examines qualitative research published in the CJA/RCV over an 18-year period, from 1995 to 2012.
We focus on qualitative research because some researchers have suggested that it is increasing in the gerontology field (e.g., Cobb & Forbes, Reference Cobb and Forbes2002; Roberto, Blieszner, & Allen, Reference Roberto, Blieszner and Allen2006; Schoenberg, Reference Schoenberg2011), but few systematic analyses have been carried out to test this assertion. Additionally, Penning (Reference Penning2011) noted that the CJA/RCV encourages aging scholarship diversity in a number of ways such as through topics and research designs, including quantitative, qualitative, and mixed methods research, but empirical evidence of this research design diversity has not been published. In our analysis, we first identify the proportion of published qualitative research, relative to quantitative research, noting trends across time (in 3-year time segments). We further examine the CJA/RCV’s scholarship diversity by looking at the topics covered in these qualitative studies, the qualitative methodologies and data collection strategies used in them, and any patterns among these three characteristics. In the discussion section, we provide recommendations for future qualitative gerontological research and for the CJA/RCV.
Literature Review
Historically, positivism and the “scientific method” have dominated gerontology research, resulting in a high proportion of quantitative research (Schoenberg, Shenk, & Kart, Reference Schoenberg, Shenk and Kart2007) relative to qualitative research, but the benefits of qualitative research in gerontology are well documented. It allows for an examination of complex aspects of aging that would be otherwise overlooked in quantitative research (Cobb & Forbes, Reference Cobb and Forbes2002; Schoenberg et al., Reference Schoenberg, Shenk and Kart2007), and qualitative methodologies are effective for accessing hard-to-reach groups. Qualitative methods are appropriate for studying older families due to the interdependence of family relationships (Matthews, Reference Matthews1993). Qualitative aging research has important applied relevance: “The privileges that qualitative researchers derive from their ‘up close and personal’ approaches provide critical insights for policy planning, appropriate service provision, and optimal provider and client/patient communication” (Schoenberg et al., Reference Schoenberg, Shenk and Kart2007, p. 9). Additionally, mixed methods research (work that includes both qualitative and quantitative components) is helpful in collaborative, practical, and applied research (Tashakkori & Teddlie, Reference Tashakkori and Teddlie2003), which describes much gerontology research.
Little is known about the prevalence of qualitative research in gerontology, however. Only three studies (Roberto et al., Reference Roberto, Blieszner and Allen2006; Schoenberg, Reference Schoenberg2011; Schoenberg et al., Reference Schoenberg, Shenk and Kart2007) have examined this topic, none of which included the CJA/RCV in their analyses. Roberto et al. (Reference Roberto, Blieszner and Allen2006) reviewed 13 journals over a 10-year period (the 1990s) for content on family gerontology theory, noting a small increase in qualitative research during that time: 9.4 per cent in the early 1990s, 15.9 per cent in the mid-1990s, and 18.5 per cent in the late 1990s. Overall, they categorized 13.5 per cent of the empirical articles in the 1990s as qualitative and 6.2 per cent as mixed methods. However, five of the 13 journals they reviewed were not gerontology journals, as their focus was on family gerontology.
In 2007, Schoenberg and her colleagues examined three years (2003 to 2005) of content in The Gerontologist, Journal of Applied Gerontology, and Journals of Gerontology Series B: Social Sciences. They found that approximately 10 per cent of the articles in those three journals during those years were qualitative in nature, but noted that because they did not know “the total distribution of qualitative versus quantitative research designs in aging studies, [they lacked] a basis for suggesting what a reasonable distribution might be” (p. 5). However, they did suggest that a 10 per cent distribution seemed low considering the utility of qualitative gerontological research. A higher number of years (i.e., not just three years) would provide a better indication of trends. Schoenberg analysed The Gerontologist once more in 2011, again covering only three years (2008, 2009, and 2010). She suggested that its qualitative content had increased since 2007, but numbers were provided in the article rather than percentages (the proportion of qualitative research published as a proportion of all the empirical studies published in that journal).
No statistics are available for qualitative gerontology research in Canada, despite the fact that researchers attempting to document the prevalence of qualitative research or other characteristics of gerontology research have called for the inclusion of Canadian and international gerontology journals in such content analyses (e.g., Alley, Putney, Rice, & Bengtson, Reference Alley, Putney, Rice and Bengtson2010; Schoenberg et al., Reference Schoenberg, Shenk and Kart2007). The only information we could locate about the prevalence of qualitative research in Canadian gerontology was in a 1988 article that reviewed Quebec social gerontology research published in a variety of periodicals. In that article, Béland stated that francophone gerontology researchers in Quebec vastly preferred qualitative research to quantitative research. Whether or not this preference still exists is unknown, but given that the CJA/RCV publishes articles in English and French, it is possible to examine the proportion of English and French articles that are qualitative versus quantitative.
Finally, although percentage trends are helpful in informing researchers about the popularity of qualitative research in academic gerontology journals, they reveal little about the features of such published work. Of the content analyses previously reviewed (Roberto et al., Reference Roberto, Blieszner and Allen2006; Schoenberg, Reference Schoenberg2011; Schoenberg et al., Reference Schoenberg, Shenk and Kart2007), only Roberto and colleagues described features such as topics, sample sizes, and sample age ranges. For example, about one third of their studies focused on caregiving issues (and health care services was not identified as a topic), but their review focused on family gerontology and included both qualitative and quantitative studies. Given that scholarship diversity is a goal of the CJA/RCV (Penning, Reference Penning2011), not only in terms of “varying topics” but also “contrasting perspectives and approaches” (p. 165), it is instructive to examine other diversity features of qualitative research. With that in mind, our goals in this study were to (a) document the prevalence of qualitative studies in the CJA/RCV, and (b) report on the topics covered in these qualitative studies, their methodologies and data collection strategies, as well as patterns among these characteristics, such as whether or not particular data collection strategies were more common in certain methodologies versus others. Although our analysis was mainly quantitative in nature, we paid careful attention to what was present as well as what was not present.
Method
We approached our analysis of the CJA/RCV as a case study, which is a “detailed and intensive analysis of a single case” (p. 66) that can be qualitative or quantitative in nature (Bryman, Reference Bryman2012). Our case study interprets descriptive quantitative information about the prevalence and characteristics of qualitative research in the CJA/RCV. Following similar content analysis methods used by others – such as (a) Humble’s (Reference Humble2012) analysis of qualitative research trends in the family studies field and the use of qualitative data analysis software, and (b) Roberto et al.’s (Reference Roberto, Blieszner and Allen2006) analysis of family gerontology theory – we reviewed all articles published in the CJA/RCV over an 18-year period (1995–2012) to identify original, empirical research reporting qualitative, mixed methods, or quantitative findings. Eighteen years was chosen as the time length as it was felt this would be a long-enough time period to be able to identify any trends.
To show trends over time, results are presented in 3-year time frames, beginning with 1995–1997 and ending with 2010–2012. The time frames of three years are consistent with how previous research has reported on qualitative gerontology research (Schoenberg, Reference Schoenberg2011; Schoenberg et al., Reference Schoenberg, Shenk and Kart2007). Articles such as editorials, essays, book reports, introductions, and rejoinders were excluded from the analysis. However, if an essay appeared to rely heavily on the presentation of quantitative data to support its argument, it was coded as quantitative (e.g., Denton & Spencer, Reference Denton and Spencer1995) – this was, at times, a judgment call on our part.
The first author, who had limited French reading capabilities, guided the analysis, and she reviewed every English article. The second author spoke and read English and French, and she reviewed every French article. Any unclear categorizations were discussed between the authors.
Hard copies of articles were reviewed if they were available in our library, otherwise we examined PDF versions accessed through our university’s databases. Identification of qualitative or mixed methods research was based on an examination of the results section of each article. In classifying an article as presenting qualitative research findings, we looked for the in-depth, rich description of a phenomenon, typically supported with participant quotes or detailed descriptions of observations. In the few cases in which an article talked about using a qualitative or mixed methods approach early on in their article, but then, in our opinion, did not present any qualitative data in their results section (i.e., they presented only numbers, percentages, or statistics such as chi squares or t tests), that article was categorized as quantitative (e.g., McGowan & Green, Reference McGowan and Green1995). Thus, our analysis differed from Schoenberg et al.’s (Reference Schoenberg, Shenk and Kart2007) approach: in addition to looking at the results section, they also categorized an article as qualitative if “qualitative” was used as a key word or if a typical qualitative method was employed. If an article did not present an easily identifiable results section, the full article was examined.
Six Microsoft Word documents (representing the six 3-year time frames) listing all articles in every journal issue within each time frame were imported into MAXQDA software (version 11), a qualitative data analysis software program. Articles presenting empirical findings were then coded as qualitative or quantitative. Articles with mixed methods findings were coded as qualitative and as mixed methods, to keep a separate tally of mixed methods articles (see Figure 1 for an example of coding). Thus, mixed methods articles were included in our overall tallies of qualitative research, unlike Roberto et al. (Reference Roberto, Blieszner and Allen2006), who kept these two categories separate.
Figure 1: Coding examples
We also kept a separate count of the French articles. It is relevant to note that in the CJA/RCV, an article can be published in both English and French, and this was the case for 10 articles during this time frame. If any of these articles presented original research, it was counted twice in whichever category it applied to (qualitative, mixed methods, or quantitative). This was the case for two articles both categorized as quantitative.
Through using MAXQDA’s Activation tool, we were easily able to compare tallies of qualitatively coded articles with tallies of quantitatively coded articles. This helped us quickly determine the percentage of qualitative articles in each time frame (the number of qualitative articles divided by the number of qualitative and quantitative articles) as well as overall for the full 18-year period. MAXQDA’s Retrieved Segments window and the Code Matrix Browser were used to review coding.
PDF copies of the qualitatively coded CJA/RCV articles were imported into a second MAXQDA project file for further analysis. In this file, we coded articles for the main topic, methodology, and data collection strategy. Footnote 1 The first author coded the English articles, then the second author coded the French articles, guided by the codes created by the first author. We consulted on any uncertain interpretations as they emerged, and the second author’s coding was reviewed by the first author. Topic coding (Richards & Morse, Reference Richards and Morse2007) was both inductive and deductive. We began with code categories decided ahead of time (such as certain topic categories) but were also open to new codes emerging in the data (e.g., different types of methodologies), and MAXQDA’s Lexical Search function helped us to search for specific words/phrases such as phenomenology and les analyses de contenu (content analysis). Several MAXQDA functions assisted us with examining code patterns across categories (such as the six time frames and selected topics). The Code Matrix Browser assisted with reviewing the presence of codes in documents. The Document Variables function allowed us to transform selected topic codes into descriptive codes (Richards & Morse, Reference Richards and Morse2007) to interrogate the data further. The Crosstabs function displayed our interrogation results (the quantitative distribution of codes across categories).
Results
We first present overall trends in qualitative research findings in the CJA/RCV. Following that, we discuss findings regarding topics, methodologies, and data collection strategies. In each section, we report overall numbers first and then note any perceived trends across time. Third, we present our interrogation findings. CJA/RCV citations are provided to show English and French examples of the examined characteristics and patterns.
Prevalence
A total of 549 articles were identified as original empirical research, with their findings presented in quantitative, qualitative, or mixed methods format. Of those 549 articles, 140 reported qualitative (or mixed methods) findings, for an overall percentage of 25.5 per cent.
An analysis of the six time periods showed a clear increase in qualitative research findings (Figure 2). In the 1995–1997 time frame, only 10 per cent of empirical articles were qualitative. Approximately one in four articles were qualitative by the 2001–2003 period, and more than two in five articles (over 40%) by the 2010–2012 period.
Figure 2: Qualitative research trends in the CJA/RCV: 1995–2012 (percentages)
Note: 1995–1997: 9/90; 1998–2000: 14/74; 2001–2003: 24/97; 2004–2006: 24/98; 2007–2009: 26/90; 2010–2012: 43/100.
French studies (n = 26) represented 18 per cent of the qualitative articles, and a higher percentage of French articles reported qualitative findings compared to English articles (Table 1). This was the trend in all time frames with the exception of the 2004–2006 period, in which none of the five French articles during that time frame were qualitative. The 2004–2006 time period was also the only period that showed no increase in qualitative research compared to the previous time frame (24.49% compared to 24.74% in 2001–2003).
Table 1: Comparison of English and French articles in the CJA/RCV reporting qualitative research findings
Twenty-six of the 140 articles (18%) presented mixed methods findings (overall, 4.7% of all the articles involved a mixed methods design), and only three of these were French. There were no discernable mixed methods trends across the six time frames. Some of these studies collected their qualitative and quantitative data simultaneously, whereas others collected their data sequentially, representing two different approaches (Creswell, Clark, Gutmann, & Hanson, Reference Creswell, Clark, Gutmann, Hanson, Tashakkori and Teddlie2003). Qualitative and quantitative data were sometimes collected with the same group of people (e.g., Quirouette & Pushkar, Reference Quirouette and Pushkar1999), a subgroup from a larger sample (e.g., Ducharme & Corin, Reference Ducharme and Corin2000), or with different samples (e.g., Wolfson et al., Reference Wolfson, Raina, Kirkland, Pelletier, Uniat, Furlini and Szala-Meneok2009).
Topics
Each article was coded once for its main topic (Table 2). Articles dealing with policy issues were coded to the particular topic the policy issue was related to, such as health care (e.g., Keigher, Reference Keigher1999). Overall, studies on health care services (45%) and caregiving (15%) were the most common topics. Articles we coded as health care services focused on paid or formal types of support and services, such as medication reimbursement policies (e.g., Chappell, Maclure, Brunt, Hopkinson, & Mullett, Reference Chappell, Maclure, Brunt, Hopkinson and Mullett1997), volunteer programs (e.g., Sévigny & Vézina, Reference Sévigny and Vézina2007), and institutional care (e.g., Lavoie, Lessard, Barylak, & Côté, Reference Lavoie, Lessard, Barylak and Côté2003). “Caregiving” articles focused on unpaid caregiving experiences (e.g., Campbell, Reference Campbell2010; Davidson, Arber, & Ginn, Reference Davidson, Arber and Ginn2000).
Table 2: Qualitative research topic trends in the CJA/RCV: 1995–2012
Note: Percentages may not not add up to 100% for some columns due to rounding up or down for individual items.
a The following topics each appeared once in 2004–2006: driving, family-work interface, gambling, and sexuality. Language and lifelong disabilities each appeared once in 2010–2012.
We coded health care services and caregiving separately, but in many articles there was overlap in the content, and it was a judgment call at times to determine the category into which the article would fit (e.g., whether the informal care or the formal care was the more predominant theme in the article). One example of this is Lavoie et al.’s (Reference Lavoie, Lessard, Barylak and Côté2003) study that explored institutional support that either empowered or hindered family caregivers. Studies examining the transition into institutional care (e.g., Belleau, Reference Belleau2007) also fell into this category, as they typically included family caregivers in their analysis. Thus, because of the common relationship between the two codes, it is useful to note the overall percentage of both categories combined, which accounted for more than 60 per cent of all the qualitative articles in the CJA/RCV. However, articles specifically categorized as health care services appeared to increase over the 18 years, accounting for over half of the articles alone by the last two time periods.
After health care and caregiving, the next most common topics, overall, were relationships (e.g., Ducharme & Corin, Reference Ducharme and Corin1997), health (e.g., Kilian, Salmoni, Ward-Griffin, & Kloseck’s Reference Kilian, Salmoni, Ward-Griffin and Kloseck2008 ethnographic study of falls), and views of aging or healthy aging (e.g., Bassett, Bourbonnais, & McDowell, Reference Bassett, Bourbonnais and McDowell2007), which accounted for another 25 per cent of the articles. The remaining 15 per cent were distributed across 11 topics, such as sexuality (e.g., Clarke, Reference Clarke2006) and driving (e.g., Rudman, Friedland, Chipman, & Sciortino, Reference Rudman, Friedland, Chipman and Sciortino2006). No trends in these topics were seen across the time frames, except that there were few studies on these topics.
Methodologies
Qualitative research consists of numerous qualitative methodologies, which follow particular philosophical underpinnings and methodological procedures that are congruent with each other. Methodologies “are concerned with the nature of question they are suited to answer, the kind of data collection consistent with this, and also the kinds of analysis and presentation of results that fit with this approach” (Holloway & Todres, Reference Holloway and Todres2003, p. 347); they have “distinctive way(s) of approaching the world with data” (Richards & Morse, Reference Richards and Morse2007, p. 35). The three most common or foundational methodologies are grounded theory, phenomenology, and ethnography, and newer ones such as narrative inquiry and discourse analysis have also been accepted by the qualitative research community (Kahlke, Reference Kahlke2014). We coded for foundational and “newer” methodologies as well as generic qualitative approaches such as interpretive description (Thorne, Reference Thorne2008) and content analysis. Although generic qualitative approaches may not be “defined strictly enough to be called methodologies” (Kahlke, Reference Kahlke2014, p. 38), many researchers do use them to guide their research. We also coded for the constant comparison process, a method associated with grounded theory (see Glaser & Strauss, Reference Glaser and Strauss1967) and often mentioned in qualitative studies.
Seventy-nine percent of the articles (n = 111) identified that a qualitative methodology was used (Table 3). Just over one in five articles did not report a methodology (e.g., Connidis, Reference Connidis2003; MacRae, Reference MacRae2002). Footnote 2 Of those that did, thematic or content analysis was mentioned most often (30%). Two examples of these were Quirouette and Pushkar’s (Reference Quirouette and Pushkar1999) mixed method study of middle-aged, university-educated women’s views of aging, and Hébert et al.’s (Reference Hébert, Nour, Durivage, Wallach, Billette and Freitas2011) exploration of how palliative care practices influenced social exclusion.
Table 3: Qualitative methodology in the CJA/RCV: 1995–2012
Note: Percentages add up to more than 100% because seven articles were coded for two methodologies and one article coded for three methodologies.
a Method, not a methodology.
b Examples of methodologies mentioned once: functional grammar, la méthode de l’autopsie psychologique, literary gerontology, narrative analysis.
The second most common methodology (17%) was grounded theory (e.g., Horton & Dickinson, Reference Horton and Dickinson2011). Ground theory’s presence was also noted in an additional eight articles that mentioned the constant comparison process. Keller, Dwyer, Edwards, and Senson’s (Reference Keller, Dwyer, Edwards and Senson2007) study about service providers’ views of their roles in promoting food security for older individuals fell into this category.
Other methodologies included phenomenology, case study, ethnography, discourse analysis, and narrative inquiry. Seven articles were coded for using two methodologies, such as Wiersma’s (Reference Wiersma2012) analysis of time following placement into a long-term care facility (case study and phenomenology). One study used three methodologies.
The overall trends seemed relatively consistent across the six time frames. That is, in all time frames, thematic or content analysis and grounded theory were the most common methods. In any given time frame, between 14 per cent (2010–2012) and 29 per cent (2001–2003) of the articles did not mention any methodology.
Data Collection
In defining data collection strategies (some might use the term “data collection methods”), our coding was guided by the types of data typically identified and used by other qualitative researchers in the field (e.g., Richards & Morse, Reference Richards and Morse2007) such as in-depth interviewing, focus groups, observations, videos, and art.
Not surprisingly, in-depth interviews were the most common way in which data was collected, used in over three-quarters of the studies (Table 4). Focus groups (28%) were the second most common way. Garceau, Vincent, and Robichaud (Reference Garceau, Vincent and Robichaud2007), for example, conducted focus groups with three groups (seniors, caregivers, and care providers and industry employees) to examine how telesurveillance services affected older adults’ social engagement.
Table 4: Qualitative data collection in the CJA/RCV: 1995–2012
Note: Percentages add up to more than 100%; articles using more than one data collection strategy were coded more than once—they were coded for each specific strategy as well as for “multiple qualitative strategies”.
A variety of other data collection strategies were used in about 10 per cent or less of the articles, such as analyzing documents (e.g., Parsons & Tindale, Reference Parsons and Tindale2001), open-ended questions from mainly quantitative surveys (e.g., Bassett et al., Reference Bassett, Bourbonnais and McDowell2007), and participant observation. Lovering, Cott, Wells, Taylor, and Wells (Reference Lovering, Cott, Wells, Taylor and Wells2002) used participant observation in their study of a garden being cared for by patients with Alzheimer’s disease, with 11 field observations lasting between one to four hours carried out over a three-month period. No arts-based data sources were identified. More than one in five articles (n = 31) involved multiple qualitative data collection strategies, such as Horton and Dickinson’s (Reference Horton and Dickinson2011) study of falls preventions in older Chinese people, which used both in-depth interviews and focus groups. Twenty-six articles also used quantitative surveys (corresponding to the 26 articles categorized as mixed methods).
Few trends were seen across the six time frames, but we did notice a large jump in the final time frame in the percentage of articles using in-depth interviews. The percentage of articles using this data collection strategy ranged from 39 per cent to 46 per cent in the first five time frames; in 2010–2012, it was over 60 per cent. No patterns were seen in the use of multiple qualitative data collection strategies or prevalence of mixed methods designs.
Patterns among Topics, Methodologies, and Data Collection
We concluded our analysis by interrogating the data (moving beyond a simple “code and retrieve” process to explore data patterns – see Silver & Lewins, Reference Silver and Lewins2014) to see if there were co-occurrences of codes that could provide additional insight into how the published qualitative research in the CJA/RCV was carried out. Numerous co-occurrences could be examined. We first looked at whether there were methodologies (the top five identified in our study – thematic analysis, grounded theory, phenomenology, case study, and ethnography) or ways of collecting data (again, the five most common: interviews, focus groups, document analysis, participant observation, and open-ended survey questions) that predominated in the five most common topics (health care services, caregiving, relationships, health, and views of aging). We then investigated whether there were any patterns of data collection among the methodologies.
Looking at topics and methodologies first, thematic analyses and grounded theories accounted for the highest percentages (between 52% and 80%) of methodologies (Table 5) in four of the five research topics. Lavoie et al.’s (Reference Lavoie, Lessard, Barylak and Côté2003) study of institutional support is an example of a thematic analysis in the category of health care services research. A grounded theory example in the caregiving category is Egdell’s (Reference Egdell2012) study of support networks in dementia care. Health research was the one category in which these two methodologies did not predominate to the degree that they did in the other topics. Kilian et al. (Reference Kilian, Salmoni, Ward-Griffin and Kloseck2008), for example, used focused ethnography in their study of family members’ perceptions of the risk of falls for elderly family members, noting that it was an appropriate methodology to use for health-related research (they did not engage in any participant observation, which Morse, Reference Morse2007, noted is acceptable for a focused ethnography).
Table 5: Qualitative topics by methodology in the CJA/RCV: 1995–2012
Note: Percentages do not add up to 100% because not all methodologies are included in this table and some studies used more than one methodology.
a Includes mention of constant comparative process.
In terms of data collection strategies used in the various topics (Table 6), in-depth interviews were by far the most common method used for data collection for research in health care services, caregiving, relationships, and health, such as MacRae’s (Reference MacRae1996) study of older women’s friendships (in the relationship category). Interviews were also used in two of every five articles studying views of aging. However, compared to the other topics, focus groups (40%) were also commonly used in health care services research, as were multiple qualitative methods (37%). In the health care services category, Roberge, Ducharme, Lebel, Pineault, and Loiselle (Reference Roberge, Ducharme, Lebel, Pineault and Loiselle2002) used four focus groups to study 21 caregivers’ perceptions of the quality of care in Geriatric Assessment Units. Lovering et al.’s (Reference Lovering, Cott, Wells, Taylor and Wells2002) study of Alzheimer patients tending a garden is an example of a study that used multiple qualitative methods: observation, analysis of architectural drawings, focus groups, and in-depth interviews.
Table 6: Qualitative topics by data collection in the CJA/RCV: 1995–2012
Note: Percentages do not add up to 100% because not all data collection strategies are included in this table and some studies used multiple qualitative methods.
An analysis of the data collection strategies used in various methodologies revealed some interesting insights (Table 7). Given that in-depth interviews were used in over three-quarters of the studies, it was not surprising that these were the most common form of data collection for all of the top five methodologies in this study. However, focus groups were used in almost 30 per cent of the grounded theory studies (n = 9) as well as in one phenomenological study (Joffres, Reference Joffres2002). Additionally, participant observation was only used in three of the eight ethnographic studies (e.g., Broome, Worrall, McKenna, and Boldy’s Reference Broome, Worrall, McKenna and Boldy2010 study of age-friendly bus systems in Australia). Compared to the other methodologies, multiple qualitative methods were more likely to be used in case study research (one third of them, such as Kaasalainen et al.’s Reference Kaasalainen, Brazil, Coker, Ploeg, Martin-Misener, Donald and Burns2010 study of pain management approaches) and ethnographic studies (half of them). Finally, a quantitative component was more likely to be included in thematic analysis (e.g., Ducharme & Corin, Reference Ducharme and Corin2000) and ethnographic studies (about one in four for each of them) compared to the other methodologies.
Table 7: Qualitative methodologies by data collection in the CJA/RCV: 1995–2012
Note: Percentages do not add up to 100% because not all data collection strategies are included in this table and some studies used multiple methods of data collection.
a Includes mention of constant comparative process.
Discussion
An analysis of a wide range of gerontology journals is necessary to obtain an adequate view of the prevalence of qualitative research in the field. This case study analysis of the CJA/RCV demonstrates that it is supportive of qualitative research, with a much higher percentage published compared to previous findings using other journal sources (Roberto et al., Reference Roberto, Blieszner and Allen2006; Schoenberg, Reference Schoenberg2011; Schoenberg et al., Reference Schoenberg, Shenk and Kart2007). Mixed methods research accounts for less than 5 per cent of all the studies, which is a lower percentage than what Roberto et al. (Reference Roberto, Blieszner and Allen2006) found in the 1990s. Although the percentage of mixed methods research is low, these results provide evidence for the CJA/RCV that it is publishing a diversity of scholarship on the basis of qualitative and quantitative research designs. Moreover, it has been increasingly supportive of qualitative research over the 18-year period.
A higher percentage of French articles were qualitative, compared to English articles. It may be that more French research is qualitative because fewer studies are carried out with French-speaking participants in Canada than with English speaking participants, and thus more French studies are exploratory in nature. However, in a 1988 review of Quebec social gerontology, Béland (Reference Béland1988) noted that qualitative methods were overwhelmingly preferred by francophone researchers in Quebec in the area of gerontology, with an emphasis on understanding rather than explanation. Thus, these findings may be an extension of an already established preference.
In terms of topics, it is not surprising that increasing numbers of articles are focusing on health care services (including policy issues) and caregiving, given Canada’s aging population and concerns about Canada’s infrastructures to support such a demographic shift. A focus on health care and policy issues and concern for the many family members and friends who will be called into action to care for an aging population are warranted, and Penning (Reference Penning2011) noted that research in the areas of health and health care would likely continue to predominate the CJA/RCV. However, such a focus on an aging tsunami (Barusch, Reference Barusch2013) means that other topics with important implications for older individuals’ health and well-being that could be qualitatively studied, such as sexuality, intimate relationships, or issues related to changing levels of independence (e.g., housing and driving), may be neglected. For example, no studies about spirituality were found, even though it appears to be connected to well-being later in life for many older adults. Additionally, very little mention was made of sexual orientation. Furlotte, Schwartz, Koornstra, and Naster’s (Reference Furlotte, Schwartz, Koornstra and Naster2012) study of older adults, HIV/AIDS, and housing was the only article that mentioned that some of their respondents were not heterosexual (outside of this 18-year review, see Humble, Reference Humble2013, for a CJA/RCV study of older couples and same-sex marriage). Keeping in mind its intended focus on scholarship diversity in topics, the CJA/RCV may want to consider having special issues or sections on some of these less qualitatively studied topics.
A variety of methodologies were used, but content analysis and grounded theory were the most common. Grounded theory’s popularity is not surprising, given its influence in qualitative research and developing micro theories, and the comfort with which researchers (particularly those with positivistic leanings) respond to its “structured” method (i.e., the “constant comparative” process, open coding, axial coding, and selective coding). It is also the most commonly misused methodology, though (Richards & Morse, Reference Richards and Morse2007). Researchers may also think that a “generic” content analysis is appropriate and feel no need to identify it as any other methodology (in fact, Sandelowski and Barroso [Reference Sandelowski and Barroso2003] have noted that findings are often presented in the same generic manner regardless of what methodology is stated as being used). Kahlke (Reference Kahlke2014) described the benefits and drawbacks of using such generic approaches.
No examples of arts-based research were found in this 18-year review, even though arts-based projects – those that use art (e.g., photovoice, drama, drawing, dance, and poetry) in “generating, interpreting, and/or communicating knowledge” (Boydell, Gladstone, Volpe, Allemang, & Stasiulis, Reference Boydell, Gladstone, Volpe, Allemang and Stasiulis2012, part. 4) – have been identified as an emerging method in health research (see Boydell et al., Reference Boydell, Gladstone, Volpe, Allemang and Stasiulis2012 for an examination of its prevalence and review of issues). Gerontologists are beginning to notice this type of research. For example, in 2011, The Gerontologist expressed an interest in promoting arts-based research in its journal, stating: “frameworks and methods of the arts and humanities focus on interpretation and expression of the multiple, the elusive, the awe inspiring, the disturbing, and even the ineffable aspects of growing older” (Kivnick & Pruchno, Reference Kivnick and Pruchno2011, p. 143). An arts-based study (Fraser, Archibald, & Nissen) about home care services does appear in the CJA/RCV in 2014.
Due to their greater presence in the data, thematic analysis and grounded theory were the most commonly used methodologies in the top five topic categories in this study. Gerontologists may want to consider asking different kinds of questions that lead them to use less commonly used methodologies to study topics such as health care services and caregiving. Phenomenology (a “foundational” methodology, but interestingly enough, only used in 10 articles), narrative inquiry, and arts-based research, for example, may reveal very different insights. Wiersma’s (Reference Wiersma2012) study of one man’s experience of time in an institution is an example of the power of a well-done and methodologically congruent (Richards & Morse, Reference Richards and Morse2007) phenomenological study. Guided by van Manen’s (Reference van Manen1997) hermeneutical phenomenology, this study focused deeply on the existential of lived time within an institution, describing in detail “biographical time, embodied time, and embedded time (including institutional time)” (p. 73).
Not surprisingly, in-depth interviews were the most common way of collecting data. Focus groups, interestingly, were the second most common. These were relatively unknown in the 1980s but have become more popular (Morgan, Reference Morgan1997; Richards & Morse, Reference Richards and Morse2007), often used in applied research and having a number of applications (for example, generating survey data or complementing other qualitative methods) (Morgan, Reference Morgan1997). Many issues related to the aging population have multiple stakeholders, and focus groups may be particularly helpful for accessing their individual views, as are multiple ways of collecting data. This may be why a higher percentage of health care services research involved focus groups and multiple methods of data collection compared to the other topics.
It is important to note that although in-depth interviews are very popular for data collection, they “may not [always] be the most appropriate way of making data” (Richards & Morse, Reference Richards and Morse2007, p. 112). Unobtrusive measures (van den Hoonaard, Reference van den Hoonaard2012) such as diaries, emails, and media content can be explored for their applicability to gerontological topics of inquiry. Participant observation was seldom used in these qualitative studies, and we encourage gerontologists to consider how they can incorporate this form of data collection into their research designs, as it has many benefits, such as revealing meanings that interviews may be unable to expose (Cobb & Forbes, Reference Cobb and Forbes2002). Participant observation can be difficult to do, however, and university review ethics board requirements may create additional challenges.
The analysis of open-ended questions from surveys was slightly more common than participant observation. With online surveys potentially becoming more common and the ease with which open-ended questions can be added to them, gerontology researchers may need to be mindful of how they are using this technology and how it affects the data quality. Adding open-ended questions to a survey is not necessarily sufficient to transform it into a mixed method design (Cobb & Forbes, Reference Cobb and Forbes2002). These studies will probably tend to have larger samples as well, and larger sample sizes do not necessarily mean better qualitative analyses. Studies with very high sample sizes (e.g., Bassett et al., Reference Bassett, Bourbonnais and McDowell2007) – typically mixed methods studies including the analysis of written open-ended survey questions – may have shallow data. Bassett et al., for example, stated that they analysed the responses of 2,783 individuals who gave a “substantive” response to the question, “What do you think makes people live long and keep well?” (p. 115), but what “substantive” meant was not defined.
Researchers collect their data in a number of ways, and many data collection strategies cut across methodologies (Holloway & Todres, Reference Holloway and Todres2003; Richards & Morse, Reference Richards and Morse2007). However, not all forms of data collection may be suited to all qualitative methodologies, which draws our attention to the concept of methodological congruence. Methodological congruence refers to “the fit between the research problem and the question, fit between the research question and the method, and of course, fit among the method, the data, and the way of handling data” (Richards & Morse, Reference Richards and Morse2007, p. 35), and it is very important in qualitative research, contributing to the trustworthiness of analyses.
Our analysis does not provide (and was not intended to provide) an in-depth analysis of methodological congruence (and we note that we did not examine the issue of how data were “handled” [i.e., analysed] in any of these studies). Nevertheless, while interrogating co-occurrences of data collection strategies and methodologies, we found several examples of grounded theory research and one example of a phenomenological study (which was actually guided by both phenomenology and grounded theory) that used focus groups to collect their data, which we found unusual, particularly for phenomenological research. Richards and Morse (Reference Richards and Morse2007) noted that the most likely primary data sources for phenomenological studies are “audio-taped, in-depth conversations, phenomenological literature” and secondary data sources are “poetry, art, and literature” (p. 33), whereas primary sources for grounded theory are “audiotaped interviews and observations”. We refer readers back to Wiersma’s (Reference Wiersma2012) article again for an excellent example of a methodologically congruent study – it is clear that Wiersma was thinking like a phenomenologist, collecting her data like a phenomenologist, and analysing it like a phenomenologist. In general, though, it appeared that most of the data collection strategies fit with the methodologies for the studies in this analysis, including ones (i.e., case studies and ethnographic studies) that collected qualitative data in multiple ways.
Finally, we note that of the small number of ethnographic studies identified in this analysis, few included participant observation. As noted earlier, focused ethnographies do not necessarily need to include participant observation (Morse, Reference Morse2007), but it is important not to lose sight of this important way of collecting data and of the central role it plays in identifying the social norms and behaviors of groups. We do note that observation was occasionally used with other methodologies. Lovering et al.’s (Reference Lovering, Cott, Wells, Taylor and Wells2002) study of gardens and Alzheimer’s patients, which used participant observation, seemed very much like an ethnographic study, but the authors labeled it a qualitative descriptive study.
Limitations and Future Research
Our percentages may be higher than previous research because we included mixed methods research in our counts. We only examined one journal, and case study research does not permit generalizations (Sandelowski, Reference Sandelowski1996). Some of the less studied topics identified in this analysis may be present in other gerontology journals, including ones with very focused topics such as the Journal of Elder Abuse and Neglect, and in non-gerontology focused journals. Alternatively, it may be that these other topics are quantitatively studied, or more likely to be quantitatively studied, but an analysis of this was beyond the scope of our project. Additionally, this review did not include other sources in which qualitative gerontology research is found, such as edited books and non-traditional forms of dissemination such as those carried out through arts-based research. However, a strength of this case study analysis is that we reviewed a full 18 years of published research, rather than just a sample of articles from those years.
Our analyses of differences or similarities across the time frames and various groups are not based on statistical tests. We recognize that in some cases we have interpreted differences between groups on the basis of very low numbers. In line with the nature of qualitative research, we have presented our interpretation of this quantitative descriptive data. We did not do an analysis of trends in quantitative research in the same journal for comparison.
Many ideas for future research arise from this study. To examine qualitative research publishing trends in further detail, studies could take a sample of articles from a range of gerontology-related journals over a long period of time or look at other sources in which qualitative research is published. Future research can examine additional characteristics of gerontological qualitative research such as sample size, age ranges, use of theory (Alley et al. [Reference Alley, Putney, Rice and Bengtson2010] noted that theory is underutilized in gerontology research), funding, and methodological congruence. Finally, more interrogations of the data can be carried out, such as whether or not certain ways of collecting data are carried out with various age groups, or if funding is more likely to be received for mixed methods research rather than for research that is solely qualitative.
Finally, both the CJA/RCV and qualitative researchers intending to submit their work to the CJA/RCV may want to take note of the less common topics, methodologies, and ways of collecting data identified in this analysis. Studies that address these gaps can contribute to the CJA/RCV’s ongoing goal of heterogeneity in scholarship.
Conclusion
Our goals in this article were to provide insight into the CJA/RCV’s commitment to scholarship diversity and contribute to ongoing conversations about qualitative gerontological research by studying the prevalence and characteristics of qualitative research in the CJA/RCV over an 18-year period. Our analysis indicates that the CJA/RCV is very supportive of qualitative research and increasingly so over this time period. Additional analyses indicate that certain topics (health care services, caregiving), methodologies (content analysis and grounded theory), and data collection methods (in-depth interviews and focus group research) predominate in these published studies. Qualitative research has much to contribute to the field of gerontology (Cobb & Forbes, Reference Cobb and Forbes2002), and we encourage researchers and the journal to consider the potential of the full range of qualitative methodologies and research strategies as well as the different topics that can be studied with them. We look forward to the continued implementation and reporting of qualitative research in this journal as well as other gerontology-focused journals.
