Employing Questionnaires in terms of a Constructivist Epistemological Stance: Reconsidering Researchers' Involvement in the Unfolding of Social Life

Abstract

In this article, I delve into what it might mean to employ questionnaires without regarding them simply as a way of attempting to discern relationships of correlation or causality between defined variables (as in positivist and post-positivist conceptions of questionnaires). I shall consider the implications of researchers using questionnaires on the basis of alternative paradigmatic orientations. I shall discuss, in particular, interpretivist stances and more constructively-oriented stances (as qualitatively-oriented paradigmatic positions) with reference to different understandings of questionnaire use. I shall also reflect on how qualitative positions that embrace a constructivist epistemological stance can lead to a redirection of questionnaires in relation to more “usual” (post-positivist-directed) usages. In the course of the discussion I make a case, drawing on a version of constructivism, for researchers taking responsibility for their involvement—no matter what methods are used—in the unfolding of the social worlds of which research is a part. Taking into account the constructivist epistemological understanding that questionnaires—as well as other research methods—contribute to the construction of responses rather than merely “finding” responses from research participants, I suggest that some responsibility needs to be taken by those employing questionnaires for the potential social impact of these on research participants as well as wider audiences.

Keywords

constructivism epistemological paradigms qualitatively-oriented research questionnaire use researcher responsibility recognition of involvement in social life

Questionnaires are usually understood as being a research method aimed at measuring and analyzing relationships between identified variables supposedly existing in social reality. Insofar as they are used in terms of this understanding, their usage can be said to be underpinned (more or less consciously) by a positivist/post-positivist position. This position, in turn, assumes that scientific studies must be directed towards capturing—as far as possible—social reality and, specifically, analyzing the regular connections that operate in social reality. Questionnaires as a quantitative research method are then regarded as offering a route to advancing our knowledge regarding relationships between certain variables. The epistemology here defines knowing as the “representation of reality.”

Regarding the links between positivism and post-positivism, Denzin and Lincoln (2003) have indicated that post-positivism is a refinement of positivism as developed in the 19^th century (pp. 12–13). Within a post-positivist stance, they noted, it is conceded that scientists should not claim to verify statements about reality and that only approximations to reality can be sought. Johnson (2009) followed this up by indicating that when applying the term “post-positivism” to an author's work, one is using a “kinder and gentler” term (than positivism), which signals that the author is cognizant of epistemological debates that have developed “over the past 75 years” (p. 450). He noted that the post-positivist philosophy of science “generally fits today's quantitative … research community” (p. 450).

Denzin and Lincoln (2000) suggested that qualitative methods (which they define as involving in-depth exploration of the dynamics of social life, and where the construction of social realities by research participants and by researchers is given prime attention) are more naturally at home “within the critical interpretive framework” than within other paradigmatic stances (p. 8). Conversely, quantitative methods are more “at home” within positivist/post-positivist stances. Nevertheless, the use of different methods, including questionnaires, but underpinned by alternative paradigms other than from where they are seen as more naturally “at home,” is not discounted by Denzin and Lincoln. In a similar vein, Hesse-Biber (2010) stated that “the deployment of a qualitative methodology does not rule out the use of quantitative methods” (p. 456). This also concurs with Flood's and my own account (1996) of the “oblique” use of methods, which involves strategically redirecting them to fulfill purposes that they were not originally designed to fulfill (p. 73). As we put it: “An oblique use … means operating [the method] through the principles of an alternative agenda” (Flood & Romm, 1996, p. 212). What is important within “oblique” usage is that, as Torlak (2001) indicated, the analyst is aware of the principles of that paradigm [which is being invoked]” (p. 307). Pollack (2006) summarized Flood and Romm's (1997) position here: “The practice of using an approach from one particular paradigm, but operated under the direction of a different paradigm, is referred to as ‘oblique’ use” (as cited in Pollack, 2006, p. 390).

In this article, I pinpoint alternative agendas/outlooks, which can be classed as qualitatively-oriented, that can be used to guide the use of questionnaires. Having first outlined the interpretivist position, I go on to discuss constructivism as an epistemological stance. I finally refer to what I call a “trusting constructivist position” as a version of constructivism (see Romm, 2001, 2002, 2007, 2010), which is based on researchers taking some responsibility for the potential impact of their inquiries on social life. I shall spell out implications of this for the design and administration of questionnaires and for the reporting of results to initial participants and wider audiences.

Traditional Usage of Questionnaires

As indicated in my introductory remarks, historically questionnaires have normally been deployed on the assumption that asking respondents questions (mainly via closed-ended questions) is a way for researchers to obtain information about certain, defined variables. The questionnaire items that respondents are asked to answer are seen as offering indicators/measurements of the variables under consideration, so that when statistical tests are applied to the answers, one can establish with some degree of confidence whether relationships of correlation exist between these variables (such that changes in one are accompanied by changes in others) (Romm, 2001, pp. 176–182). Questionnaires have also been presented as advancing our knowledge about relationships of causality by forwarding theoretical inferences about the direction of influence of the variables (Romm, 2001, 2007). In other words, they are used with the overall intention of locating, as nearly as possible, the regular causal patterns that are operative in the workings of social reality. The use of questionnaires underpinned by (post-positivist) conceptions of social scientific research as geared to grasping the operation of correlation and causality in social reality is well documented in the research literature (cf. Babbie, 2011; Bryman, 1992; Punch, 2003; Sapsford, 2007).

However, Scott (2010) has cautioned that the traditional understanding of questionnaires as providing a source of more or less neutral information is not adhered to by all who practice quantitative research. Indeed, she claimed the following:

There are very few adherents to epistemologies of objective knowledge. Quantitative researchers are not naïve positivists. They acknowledge the role of social construction in measures and are wary of quantification being seen as the equivalent of scientific reasoning. They know better than most that “statistics can lie.” (p. 233)

I return to the issue of the ways in which questionnaires can be regarded as constructing what they deem to measure (as alluded to by Scott) in the sections below and, particularly, in the section that elucidates a constructivist perspective.

Aside from the “traditional” rationale of using questionnaires as a route to obtaining information regarding measurable variables, various authors have pointed out that they can also be used in terms of alternative understandings of the research process and of the results of such research.

Weberian Interpretivism: Implications for Questionnaire Use

At the turn of the 20^th century, Max Weber offered an alternative to positivist accounts of what social scientific inquiry should involve (Weber, 1949), which became labeled as “interpretivism.” For Weber, social scientific models have to be rooted in an understanding of some “complex of meaning” that is ascribed to the actors involved. The explanation of sequences of events must be rooted in the understanding (or interpretation) of meaning. Social scientists must try to develop plausible accounts of the motivating meanings that constitute social existence. For example, in his book The Protestant Ethic and the Spirit of Capitalism (1930), he argued that, in the process of the emergence of capitalism, wealth accumulation for Protestants was a sign that God favored them. He thus explained the emerging social formation in terms of the meanings that people attributed to their own and to others' actions. Nevertheless, although Weber broke with positivist ontological assumptions (with his ontology of social reality as consisting of meaning-making of participants rather than of nature-like cause-effect relationships), he still adhered to positivist epistemology as far as the search for value-freedom (objectivity) on the part of social scientists is concerned. He agreed with the positivist view that scientists must, as far as possible, divest their personal concerns when acting as scientists trying to study (in this case to understand) the social world (Weber, 1949). Weber insisted (in line with positivist-oriented views of science) that, in their attempts to understand social life, social scientists should be dedicated to the task of “recognizing facts, even those which may be personally uncomfortable, and to distinguish them from [their] own evaluations” (p. 5).

What he offered as distinct was his view that social scientists need to concentrate on words/language as expressions of meaning (and not primarily on numbers) to create insightful accounts of social reality. This of course has implications for the use of questionnaires: insofar as interpretively-oriented researchers wish to use them, they are used in an effort to grasp social life on the level of “causality” by locating sequential patterns, combined with providing explanations for these located patterns, based on an exploration of meaning-motivations—using alternative methods.

In elucidating Weber's epistemological position, Hammersley (1995) stated that in his account of the practice of social scientific research, Weber rightly recognized the “need for detachment from political commitments,” that is, the need for “objectivity” (p. 115). This does not mean that objectivity is always achieved because, as Hammersley (1995, 2003) has pointed out, research into the patterning of social life is an active process, in which accounts of the world are produced through somewhat selective observation (on the basis of whatever research methods are deployed). Nevertheless, the ultimate goal is to make advancements in our knowledge (Hammersley, 2003).

As far as implications for policy-making are concerned, Fielding (2010) pointed to developments in the “policy community,” especially in Europe and the United States, towards encouraging qualitative research alongside quantitative methods (e.g., surveys). He indicated that this is increasingly considered as “bringing alive policy issues with an immediacy sometimes lacking in quantitative data” (p. 130). But he accepts that, when surveys are administered, these are to be judged—by colleagues and others—in terms of their associated “quality standards.” What this means is that, at the point at which questionnaires are used within some mixed-method projects, the standards for judging research normally associated with quantitative research are not put under scrutiny. Quantitative and qualitative research (with their associated standards of judging quality research) are seen as capable of co-existing within a mixed-method approach, whereas mono-method quantitative research is seen as ill-equipped on its own to come to grips with social meaning-making. This is consistent with an interpretivist position, which makes provision for using a range of methods, appropriately applied (including questionnaires) as part of the researcher's repertoire that may be drawn upon in attempts to produce knowledge.

Other researchers, however, have expressed concern that, once social research is seen as a process of making advancements in grasping more adequately the patterning of social life, the unique ability of the qualitative paradigm to appreciate that social reality is subject to multiple interpretive frameworks (as applied by research participants and by researchers) becomes lost. This kind of concern implies the adoption of an alternative (more constructivist-oriented) stance, as expounded below (see also Romm, 2001, pp. 99–125.)

A More Constructivist-Oriented (Qualitative) Position: Implications for Questionnaire Use

Hesse-Biber (2010), in forwarding a constructivist argument, made the point that “a qualitative approach encompasses several research traditions that hold as their core assumption that reality is socially constructed and multiple” (p 455). Here Hesse-Biber concurs with Denzin and Lincoln (2003) who, in summarizing the major differences between qualitative and quantitative approaches to research, argued that qualitative researchers “stress the socially constructed nature of social reality” and also emphasize the “value-laden nature of inquiry” (p. 13). Their suggestion is that qualitative researchers regard the social world as being (re)created partly via the language constructions employed by people during social interactions. Professional social researchers are likewise engaged in processes of co-constructing—with others—visions of social reality as they conduct social inquiries. The accounts generated are considered as having the epistemological status of being stories that provide value-imbued interpretive lenses—rather than offering more or less neutral renditions of how “social reality” operates.

When those conducting qualitative inquiries stress the constructed character of social life, and the constructed character of researchers' engagement with others in creating visions, they contest the notion that social scientists should strive to bracket their values from the processes of research. As Johnson (2009) put it, “[such] qualitative research tends to view facts and values as … entangled” (p. 452). He added that, according to this position, “raw data, especially social science data, cannot be interpreted in the absence of values. Human beings cannot fully reason on or about ‘facts’ without concurrently reasoning and relying on values” (p. 452).

To cater for the assessment of the quality/worth of research reports that might be proffered by researchers, different criteria have been developed over time within this tradition (see, for instance, Chilisa, 2007; Collins, 1990, 2000; Demerath, 2006; Hsiung, 2012; Kenny, 2002; Ladson-Billings, 2003; Lincoln & Denzin, 2003; Lincoln & Guba, 2003; McIntyre-Mills, 2008; Midgley, 2000; Naidoo, 2008; Onwuegbuzie & Leech, 2007; Rajchman, 2007; Romm, 1997, 2007, 2010, 2011). These criteria are put forward as a way of assessing accounts/stories in terms of an understanding of validation that is not premised on the view that scientific exploration necessarily amounts to seeking truth as the representation of reality it (supposedly) is. Lincoln and Guba (2003) have called this an “extended agenda” for understanding issues of validity (p. 274).

Lincoln and Guba (2003) did not delve into the question of what this implies in relation to the use of quantitative methods. But one of the implications of their argument is that, to the extent that “qualitative researchers” do make use of statistical measures and methods, these would be seen as needing to be linked to an exploration of multiple ways of viewing any “observed” patterns. This would then still allow for an epistemological understanding that there is no univocal way of envisaging the patterns and, furthermore, that there is, as they put it, “no royal road to ultimate knowledge” (p. 274). Lincoln and Guba's (2003) reference to alternative “roads” to knowing admits the possibility (at times) of employing both quantitative and qualitative methods as part of the knowing process, with the primary focus still being on developing context-grounded interpretations of social life, based on “prolonged engagement” with research participants (p. 275). This would seem to fit in with what Johnson, Onwuegbuzie, and Turner (2007) called assigning “dominant status” (p. 124) to qualitative methods—and associated philosophical underpinnings—in certain mixed-method research designs. They explained that here the primary reliance is on a “qualitative-constructivist approach,” while it is “concurrently recognized that the addition of quantitative data and approaches are likely to benefit most research projects” (p. 124). This is also consistent with discussions of what Creswell (2003), Creswell, Slope, Plano Clark, & Green (2006), and Hall & Ryan (2011) have called qualitatively-driven usages of mixed methods.

What I wish to draw attention to is that in the account as given by Johnson, Onwuegbuzie, and Turner (2007), it appears (from the phrasing they use) that they are viewing “quantitative data and approaches” as inseparable—without considering the possibility that quantitative data could be conceptualized by invoking a different paradigmatic approach. In other words, they do not create openings for what I am arguing for in this article—namely, that questionnaires themselves could be used in a project with reference to a qualitative-constructivist outlook. I suggest that this is another (less usual) way of treating questionnaires, and that this way of treating them can apply both in so-called mono- and in mixed-method research, as elucidated further below. This option when applied in mixed-method research would be an extension of Johnson, Onwuegbuzie, and Turner's (2007) suggestion that mixed designs can become an opportunity for researchers to think about philosophical questions in relation to ontology (views about reality), epistemology (views about knowing), and axiology (views about the role of values in research) (p. 118).

Interestingly, in her outline of her understanding of the transformative paradigm, Mertens (2010) too hinted at “alternative” (non-traditional) ways of treating questionnaires when viewed through the lens of power considerations—that is, of power relations in the research process. She indicated that, when working with a transformative paradigm as one's paradigmatic basis, one would “raise questions about how decisions are made about what is ‘researchable,’ which variables to study, and the definitions and measurement of those variables” (p. 13).

Space constraints do not permit a full consideration of the links between the specific qualitative constructivist position being forwarded in this article and the transformative paradigm. Suffice to say that what I call a “trusting constructivist approach” (Romm, 2001, 2002, 2010) appreciates Mertens' (2010) view that in working with a transformative research agenda, if and when using questionnaires, one would focus on the manner in which “variables” become measured and defined. This means that the power to construct realities becomes itself subject to review. Mertens (2010) sees this as in keeping with the adoption of an epistemology which does not assume that “a morally neutral, objective observer will get the facts right” (p. 13). In the next section I spell out what I see as some implications of understanding questionnaires in terms of such an epistemology.

Reflexivity, Transparency, and Discursive Accountability

When operating with a constructivist epistemology, it is understood that whatever becomes “found” during questionnaire administration does not measure something outside of the interaction between researchers (or their research instruments) and research participants. Hence the interaction between participants and the questionnaire is seen as key to the results that become generated (cf. Galasinski & Kozlowska, 2010, pp. 271–272). This radically queries the idea that questionnaires can and should be designed to minimize the so-called “reactivity effect,” defined as the effect of the manner of asking questions of participants. For example, see also Speer and Hutchby's (2003) accounts of the unavoidability of the reactivity effect in social research.

One way of accounting for inevitable researcher effects is for researchers to make transparent to themselves and to others that the questionnaire method (like any method) cannot claim to be a “neutral mechanism to collect people's views and opinions” (as put across also by Speer, 2002, as cited in Galasinski & Kozlowska, 2010, p. 272). Recognition on the part of researchers of the manner in which research generates rather than finds data, in turn, can be seen as implying a stance of reflexivity. But this requires some further qualification. For example, in considering Hammersley's reference to the adoption of a “reflexive” approach, Speer and Hutchby (2003) remarked that “he does not provide any indication of what such an orientation would look like in his own or others' work” (p. 353). In Romm (2010), I summarized the debate around this as seen from a constructivist point of view (pp. 242–244). I argued that Hammersley's position does not make sufficient provision for researchers taking into account the consequences of their research framings.

As De Souza (2004) explained, reflexivity (in a constructivist-oriented position) involves an admission by researchers that their “social position, personal histories, and lived experiences matter” in the way in which the research becomes constituted (p. 473). Reflexivity in this sense involves an admission that these histories and experiences “matter” in our framing of the questions/issues we raise as inquirers.

De Souza's (2007) position on reflexivity is accompanied by her proposal that it is incumbent upon researchers when setting up research projects to consult with (key) participants in defining appropriate methodologies, including how any research instruments will be designed (p. 12). She considers this indeed as a specific contribution that Maori research adds in its critique of Western-oriented research approaches, which do not generally include such consultation. An exception would of course be participatory action research; but as Reason (2006) indicated, research defined as a process of acting in relationship with others is not considered as part of mainstream social research in either North America or Europe—or indeed in other parts of the globe (p. 188). In this article I have not tried to explore the arguments of those advocating for action research, but in McKay and Romm (2008b) we considered the points of connection between what we call active research and traditional understandings of action research.

As De Souza (2007) sees it, in any research project, consultation needs to occur “at all stages of the research project lifecycle” (p. 12). As far as “results” generated from research projects are concerned, she has advocated the need to open spaces for interpreting these via, for example, some community meetings prior to “wider dissemination” (p. 14). This is all to avoid researchers' pre-framing of issues, and attendant explorations and analyses, being controlled unduly by their initial inclinations. In short, in De Souza's (2007) conception, insofar as questionnaires (as a research tool) are used within the research process, an alternative relationship with participants and wider audiences needs to be brought to the fore as one of the indicators of “research quality” (p. 14).

Acquah (2007), writing in the context of Botswana, expressed a similar position when he compared the “standard” use of surveys with more participatory approaches. He pointed out that “in a situation where survey research is to be done” (with the advantage that scope is created for generalizing from a sample to a population), the use of a participatory approach “enables [some] members of the community to discuss the issues being [or to be] investigated” (p. 130). Furthermore, he insisted that once research results are “available,” opportunities need to be set up for dialogue around what people in the community (including interested research participants) may expect and what the “results” of the survey show (p. 130). As he put it:

Sometimes the results of a survey may turn out to be different from what the community expected. A workshop is an opportunity for discrepancies in the results to be discussed and compromise solutions arrived at. (p. 131)

Here Acquah (2007) argued that when results do not make sense to (or resonate with the lived experiences of) members of the communities where the research has been conducted, provision needs to be made for reaching compromises in relation to the “identifications of problems [that is, issues defined as problematic] and causes” (p. 131). Acquah (2007) thus stated his view that researchers should not present survey findings with a scientific authority that renders them undiscussable in communities: through discussion it is possible that alternative ways of seeing and treating the results may be produced.

De Souza's (2007) and Acquah's (2007) proposals regarding liaison with community members clearly go beyond Scott's (2010) announcement that many quantitative researchers may be aware that “statistics can lie” (p. 233). Their considerations call for researchers to take this into account by liaising with (samples of) participants and wider communities, so as to exercise what can be considered as “discursive accountability” as part of the process of developing trust (Romm, 2001, pp. 124–125).

Midgley and Shen (2007) have added a further angle to these considerations when they noted that investigating “causality” itself need not be tied to seeking linear cause-effect relationships. They outline what they call a Buddhist Systems Methodology (BSM), where emphasis is on “awareness of the systemic, contextual and interpretive nature of both causation and consequences” (p. 204). As they see it,

the cause-condition-effect concept [understood as linked to human interpretation] can promote more careful decision making than thinking in terms of linear cause-effect relationships. (p. 204)

Midgley and Shen (2007) defined BSM as an “exploratory questioning approach” (p. 195). This means, in practice, that when considering causality in social life we do not try to close (or unduly limit) discussion about how the links between factors/events in the social world are to be conceptualized by referring to statistically derived inferences.

I have labeled the considerations of authors such as De Souza, Acquah, and Midgley and Shen as constructivist-oriented in the sense that they treat, and implore others to treat, any research “results” as being constructed and hence as open to reconstruction. Indeed, I would suggest that their considerations can be labeled as complying with a trusting constructivist position, where the focus is on researchers earning trust in their manner of proceeding by showing that they are willing to engage in discourse around their practices (Romm, 2001, pp. 113–125).

Designing/Re-tuning Questionnaires to Encourage Rethinking Around “Restrictive” Cultural Constructions

I have suggested above that if one uses questionnaires while embracing a constructivist/trusting constructivist outlook, a different relationship with research participants and wider audiences can come into play (than in post-positivist or indeed interpretively-oriented research). The construction of research framing(s) then involves more participation on the part of participants, and any research “results” are forwarded as invitations for further dialogue with participants and wider audiences.

I now wish to add yet another option for consideration—that questionnaires be designed/re-tuned with the intention that they initiate learning opportunities for participants and others. This “retuning” is another way of researchers taking some responsibility for the potential impact of research on participants and wider audiences.

My suggestion here is that researchers can try to design questionnaires such that avenues for rethinking cultural constructions that are arguably restrictive can be explored. I provide two illustrative examples below—(a) one from a mono-method research design and (b) one from a mixed-method design. (I was personally involved in the latter.)

Using a Questionnaire for Pedagogical Purposes (Canada): Taylor and Hoechsmann

The first example to which I refer is taken from the work of Taylor and Hoechsmann (2011). I see Taylor and Hoechsmann's way of deploying their questionnaire as an indication of their acknowledgement of the potential influence of questions on respondents. Taylor and Hoechsmann (2011) explained that they expressly incorporated the aim of using questionnaires to perform an educative function. They reported on a national survey of high school students in five Canadian provinces, which was aimed at investigating “what young people know about the struggles as well as the intellectual, social, political, and cultural contributions of racialized peoples globally and nationally and where they learned it (school, media, family and community)” (p. 220). Taylor and Hoechsmann (2011) indicated that they intentionally changed the usual format of questionnaires and transformed it “from the traditional format where questions are posed and right answers must be selected, to one which offers some information and then asks the student to respond if they already knew this, and where they had learned it” (p. 226).

Taylor and Hoechsmann (2011) offered a sample question from the survey to illustrate their approach (using historical data typically taught in schools):

Harriet Tubman, a slave from Maryland, escaped to freedom in Canada in 1849. Later on she helped hundreds of slaves to escape the U.S. through the “Underground Railroad.”

I didn't know about her

I learned about her at school

I learned about her from a family member

I learned about her from the media (newspaper, magazine, TV, radio, Internet, etc.)

I learned about her from friends or community members

They pointed out in relation to this question that:

As can be seen, students are not “tested” and the goal is one of education and empowerment. On this question, for example, 79.2% of participants had heard of Harriet Tubman and felt they knew some of the information presented. Of these students, an overwhelming majority claimed to have learned about her in school (67.6%), while some had learned about her from the media (34.9%), far fewer from their family (10.1%) or from friends or community members (6.8%). Of the students who had known nothing of her, the survey may serve as an introduction and invitation to learn more. The particular structure of the survey, then, is such that it teaches as it learns. (p. 226)

Taylor and Hoechsmann (2011) stated that they are thus able to use a “quantitative research instrument inspired by a pedagogical impulse and qualitative interest in knowledge not as static or inert but negotiated and dynamic” (p. 226).

They therefore suggested that they treated the questionnaire as an instrument that—at the moment of administration—is already influencing people's conceptions, in that the students' understandings of Harriet Tubman as an escaped slave could become developed simply by their exposure to the questionnaire. This also is linked to Harris-Lacewell's (2003) point (made in the context of examining racial politics) that respondents' attitudes are often not well thought through and indeed become “formed” in the process of their filling in questionnaires, especially if their views are relatively unconsidered (p. 235).

Taylor and Hoechsmann (2011) believe that their (consciously educative) design of the questionnaire can be of potential interest to

numerous other fields of inquiry (e.g., health education, legal or civic education, adult education, education for human rights or civic engagement) in which the goals of engagement, discussion, collective deliberation, or “empowerment” outweigh concerns for experimental measurement alone. (p. 227)

Taylor and Hoechsmann (2011) call on researchers in these various fields to consider the possibilities of empowerment when they design the tools/instruments to be used in the research process. I would add that, once empowerment is seen as outweighing “measurement” of some pre-existing social reality, one could claim that an alternative epistemological paradigm is already being used. A constructivist epistemology—which recognizes that the views of respondents/participants are constructed in relation to questions asked—can be said to be invoked here.

Taylor and Hoechsmann (2011) have chosen to use their questionnaire to disrupt cultural conceptions which treat slaves as having passively accepted their condition and which prioritize the efforts of White abolitionists over and above the resistance of slaves. Their questionnaire itself makes way for alternative “ways of seeing” on the part of respondents to come to the fore—alternatives which may not come to the fore unless emphasized in the design of the questionnaire. Audiences, too, are invited to engage in “collective deliberation” as initiated by the questionnaire results. Taylor and Hoechsmann (2011) can thus be said to be shifting/re-tuning the more traditional questionnaire design.

As I argued in Romm (2010), those constructing (re-tuning) and interpreting questionnaires in this way are specifically concerned with

the problem of questionnaires possibly sedimenting and crystallizing … social categories and attendant thinking (arguably leading to further polarization in social life) – rather than potentially undercutting this thinking. (p. 54)

Attempts can thus be made to use questionnaires so that they undermine socially rigidified categories and meaning-making that are unnecessarily limiting in terms of their social consequences. That is, questionnaires can intentionally enable respondents/participants to rethink issues that they may not have considered before and expose them to ways of seeing by asking them to respond to questions phrased in a certain way. This can also open more spaces for audiences (reading “reports”) to re-reflect on issues raised and to participate in (further) “collective deliberation.”

Using a Questionnaire to Raise Awareness of “Taboo” Topics (Zambia): McKay and Romm

Another example that casts light on how questionnaires can be used in order to create “dynamic knowledge,” as Taylor and Hoechsmann (2011) put it, comes from the work of McKay and Romm (2008a, 2008b). In this case, we intentionally deployed a questionnaire exploring HIV and AIDS in the informal sector in Zambia with the idea that the questionnaire itself could evoke new ways of seeing (restrictive) cultural practices. For the research design, we took our cue from McKay's (2003) previous four-country study—so that the research process included four (sequential) stages (which are discussed in detail in McKay & Romm, 2006, pp. 66–82):

Knowledge, Attitudes, Perceptions and Behaviours (KAPB) questionnaire administered to a sample of 407 respondents/participants in the informal economy from across four targeted sites in Zambia. (During the KAPB questionnaire administration, the research assistants also elicited participants' interest in attending rapid assessment workshops and potentially operating as peer educators in the informal economy).

Rapid assessment (RA) workshops undertaken in the same areas (with about 50 participants each, selected from participants in the KAPB survey and from relevant organizations—relevant as defined by participants and the research assistants).

Peer education implemented, which made use of a training brochure developed by McKay and Morr (a medical doctor) and which emerged from the earlier stages of the research.

A national workshop set up with the aim of disseminating and discussing the research and accompanying recommendations to date and extending these—while also clarifying respective roles of agents/actors in implementing these recommendations.

We designed the study so that we could acknowledge our own responsibility for the potential consequences of the way the survey questions were framed and closed- and open-ended responses invited; for the way research assistants were encouraged to engage with respondents/participants; for the way themes were developed and pursued in rapid assessment workshops; for the way learning processes could be (further) engendered via the peer education process; for the way that the national workshop was set up; and finally, for the way in which our report might be received (McKay & Romm, 2008b, p. 154).

Below I concentrate on offering an account of what this meant in relation to the “administration” of the KAPB questionnaire.

Although the KAPB instrument was similar to conventional KAPB surveys in that it contained the standard indicators employed in KAPB surveys, it was expanded to include a range of open-ended questions. In addition, some questions were included with the intention of opening up new ways of considering certain cultural practices (e.g., men alone deciding when to use a condom). We thus tried to incorporate what we call an “active” research agenda where we acknowledged our attempt to “make a difference” in the social arena via the use of the questionnaire, along with the later stages of the research. Our fixed-choice options in relation to condom use included a large variety of reasons for not using a condom—25 in all—as well as an “other” category, where participants could add other reasons. For all our options presented in the instrument, participants could state that they agreed, disagreed, or were unsure. Four of the options provided were developed (drawing upon and elaborating previous KAPB surveys) to provide participants with the opportunity to express whether they had had the experience of being forced and/or not had the skills/power to negotiate condom use with their sexual partners. The exact phrasing was as follows:

I sometimes feel forced to have sex without a condom

I can't negotiate with my partner to use a condom

My partner doesn't like me to use a condom

I don't know how to ask my partner to use a condom

Our questions here, in the KAPB survey, can be said to already serve the purpose of “raising for attention topics normally considered not easy to speak about or raise (and not normally dealt with in behavioural change campaigns)” (McKay & Romm, 2008b, p. 156, italics mine).

Our analysis of the responses included a gender breakdown. The results suggested that, while men felt that they could not easily open negotiations on condom usage, women were more likely to feel forced, to state that their partner was unwilling to use a condom (and therefore they engaged in condomless sex), and to indicate that they did not know how to request their partner to use a condom. This was also expressed in some of the responses people gave when asked in an open-ended question about their “risk behavior.”

As the research assistants interacted with the respondents/participants during the KAPB survey (with some opportunities for this provided in open-ended questions), they learned about different expressions of gender relations. This, in turn, informed their co-facilitation (with participants) of the rapid assessment workshops, where gender relations were discussed and people had the opportunity to consider others' sense of their felt vulnerabilities and how this created risks for both parties. What I would like to highlight here is that some openings were already created via the questionnaire for altering the so-called identifiable “connection” between the variables of “gender” and “attitudes to condom negotiation” via the very way in which the closed- and open-ended questions were posed and raised in the questionnaire for participants to think about. Significantly, then, the questionnaire could serve as a spark for thoughtful reconsideration in relation to more or less “taboo” topics, thus contributing to the way that gender dynamics become viewed and enacted.

It is worth clarifying that I am not suggesting that the presence of open-ended questions in a questionnaire implies that the procedure becomes “qualitative” in character: Instead, I am suggesting that a constructivist epistemological stance allows one to consider both closed- and open-ended questions as opportunities for enabling people to develop their opinions as they engage with the questionnaire. Having pointed to these various options for re-conceptualizing and re-designing the traditional use of questionnaires (designed as tools to “find out” about variables), I now am in a position to summarize, in tabular form, the various paradigmatic underpinnings discussed above in terms of their implications for questionnaire usage.

Table 1

Comparing Understandings of Questionnaire Use

	Rationale for Using Questionnaires	Relationship to Respondents/Participants	Relationship to Wider Society
Positivist/Postpositivist	Questionnaires are designed as a method of advancing knowledge of relationships (correlation and causality) between measurable variables in social reality.	Respondents' responses ideally offer a window into their biographical information, their social positioning, their behavior (as they recall it), and their opinions/attitudes/views. These responses, when analyzed statistically, allow researchers to test for relationships between the variables so that certain statements can be made (with some degree of confidence).	Readers of the results, including policy makers, become more informed about relationships between variables than they were before the studies were undertaken (using scientific protocols).
Interpretivist	Questionnaires are designed as a possible method of advancing knowledge of causal sequences that can be explained on the level of meaning (when combined with additional methods directed at understanding meaning-making).	Respondents' responses ideally offer some window into their biographical information, their social positioning, their behavior (as they recall it), and their opinions/attitudes/views. These responses, when analyzed by researchers (with a view to locating causal sequences operative in specific social contexts), can provide some information, which is complemented by the exploration of meaning via alternative methods.	Readers, including policy makers, of the researchers' interpretations become more informed about the patterning of social life than they were before the studies were undertaken (using the protocols of social science).
Constructivist	Questionnaires can be considered as one method, amongst others, of creating constructed knowledge, which is recognized to be a social construction (created here through the interaction between researchers and participants).	Responses given in questionnaires are admitted to bear the mark of the context of interaction between “the instrument” (as interpreted by participants) and the participants. This context must be made visible at the point at which results are interpreted by researchers. Multiple interpretations can be offered in terms of what researchers believe participants are expressing in response to the various questions asked.	Audiences of (draft) results are invited to offer interpretations of any “connections” between variables. Draft reports need not express a univocal account of what the “results” indicate. Processes (e.g., workshops) should be initiated to reconsider the report so that a variety of interpretations can be discussed.
Trusting constructivist emphasis	Questionnaires are constructed with researchers bearing in mind that the social consequences of the instrument itself will not be neutral; researchers appreciate that the questionnaire can be a tool for forming people's ways of envisaging/framing “problematic” issues.	Researchers make provision for listening to key participants' conceptions of how issues should be framed (before the final questionnaire items are set). Researchers make it clear to participants (e.g., at the top of the questionnaire and/or via research assistants) that they do not regard the instrument as a neutral tool for “data extraction.” Questionnaires can also be aimed at generating a pedagogical context enabling participants as well as researchers to develop enriched perspectives.	Researchers acknowledge that any results presented (tentatively) are a product of the way the questions were framed and the categories created. Any “reporting” is therefore subject to revision by inviting feedback from original participants and wider audiences. Trust can be earned by researchers showing that they are willing to take this feedback seriously as part of the discussion on presented reports. These discussions include considering extended action options in view of various interpretations of the research import.

Conclusion

In this article I have explored possibilities of using questionnaires without operating from an epistemological position where valid knowing is defined as geared to the representation of external realities. Instead, I suggested that questionnaires can be employed within a qualitative-constructivist outlook, where alternative standards for rendering researchers accountable are invoked. Researchers are then tasked to take into account the manner in which research framings can make a difference to the way in which participants and wider audiences envisage social life. In elaborating and extending the constructivist outlook, I made a case for why researchers can be expected to take some responsibility for the potential impact of the construction and administration of any questionnaire.

I have concentrated on fleshing out what it might mean in the context of questionnaire construction to cater expressly for a link between “knowing” and “acting”—so that knowing is not seen as divorced from its involvement in the continuing unfolding of our social existence. This implies being attuned to taking on board “the possible consequences of using particular theories and methods, and the possible consequences of recommendations that might arise from using those methods” (Midgley & Shen, 2007, p. 205).

What I have tried to express in this article is the need for self-reflective consideration on the part of social inquirers of the possible consequences of ways of developing questionnaires, including ways of writing up and eliciting discussions around statements of “results.”

References

Acquah

B. K.

(2007). Community involvement in social research in Botswana. In Rwomire

& Nyamnjoh

F. B.

(Eds.), Challenges and responsibilities of social research in Africa: Ethical issues (pp. 127–132). Addis Ababa, Ethiopia: The Organization for Social Science Research in Eastern and Southern Africa (OSSREA).

Babbie

E. R.

(2011). The basics of social research (5th ed.). Belmont, CA: Wadsworth.

Bryman

(1992). Quantity and quality in social research. London, United Kingdom: Routledge.

Chilisa

(2007). Decolonizing ethics in social research: Towards a framework for research ethics in African contexts. In Rwomire

& Nyamnjoh

F. B.

(Eds.), Challenges and responsibilities of social research in Africa: Ethical issues (pp. 199–208). Addis Ababa, Ethiopia: The Organization for Social Science Research in Eastern and Southern Africa (OSSREA).

Collins

P. H.

(1990). Black feminist thought: Knowledge, consciousness and the politics of empowerment. London, United Kingdom: Harper Collins.

Collins

P. H.

(2000). Black feminist thought: Knowledge, consciousness and the politics of empowerment (2nd ed.). London, United Kingdom: Harper Collins.

Creswell

J. W.

(2003). Research design: Qualitative, quantitative, and mixed methods approaches (2nd ed.). Thousand Oaks, CA: Sage.

Creswell

J. W.

Slope

Plano Clark

V. L.

, & Green

D. O.

(2006). How interpretative qualitative research extends mixed methods research. Research in the Schools, 13(1), 1–11.

Demerath

(2006). The science of context: Modes of response for qualitative researchers in education. International Journal of Qualitative Studies in Education, 19(1), 97–113.

10.

Denzin

N. K.

, & Lincoln

Y. S.

(2000). Introduction: The discipline and practice of qualitative research. In Denzin

N. K.

& Lincoln

Y. S.

(Eds.), Handbook of qualitative research (2nd ed., pp. 1–29). Newbury Park, CA: Sage.

11.

Denzin

N. K.

, & Lincoln

Y. S.

(2003). Introduction. In Denzin

N. K.

& Lincoln

Y. S.

(Eds.), The landscape of qualitative research: Theories and issues (2nd ed., pp. 1–45). London, United Kingdom: Sage.

12.

De Souza

(2004). Motherhood, migration and methodology: Giving voice to the “other.” The Qualitative Report, 9(3), 463–482. Retrieved from http://www.nova.edu/ssss/QR/QR9-3/desouza.pdf

13.

De Souza

(2007). Walking a tightrope: Asian health research in New Zealand. Health and Social Care, 4, 9–20.

14.

Fielding

(2010). Mixed methods research in the real world. International Journal of Social Research Methodology, 13(2), 127–138.

15.

Flood

R. L.

, & Romm

N. R. A.

(1996). Diversity management: Triple loop learning. Chichester, United Kingdom: Wiley.

16.

Flood

, & Romm

N. R. A.

(1997). From metatheory to multimethodology. In Mingers

& Gill

(Eds.), Multimethodology (pp. 291–322). Chichester, United Kingdom: Wiley

17.

Galasinski

, & Kozlowska

(2010). Questionnaires and lived experience: Strategies of coping with the quantitative frame. Qualitative Inquiry, 16(4), 271–284.

18.

Hall

J. N.

, & Ryan

K. E.

(2011). Educational accountability: A qualitatively driven mixed-methods approach. Qualitative Inquiry, 17(1), 105–115.

19.

Hammersley

(1995). The politics of social research. London, United Kingdom: Sage.

20.

Hammersley

(2003). “Analytics” are no substitute for methodology: A response to Speer and Hutchby. Sociology, 37(2), 339–351.

21.

Harris-Lacewell

(2003). The heart of the politics of race: Centering Black people in the study of White racial attitudes. Journal of Black Studies, 34(2), 222–249.

22.

Hesse-Biber

(2010). Qualitative approaches to mixed methods practice. Qualitative Inquiry, 16(6), 455–468.

23.

Hsiung

P-C.

(2012). The globalisation of qualitative research: Challenging Anglo-American domination and local hegemonic discourse. Forum: Qualitative Research, 13(1).

24.

Johnson

R. B.

(2009). Toward a more inclusive “scientific research in education.” Educational Researcher, 38(6), 449–457.

25.

Johnson

R. B.

Onwuegbuzie

A. J.

, & Turner

(2007). Toward a definition of mixed methods research. Journal of Mixed Methods Research, 1(2), 112–133.

26.

Kenny

(2002). North American Indian, Métis and Inuit women speak about culture, education and work. Ottawa, ON: Status of Women Canada.

27.

Ladson-Billings

(2003). Racialized discourses and ethnic epistemologies. In Denzin

N. K.

& Lincoln

Y. S.

(Eds.), The landscape of qualitative research: Theories and issues (2nd ed., pp. 398–432). London, United Kingdom: Sage.

28.

Lincoln

Y. S.

, & Denzin

N. K.

(2003). The seventh moment: Out of the past. In Denzin

N. K.

& Lincoln

Y. S.

(Eds.), The landscape of qualitative research: Theories and issues (2nd ed., pp. 611–640). London, United Kingdom: Sage.

29.

Lincoln

Y. S.

, & Guba

E. G.

(2003). Paradigmatic controversies, contradictions, and emerging confluences. In Denzin

N. K.

& Lincoln

Y. S.

(Eds.), The landscape of qualitative research: Theories and issues (2nd ed., pp. 253–291). London, United Kingdom: Sage.

30.

McIntyre-Mills

J. J.

(2008). Systemic ethics: Expanding the boundaries of rights and responsibilities. Systems Research and Behavioral Science, 25, 147–150.

31.

McKay

V. I.

(2003). Consolidated report on HIV/AIDS interventions in the informal sector: A four-country study. Geneva, Switzerland: ILO.

32.

McKay

V. I.

, & Romm

N. R. A.

(2006). Report on HIV/AIDS in relation to the informal sector in Zambia. Retrieved from http://www.academia.edu/1391794/Report_on_HIV_AIDS_in_Relation_to_the_InformalSector_in_Zambia

33.

McKay

V. I.

, & Romm

N. R. A.

(2008a). A systemic approach to addressing HIV/AIDS in the informal economy in Zambia: Methodological pluralism revisited. International Journal of Applied Systemic Studies, 1(4), 375–397.

34.

McKay

V. I.

, & Romm

N. R. A.

(2008b). Active research toward the addressal of HIV/AIDS in the informal economy in Zambia: Recognition of complicity in unfolding situations. Action Research, 6(2), 149–170.

35.

Mertens

(2010). Philosophy in mixed methods teaching: The transformative paradigm as illustration. International Journal of Multiple Research Approaches, 4, 9–18.

36.

Midgley

(2000). Systemic intervention: Philosophy, methodology and practice. New York, NY: Kluwer Academic/Plenum Publishers.

37.

Midgley

, & Shen

C. Y.

(2007). Toward a Buddhist systems methodology 2: An exploratory, questioning approach. Systemic Practice and Action Research, 20, 195–210.

38.

Naidoo

(2008). Researching reproduction: Reflections on qualitative methodology in transforming society. Forum: Qualitative Social Research, 9(1).

39.

Onwuegbuzie

A. J.

, & Leech

N. L.

(2007). Validity and qualitative research: An oxymoron? Quality and Quantity, 41, 233–249.

40.

Pollack

(2006). Pyramids or silos: Alternative representations of the systems thinking paradigms. Systemic Practice and Action Research, 19, 383–398.

41.

Punch

K. F.

(2003). Survey research: The basics. London, United Kingdom: Sage.

42.

Rajchman

(2007). Introduction. In Foucault

& Lotringer

(Ed.), The politics of truth (pp. 9–28). Los Angeles, CA: Semiotext(e).

43.

Reason

(2006). Choice and quality in action research practice. Journal of Management Inquiry, 15(2), 187–203.

44.

Romm

N. R. A.

(1997). Becoming more accountable: A comment on Hammersley and Gomm. Sociological Research Online, 2(3). Retrieved from http://www.socresonline.org.uk/2/3/2.html

45.

Romm

N. R. A.

(2001). Accountability in social research: Issues and debates. New York, NY: Kluwer Academic/Plenum.

46.

Romm

N. R. A.

(2002). A trusting constructivist approach to systemic inquiry: Exploring accountability. Systems Research and Behavioral Science, 19(5), 455–467.

47.

Romm

N. R. A.

(2007). Issues of accountability in survey, ethnographic, and action research. In Rwomire

& Nyamnjoh

(Eds.), Challenges and responsibilities of social research in Africa: Ethical issues (pp. 51–76). Addis Ababa, Ethiopia: The Organization for Social Science Research in Eastern and Southern Africa (OSSREA).

48.

Romm

N. R. A.

(2010). New racism: Revisiting researcher accountabilities. New York, NY: Springer.

49.

Romm

N. R. A.

(2011). Reconsidering methodological arguments: A commentary on Stephen L. Smith's paper “Naïve expertise: Spacious alternative to the Standard Account of Method (SAM).” Philosophy of Management, 10(2), 75–90.

50.

Sapsford

(2007). Survey research (2nd ed.). London, United Kingdom: Sage.

51.

Scott

(2010). Quantitative methods and gender inequalities. International Journal of Social Research Methodology, 13(3), 223–236.

52.

Speer

S. A.

(2002). “Natural” and “contrived” data: A sustainable distinction? Discourse Studies, 4, 511–525.

53.

Speer

S. A.

, & Hutchby

(2003). Methodology needs analytics: A rejoinder to Martyn Hammersley. Sociology, 37(2), 353–359.

54.

Taylor

, & Hoechsmann

(2011). Beyond intellectual insularity: Multicultural literacy as a measure of respect. Canadian Journal of Education, 34(2), 219–238.

55.

Torlak

G. N.

(2001). Reflections on multimethodology: Maximizing flexibility, responsiveness, and sustainability in multimethodology interventions. Systemic Practice and Action Research, 14(3), 297–337.

56.

Weber

(1930). The Protestant ethic and the spirit of capitalism ( Parsons

, Trans.). London, United Kingdom: Unwin Hyman.

57.

Weber

(1949). The methodology of the social sciences ( Shils

E. A.

& Finch

H. A.

, Eds. & Trans.). New York, NY: The Free Press.