can you help?

260 IEEE TRANSACTIONS ON PROFESSIONAL COMMUNICATION, VOL. 53, NO. 3, SEPTEMBER 2010

Ethos, Pathos, Logos, Kairos: Using a Rhetorical Heuristic to Mediate Digital-Survey Recruitment Strategies —MARTINE COURANT RIFE

Abstract—How might the rhetorical strategies of ethos, pathos, logos, and kairos play a mediational, intervening role in the successful administration of online surveys? What are the general costs and benefits of conducting survey research? Based on the activity of administering an online survey (� � 334) testing knowledge and understanding of US copyright law among digital writers (both students and teachers) in US technical and professional writing (TPW) programs, I blend Rhetorical Theory with Activity Theory by conducting a rhetorical analysis within an Activity Theory paradigm. I posit that a rhetorically informed heuristic mediates between the researcher and potential participants when the researcher attempts to recruit individuals to respond to an online survey.

Index Terms—Activity Theory, attrition, community, digital survey, ethos, heuristic, kairos, logos, mediation, pathos, response rates, rhetoric, survey design.

In this piece, I blend Rhetorical Theory with Activity Theory by conducting a rhetorical analysis within an Activity Theory paradigm to posit that a rhetorically informed heuristic acts as a mediational means between the researcher and potential participants when the researcher attempts to recruit individuals to respond to an online survey. The concept of “mediational means” draws from Vygotsky’s Activity Theory and indicates means, such as a rhetorically informed heuristic, that intervene in the given activity, in this case the administration of an online survey. “Mediation” occurs when an actor coordinates between the task environment and some other means in order to achieve a goal. In this case, the researcher’s activity of administering an online survey was mediated by a rhetorically informed heuristic. By blending Rhetorical Theory with Activity Theory in executing research, the question becomes: How might the rhetorical strategies of ethos, pathos, logos, and kairos play a mediational, intervening role in the successful administration of online surveys? This is an important question as researchers increasingly conduct online-survey research due to the efficacies of time and money. The administration of online surveys can easily cross international borders and is economical. Also, if online surveys are designed properly and achieve adequate participation, they can be an extremely efficient means of collecting

Manuscript received August 17, 2008; revised January 31, 2009; accepted April 28, 2009. Date of current version August 25, 2010. The author is with the Communication Department, Lansing Community College, Lansing, MI 48901-7210 USA (email: [email protected]). Color versions of one or more of the figures in this paper are available online at http://ieeexplore.ieee.org.

IEEE 10.1109/TPC.2010.2052856

data for multiple purposes. Yet, we might ask what the challenges are to using such methods and how those challenges might be addressed with the use of rhetoric as a mediational means.

This issue arose in a two-year, mixed-methods study on rhetorical invention in copyright-imbued environments wherein knowledge and understanding of US copyright law among digital writers—both students and teachers—in US technical and professional writing (TPW) programs were examined [1]–[4]. The research was designed in two phases: a quantitative phase (survey) and a qualitative phase (discourse-based interviews with digital writers) [5]. The quantitative phase used an online survey that was administered to individuals in a randomly selected population of TPW programs. Participants in the study gave informed consent before participating.

In order to examine the two questions posed in this introduction—how rhetoric might intervene in survey research and what the general costs and benefits to conducting survey research are—this paper focuses exclusively on the study’s first phase: the digital survey ( 334). I begin with a description and discussion of rhetorical concepts: ethos, pathos, logos, and kairos. Next, I describe Activity Theory as a framing paradigm. The discussion of Activity Theory is followed by a brief contextualization of surveys, online-survey collection methods, survey-design issues as they relate to attrition, and the benefits and limitations of online-survey research. Finally, a rhetorical analysis of ethos, pathos, logos, and kairos as mediational means in the research recruitment process is followed by a discussion of implications and a conclusion.

0361-1434/$26.00 © 2010 IEEE

RIFE: ETHOS, PATHOS, LOGOS, KAIROS 261

RHETORICAL THEORY Although there are not conclusive agreements on the meaning of rhetoric or rhetorical analysis [6], I define rhetoric as the art of persuasion as well as a particular epistemological outlook that emphasizes the ability to generate probable knowledge. Rhetorical analysis is defined as “an effort to understand how people within specific social situations attempt to influence others through language” [6, p. 281]. Here, I leveraged the power of rhetoric as mediational means in an attempt to recruit (influence) possible digital-survey participants.

Ethos, pathos, logos, and kairos—the first three are commonly known as inventional strategies or modes of persuasion. Kairos is not discussed as often in the literature, but it has recently experienced a revival in interest [7]. Kairos’s interplay with ethos, pathos, and logos may provide some idea of how one might employ rhetoric in order to successfully administer an online survey.

The five well-known canons of rhetoric are invention, arrangement, style, memory, and delivery. Ethos, pathos, and logos are usually considered to fall under the canon of invention. Under invention, two kinds of proofs are the inartificial proof (hard evidence) and artificial proofs, also known as inartistic and artistic proofs [8]. Atechnic (inartistic) proofs are those already existing, and entechnic (artistic) proofs are those invented and are the rhetor’s techne [9], [10]. The three artificial/artistic/entechnic proofs are ethos, pathos, and logos. While some scholarship argues the researcher is a rhetorician [10], [11], or a persuader, and other literature argues that methodology choices themselves are rhetorical [12], there is little in our literature pointing specifically to the use of rhetoric as mediational means in the research process itself, particularly in the context of administering an online survey.

Ethos Ethos is “establishing the persuader’s good character and hence credibility” [13, p. 166]. Constructing effective researcher ethos is important when conducting online studies [14]. Ethos should be developed as a “valid and vital focus for compositionists” [15, p. 149]. Henry argues ethos should be central to composing, including composing in research [15, p. 149]. It is helpful to think of Henry’s notion of the “implied author” as constructed in a technical document, and the writer’s ability to shape the reader’s perception. Elements of ethos, including credibility, good character, personal qualities, connotations, and

affective influence, can be constructed through writing. In digital contexts, the construction of ethos through writing is crucial because of writing’s centrality in representing the author. As I learned in my unplanned experimentation with constructing ethos in my recruitment email, professional writers need to self-consciously employ technology as they create implied authors [15, p. 153].

Pathos Pathos is “putting the audience in the appropriate mood, by playing on its feelings” [13, p. 166]. Appeals to pathos are the “persuasive reasons in an argument that derive from a community’s mostly deeply and fervently held values” [6, p. 284]. Formal instructive materials on the art of rhetoric and use of pathos emerged around the fifth century [16]. According to Kennedy, rhetoric developed a metatheory of itself due to the “application of the democratic process on a large scale to judicial procedure” in fifth-century Athens [16, p. 27]. Since regular citizens were the rhetors who argued their cases before the jury (the legal system along with the notion of “lawyers” had not yet developed) and because of the “popular and unprofessional nature of the jury,” appeals to logic were not as effective and presented the rhetor opportunities for increased “brilliant digressions and emotional appeal” [16, p. 28]. In a landmark but ancient (370 BCE) text in rhetoric studies, Plato’s Phaedrus, a famous dialogue takes place between two characters: Phaedrus and Socrates. Socrates says that through the “field of pathetic expressions” a speaker can acquire the “ability to stir a whole crowd of people at once and the same time to frenzy and then to charm them out of it by magic” [16, p. 56]. Ethos works with pathos because ethos first secures the audience’s sympathy through the speaker’s credibility, verifying the speaker’s conviction [16, pp. 93–94]. Once the audience feels secure, this sense of security is played upon by the emotional appeal, efficiently increasing the level of persuasion. Due to this power, our composing practices should centralize ethos [15]. Obviously, with research, one must always act ethically and responsibly toward participants. It is important to keep this in mind due to rhetoric’s power.

Logos Logos is logic, reasoning, or “rational argument” [13, p. 166]. The Greek sense of logos is “reasoned account” [17, p. 98]. Logos can be both “ambiguous” and “mystical” [16, p. 8]. Word choices, logic choices, and readable sentence structures are attributable to the logos one constructs. Logos also blends with ethos because a well-developed and admirable sense of logos can help establish ethos.

262 IEEE TRANSACTIONS ON PROFESSIONAL COMMUNICATION, VOL. 53, NO. 3, SEPTEMBER 2010

Kairos Kairos is timing. Kairos is “the Greek word for time, place, circumstances of a subject” [13, p. 94]. Kairos can be employed to evaluate an argument’s “contextual appropriateness or situational fit” [18, p. 230]. The Greeks understood two senses of time: the quantitative sense, or chronos, and the qualitative sense, or kairos. Chronos is important to online surveys simply because individuals do not want to spend lots of time on surveys. Kairos is application, something done at the appropriate season, not just at any time but at the right time [18]. In this study, kairos was relevant to an application in an open system. Sending the survey-recruitment email had to be timed properly, considering the audience’s local needs. From the email recipient’s perspective, kairos influences whether the recipient chooses to save, delete, or take the time to click the survey link:

In Hellenistic depictions, the god Kairos often holds a pair of scales, and sometimes the god is portrayed surreptitiously depressing one pan of the scales in his favor. [17, p. 57]

The application of any given activity where kairos is involved will be temporal. The term also “carries a spatial metaphor, that of a critical opening” and in “both archery and weaving, referred to a ‘penetrable opening, an aperture,’ through which an arrow or a shuttle must pass” [7, pp. 83–84]. The effective interaction between pathos and ethos, and even logos, as they construct the implied author in a text is informed by kairos: placing a well-balanced appeal to emotion at the right time, in the right place of a text, with the right ethos. In some sense, kairos trumps the other three rhetorical strategies because no matter how perfect the ethos, no matter how strong the pathos, and no matter how logical the argument, if the timing is wrong, the audience will not listen and will not take the survey.

ACTIVITY THEORY In this section, Activity Theory is blended with Rhetorical Theory to show how ethos, pathos, logos, and kairos might function as mediational means in the research process. Activity Theory has become increasingly popular among researchers across disciplines [19] and is increasingly used to study composing processes [20], [21]. Hart-Davidson justifies using Activity Theory in his writing research for all the reasons it helps frame my research. Activity Theory addresses processes that

unfold in an improvisational way in response to a rhetorical situation take place over

relatively long periods of time and are mediated by multiple technologies. [21, p. 156]

Activity Theory helps because existing literature on effective survey recruitment strategies is largely situated in the behavioral sciences where variables are isolated and tested. Isolating and testing variables prevents a more ecological research model where research “unfolds in an improvisational way in response to a rhetorical situation” [21, p. 156]. Isolating and testing variables can erase rhetorical strategies’ ability to become agents in the successful recruitment of survey participants. In my analysis, a heuristic informed by rhetorical strategies of invention serves as one of the mediating “multiple technologies” of the researcher, one that assists the researcher with the improvisation that often occurs during the research process but is rarely discussed in the study write-up. In Fig. 1, a traditional Activity Theory triangle shows how one might visualize a rhetorically informed heuristic mediating digital-survey recruitment strategies.

EXISTING RESEARCH ON ISSUES IN SURVEY RESPONSE RATES

Best practices with surveys are to avoid leading questions, keep questions short, be strategic with the order of questions, and pretest and thoughtfully organize questions [22]. The survey design should include categories and subcategories of what one wants to measure. It has been found that a higher response rate is obtained by combining postal and email methods, piloting the survey first, ensuring the digital survey is supported in multiple browsers, using cash incentives or prizes, telling individuals the survey will not take much time, providing email prenotifications, and sending reminder emails. Higher response rates were also obtained when there was preexisting social cohesion in the surveyed group, when the request to participate drew on respondents’ interests and/or beliefs, when the survey came from an authority, when respondents felt the chance to participate was somewhat exclusive, when participants felt their voices could be heard, when the researcher was likable, and when they understood that they would be “helping” by taking the survey [23], [24]. Survey attrition was higher when questions were crowded visually, placed in tables, and when the survey contained many open-ended questions requiring multiple answers [23], [24].

In my digital survey, I carefully planned the survey recruitment and design so that they would meet all

RIFE: ETHOS, PATHOS, LOGOS, KAIROS 263

Fig. 1. Traditional Activity Theory triangle.

of the enumerated best practices. While funding was not available to either provide cash incentives or use US mail, each additional survey best practice was carefully considered, and every attempt was made at implementation. The survey was piloted in the spring of 2006 and again in the summer of 2007. The pilot phase provided feedback from test participants so that language in the survey as well as functionality of the survey itself could be tested and revised. Fig. 2 shows the initial email that was sent to program directors after the pilot phase.

I attempted to implement all best practices. I informed potential participants that the survey was “a short, 20-minute online survey.” The email itself, since it did not contain a link to the survey, served as a prenotification. I also sent a reminder email. In my original email, I tried to draw on what I believed were participants’ interests and beliefs by informing them that

I will be posting aggregate survey results online soon and will share this information with all participants as well as the technical/professional writing community.

By positioning myself in alliance with the Writing in Digital Environments (WIDE) Research Center at Michigan State University, I tried to create an ethos of authority. Notice that while my signature block reflected my status as “Ph.D. Candidate,” I did not call any attention to the fact that the survey was needed in order for me to complete my dissertation research. I thought doing this would make me less authoritative as a researcher, and the social science literature indicated that response rates would be higher if the researcher was an authoritative figure. By informing the participants they had been randomly selected, I attempted to make them feel the chance to participate was somewhat exclusive. I attempted to make them understand their voices would be heard and that they would be helping out when I stated

Participation from both faculty and students is crucial to the success of this study, and will add important field knowledge If you think you don’t have much knowledge on this, that’s OK. I would still love you to participate, and that would indeed provide exceedingly useful information.

264 IEEE TRANSACTIONS ON PROFESSIONAL COMMUNICATION, VOL. 53, NO. 3, SEPTEMBER 2010

Fig. 2. Initial email.

I tried to sound friendly and likable in the recruitment email, as specified in the literature.

As far as survey design, questions in the digital survey were generally one per browser page in order to avoid visual crowding. No questions were in tables, nor did I use any open-ended questions other than ones that asked for contact information for those interested in participating in the interviews. I anticipated survey administration would run smoothly because I had done everything specified in the social science literature. This assumption turned out to be incorrect, spurring my need to employ the rhetorical strategies of ethos,

pathos, logos, and kairos as mediational means in the research process.

ONLINE SURVEYS IN CONTEXT

Online digital surveys are becoming increasingly popular, in part because of their value. The online survey in this case was inexpensive—just US$20 per month for a subscription to surveymonkey.com. Another benefit to using an online survey service such as surveymonkey.com is that it instantly aggregates all of the data for the researcher and allows the data to be instantly downloaded to

RIFE: ETHOS, PATHOS, LOGOS, KAIROS 265

Microsoft Excel or other data-processing software, such as SPSS.

The digital survey took place in late fall 2007 as part of a larger dissertation study. The study used a mixed-methods, sequential research design consisting of a survey ( 334) followed by interviews with seven digital writers about whether and how knowledge and understanding of copyright law was mediating digital-composing choices. The interviews are not discussed in this paper because of space and because in the qualitative interviews, recruitment was not an issue. Individuals selected for interviews were professional writers with whom I already had some contact either as friends or colleagues. (For details of the study findings, see [3].)

While online survey methods are relatively new, surveys as instruments of quantitative-data collection have been viewed as legitimate research methods for some time in the social sciences. One of the earliest known uses was documented in the Bible, Numbers 26: 1–2:

After the plague the Lord said to Moses and to Eleazar the son of Aaron, the priest, “Take a census of all the congregation of the people of Israel, from twenty old and upward.” [22, p. 243]

History also situates the survey in other little-known contexts. In 1880, a German political sociologist mailed 25,000 questionnaires to French workers to determine the extent of their exploitation by employers. The researcher was Karl Marx, but while we have records of the survey dissemination, no record exists of any response rate [22, p. 243].

Both online and offline surveys are good at accommodating and measuring characteristics in large populations and providing flexibility in questions asked. But surveys can be superficial in their coverage of complex topics; they cannot deal well with contexts. In some ways, surveys are inflexible because once launched, they must remain unchanged (or you risk complicating your results). They are artificial, like experiments. According to Babbie, surveys are generally weak on validity (as in testing what it is you want to test) and strong on reliability (as in being replicable) [22, p. 278]. As Babbie points out, online surveys are good because they are inexpensive and easy, but they are less likely to represent a more general population. Andrews, Nonnecke, and Preece specifically point to the lack of equivalence between online and offline populations [23]. Lack of equivalence might not be relevant, however, depending on your research questions. The research questions in my survey

specifically examined only individuals who had published on the web, so the differences between online and offline populations were not important.

PROBLEMS WITH DIGITAL SURVEY RESPONSE RATES

Recent research indicates that online surveys generally have much lower response rates than mailed surveys [25]–[28]. According to Andrews, Nonnecke, and Preece:

Email response rates of 20% or lower are not uncommon and, although rates exceeding 70% have been recorded, they are attributed to respondent cohesiveness (e.g., an existing workgroup) as often occurs in organizational studies. [23, p.11] (See also [29] and [30].)

Three main reasons for low response rates with digital surveys as compared to paper or in-person surveys are (1) lack of a cohesive community in online settings; (2) the fact that email surveys might be perceived as a harassment, or “spam”; and (3) possible lack of technological savvy among recipients.

Regarding the issue of community, when comparing paper and electronic survey response rates, it has been found that those who return online surveys are more connected to their profession and have more contact with colleagues compared to those who returned paper surveys [24], [31]. With respect to the spam problem, studies have indicated that survey emails felt like an invasion of the potential participant’s computer space and were viewed as rude [32], [33]. Finally, regarding a possible lack of technological savvy among recipients, research validating this possibility also suggests low response rates might be due to low self-confidence regarding technology among digital-survey recipients [31].

Another problem with online survey research is that certain biases occur when using online data-collection methods since, currently, offline and online populations are not equivalent due to the digital divide. Research shows online survey respondents were younger and more confident about their abilities to navigate the web and were online more often than those who responded to paper surveys [34]. While some research indicates that the content results from postal and digital surveys are becoming similar,

to infer for a general population based on a sample drawn from an online population is not

266 IEEE TRANSACTIONS ON PROFESSIONAL COMMUNICATION, VOL. 53, NO. 3, SEPTEMBER 2010

as yet possible and will not be possible until the online and offline populations reflect each other. [23, p. 10]

The social science survey research that informed the drafting of the initial recruitment email is epistemologically situated in a causal-process model, where variables are isolated and tested in a traditional stimulus-response framework. Chains of causality are then theorized based on the presence or absence of certain variables. The blending of Rhetorical Theory with Activity Theory in this paper is in response to what those of us with a rhetorical view of the world may see as a limited perspective provided by such cause-and-effect-based methodologies. The use of the Activity Theory paradigm is particularly appropriate since Vygotsky explicitly developed it as a refutation of stimulus-response methodologies [35]. Vygotsky’s concept of mediation, embodied in the Activity Theory paradigm as illustrated in Fig. 1, acts as intervention in the typical stimulus-response model. Similarly, I argue that rhetorical invention strategies, or modes of appeal, such as ethos, pathos, logos, and kairos can mediate and facilitate the work of researchers conducting online-survey data collection.

ONLINE SURVEY RESEARCH AND ACCOMPANYING PROBLEMS During the administration of the digital survey, I, as the researcher, encountered some serious problems that had I not relied on the intervention of rhetorical strategies as mediational means, would have caused my project to fail. To provide context, here are the research questions the survey explored:

(1) What do writers know and understand about copyright and fair use?

(2) How confident are writers in this knowledge? (3) Are writers unable to express themselves fully

because of fear of copyright liability (i.e., is their digital “speech” chilled)?

(4) How important do writers think knowledge of copyright and fair use is to their work?

(5) What is the relationship, if any, among knowledge of copyright/fair use, confidence in that knowledge, and levels of “chilled speech” in digital environments?

Between September 17, 2007, and November 2, 2007, a total of 1,935 recruitment emails was sent, with a total response rate before survey attrition of 446, for an overall response rate of 23%. (See Fig. 3 for attrition.) The final email reminder was

sent out November 19, 2007, and the survey closed November 30, 2007. Therefore, survey data were collected over 11 weeks. I contacted students and teachers in 155 TPW programs and/or writing majors; I received responses from 64 TPW programs, for a programmatic response rate of 41%. The 64 programs that had some representation in the survey are listed in Appendix A.

Without using the rhetorical strategies of ethos, pathos, logos, and kairos as intervening mediational means, it is likely the overall response rate would have been much lower. This paper addresses problems I encountered in hope of calling for further research in this area. It must be noted, as explained below, that the research design in the larger study did not expressly anticipate the necessity of employing rhetoric in order to recruit participants. The need to use rhetoric for purposes of recruiting digital-survey participants arose out of necessity during the research process, a process that bordered on total failure early in the survey administration.

As far as response rates, or the measure of whether or not recruitment worked, it is certainly possible and even probable that many participants deleted the email without opening it. They may have considered it spam, as the social science literature posits. Without further research, the reasons for low response rates cannot be known for sure, but we can look at the problems that arose in my research and theorize how rhetoric as mediational means might address those problems for future researchers.

Problems and Solutions for Population Sampling In order to theorize whether there could have been problems in response rates due to the lack of community [24], [31], it is useful to understand the population sampled and subsequently contacted via a recruitment email. Knowing the population might help determine whether a lack of community contributed to low response rates. Using traditional social science methods as explained by Babbie [22], the population surveyed was a stratified, random selection of 155 TPW programs and National Council of Teachers of English (NCTE) writing majors appearing on Society for Technical Communication (STC) and Association of Teachers of Technical Writing (ATTW) membership lists as well as on NCTE writing major lists as of September 2, 2007.

I noted discrepancies among the ATTW, STC, and NCTE lists, possibly pointing to a lack of community

RIFE: ETHOS, PATHOS, LOGOS, KAIROS 267

or cohesion in the TPW community. That it was necessary to piece together a membership of TPW programs from three lists indicates the lack of a definitive document uniting these professionals as one community. Key technical writing programs were missing from the STC list and the ATTW list but were present on the NCTE four-year program list. Once a master list was created from these lists, programs were divided into five categories: Ph.D. programs, M.A. programs, four-year programs, two-year programs, and certificate programs. The stratified sample that received my email solicitation was randomly selected from the master list. From the selected programs, I obtained email addresses for potential survey participants.

Difficulties Attempting to Locate Participant Emails To collect data, email addresses were gathered from program websites. The difficulty that this task presented due to problems with program website design and with who was and was not permitted visibility on a webpage is discussed by Knight et al. [36]. If student emails were unavailable, students from those particular programs were not contacted. If a program website did not include student contact information, which was the norm, college/university directories were examined. If students’ names but not emails were provided on program websites, online college directories were used to locate student emails where available. Therefore, the survey population contains a bias in favor of websites that make student as well as faculty contact information visible. If neither faculty nor student emails were available, I could only contact the program director to pass along my survey.

Attrition and Survey Design Recall that in my survey design, questions in the digital survey were one per page in order to avoid visual crowding. No questions were in tables, nor did I use any open-ended questions other than ones that asked for contact information for those interested in participating in the interviews. Yet I experienced attrition of participants, as illustrated in Fig. 3.

In Fig. 3, the axis represents the number of survey respondents, and the axis represents the survey questions as they progressed over time. Notice the plateau at 373 on the axis and between 5 and 10 on the axis, which I discuss below. With respect to attrition, 446 participants began the survey (i.e., consented to take the survey), while 334 finished the entire survey. These data are automatically collected by the survey software used. Note that rhetoric used as mediation in the

recruitment process is a problem different from, but connected to, the problem of survey attrition due to factors such as survey design, length of the survey, difficulty of the questions, and other variables out of the control of the researcher. The issue of survey attrition is worth mentioning here, however, because it occurred during this study. Although overall attrition occurred during the survey, there were six questions, as depicted in Fig. 3 on the axis between 5 and 10, where no attrition occurred. In other words, 373 participants stayed with the survey consistently for these six questions. That there was no attrition for six consecutive questions in the survey is something worth noting for those who wish to design surveys. The six questions were short and concise and asked participants about their copyright behavior (rather than their copyright knowledge). Participants may have found these particular questions easy to answer and interesting.

Fig. 3 shows that while 446 participants began the survey, 334 finished the entire survey, for a survey attrition rate of 112, or 25%. (Only 437 participants are illustrated in Fig. 3 because between the informed consent agreement and the first official question, 9 participants dropped out of the survey.) Note that the first two questions, depicted in Fig. 3 on the axis between 1 and 2, were screening questions on whether the respondent had published to the web. Therefore, “attrition” in the technical sense might be measured between question 2 and the last multiple-choice question. In this case, the participants start at 395 and end at 334, for an attrition rate of only 15%.

From the 64 TPW programs and/or writing majors, 334 students (41%), teachers (47%), and others (12%) completed the entire survey. “Others” were participants who did not identify themselves as either students or teachers but who were available through programmatic emails. It is likely that “others” were former students, alumni, or teachers who had moved out of academia; however, the data do not make it possible to know for sure who they were. Table I shows the stratification in the survey responses for those who completed the entire survey.

The main variation between the actual population that was contacted with the recruitment email and the participating population was that Ph.D. programs participated in the survey on a significantly higher level than how they appear in the entire population, while two-year programs participated at a significantly lower level. This difference might be expected since, in general,

268 IEEE TRANSACTIONS ON PROFESSIONAL COMMUNICATION, VOL. 53, NO. 3, SEPTEMBER 2010

Fig. 3. Survey attrition.

TABLE I STRATIFICATION-REPRESENTATION OF

PROGRAMS IN STUDY POPULATION

individuals in Ph.D. programs tend to interact with research more often than those in two-year programs. So, the explanation could be cultural, but other reasons might explain this as well. Perhaps Ph.D. programs have more technology available or have more technology integrated into the curriculum, making the survey more accessible and more relevant. The reasons for the difference between the actual population and the respondents ultimately remain unknown. But the difference must be taken into account for purposes of generalizability. The studied population is biased in favor of Ph.D. programs and does not have adequate representation of two-year programs.

The survey had 6 parts and 29 questions total—15 one-part and 14 two-part questions. The questions were organized based on hypotheses. Each variable in a hypothesis necessitated survey questions as indicators. While variable isolation and testing of hypotheses is not necessarily the ultimate way to create knowledge, surveys lend themselves to this kind of exploration. The isolation of variables helps to organize any kind of writing—an outline, or the use of subheadings, for example, is a kind of isolation of variables. It is worth noting that to fill gaps left by the survey design and its isolation of variables, the overall mixed-methods research design of the entire study included interviews with seven survey participants: digital writers who volunteered their contact information. In the face-to-face interviews, I explored the same as well as additional research questions, and web texts were collected from interviewees to formulate discourse-based interview questions to elicit tacit knowledge about their digital composing choices. (For the full study report, see [3].)

Why No Attrition on the Chilled Speech Questions? As mentioned, there were six questions in the survey where no attrition occurred. In order to offer a possible explanation for this, a brief overview describing the kinds of elements surveys can measure is necessary. Generally, a survey can gather information about individuals

RIFE: ETHOS, PATHOS, LOGOS, KAIROS 269

and take measurements of their attitudes, cognition, and behavior [22], [37]. For example, the variables measured in the digital survey were slotted into measurements such as demographics (educational level, student/teacher/other, institutional affiliation); attitude toward importance of knowledge of copyright/fair use; cognitive abilities (knowledge of copyright/fair use on a 14-point index and confidence in one’s knowledge of copyright/fair use on a 5-point scale); attitude towards obeying one’s conscience versus obeying the law; and behavior (chilled speech on a 6-point index).

The following is an example of a survey question that tested participants’ cognitive abilities, or their knowledge of how fair use applies only in the case of unauthorized use:

If you ask a copyright holder for permission to use their material in your own original web composition, and they say no, this means you absolutely cannot use the copyright holder’s material under fair use.

Note that the best answer was “false” because fair use does not require permission and is often relied upon when permission is not given. For example, with respect to seeking permission to create a parody, copyright holders are unlikely to agree to permit someone to use their materials in order to make fun.

After opening with some typical informed-consent language, the survey’s first part contained three questions and screened the individuals, noting whether they accepted the terms and whether they had ever composed for and published to the web. The second part asked one question about whether it was more important to obey the law or follow one’s conscience. The third part asked a single question on educational level. The questions in the survey where there was no attrition appeared in the fourth part of the survey. This part contained six questions measuring behavior, that is, whether or not one’s digital speech was chilled. As a possible explanation as to why there was no attrition during these six questions, it could be because the questions were short: They only asked whether or not a participant had experienced something instead of asking the participant to analyze a scenario or exercise analytical skills. Also, the questions may have been particularly interesting to the participants. Ultimately, however, without further research, it is unknown why there was no attrition during these six questions. These questions are noted in Table II.

In the fifth section of the survey, participants were asked a total of 14 substantive questions on basic copyright law and the fair use four-factor test. In order to answer these questions, some of which were quite difficult, participants had to know something about copyright law. They had to have knowledge outside their own experiences to answer. The difficulty of answering the cognitive copyright questions likely explains why much survey attrition occurred after the chilled-speech questions. Some questions in this part were based on existing online quizzes, but many also echoed Waller, Bryan, and Howard’s fair-use inquiries [38]–[40]. In formulating the questionnaire, I also used my training as a lawyer specializing in intellectual property (particularly fair use), as I am admitted to practice in two states, with an active license in one.

In the final four-question section of the survey, participants were asked how important they thought knowledge and understanding of copyright law was to their work as digital writers. They were also asked how they primarily described themselves, as students, teachers, or others; they were asked to specify their institution; and, finally, students were asked if they were willing to participate in an interview. Survey responses were anonymous unless participants volunteered identities.

The scope of the survey was limited because it is unfair to subject research participants to long, complicated surveys [22], [37]. Also, survey participants were offered no payment or other remuneration, although as an attempt at reciprocity, I promised to share research results and resources (survey answers) with those who identified themselves.

DISCUSSION The social science literature indicated that groups that have a high sense of cohesion or community tend to respond at higher rates to surveys. The email-recruitment strategy was designed with the idea that the TPW community was well developed as a community. Even though some scholarship in the field discussed a kind of identity crisis with TPW, I thought that since many regarded members of the TPW educational community as digital-writing experts who are interested in digital research (see [41]), my online survey would be welcome. While I initially thought that the academic community of TPW was unified enough to readily respond to a digital survey about itself, that assumption turned out to be incorrect. After all, the initial response

270 IEEE TRANSACTIONS ON PROFESSIONAL COMMUNICATION, VOL. 53, NO. 3, SEPTEMBER 2010

TABLE II CHILLED SPEECH INDEX: � � 373

rate was only 23%, and there was only an overall response rate of 17% for those who finished the entire survey.

Even though the response rate was low, the overall number of 334 participants yields, to the extent that the responding population represents the whole, a confidence level of approximately 94% (plus or minus 5.6%). Lauer and Asher explain that with a population of 30 students, if it is discovered that 60%

have problems with organization, the researcher can conclude that in the entire population, 60% plus or minus 18% also have problems with organization. [42, p. 59]

So with a randomly selected population of 30, with all 30 responding, if 60% have a certain variable, 42% to 78% have this same variable in the larger population. Since increased certainty in the research results comes with larger populations, larger populations are desirable when administering digital surveys. Larger populations are only obtainable in “live” research if one gets people to participate. The ability to obtain participation is where rhetoric as mediational means in the research process may play an important role. If the researcher thinks of a

rhetorically informed heuristic comprising the persuasive strategies of ethos, pathos, logos, and kairos as mediating the research process, it might make the researcher more successful when attempting to administer an online survey.

Since the population that responded to the survey has some equivalence to the overall population (i.e., the population on the master list), the data collected can provide some additional understanding of the research questions being explored. Babbie states that from a social scientist’s perspective and as a general rule, a response rate of 50% is acceptable to analyze and publish; 60% is good, and 70% is very good. But one should not discard data yielded from lower response rates if other measures show reliability [22, pp. 260–261]. As Babbie points out, there is usually something more going on than unwillingness to take the survey when a low response rate is obtained. Bolstein, for example, found that those who did not respond to a preelection political poll were less likely to vote than those who did participate [43]. Babbie does concede that the literature varies widely on what is an acceptable response rate. The issue of why individuals do not participate cannot be addressed by rhetoric alone. However, focusing on rhetoric as mediation can facilitate success for the researcher.

RIFE: ETHOS, PATHOS, LOGOS, KAIROS 271

While administering the survey, as the researcher, I fortunately began thinking of rhetoric as mediational in the recruitment process and subsequently saved the data-collection process from even more dismal results. After sending the original recruitment email directly to program directors, almost no survey responses were received in return. That initial email was sent to 50 program directors. Recall that I positioned myself as a researcher with the WIDE Research Center at Michigan State University, de-emphasizing my student status. After that email was sent to the 50 program directors, 6 survey responses were received. Trying to address this problem, the recruitment email was shortened and a link to the digital survey was included. This shortened version containing the survey link was sent to a second set of 50 program directors. Again, the response rate was very low—another 6 responded to the survey, giving a total of 12 survey respondents from a pool of 100. Simply shortening the language and including the link to the digital survey in the recruitment email did not increase the overall response rate.

Faced with the possibility of failing to obtain responses to the online survey, I formulated a different approach to the research. I decided that drawing upon both Rhetorical Theory and Activity Theory in administering the online survey might prove helpful. Administration of the online survey could then be conceptually mediated by my use of the rhetorical strategies of ethos, pathos, logos, and kairos. With respect to kairos, contacting the right person and at the right time, I thought program directors might feel like they were endorsing the research or the survey by passing it along. They might feel this was an undue exercise of their authority. Two program directors returned the initial recruitment email noting this kind of reaction. But this small indication is certainly not enough to generalize the entire population of program directors without further research. Yet, in the success phase of the survey, instead of contacting program directors, every program webpage was visited in an attempt to locate faculty and student emails in order to contact them directly. Also, at this point, writing strategies were used to change the persona of the implied author constructed in the recruitment email. Thinking of ethos, pathos, logos, and kairos as mediating the rhetorical process informed these revision decisions. I decided to change my ethos from that of an authoritative researcher to that of a Ph.D. student trying to finish her dissertation. While true, it was also a conscious use of pathos—appeal

to emotion, even pity. This use of pathos worked together with the ethos I established in the revised email that clearly positioned me as a student rather than an established researcher. Regarding kairos and considerations of an audience, the student-to-student email request was likely more “seasonal” or fitting as a persuasive strategy. Since I too was a student (lending me ethos via a connection to the audience), perhaps the student recipients would be more likely to take the survey.

For faculty, I used more casual-sounding language that made me sound more like a student needing help with research rather than a professional researcher: “If you would share this survey with students in your program that would be great as well.” I tried to connect to faculty using logos, or reasoning, that would relate to faculty goals, such as student learning: “The idea is that this survey is supposed to be a learning experience.” I decided that the original logos I provided, the reasoning that I wanted to check individuals’ knowledge of fair use, might make some individuals (perhaps especially program directors) defensive about their knowledge levels. So I thought focusing my justification on completing my dissertation might be a better use of logos. Therefore, I thought that just “being myself,” that is, a Ph.D. student trying desperately to finish her dissertation research, in the recruitment email might help improve the response rate. The revised recruitment email is provided in Fig. 4.

Once I consciously employed the rhetorical strategies of ethos, pathos, logos, and kairos in my recruitment emails, I started experiencing success in response rates. Some of this success was, of course, attributable to the fact that more people were contacted. But some of the increased response rate was due to employing rhetorical strategies as mediational means in the research process. Ultimately, an overall response rate of 446 was achieved. After I stopped treating best practices as isolated variables and started using rhetoric as mediational means, I was able to collect an adequate amount of data to complete my study. Since this phenomenon arose indirectly from a problem encountered in administering a survey in a larger study, I cannot know for certain the extent to which rhetoric-as-mediation saved the study. At the very least, this issue is worthy of further research.

And while the results are inconclusive, the issues that arose with this informal experiment may present hypotheses for further study. The following are ideas that arose during the administration of the online survey that could be explored.

272 IEEE TRANSACTIONS ON PROFESSIONAL COMMUNICATION, VOL. 53, NO. 3, SEPTEMBER 2010

Fig. 4. Revised recruitment email.

USE OF ETHOS IN RECRUITMENT After clearly positioning myself, the researcher, as a student rather than a teacher, I received more responses to the survey. I also achieved increased response rates by contacting students and faculty directly rather than through program directors. Thinking of ethos, pathos, and logos as mediational means, it is possible that the connection students felt with another student researcher may have compelled some to participate in the survey. This finding highlights the importance of understanding one’s audience and the connection one might establish through a rhetorically informed mediational heuristic with that audience. Such issues will have to be determined on a case-by-case basis, in context.

USE OF PATHOS IN RECRUITMENT When I edited the recruitment email so that it informed potential participants that I, the researcher, needed the data to complete my dissertation, the response rates increased. A prior study showed a 19% increase in participation when the line “it would really help us out” was added to the request [44, p. 254]. Some of the students who responded to my survey were

likely conducting research and may have had an emotionally supportive reaction to receiving a request from another student (me) who needed help. Faculty may have had an affective response to their internalized wish to help students. After this study was completed, I presented my research at the 2008 ATTW conference in New Orleans, Louisiana, explaining my problems with recruitment. A member of the audience identified herself as a professor and said that while she did not normally respond to digital-survey requests, she always did so if it was a student who was trying to complete a dissertation. This is an issue worth further exploration. It points to the problem within the social science literature that found individuals more likely to respond to a survey if it comes from an authority figure. In the context of my survey, this was not verified. Instead, in the TPW community, it appears digital surveys are more likely to be responded to when they meet what might be key values of the community—supporting the formal educational and research goals of newcomers in the discipline.

One problem when enacting pathos is that it should appeal to the community’s values [6]. But

RIFE: ETHOS, PATHOS, LOGOS, KAIROS 273

what if the community lacks adequate cohesion? This question is important because it may be the case that recipients of an email solicitation have little in common; then, it might be difficult, if not impossible, to appeal to their community values through pathos, logos, ethos, or even kairos. The community must have some similarity in relation to what appeals will be effective with them. Without this similarity, rhetoric as mediational means will not be an effective tool.

My experience as a researcher administering this survey supports the idea that TPW lacks cohesiveness or community in three ways. One, as I constructed the study population in order to obtain a master list of programs, a bricolage among three membership lists had to be created and, of course, the lists used were not all-inclusive. The lack of a definitive list of TPW programs is significant. Two, the problem of simply trying to locate contact information for faculty, and especially students, also points to a potentially serious problem with the “community” and who is allowed representational space in the field. (See [36] for a detailed discussion of this.) And three, as noted before, there is some evidence that response rates are low when a sense of cohesion is low among the community studied [31]. The survey had a low response rate, so this third point is supported.

Grabill defines community as a group that has shared values, networks, relationships, shared meanings, and socially constructed boundaries created by shared norms [45]. But in the field of TPW, there are debates on whether these shared norms and values exist. For example, Connors, Dobrin, Durack, Lay, and Rutter (among others), outline debates within the field of TPW regarding how we define ourselves as a profession, what matters, and what we study, while writers such as Charney, Haswell, Johanek, North, and Phelps document debates on appropriate methodologies and resistance to the use of typical “social science” and/or “empirical” research design in TPW or composition studies [12], [18], [46]–[54]. The low response rate at least tentatively supports the notion that we are not cohesive as a field, that there is little sense of community, and that, perhaps, social science methods are not supported. Since pathos is supposed to be developed to appeal to the community’s values, it will be very difficult to employ pathos where there is a weak sense of community.

USE OF LOGOS IN RECRUITMENT Again, when the survey was contextualized as one wherein the researcher was a student conducting

dissertation research rather than an authoritative researcher associated with a high-level research institution, I received more responses. This illustrates the potential of using logos as a means of mediation in the research process.

USE OF KAIROS IN RECRUITMENT

Contacting the right people at the right time in the research process is a matter of kairos intervening as mediational means. While this issue, in particular, needs further exploration, I achieved a quicker response rate when I sent out the survey on a Thursday afternoon rather than on a Monday morning. This observation makes sense because on Mondays, faculty and students are sometimes overwhelmed by the oncoming workweek. The timing of an email solicitation should be tailored to the particular audience. I also gained some responses when I began including a link to the survey in the email solicitation. The link may have been perceived as more time efficient since it did not require the participant to first respond to the email and then wait to receive a second email with the link. Plus, including the link in the initial recruitment email also protected the participant’s privacy. I did not need to know whether someone was interested in participating or not.

IMPLICATIONS AND CONCLUSION

The implications here can be categorized in two ways: (1) issues with rhetoric as mediational means in the recruitment process of an online survey and (2) issues with digital-survey design and its relationship to survey attrition.

As depicted in Fig. 1, thinking of a rhetorically informed heuristic as mediational provides a paradigmatic way of thinking about research outside the typical social science stimulus-response model. Activity Theory is especially appropriate to use in this case because Vygotsky developed it particularly to refute the stimulus-response model. Of course, visualizing a rhetorically informed heuristic as influential during the recruitment process of administering an online survey complicates the research process as well. While further research is needed, the inventional strategies of ethos, pathos, logos, and kairos work together in a mutually informing or ecological manner. Thus, it might be difficult to operationalize a single rhetorical appeal, like ethos, to see how it alone influences the research process, but it might be interesting to attempt this. Further

274 IEEE TRANSACTIONS ON PROFESSIONAL COMMUNICATION, VOL. 53, NO. 3, SEPTEMBER 2010

research into how rhetoric mediates the research process, using an Activity Theory paradigm, could generate valuable results for the professional writing community.

Due to the relative value and ease of assembling an online survey, like individual citizens arguing their cases before the jury in fifth-century Athens, there may be fewer barriers between the researcher and the participants than what existed prior to internet research methods. Instead of hiring individuals to go door-to-door, make thousands of telephone calls, or stuff envelopes, I had a direct line to each potential participant via email. As in ancient Greece, where individual citizens were trained in the skills of rhetoric to make them more effective orators, it may be that the ability to administer an online survey will become available to more people than when we could only complete either mail, telephone, or in-person surveys. These more traditional survey methods are more costly and more time consuming. The art of rhetoric was, in part, developed for use by the “common person,” so it might be worth revisiting for the “common researcher,” not just those doing formal dissertations like I was. Rhetoric could be used by individuals completing short surveys in the workplace, for marketing reasons, or to conduct smaller, individual research projects related to technical writing and documentation.

With respect to the second implication, that survey-design issues are related to attrition, the phenomena that emerged in this study are inconclusive but do raise questions for further research. The six chilled-speech questions experienced no attrition. They were short questions that did not require any outside knowledge but instead tested behaviors or whether something had happened to the participant. Survey participants, for some reason, were engaged with these particular questions. It would be useful to develop further research that explores what makes survey questions interesting and effective for participants so that they are willing to stay in the survey.

Regarding the more serious issue of community in TPW, this too is something worth pondering, researching, and perhaps addressing in our scholarship. Grabill implies that a researcher may facilitate the construction of community [45]. Perhaps during the 11 weeks when data were collected, the act of researching the community created ad hoc networks, connecting individuals—responses to my survey did reach a critical mass and then tapered off near the final weeks (i.e., perhaps the population became saturated). Future research might situate the survey itself as mediational in a larger Activity Theory-based framework, rather than situating the rhetorically informed heuristic as mediational in the research process, as depicted in Fig. 1. This research might examine the networks in which the researcher, the potential respondents, and the various institutions and professional roles (i.e., students, teachers, alumni) are embedded and the ways those networks influence recipients’ responses to the researcher’s request for participation.

This analysis might further explore the issue of community in TPW as well as the ways in which institutions value research. For example, as stated earlier, the population that responded to the survey included more Ph.D. programs and fewer two-year programs than are represented in the actual population; without further research, we cannot know how and whether institutional culture and its valuing of research play a role in a more global view of the TPW community. Learning what factors influence a system that ideally would support research about itself might prove immensely helpful to the TPW community. Ultimately, learning about our field as a whole will prove difficult if we do not participate in more studies that examine our own collective practices. Online surveys lend themselves to obtaining views of the field’s collective practices. Leveraging rhetoric as mediational means in an Activity Theory paradigm as we move forward in this effort may prove useful.

APPENDIX A

PROGRAMS PARTICIPATING IN THE SURVEY

Arizona State University Arizona State University East Austin Peay State University Ball State University

Baylor University Boise State University Briar Cliff University Clarkson University Clemson University Coe College

RIFE: ETHOS, PATHOS, LOGOS, KAIROS 275

Eastern Michigan University Edison Community College Georgia Institute of Technology George Mason University Grand Valley State University, Michigan Hilbert College Illinois Institute of Technology Illinois State University Iowa State University Ithaca College Kutztown University of Pennsylvania Lebanon Valley College Louisiana Tech University Massachusetts Institute of Technology Mercer University Miami University Michigan State University Milliken University Minnesota State University, Mankato Missouri State University Missouri Western State University Montana Tech of the University of Montana Murray State University Nazareth College of Rochester New Jersey Institute of Technology New Mexico State University North Carolina State University Oklahoma State University Penn State University Purdue University Rensselaer Polytechnic Institute Rocky Mountain College of Art and Design

San Francisco State University San Jose State University Southern Polytechnic State University Texas State University University of Arkansas at Little Rock University of Baltimore University of Central Florida University of Cincinnati University of Hartford University of Illinois University of Massachusetts University of Massachusetts-Dartmouth University of Minnesota University of New Mexico University of North Carolina Wilmington University of Pittsburgh University of Texas at San Antonio University of Wisconsin-Milwaukee University of Washington University of Wisconsin Utah State University Washington State University

ACKNOWLEDGMENTS

This work was supported in part by a Fellowship Award from the College of Arts and Letters at Michigan State University and in part by the WIDE Research Center. The author would like to thank Professor William F. Hart-Davidson for his advice throughout the study.

REFERENCES [1] J. W. Creswell, Research Design: Qualitative, Quantitative, and Mixed Methods Approaches, 2nd ed. Thousand

Oaks, CA: Sage Publications, 2003. [2] H. Ding, “Confucius’s virtue-centered rhetoric: A case study of mixed research methods in comparative

rhetoric,” Rhetoric Rev., vol. 26, no. 2, pp. 142–159, 2007. [3] M. C. Rife. (2008). Study report: Knowledge and influence of copyright law for U.S. professional writers

working in educational-context digital environments. SSRN Electronic Paper Collection. [Online]. Available: http://papers.ssrn.com/sol3/papers.cfm?abstract_id=1280525

[4] M. C. Rife and W. Hart-Davidson. (2006). Is there a chilling of digital communication? Exploring how knowledge and understanding of the fair use doctrine may influence web composing. Pilot study report. SSRN Working Paper Series. [Online]. Available: http://papers.ssrn.com/sol3/papers.cfm?abstract_id=918822

[5] L. Odell, D. Goswami, and A. Herrington, “The discourse-based interview: A procedure for exploring the tacit knowledge of writers in non-academic settings,” in Research on Writing, P. Mosenthal, L. Tamor, and S. Walmsley, Eds. New York: Longman, 1983, pp. 221–236.

[6] J. Selzer, “Rhetorical analysis: Understanding how texts persuade readers,” in What Writing Does and How It Does It: An Introduction to Analyzing Texts and Textual Practices, C. Bazerman and P. Prior, Eds. Mahwah, NJ: Lawrence Erlbaum, 2004, pp. 297–307.

[7] C. Miller, “Opportunity, opportunism, and progress: Kairos in the rhetoric of technology,” Argumentation, vol. 8, no. 1, pp. 81–96, 1994.

[8] M. C. Rife, “Ideas towards a fair use heuristic: Visual rhetoric and composition,” in Composition & Copyright: Perspectives on Teaching, Text-making, and Fair Use, S. Westbrook, Ed. Albany, New York: SUNY Press, 2009, pp. 133–153.

[9] R. L. Enos and J. M. Lauer, “The meaning of heuristic in Aristotle’s rhetoric and its implications for contemporary rhetorical theory,” in A Rhetoric of Doing: Essays on Written Discourse in Honor of James L. Kinneavy, S. P. Witte, N. Nakadate, and R. D. Cherr, Eds. Carbondale, IL: Southern Illinois Univ. Press, 1992, pp. 79–87.

276 IEEE TRANSACTIONS ON PROFESSIONAL COMMUNICATION, VOL. 53, NO. 3, SEPTEMBER 2010

[10] S. Doheny-Farina, “Research as rhetoric: Confronting the methodological and ethical problems of research on writing in nonacademic settings,” in Writing in the Workplace: New Research Perspectives, R. Spilka, Ed. Carbondale, IL: Southern Illinois Univ. Press, 1993, pp. 253–267.

[11] M. B. Debs, “Corporate authority: Sponsoring rhetorical practice,” in Writing in the Workplace: New Research Perspectives, R. Spilka, Ed. Carbondale, IL: Southern Illinois Univ. Press, 1993, pp. 158–170.

[12] C. Johanek, Composing Research: A Contextualist Paradigm for Rhetoric and Composition. Logan, UT: Utah State Univ. Press, 2000.

[13] R. A. Lanham, A Handbook of Rhetorical Terms, 2nd ed. Berkeley, CA: Univ. of California Press, 1991. [14] F. Sapienza, “Ethos and research positionality in studies of virtual communities,” in Digital Writing Research:

Technologies, Methodologies, and Ethical Issues, H. A. McKee and D. N. DeVoss, Eds. Cresskill, NJ: Hampton Press, 2007, pp. 89–106.

[15] J. Henry, Writing Workplace Cultures: An Archaeology of Professional Writing. Carbondale, IL: Southern Illinois Univ. Press, 2000.

[16] G. Kennedy, The Art of Persuasion in Greece. Princeton, NJ: Princeton Univ. Press, 1963. [17] J. M. Atwill, Rhetoric Reclaimed: Aristotle and the Liberal Arts Tradition. Ithaca, NY: Cornell Univ. Press, 1998. [18] L. W. Phelps, Composition as a Human Science: Contributions to the Self-Understanding of a Discipline. New

York: Oxford Univ. Press, 1988. [19] W. M. Roth and Y. J. Lee, “Vygotsky’s neglected legacy: Cultural-historical activity theory,” Rev. Educ. Res.,

vol. 77, no. 2, pp. 186–232, 2007. [20] C. Geisler and S. Slattery, “Capturing the activity of digital writing: Using, analyzing, and supplementing video

screen capture,” in Digital Writing Research: Technologies, Methodologies, and Ethical Issues, H. A. McKee and D. N. DeVoss, Eds. Cresskill, NJ: Hampton Press, 2007, pp. 185–200.

[21] W. Hart-Davidson, “Studying the mediated action of composing with time-use diaries,” in Digital Writing Research: Technologies, Methodologies, and Ethical Issues, H. A. McKee and D. N. DeVoss, Eds. Cresskill, NJ: Hampton Press, 2007, pp. 153–170.

[22] E. Babbie, The Practice of Social Research, 10th ed. Belmont, CA: Wadsworth, 2004. [23] D. Andrews, B. Nonnecke, and J. Preece. (2003). Conducting research on the internet:

Online survey design, development and implementation guidelines. [Online]. Available: http://auspace.athabascau.ca:8080/dspace/bitstream/2149/1336/1/Online_survey_design_IJHCI04.pdf

[24] R. M. Groves, R. B. Cialdini, and M. P. Couper, “Understanding the decision to participate in a survey,” Public Opin. Quart., vol. 56, no. 4, pp. 475–495, 1992.

[25] M. P. Couper, “Review: Web surveys: A review of issues and approaches,” Public Opin. Quart., vol. 64, no. 4, pp. 464–494, 2000.

[26] M. P. Couper, J. Blair, and T. Triplett, “A comparison of mail and e-mail for a survey of employees in federal statistical agencies,” J. Official Statist., vol. 15, no. 1, pp. 39–56, 1998.

[27] M. P. Couper, M. W. Traugott, and M. J. Lamias, “Web survey design and administration,” Public Opin. Quart., vol. 65, no. 2, pp. 230–253, 2001.

[28] D. R. Schaefer and D. A. Dillman, “Development of standard email methodology: Results of an experiment,” Public Opin. Quart., vol. 62, no. 3, pp. 378–397, 1998.

[29] J. P. Walsh, S. Kiesler, L. S. Sproull, and B. W. Hesse, “Self-selected and randomly selected respondents in a computer network survey,” Public Opin. Quart., vol. 56, no. 2, pp. 241–244, 1992.

[30] D. F. Witmer, R. W. Colman, and S. L. Katzman, “From paper-and-pencil to screen-and-keyboard,” in Doing Internet Research: Critical Issues and Methods for Examining the Net, S. Jones, Ed. Thousand Oaks, CA: Sage, 1999, pp. 145–161.

[31] G. W. Yun and C. W. Trumbo. (2000, Sep.). Comparative response to a survey executed by post, e-mail, & web form. J. Comput. Mediated Commun., vol. 6, no. 2. [Online]. Available: http://jcmc.indiana.edu/vol6/issue1/yun.html

[32] T. Duhamel, F. Langerak, and N. Schillewaert, “Non-probability sampling for WWW surveys: A comparison of methods,” Int. J. Market Res., vol. 40, no. 4, pp. 307–313, 1998.

[33] S. J. Swoboda, N. Muehlberger, R. Weitkunat, and S. Schneeweiss, “Web-based surveys by direct mailing: An innovative way of collecting data,” Social Sci. Comput. Rev., vol. 15, no. 3, pp. 242–255, 1997.

[34] Y. Zhang, “Using the Internet for survey research: A case study,” J. Amer. Soc. Inf. Sci., vol. 51, no. 1, pp. 57–68, 2000.

[35] L. S. Vygotsky, Mind in Society: The Development of Higher Psychological Processes. Cambridge, MA: Harvard Univ. Press, 1978.

[36] A. Knight, M. C. Rife, P. Alexander, L. Loncharich, and D. N. DeVoss, “About face: Mapping our institutional presence,” Computers and Composition, vol. 26, no. 3, pp. 190–202, 2009.

[37] C. Backstrom and G. D. Hursh-Cesar, Survey Research, 2nd ed. Evanston, IL: Northwestern Univ. Press, 1963.

[38] R. D. Waller, “Fair use. Ethics case,” Intercom, vol. 53, no. 4, pp. 40–41, 2006. [39] J. G. Bryan, Ed., “Responses to ‘Fair use’ ethics case by R. D. Waller,” Intercom, vol. 53, no. 8, pp. 39, 41, 2006. [40] T. W. Howard, “Who ‘owns’ electronic texts,” in Central Works in Technical Communication, J. Johnson-Eilola

and S. A. Selber, Eds. New York: Oxford Univ. Press, 2004, pp. 397–406. [41] H. A. McKee and D. N. DeVoss, Eds., Digital Writing Research: Technologies, Methodologies, and Ethical

Issues. Cresskill, NJ: Hampton Press, 2007. [42] J. M. Lauer and J. W. Asher, Composition Research: Empirical Designs. New York: Oxford Univ. Press, 1998.

RIFE: ETHOS, PATHOS, LOGOS, KAIROS 277

[43] R. Bolstein, “Comparison of the likelihood to vote among preelection poll respondents and nonrespondents,” Public Opin. Quart., vol. 55, no. 4, pp. 648–650, 1991.

[44] J. C. Mowen and R. B. Cialdini, “On implementing the door-to-face compliance technique in a business context,” J. Market. Res., vol. 17, no. 2, pp. 253–258, 1980.

[45] J. T. Grabill, Community Literacy Programs and the Politics of Change. Albany, NY: State Univ. of New York Press, 2001.

[46] R. J. Connors, “The rise of technical writing instruction in America,” J. Tech. Writing Commun., vol. 12, no. 4, pp. 329–352, 1982.

[47] D. N. Dobrin, “What’s technical about technical writing?,” in New Essays in Technical and Scientific Communication: Research, Theory, Practice, P. V. Anderson, R. J. Brockmann, and C. R. Miller, Eds. Farmingdale, NY: Baywood, 1993, pp. 227–250.

[48] K. T. Durack, “Gender, technology, and the history of technical communication,” Tech. Commun. Quart., vol. 6, no. 3, pp. 249–260, 1997.

[49] M. M. Lay, “Feminist theory and the redefinition of technical communication,” J. Bus. Tech. Commun., vol. 5, no. 4, pp. 348–370, 1991.

[50] R. Rutter, “History, rhetoric, and humanism: Toward a more comprehensive definition of technical communication,” J. Tech. Writing Commun., vol. 21, no. 2, pp. 133–153, 1991.

[51] D. Charney, “From logocentrism to ethnocentrism: Historicizing critiques of writing research,” Tech. Commun. Quart., vol. 7, no. 1, pp. 9–32, 1998.

[52] D. Charney, “Empiricism is not a four-letter word,” in Central Works in Technical Communication, J. Johnson-Eilola and S. A. Selber, Eds. New York: Oxford Univ. Press, 2004, pp. 281–299.

[53] R. Haswell, “NCTE/CCCC’s recent war on scholarship,” Written Commun., vol. 22, no. 2, pp. 198–223, 2005. [54] S. North, The Making of Knowledge in Composition: Portrait of an Emerging Field. Portsmouth, NH:

Boynton/Cook, 1987.

Martine Courant Rife is a professor at Lansing Community College, Lansing, MI, and also an attorney. She is currently working on two book projects: Invention, Copyright, and Digital Writing: Toward a Rhetoric-Based Theory of Composing Process for the 21st Century, as well as an edited collection with D. N.

DeVoss and S. Slattery, Copy(write): Intellectual Property in the Writing Classroom. Her research examines intersections between law, rhetoric, and professional communication. She is the winner of the Society for Technical Communication 2007 Frank R. Smith Outstanding Journal Article Award.