Sentence Combining and Pattern Practice Revisited - CiteSeerX

40
Articles Teaching the IMRaD Genre: Sentence Combining and Pattern Practice Revisited Joanna Wolfe, 1 Cynthia Britt, 1 and Kara Poe Alexander 2 Abstract The authors describe two pedagogical strategies—rhetorical sentence combining and rhetorical pattern practice—that blend once-popular teach- ing techniques with rhetorical decision making. A literature review identi- fied studies that associated linguistic and rhetorical knowledge with success in engineering writing; this information was used to create exercises teaching technical communication students to write Introduction, Methods, Results, and Discussion (IMRaD) reports. Two pilot studies report promis- ing results: Preliminary findings suggest that students who were taught this method wrote essays that were perceived as significantly higher in quality than those written by students in a control section. At the same time, how- ever, the pilot studies point to some challenges and shortcomings of exercise-oriented pedagogies. 1 University of Louisville, Louisville, KY, USA 2 Baylor University, Waco, TX, USA Corresponding Author: Joanna Wolfe, Bingham Humanities Room 315, Louisville, KY 40292, USA Email: [email protected] Journal of Business and Technical Communication 25(2) 119-158 ª 2011 SAGE Publications Reprints and permission: sagepub.com/journalsPermissions.nav DOI: 10.1177/1050651910385785 http://jbtc.sagepub.com at PENNSYLVANIA STATE UNIV on September 11, 2016 jbt.sagepub.com Downloaded from

Transcript of Sentence Combining and Pattern Practice Revisited - CiteSeerX

Articles

Teaching theIMRaD Genre:Sentence Combiningand Pattern PracticeRevisited

Joanna Wolfe,1 Cynthia Britt,1 andKara Poe Alexander2

AbstractThe authors describe two pedagogical strategies—rhetorical sentencecombining and rhetorical pattern practice—that blend once-popular teach-ing techniques with rhetorical decision making. A literature review identi-fied studies that associated linguistic and rhetorical knowledge withsuccess in engineering writing; this information was used to create exercisesteaching technical communication students to write Introduction, Methods,Results, and Discussion (IMRaD) reports. Two pilot studies report promis-ing results: Preliminary findings suggest that students who were taught thismethod wrote essays that were perceived as significantly higher in qualitythan those written by students in a control section. At the same time, how-ever, the pilot studies point to some challenges and shortcomings ofexercise-oriented pedagogies.

1 University of Louisville, Louisville, KY, USA2 Baylor University, Waco, TX, USA

Corresponding Author:

Joanna Wolfe, Bingham Humanities Room 315, Louisville, KY 40292, USA

Email: [email protected]

Journal of Business and TechnicalCommunication

25(2) 119-158ª 2011 SAGE Publications

Reprints and permission:sagepub.com/journalsPermissions.nav

DOI: 10.1177/1050651910385785http://jbtc.sagepub.com

at PENNSYLVANIA STATE UNIV on September 11, 2016jbt.sagepub.comDownloaded from

KeywordsIMRaD reports, engineering writing, pedagogy, linguistic studies, sentencecombining, rhetorical pattern practice

Recent scholarship has lamented the decline of sentence-focused research

in composition studies. MacDonald (2007) charted a dramatic decline in

sessions of the Conference on College Composition and Communication

devoted to language since the 1970s, stating that ‘‘the more elite parts of the

profession still consider language study, grammar, or work on style to be

remnants of the past rather than vital subjects for current professional

research’’ (p. 612). Connors (2000) and Myers (2003) both lamented what

Connors called the ‘‘erasure of the sentence’’ as an area of critical attention

in composition pedagogy, with Myers bemusedly wondering why

‘‘acknowledging the importance of writing sentences is so often construed

as diminishing the importance of other forms of discourse’’ (p. 610). All

three authors criticized the failure of mainstream composition studies to

adapt into our own research recent theories and insights from linguistics.

In technical communication, sentence-level research and pedagogy have

had a similar, if less dramatic, decline. Few journal articles in the last

decade have examined methods for teaching syntax, sentence structure,

or style. Meanwhile, textbook sections devoted to sentence-level language

instruction have remained stagnant even as other sections of these books

have dramatically expanded. Reporting Technical Information by Houp and

Pearsall (1973, 1984) provides a case in point. In the 1973 edition, this text-

book had a 20-page chapter devoted to style and included a 27-page hand-

book on common errors, usage, and conventions. Thus, discussion of

sentence-level instruction constituted 47 of its 438 pages, or 10.7% of the

book. In the 1984 edition, the handbook had expanded to 35 pages, with the

chapter on style remaining basically identical. In the 2002 edition (Houp,

Pearsall, Tebeaux, & Dragga), the handbook comprised 32 pages with the

chapter on style still more or less identical to the 1973 version. But other

sections of the book had expanded so that sentence-level instruction

accounted for 52 of 689 pages, or just 7.5% of the text.

Similarly, the 1979 edition of Lannon’s Technical Writing had 78

(13.0%) of its 601 pages devoted to style, mechanics, and usage, but the

2003 edition had only 75 (9.9%) of its 760 pages treating this topic. Of the

five textbooks for which we were able to find substantially older editions,

only Pfeiffer’s text increased the proportion of space dedicated to

sentence-level instruction (from 9.2% in the 1991 edition to 9.5% in the

120 Journal of Business and Technical Communication 25(2)

at PENNSYLVANIA STATE UNIV on September 11, 2016jbt.sagepub.comDownloaded from

2006 edition). Meanwhile, Burnett’s 2005 edition completely dropped the

44-page handbook on usage included in earlier editions, even as the text

itself increased to 827 pages (compared to 738 pages in the 1997 edition).

At the same time, Olsen and Huckin’s textbook, which arguably contains

one of the most helpful and comprehensive discussions of language and

style in technical communication textbooks, has been out of print since

1991. In short, although sentence-level instruction has always been a part

of our field, compared to other areas, it seems to have diminished in relative

importance.

But without effective theories and strategies for teaching sentence-level

issues, writing instructors are largely left with reactive methods for teaching

style, syntax, and mechanics. In other words, language instruction occurs

most often when problems in an individual student’s writing compel us to

comment. Although such language-in-context instruction is often cited as

a best practice in writing instruction, it has also been criticized for empha-

sizing prescriptive over inventive qualities of grammar. For instance, Sams

(2003) noted that language-in-context instruction focuses on telling stu-

dents what they did wrong rather than on teaching them about the range

of linguistic options they have at their disposal. Sams advocated comple-

menting language-in-context instruction with language-based lessons that

help students develop a vocabulary for discussing the grammatical options

at their disposal in various rhetorical situations.

Another problem with language-in-context instruction is that, as many

teachers believe, too many markings on a paper may discourage students

and reinforce negative attitudes toward writing (Davies & Cousin, 2002).

To avoid such negative consequences, instructors often selectively com-

ment, which, for many students, means that numerous sentence-level prob-

lems are never addressed. Such selective commenting practices can even be

viewed as politically regressive because students who are not fully accultur-

ated into mainstream academic or professional English practices may not

receive the instruction necessary for acquiring these sentence-level compos-

ing skills. Instructors, then, are faced with a tricky pedagogical balancing

act of trying to offer comments that will help students acquire these neces-

sary skills without overwhelming them or being too prescriptive. Thus,

proactive strategies for sentence-level instruction, such as Sams (2003)

advocated, would seem to be a necessary and valuable part of writing

instruction.

In contrast to reactive, corrective approaches to teaching language, sen-

tence combining and pattern practice (forms of imitation or modeling exer-

cises) are proactive methods for teaching grammar, style, and invention.

Wolfe et al. 121

at PENNSYLVANIA STATE UNIV on September 11, 2016jbt.sagepub.comDownloaded from

Both methods were popular in the 1970s and early 1980s but rapidly fell out

of favor in a decline well documented by Connors (2000). Sentence com-

bining, at its most basic, is the process of combining two or more short,

simple sentences into a longer sentence using embedding, deletion, subor-

dination, or coordination. Pattern practice is a variation on ancient Greek

and Roman imitation exercises in which students choose sentences or

groups of sentences to use as patterns for designing their own sentences.

The goal of pattern practice is not to follow specific patterns for words but

to call attention to the structure of the model and how it functions so that

students can then implement more advanced patterns and structures

into their own writing (Corbett, 1971). Both sentence-combining and

pattern-practice pedagogies have traditionally involved repeated practice

to encourage students to embed the targeted patterns into their linguistic

repertoire.

Although sentence combining and pattern practice have been critiqued

for their behaviorist underpinnings, the theory behind both methods is pro-

gressive in many ways. These techniques were once popular because they

provide a way to teach sentence and paragraph structure but avoid the draw-

backs of traditional grammar instruction. Rather than requiring students to

memorize complex taxonomies of different parts of speech or search for

errors in already completed sentences, both sentence combining and pattern

practice encourage sentence production and experimentation. This focus on

production helps writers to develop language ‘‘chunks,’’ or multiword

phrases, which many linguists believe is an important step in lexis and lan-

guage learning (Myers, 2003, p. 613). Through repeated sentence combin-

ing and pattern practice, writers embed in their memories these chunks of

effective standard phrases and lexical units. Writers can later draw on this

repertoire of language chunks when composing, using them as building

blocks in developing their own texts. Through such methods, grammar, syn-

tax, and contextual vocabulary can be taught without the stigma of error

correction.

Both sentence combining and pattern practice, then, focus on the inven-

tion of future discourse rather than on the correction of past discourse.

Although the formulaic nature of these tools may feel restrictive to many,

all communication relies on shared codes and formulas. As Graff (2003/

2008), a proponent of imitation exercises, noted, ‘‘formulas can enable crea-

tivity and complication as often as they can stifle them.’’ He further sug-

gested that when ‘‘we refuse to provide such formulas on the grounds

that they are too prescriptive or that everything has to come from the stu-

dents themselves, we just end up hiding the tools of success’’ (p. 40).

122 Journal of Business and Technical Communication 25(2)

at PENNSYLVANIA STATE UNIV on September 11, 2016jbt.sagepub.comDownloaded from

Likewise, Macken-Horarik (2002) reminded us that ‘‘making language

explicit does not mean separating it from a more generalized rhetorical

competence’’ (p. 41). In other words, explicit and formulaic approaches

to language instruction need not be inflexible or arhetorical.

Not only do sentence combining and pattern practice make theoretical

sense, but they have also been experimentally validated. At least four

systematic meta-analyses concluded that sentence combining has signifi-

cant positive effects on student writing. Unfortunately, as Connors (2000)

noted, such research is often misrepresented—sometimes to such a degree

that scholars mistakenly believe that these methods have been disproved or

that all sentence-combining research is inherently flawed. Although there is

much to criticize in sentence-combining research, especially in studies

looking only at changes in syntactic complexity as an indicator of writing

improvement, several reliable and persuasive studies warrant continued

exploration of sentence-combining pedagogy. To cull the good from the bad

research, Hillocks (1984) applied rigorous selection criteria, such as includ-

ing holistic and reliable measures of writing quality, examining multiple

writing samples, and controlling for teacher effects, to identify five studies

that together suggest that sentence combining yields significant measurable

gains in the holistic assessment of student writing. Similarly, Graham and

Perin (2007), using different but no less rigorous selection criteria, analyzed

five studies (two of which overlap with Hillocks’s sample) to conclude that

sentence-combining instruction positively influences overall writing qual-

ity. Two other meta-analyses with less rigorous criteria likewise concluded

that sentence combining is an effective instructional technique (Andrews

et al., 2004; Asher, 1990). These meta-analyses taken together provide com-

pelling grounds for continued research in sentence combining.

Pattern practice has been studied far less than sentence combining but is

potentially even more promising. Hake and Williams (1979) found that stu-

dents given imitation-based exercises experienced even greater gains in

organization, meaning and logic, and style than did students who completed

sentence-combining exercises. Similarly Vande Kopple (1997) reported

success using pattern-practice exercises to teach key writing concepts, such

as the given–new contract, and both Myers (2003) and Connors (2000)

explicitly advocated the practice. We also see pattern practice at work in the

‘‘writing templates’’ provided by Graff and Birkenstein (2007) in their pop-

ular textbook They Say/I Say, and the imitation of models is central to the

genre-based approaches to language instruction espoused by the Sydney

school of genre (cf. Feez, 2002; Knapp & Watkins, 2005; Macken-

Horarik, 2002; Paltridge, 2001). As Myers (2003) indicated, pattern practice

Wolfe et al. 123

at PENNSYLVANIA STATE UNIV on September 11, 2016jbt.sagepub.comDownloaded from

is effective not only because it teaches specific sentence structures but also

because it provides contextual vocabulary instruction.

Yet, despite the evidence in favor of sentence combining and related

pedagogies, it is difficult to look at previously popular sentence-

combining workbooks, such as Strong’s (1973) or DeGeorge, Olson, and

Ray’s (1984), and not feel overwhelmed by the sheer quantity of exercises.

Is this much practice really necessary? More to the point, echoing Elbow’s

(1985) criticism, we find these exercises to be disturbingly arhetorical.

These workbooks never discuss why students might want to use these par-

ticular sentence patterns or provide examples of situations in which a given

sentence pattern might be more or less effective to achieve a particular

rhetorical goal. To all appearances, form is presented and taught for form’s

own sake.

But even though sentence-combining exercises have traditionally been

presented arhetorically, they are not inherently arhetorical. Form and struc-

ture can be taught in tandem with rhetorical instruction. Altman, Caro,

Metge-Egan, and Roberts (2001) recently developed a workbook that

begins to combine rhetorical instruction with sentence-combining exer-

cises. Rather than encouraging lengthy clauses or complex sentences for

their own sake, this workbook focuses on strategies for building coherence.

To this end, major sections give students practice in subordination, transi-

tions, and parallelism. Moreover, Altman et al. instructed students on the

rhetorical consequences of different sentence combinations. For instance,

the unit ‘‘Joining Sentences to Show Concession’’ requires students to con-

sider the different rhetorical impacts of statements such as ‘‘Although com-

pressed natural gas is clean and cheap, it has a low cruising range’’ versus

‘‘Although it has a low cruising range, compressed natural gas is clean and

cheap’’ (p.75). Such instruction can sensitize students to the rhetorical

effects of different sentence combinations and emphasize the tight relation-

ship between rhetorical goals and linguistic structures.

To distinguish exercises that teach sentence combining or pattern practice

in tandem with rhetorical decision making from the less contextualized and

less rhetorical practices of traditional sentence combining or imitation, we

call these exercises rhetorical sentence combining or rhetorical pattern prac-

tice. We first became convinced of the potential of rhetorical sentence com-

bining when one of us (Wolfe) used some of Altman et al.’s (2001) exercises

in a first-year composition (FYC) class to help students manage the difficult

rhetorical task of responding respectfully to opposing viewpoints. Students in

this class not only made effective use of subordination to refute and concede

points, but also made far-fewer errors in sentence structure than did students

124 Journal of Business and Technical Communication 25(2)

at PENNSYLVANIA STATE UNIV on September 11, 2016jbt.sagepub.comDownloaded from

in the instructor’s previous FYC classes. Moreover, students spoke highly of

the exercises, with many citing them as the most influential activity of the

semester in improving their writing.

Buoyed by this success, we sought to integrate similar exercises into

introductory technical communication classes. Previous research has

reported success using modeling and explicit instruction on syntactic struc-

tures to teach scientific genres (Macken-Horarik, 2002). In this study, we

were particularly interested in developing a pedagogy that would benefit

engineering students because employers often report that engineering grad-

uates lack sufficient workplace communication skills (see Reave, 2004, for

an overview of this research), and our own research has suggested that tra-

ditional technical communication textbooks insufficiently prepare engi-

neering students for the writing situations they will face (Wolfe, 2009).

Moreover, we have found engineering students to be particularly resistant

to traditional technical communication pedagogy. Other research confirms

our impressions: Compared to students in other majors, engineering stu-

dents are more likely to report that they dislike writing (Charney, Newman,

& Palmquist, 1995) and be highly critical of their technical writing instruc-

tion (Ford, 2004). They may even see fluency in writing as being at odds with

an engineering identity. For instance, recent research has found a negative

correlation between students’ verbal skills and their success in engineering

course work (Zhang, Anderson, Ohland, & Thorndyke, 2004). We therefore

wondered if a pedagogy that includes relatively short exercises focused on

discipline-specific syntax might succeed in reaching these students in ways

in which other pedagogies have failed. To this end, we received a grant from

the Engineering Information Foundation to test our theories.

In this article, we describe our rationale and process for developing the

rhetorical sentence-combining and pattern-practice exercises. Then we

present two pilot studies, conducted at separate universities, that assess the

impact of these exercises on student writing and rhetorical knowledge. Our

results are preliminary: Our pedagogy has evolved over the course of this

project, and we have consistently chosen to do what we believe would best

serve students rather than make the strongest methodological design. None-

theless, we believe that our pedagogy is sufficiently innovative and our

research questions are sufficiently provocative for this project to be of inter-

est to the technical communication community. This project addresses three

basic research questions:

1. What might a pedagogy that attempts to relate linguistic form to

rhetorical meaning making in the context of engineering discourse

Wolfe et al. 125

at PENNSYLVANIA STATE UNIV on September 11, 2016jbt.sagepub.comDownloaded from

look like, and what specific forms and meaning-making structures

would such a pedagogy emphasize?

2. Would a pedagogy that uses rhetorical sentence-combining and

pattern-practice exercises to reinforce these relationships between

form and meaning lead to improvements in students’ writing, and

would students be able to transfer the skills that these exercises

emphasize to their own independent written products?

3. Would such a pedagogy have any discernable effect on students’

rhetorical knowledge or reasoning processes?

Developing the Exercises: What Do Engineers Needto Know About Language?

The first step in developing effective exercises was to find out which pat-

terns and grammatical structures we wanted students to acquire. A key

assumption in our curriculum is that rhetorical goals and specific lexical

structures are often closely intertwined. Because linguists suggest that

grammatical use in different contexts and registers can noticeably vary,

we wanted to make sure we were advocating sentence structures that are

actually common to engineering. To this end, we conducted a literature

review of relevant research, looking both at research identifying the rheto-

rical considerations engineering writers must negotiate and at research

defining the linguistic structures most common to various kinds of engi-

neering documents. We supplemented this literature review with our own

analyses of professional, published academic, and student engineering doc-

uments (see Wolfe, 2009, for an additional discussion of this research).

Our research suggested that one of the most important rhetorical tasks

facing engineers is making arguments based on numbers and data. In retro-

spect, such a conclusion may seem obvious, but students rarely receive

direct instruction in what one experienced engineer called ‘‘taking the gob-

ledygook of numbers and putting it with some verbiage to make it clear

what happened’’ (Winsor, 1996, p. 22). According to Barabas (1990), the

ability to effectively interpret data is the primary rhetorical skill that

separates good engineering writers from poor ones. She found that good

engineering writers included more conclusions and speculated on what their

findings meant whereas poor writers included fewer conclusions and over-

loaded their writing with unnecessary details about what they did. Walker

(1999), Herrington (1985), and Kryder (1999) all found that engineering

faculty they worked with considered discussing test results and data to be

a major part of engineering communication. As Winsor (1996) and

126 Journal of Business and Technical Communication 25(2)

at PENNSYLVANIA STATE UNIV on September 11, 2016jbt.sagepub.comDownloaded from

Ornatowski (1998) both showed, translating numbers into persuasive argu-

ments can be a rhetorically complex task, yet such complexities are rarely

addressed in technical communication classes (Wolfe, 2009).

Moreover, our own analyses of engineering writing suggest that as

important as writing about data is to engineering communication, many

engineers, in fact, do it poorly. For instance, we initially could not under-

stand why one lengthy report by the National Highway Transportation

Safety Administration (1995) recommended that auto manufacturers use

particular car window glazings when their executive summary clearly

showed that these glazings increased the risk of head injuries during a crash.

Only after reading the body of the 144-page report did we realize that the

new window glazings increased head injuries because—unlike traditional

windows—they kept passengers safely inside the car instead of from being

ejected into traffic. Our initial misreading (one that other readers concurred

with) was due to the report authors’ failure to prioritize information by sub-

ordinating the data on head injuries to the more important data on ejection

rates, which are not mentioned at all in the executive summary. Other pro-

fessional reports similarly made us question how authors moved from evi-

dence to conclusions until we did the heavy work of interpreting the data

ourselves. Many of the student reports we read were even worse in that the

writers failed to stress, or in some cases even mention, their research con-

clusions. One experienced engineering instructor told us that novice engi-

neers tend to avoid making clear recommendations or generalizations for

fear of being wrong.

The task of moving from data to conclusions is not only rhetorically

important but linguistically complex. For instance, Broadhead (1985) found

that the discussion sections of engineering reports (which emphasize impli-

cations, recommendations, and conclusions) use more complex rhetorical

structures involving greater proportions of cohesive devices than do other

report sections. In particular, cohesive signals showing qualification or con-

cession (e.g., however) and consequence (e.g., therefore) are especially fre-

quent. Broadhead also noted that because discussion sections require

complicated syntax, they tend to contain a disproportionately high percent-

age of usage and punctuation errors. Herrington (1985) further noted that

acknowledging errors is a higher level rhetorical skill expected of

advanced engineering students, and Barabas (1990) suggested that engi-

neers need linguistic and rhetorical structures that will help them couch

negative findings and results as positives. Our literature review, then,

underscores the need to reinforce grammatically correct linguistic patterns

for arguing about data.

Wolfe et al. 127

at PENNSYLVANIA STATE UNIV on September 11, 2016jbt.sagepub.comDownloaded from

Based on this preliminary literature review and research, we identified

four main rhetorical goals for creating arguments from data:

� Prioritize more important conclusions and subordinate less important

ones.

� Explain the reasoning process that leads to conclusions.

� Acknowledge errors, problems, and aberrant findings in a way that

does not detract from the main conclusions readers should draw.

� Adapt the discussion of data to fit readers’ goals and backgrounds. In

particular, writers should be able to meet the needs of both technical

readers, who will scrutinize their results and analysis carefully for

errors, and nontechnical or managerial readers, who will want to

quickly access the bottom-line messages.

Of course, the types of linguistic and textual structures needed to support

these four subgoals overlap. With this caveat in mind, Table 1 draws rela-

tionships between these rhetorical goals and the types of linguistic and tex-

tual structures they require.

To teach the rhetorical goals and linguistic and textual structures specified

in Table 1, we developed a set of instructional materials to teach students how

to write a technical report in the IMRaD (Introduction, Methods, Results, and

Discussion) genre. These instructional materials include seven sets of rheto-

rical sentence-combining exercises, six pattern-practice exercises, additional

material explaining the IMRaD genre, and prompts asking students to analyze

and evaluate sample prose written in this genre.

Sentence-Combining Exercises

The seven sets of rhetorical sentence-combining exercises are designed to

teach students how to write grammatically correct complex sentences and

to guide students through rhetorical decision making so that they can

become strong writers who are able to adapt information for particular

rhetorical ends. Each exercise set contains 5 to 20 sentence-combining

prompts. The topics covered include

� writing grammatically correct subordinate clauses

� using subordination to make concessions

� using subordination, coordination, or conjunctive cohesion (includ-

ing transitional devises) to connect ideas while emphasizing the

appropriate topic

128 Journal of Business and Technical Communication 25(2)

at PENNSYLVANIA STATE UNIV on September 11, 2016jbt.sagepub.comDownloaded from

Table 1. Linguistic and Textual Structures Needed to Create Arguments FromData

Prioritizing conclusions and explaining the reasoning process that leads to them

Deductive paragraphstructure

The typical academic paragraph structure of topicsentence–support–conclusion helps writers conveydeductive reasoning, such as drawing conclusionsfrom evidence. In terms of data analysis, the mostbasic paragraph structure would include a topicsentence summarizing a trend in the data, numbersand data cited as support for this trend, and aconcluding sentence connecting the trend to read-ers’ goals. Writers need to analyze readers’ goals inorder to place the most important conclusionsdrawn from data in the topic position.

Conjunctive cohesion Cohesion is, of course, important to all writing butmay be especially important when linking data toconclusions or otherwise detailing the reasoningprocess. Conjunctive cohesion (Halliday & Hasan,1976) involves developing a range of lexical cues(e.g., for example, in addition, however) that explicitlysignal relationships between ideas. Our analysis ofconjunctive ties in engineering academic articlessuggests that engineers employ similar numbers andtypes of conjunctive ties as do other disciplines.1

Referential and lexicalcohesion

Repeating words or using demonstratives such as this,that, or these helps writers to manage given and newinformation, providing readers with the referentialand lexical cohesion that helps them to follow acomplex topic. The placement of given informationat the beginning of sentences (topic position) andnew information at the end of sentences (theemphasis position) may be particularly crucial forreaders who lack technical expertise and thus theknowledge base to fill in cohesive ties left implicit.McKenna (1997) suggested that engineering writersunderuse demonstratives (e.g., this, that, these) thatenhance textual cohesion by referring to informa-tion in the previous sentences.

(continued)

Wolfe et al. 129

at PENNSYLVANIA STATE UNIV on September 11, 2016jbt.sagepub.comDownloaded from

Table 1 (continued)

Analytic methods forreferring to data

Both Barabas (1990) and McKenna (1997) noted thatpoor engineering writers tend to overemphasizeprocedural information and use narrative ratherthan analytic structures to organize information.This overreliance on narrative organizational pat-terns is manifested in sentence openers, such as Aswe show in Figure 1 . . . or The control dial wasinspected and the following noted . . . .Such sentences draw readers’ attention to the actsof producing or gathering data when their attentionshould be on the conclusions. Openers such as Fig-ure 1 shows . . . or The control dial indicated . . .appropriately focus attention on what the datamean rather than how the data were produced.

Acknowledging procedural errors, problems, and aberrant findings

Subordination Subordinate clauses are crucial to conceding problemsor to subordinating negative messages to positiveones. Incorrect subordination is often a cause ofsentence structure errors, suggesting that studentsmay need practice in using this structure correctly.

Qualifiers Phrases or words such as may, might, could help writersmitigate claims or express less than complete cer-tainty about conclusions.

Adapting discussions to readers’ goals and backgrounds

Modifying phrases Modifiers can help writers move technical or statisticalinformation from the subject position to lessimportant positions in sentences so that readerswho do not care about this information can easilyignore it. For example, the following sentence uses amodifier to de-emphasize numerical details: Thestrength of the concrete cylinder increased 68%, from5931.17 psi on day 7 to 6593.73 psi on day 28.

Voice Moving between active and passive voices can helpwriters adjust to readers’ needs. Passive sentencestructures are appropriate for technical audiencesbecause such sentence structures help writers focuson things and procedures (Ding, 2001; Walker,1999) whereas active sentence structures may bemore appropriate for managerial or nontechnicalaudiences because such structures allow writers toemphasize clients, institutions, and other agents.

130 Journal of Business and Technical Communication 25(2)

at PENNSYLVANIA STATE UNIV on September 11, 2016jbt.sagepub.comDownloaded from

� using conjunctive cohesion and subordination to write paragraphs

discussing data

� using nonrestrictive modifiers (a phrase or clause adding information

that is not essential for understanding the sentence) to subordinate

technical information to bottom-line messages

� moving between active and passive voice for different audiences and

purposes

� using parallelism to write recommendations

Most of our exercises differ from more standard sentence-combining

exercises in that our prompts require students to make rhetorical deci-

sions as they combine sentence kernels. For example, an exercise on

using subordination to make concessions might ask students to com-

bine the following sentence kernels to emphasize the spam filters

drawbacks:

The spam filter eliminated the majority of junk e-mails.

The spam filter had many false positives.

Combined: Although the spam filter eliminated the majority of junk

e-mails, it had many false positives.

Similarly, an exercise on using nonrestrictive modifiers might ask stu-

dents to combine the following sentence kernels to emphasize the most

important, or bottom-line, information without sacrificing technical

precision:

The 7-day strength of the concrete cylinder was 5931.17 psi.

The 28-day strength was 6593.73 psi.

This represents an increase of 68%.

Combined: The strength of the concrete cylinder increased 68% between

day 7 and day 28, from 5931.17 to 6593.73 psi.

Exercises on using passive and active voice ask students to combine sen-

tences either for a managerial audience interested in people and agents or

for a technical audience primarily interested in things and processes.

Although some exercises are less rhetorically motivated than others—for

instance, it is difficult to teach parallel structure in the context of rhetorical

decision making—we at least present sentence structures alongside the con-

text of their actual use. Thus, we emphasize that parallel structure is

Wolfe et al. 131

at PENNSYLVANIA STATE UNIV on September 11, 2016jbt.sagepub.comDownloaded from

particularly important in bulleted lists, which are often used to state recom-

mendations, and in subheadings because hurried readers scan reports

quickly for the information they need.

Pattern-Practice Exercises

The pattern-practice exercises consist of giving students a model paragraph,

or pattern, for discussing certain types of data in a particular rhetorical sit-

uation, providing them with three to six data tables or figures, and asking

them to write a paragraph about these data based on the pattern. Before writ-

ing about the data, students are first prompted to consider why someone

would look at these data and what that person would most want to know.

The basic formula for writing about simple data that shows a single trend

consists of three steps:

1. Include a statement discussing a main finding in the data. This state-

ment should (a) refer concisely to a data source and (b) use language

that avoids making inferences that are not fully supported by the

data. This first statement establishes the writer’s ethos as a reliable

reporter (e.g., Table 1 indicates that the O-ring failed to maintain

contact at low temperatures).

2. Include a statement pointing out critical data points in support of

the first statement. Because we want to reinforce the earlier les-

sons on using subordination, we recommend the basic pattern use

of a subordinator that emphasizes the clause describing the data

points that the writer most wants to point out (e.g., Although the

O-ring maintained contact at 100F, it lost and failed to reestab-

lish contact at 50F).

3. Include a concluding statement that makes either a direct or an

indirect recommendation connecting to the reader’s goals. The

force of this recommendation depends on the reader’s preferences

and the rhetorical situation: If safety is an issue or if the reader

has explicitly asked for a recommendation, then this statement

should clearly state what should follow from the data; otherwise,

writers should default to indirect recommendations. This conclud-

ing statement should also use both lexical and conjunctive cohe-

sion to emphasize its connection to the preceding statements (e.g.,

These findings therefore show that the O-ring seal should not be

used because it is unreliable at low temperatures).

132 Journal of Business and Technical Communication 25(2)

at PENNSYLVANIA STATE UNIV on September 11, 2016jbt.sagepub.comDownloaded from

We added two further steps to this basic formula to help students deal with

data showing secondary trends and data showing contradictory trends, or

aberrant results:

4. If the data contains secondary trends worth pointing out, the writer

should repeat Steps 1 and 2 with the new trend, using a transition

showing accumulation (e.g., In addition, the secondary O-ring may

not seal quickly enough to prevent joint leakage).

5. If the data contained a trend contradicting the main finding that the

writer was discussing, the writer should use appropriate subordina-

tion and cohesion to acknowledge, explain, and subordinate these

aberrant results to the main conclusion (e.g., However, previous

launches have experienced minimal O-ring erosion. This lack of

erosion is likely due to the warm temperatures on the days of these

launches. Overall, these findings suggest that the O-ring seal cannot

be guaranteed, even though O-ring erosion has not been a problem

on past launches conducted during ideal conditions).

Additional exercises teach students how to modify these patterns to include the

results of statistical tests or to transform the data with mathematical operations

in order to emphasize the percentage of increase, decrease, or difference

between groups. We point out, for example, that the same 5-lb weight differ-

ence could be accurately reported as either Y is 25% larger than X or X is 20%smaller than Y if Y weighs 25 lb and X weighs 20 lb. The choice of which

version to use depends on the writer’s rhetorical goals.2

After students receive both written and in-class instructions for analyz-

ing and reproducing the patterns they have been given, they complete exer-

cises in which they analyze a data set and use the pattern to digest the data

for a particular reader. The word digest comes from Barabas’s (1990)

study of managers who used this word to describe effective data reporting.

Figure 1 shows three sample exercises. So that students would not have to

wait multiple days to receive feedback from the instructor, we gave students

an electronic copy of the exercises completed by an instructor and asked

them to complete a self-evaluation by comparing their responses to those

of the instructor.

Many of the prompts for these rhetorical pattern-practice exercises are

specifically designed to emphasize the importance of interpreting, or digest-

ing, data for particular readers. For instance, Example B in Figure 1 shows

truck accident rates on three different highways. Students quickly discover

that how they interpret these data depends on whether their audience is a

Wolfe et al. 133

at PENNSYLVANIA STATE UNIV on September 11, 2016jbt.sagepub.comDownloaded from

Example A

Table 1. Mean target hit percentage and time to first shot by military equip-ment configuration

Equipment configuration Hit percentage Mean time to first shot (in seconds)

Current .77 3.71Proposed .82 3.75

What does the person reading these data most likely want to know?What is the main trend you want to emphasize? Why?Digest these data in a short paragraph:

Example B

Table 2. Truck accidents per vehicle miles for three different interstatehighways

Interstate Accidents per 100,000 vehicle miles

I-53 1.94I-55 1.98I-62 3.19

What does the person reading these data most likely want to know?

Example C

Digest Table 3 in a well-written paragraph that answers the question ‘‘Are fogwarning systems worth implementing?’’

Table 3. Average speed and standard deviation in speed of vehicles underadverse weather conditions before and after a fog warning and education systemwas implemented

ImplementationBefore

implementationAfter

implementation

Average vehicle speed 45.5 mph 48.0 mphStandard deviation in vehicle speed 14.3 mph 7.1 mph

NOTES:High standard deviations in vehicle speeds are a major cause of accidents. Reducing thestandard deviation in vehicle speed will lead to fewer accidents.The speed limit on the road was 50 mph.

Figure 1. Three Rhetorical Pattern-Practice Exercises.Example A prompts students to consider the rhetorical situation before writing aboutthe data. Example B prompts students to consider how the paragraph should changebased on the reader’s needs (e.g., anengineeror politician would require different inter-pretations from a truck driver). Example C shows students how important interpreta-tion is to a correct understanding of the data.

134 Journal of Business and Technical Communication 25(2)

at PENNSYLVANIA STATE UNIV on September 11, 2016jbt.sagepub.comDownloaded from

truck driver trying to choose the safest route or a civil engineer deciding

where to make safety improvements. Similarly, Example C emphasizes the

need for knowledgeable writers (i.e., who understand the information in the

accompanying notes) to interpret and digest data for the less informed.

Many uninformed readers may initially look at the table and erroneously

conclude that the system failed because vehicle speeds increased. Such data

clearly does not speak for itself but needs an informed writer who can instruct

readers to focus their attention on standard deviations as opposed to speed.

We tested the effectiveness of the exercises in pilot studies at two differ-

ent institutions. At the first institution, we gathered the instructors’ and stu-

dents’ impressions of the exercises whereas at the second institution, we

used a quasi-experimental design to provide preliminary assessments of our

pedagogy’s effectiveness. We followed all the regulations for human-

subjects research at both our universities.

Pilot Study 1

Two of us (Wolfe and Britt) piloted these exercises in our Scientific and

Technical Writing classes at a Midwestern research I university. This

class is usually taken by sophomore or junior students, and two seme-

sters of FYC or the equivalent are prerequisites for the class. Approxi-

mately 30% of students in the class come from engineering, 30% from

computer information systems (part of the business school), 30% from

math and natural sciences, and 10% from other majors across the uni-

versity. This class fulfills a required writing component at the university

and is required for industrial engineering and computer information

systems majors.

The previous semester, we had codesigned and cotaught science-based

sections of this course using Writing in the Sciences by Penrose and Katz

(1998)—one of the few textbooks available that treat IMRaD genres with

substantial detail. Although we found much to admire in this textbook,

including its incorporation of linguistic and genre-based research, we ulti-

mately decided it was too advanced for our students. Moreover, although

the text provides substantial material emphasizing academic introduction,

method, and discussion sections, its coverage of how to write results is lim-

ited. Our students had difficulty understanding what to say about the data

they collected, and they often conflated methods with results—a problem

that McKenna (1997) and Barabas (1990) also noted with their students.

The problems we experienced in teaching with this text ultimately moti-

vated the exercises we discuss here.

Wolfe et al. 135

at PENNSYLVANIA STATE UNIV on September 11, 2016jbt.sagepub.comDownloaded from

In the 3rd week of the semester in which we conducted the first pilot

study, we began assigning these exercises after students had completed

an introductory unit on instructions. After focusing on these exercises

exclusively for 2 weeks, we asked students to (a) write a short recommen-

dation memo in IMRaD format based on data we gave them in an Excel

spreadsheet and (b) work collaboratively in groups to design, execute, and

write about an experiment on a usability or communication topic. We pro-

vided students in both classes with instructions on collaborative writing

techniques.

Although we initially worried about student resistance, students seemed

focused and intent as they completed the exercises in class. And when they

collaborated, we often overheard conversations that were specific and lan-

guage focused. For instance, a student trying to figure out what to write

about a set of graphs might say, ‘‘We’ll need to use a however to explain

this,’’ or students might debate which of two trends would be more impor-

tant to their audience. Most important, we noticed that the writing of these

students was markedly better than that of the students in the previous seme-

ster when we had tried to teach the IMRaD report using the Penrose and

Katz’s (1998) text: The initial IMRaD drafts of our weakest students this

semester were on a par with or superior to the initial drafts of our strongest

students the semester before.

At the end of the semester, we surveyed students to obtain their

impressions about these exercises. Overall, the students shared our impres-

sions of the effectiveness of the exercises (see Table 2). They generally

agreed that the exercises improved their writing and helped them see that

data need to be interpreted, averaging agreement scores of 3.9 and 4.0,

respectively, out of a 5-point Likert scale. Students in engineering and in

math and natural sciences were more positive overall than were other

students. This finding is not surprising since IMRaD is central to these

disciplines whereas the genre is much less pervasive in computer informa-

tion systems (a fact that might make us question the common practice of

teaching such diverse majors in a single course). Students also tended to

agree that the exercises were tedious, a finding that did not surprise us.

A great many things are tedious yet quite useful. For instance, many people

consider jogging or lifting weights as tedious, yet they do these activities

religiously. In fact, we used this analogy of going to the gym as a way

to sell students on the exercises: Short-term tedium is often a trade-off

we make for long-term results.

The few students who opted to respond to our open-ended request for

comments about the exercises generally stressed their helpfulness. For

136 Journal of Business and Technical Communication 25(2)

at PENNSYLVANIA STATE UNIV on September 11, 2016jbt.sagepub.comDownloaded from

Tab

le2.

Ave

rage

Agr

eem

ent

(and

Stan

dar

dD

evia

tion)

Score

sof

Studen

tsin

Pilo

tSt

udy

1in

Res

ponse

toSu

rvey

Stat

emen

tsA

bout

the

Effec

tsofth

eExer

cise

son

Thei

rR

het

ori

calT

hin

king

and

Wri

ting

Skill

s

Surv

eySt

atem

ent

Studen

tsin

Com

pute

rIn

form

atio

nSy

stem

s(n¼

11)

Studen

tsin

Engi

nee

ring

(n¼

7)

Studen

tsin

Mat

han

dN

atura

lSc

ience

s(n¼

6)

Studen

tsin

Oth

erM

ajors

(n¼

6)

Tota

l(N¼

30)

Impro

ved

my

wri

ting

3.8

(0.8

)4.1

(0.7

)4.3

(0.5

)3.5

(1.2

)3.9

(0.8

)In

crea

sed

my

confid

ence

inm

yw

riting

abili

ty3.5

(0.8

)3.4

(1.1

)4.3

(0.5

)3.3

(1.2

)3.6

(1.2

)

Hel

ped

me

see

that

dat

anee

dto

be

inte

rpre

ted

3.7

(0.6

)4.1

(0.4

)4.5

(0.8

)3.7

(0.8

)4.0

(0.7

)

Seem

eda

was

teoftim

e2.3

(0.5

)2.6

(0.8

)1.8

(0.8

)2.5

(1.2

)2.3

(0.8

)Se

emed

tedio

us

3.5

(1.0

)3.4

(1.4

)3.2

(1.5

)4.2

(0.4

)3.6

(1.1

)

Not

e:1¼

stro

ngly

disa

gree

;5¼

stro

ngly

agre

e.

137

at PENNSYLVANIA STATE UNIV on September 11, 2016jbt.sagepub.comDownloaded from

example, one student commented, ‘‘These exercises helped to improve my

writing abilities and the repetition helped me remember and learn what I

was doing.’’ But others commented that the exercises, while helpful, were

‘‘too involved’’ or ‘‘overpracticed.’’ Based on this feedback, we tried to

streamline some of the exercises prior to implementing the second pilot

study.

After the semester was over, two students contacted us to provide addi-

tional feedback. One student let us know that an instructor in her business

class had singled out her writing as the best in the class; she attributed her

success on these business assignments to the formulas we had taught her. A

second student, a dyslexic engineering graduate student, told us that his

writing scores in his engineering course work had improved from in the

60s at the beginning of the semester to 97 at the end of the semester. He also

attributed this improvement to what he learned from the exercises:

This increase was mainly due to what I got out of the class. The idea of a

standard format was never really introduced to me. Now that I know of

a breakdown method, I can think of small parts and put it all together, much

more efficiently, for a better paper.

This anecdotal evidence encouraged us to test the exercises using more

formal methods.

Pilot Study 2

The second pilot study used a more formal quasi-experimental design to

provide preliminary assessments of the exercises’ effectiveness. The third

author of this manuscript, Alexander, taught a technical writing class to stu-

dents in the control group before she began implementing the exercises.

Then in her technical writing class the following four semesters, she

assigned the exercises to the students in the experimental group. We col-

lected and assessed in this study surveys and reports from students in both

groups.

Background and Methods

Alexander used these materials for four consecutive semesters in her Tech-

nical and Professional Writing class at a southern, Baptist-affiliated

research university. Before this study began, she had taught technical writ-

ing two semesters and had 6 years of teaching experience overall. She had

also participated as a research assistant in an observational study of

138 Journal of Business and Technical Communication 25(2)

at PENNSYLVANIA STATE UNIV on September 11, 2016jbt.sagepub.comDownloaded from

technical writing classrooms funded by the National Science Foundation

(see Wolfe & Alexander, 2005).

The course. The Technical and Professional Writing course is primarily

taken by students at the junior or senior level; two semesters of FYC are

a prerequisite for the class. Approximately 40% of the students in the class

come from engineering, 30% from science, 10% from business, 10% from

professional writing, and 10% from other majors across the university. This

class fulfills an upper-level writing requirement at the university and is a

required course for students majoring in engineering, computer science,

natural science, and professional writing.

Students. The control group consisted of the 15 students—6 women and

9 men—who took Technical and Professional Writing in the fall 2006

semester. Of these students, 4 were juniors and 11 were seniors. All students

spoke English as their native language.

The experimental group consisted of the 55 students who took Technical

and Professional Writing from this instructor during the four semesters from

spring 2007 through fall 2008. The large difference between the sizes of the

control group and the experimental group is due to our collecting experi-

mental data for four semesters as we continued to refine the exercises

and their implementation. The experimental group contained 19 females

and 35 males and 1 student who failed to complete the demographic section

of the survey. In all, 9 of the experimental students were sophomores,

20 were juniors, and 25 were seniors. All but two of the students spoke

English as their native language.

Pedagogy. To teach students in the control group about technical writing,

Alexander used Lannon’s (2005) Technical Communication. Her class ses-

sions included lectures, discussions on technical writing concepts, and

peer-review workshops. She also discussed theories and techniques of col-

laborative writing. The control classes had five assignments: introductory

memo, technical definition, career profile, instruction manual, and technical

report.

In the first semester of the experimental classes, Alexander used Gurak

and Lannon’s (2006) A Concise Guide to Technical Communication and

implemented the exercises; in subsequent semesters, she did not use a text-

book but instead relied on a course packet of the exercises and additional

readings distributed through the Blackboard course-management system.

In the experimental classes, the exercises were assigned throughout the

Wolfe et al. 139

at PENNSYLVANIA STATE UNIV on September 11, 2016jbt.sagepub.comDownloaded from

semester, and substantial class time was dedicated to completing them. The

proportion of students’ grades specifically dedicated to completing these

exercises grew from 15% in the first two semesters to 25% in the last two

semesters. In most of the semesters in which the exercises were assigned,

students each wrote a short IMRaD report and then collaborated on a major

IMRaD report on a communication or usability topic suitable for publica-

tion in an online academic journal. In addition to the IMRaD assignments,

students wrote an instruction manual and a career profile report. Students

also learned collaborative writing techniques. Because Alexander’s peda-

gogy was evolving during the course of this study, we did not want to hinder

her from implementing what she considered the best pedagogy at the time.

Thus, we could not control some of the differences between the control and

the experimental groups.

Data collection and analysis. To see whether our pedagogical approach had

any influence on students’ rhetorical knowledge, we asked students to com-

plete a questionnaire ranking the relative importance of five types of infor-

mation that Barabas (1990) identified as appearing in IMRaD reports: what

I did, how I did it, what I found, what my findings mean, and what my find-

ings are good for. For the purposes of our analysis, we grouped Barabas’s

five types of information into three groups: details (what I did and how I

did it), findings (what I found), and implications (what my findings mean and

what my findings are good for). Barabas found that managers gave higher rat-

ings to writers whose progress reports included more information about

implications. We hypothesized that students exposed to our exercises would

value implications information as more important and details information as

less important to a good IMRaD report than would a control group.

We also collected IMRaD reports from both the control and the experi-

mental classes. In both types of classes, the reports were collaboratively

composed as part of a group project. In the control course, students were

instructed to design and conduct their own experiment in which they col-

lected survey data on a topic of their choosing. In the first two semesters

of the experimental classes, students designed and implemented a usability

experiment (involving usability tests and surveys), whereas in the last two

semesters, students could choose whether to conduct a usability experiment

or to use a survey for data collection.

All five of the control-group reports and a random sample of 12 of the

experimental-group reports were given to an experienced technical writing

instructor (unaffiliated with this study and unaware of the experimental

pedagogy), who used a 4-point scale to rate them on seven criteria: analysis

140 Journal of Business and Technical Communication 25(2)

at PENNSYLVANIA STATE UNIV on September 11, 2016jbt.sagepub.comDownloaded from

of data, adherence to the IMRaD format, organization and coherence, use

of concessions, errors, style, and overall quality. The Appendix contains a

detailed description of these criteria. To assess the reliability of these

ratings, one of us (Wolfe) also evaluated them (these evaluations were only

used as a double check of reliability and are not reflected in any of the

results reported here). Agreement levels using Pearson’s r were r > .75 for

overall quality and adherence to the IMRaD format, which represents

strong agreement above chance; r > .60 for analysis of data and use of

concessions, which indicates fair agreement above chance; and r > .50 for

all other measures, which indicates fair agreement above chance. To ensure

that the results do not reflect researcher bias, only the scores by the inde-

pendent rater are reported in the following analysis below.

Results

The results of pilot study 2 show that the exercises had effects both on

students’ writing and on their rhetorical knowledge.

Effects of the exercises on students’ writing. The reports written by students

in the experimental group were rated as significantly higher in quality than

were those written by students in the control group (see Table 3). In partic-

ular, the reports from the experimental group were rated significantly higher

than the reports from the control group in adhering to the IMRaD format,

F(1, 16) ¼ 12.85, p < .01; in organization and coherence, F(1, 16) ¼8.19, p < .05; and in overall quality F(1, 16) ¼ 5.51, p < .05. The

experimental-group reports were not only rated higher than the control-

group reports for genre and coherence criteria; these reports were also rated

marginally higher in content criteria, such as the quality of their data anal-

ysis, F(1, 16) ¼ 3.33, p < .10, and their use of concessions acknowledging

problems with their methods or contradictory trends in their results, F(1, 16)

¼ 3.13, p < .10. These findings lend support to the arguments of Witte

and Faigley (1981) and Graff (2003/2008) that instruction in cohesion and

other linguistic structures can lead to improvements in invention and

elaboration.

In general, the reports written by students in the control group were less

successful than were those written by students in the experimental group in

stressing the bottom-line implications of their findings. As a case in point,

the following paragraph from a control-group report comparing conditions

across campus dormitories requires quite a bit of effort to discern its main

point:

Wolfe et al. 141

at PENNSYLVANIA STATE UNIV on September 11, 2016jbt.sagepub.comDownloaded from

When presented the statement ‘‘The dormitory was adequately maintained’’

to evaluate the traditional dormitories, the average of the respondents’

answers yielded a 2.88. This value gives evidence that while the respondents

agreed with the statement, it was to a lesser degree. Furthermore, when pre-

sented the same statement concerning North Village, respondents averaged a

3.68; thus this shows the respondents agreed very strongly with this state-

ment. The great difference between the two averages shows that there was

a significant difference in the level of agreement when it came to traditional

dormitories and North Village.

This paragraph foregrounds numbers at the expense of the bottom-line

story, making it difficult for readers to determine which dormitory the

survey respondents preferred. By contrast, the following excerpt from

an experimental-group report comparing the usability of travel Web sites

states a bottom-line trend in the data and then supports it with numeric

details:

Table 1 shows users preferred Expedia and Travelocity out of all the websites

tested, indicating that users prefer all-purpose travel websites over airline

websites. Websites were ranked by users on a scale of 1-4 with 1 ¼ best and

4 ¼ worst. Expedia and Travelocity had the two highest average rankings

at 2.1 and 2.4 respectively while Delta and Northwest were ranked 2.5 and

3.0 respectively. This data indicates that all-purpose travel websites are gen-

erally preferred to the actual airline websites.

Table 3. Average Ratings (and Standard Deviations) That an IndependentEvaluator Gave to the Control-Group and the Experimental-Group Reports

CriteriaControl Group

(n ¼ 5)Experimental Group

(n ¼ 12)

Analysis of data 2.60 (0.89) 3.42 (0.82)yAdherence to IMRaD

genre2.60 (0.55) 3.75 (0.62)**

Organization andcoherence

2.10 (1.14) 3.46 (0.78)*

Use of concessions 2.40 (0.55) 3.29 (1.05)yStyle 2.40 (1.14) 3.29 (0.75)yErrors 2.50 (0.71) 3.00 (0.74)Overall quality 2.50 (0.71) 3.33 (0.65)*

yp < .10; *p < .05; **p < .01.Note: IMRaD ¼ Introduction, Methods, Results, and Discussion; 1 ¼ weak; 4 ¼ strong.

142 Journal of Business and Technical Communication 25(2)

at PENNSYLVANIA STATE UNIV on September 11, 2016jbt.sagepub.comDownloaded from

In contrast to the first paragraph, readers of this excerpt should be able to

easily grasp which websites survey respondents preferred. We believe that

the experimental-group reports had higher marks in both organization and

analysis, both of which affect readers’ perceptions of overall quality, largely

because they had greater success in emphasizing the bottom-line story of

the data.

The writers of the control-group reports also had trouble determining

whether information should appear in the method, results, or discussion sec-

tion—a distinction that both Barabas (1990) and McKenna (1997) stated is

a problem in engineering writing. The following excerpt from the results

section of a control-group report illustrates the problems these writers had

in understanding where to place various types of information:

Our final question was intended as the main point of emphasis. We wanted to

know whether or not students would want to live at Brooks College after

being informed of all of its proposed traditions. The average on this question

was very low (1.9), with a resounding 16 students answering a 1 (Strongly

Disagree). It is clear that the majority of students surveyed have no desire

to live at Brooks. Table 3 (below) displays all of the results. Graphic results

for all of the survey questions can be found in Appendix B.

The first two sentences of the paragraph discuss the rationale for including

particular questions on a survey—a rationale belonging more appropriately

in the method section. The writers have also placed all of their tables and

graphs in an appendix rather than more appropriately integrating them into

the body of the results section. In addition, many of the control-group

reports also misunderstood the purpose of the discussion section, using it

to reiterate the findings that were in the results section. Such confusion was

largely missing in the experimental group, which had done pattern-practice

exercises modeling the types of linguistic and rhetorical moves common in

each of the main IMRaD sections.

Also, the experimental-group reports show a strong use of the cohesive

strategies that the exercises were designed to instill. Cohesive devices—

including conjunctive, referential, and lexical cohesion—are underlined

in the following excerpts from experimental-group reports:

Figure 1 tabulates these results, and shows that 53% of students strongly

agree that gaming distracted them from studying. Of this number, 42% were

males and only 11% were females. Overall, this study does show that video

games are a daily distraction to students. These results agree with the Pew

Wolfe et al. 143

at PENNSYLVANIA STATE UNIV on September 11, 2016jbt.sagepub.comDownloaded from

Internet Research, which found that 49% of the students polled felt that gam-

ing was a distraction to their study habits.

Our findings show that attendance policies do not greatly influence the

majority of college students’ attendance habits . . . . The attendance policy

was not noted to have any universal benefit (e.g. an increase in all student

GPAs). However, it was noted to have drawbacks (e.g. added stress and risk

of illness). We therefore recommend that removal of the compulsory atten-

dance policy may be best for the university. Nevertheless, more research must

be done on a larger sample of students before such an important decision can

be made. Our data also indicates that personal motivation and drive are most

indicative of high GPA and strong classroom performance, which agrees with

previous research. Based on our findings, we therefore recommend that some

kind of program to teach good study habits and increase personal motivation

would be most beneficial in place of or in addition to a compulsory atten-

dance policy.

The first excerpt, from a results section, shows a deductive paragraph struc-

ture that starts by stating the main claim and then follows by elaborating on

and providing support for that claim. Sentences in this paragraph progress

according to the given–new contract by referring to the data or results (the

given information) in the topic position and introducing new claims in the

emphasis position of the sentences. The second excerpt, from a discussion

section, uses not only referential cohesion but a number of conjunctive

adverbs that help the writers concede points and make recommendations.

Although the style of these excerpts is not perfect, they both use a number

of different cohesive devices to connect data and conclusions. The fre-

quency of such cohesive devices in the experimental-group reports helps

explain why these reports were evaluated as having stronger organization

and coherence than do the control-group reports.

Effects of the exercises on students’ rhetorical knowledge. Consistent with

their writing more rhetorically effective reports than did the students in the

control group, students in the experimental group ranked the relative impor-

tance of different types of information more in line with their readers’ val-

ues than did the students in the control groups (see Figure 2). The

questionnaire results show an interaction between the experimental condi-

tion and the relative rankings of different types of information, F(2, 67) ¼3.19, p < .05, using Wilks’ lamba. In particular, students in the experimental

group perceived discussing the implications of their findings (e.g., what the

findings mean and how they are useful) as more important than reporting

the findings themselves. By contrast, students in the control group

144 Journal of Business and Technical Communication 25(2)

at PENNSYLVANIA STATE UNIV on September 11, 2016jbt.sagepub.comDownloaded from

perceived reporting the findings as more important than discussing their

implications—a belief reflected in many of their reports, in which

bottom-line implications were often buried beneath masses of data. In other

words, the types of information that students in the experimental group

ranked as most important were more in line with the types of information

that the managers in Barabas’s (1990) study valued most. These findings

suggest that students who have been trained by the exercises are more likely

to recognize the importance of emphasizing the bottom-line implications of

their data than are students who have not had such training.

Impressions of the exercises’ effects. At the end of each semester, we sur-

veyed the students in the experimental group to get their impressions about

the exercises. The results of these surveys show that students in general

agreed that the exercises were helpful although their enthusiasm was some-

what lukewarm (see Table 4). They agreed that the exercises helped

improve their writing and helped them see that data need to be interpreted,

averaging agreement scores of 3.6 and 3.8, respectively, on a 5-point scale.

But overall, the students in Pilot Study 2 had slightly less favorable attitudes

than did the students in pilot study 1 (see Table 2). Perhaps the students in

pilot study 1 were more invested in the exercises’ success because they

knew that their instructors had written the exercises.

Figure 2. Control-Group and Experimental-Group Students’ Average Rankings ofthe Relative Importance of Types of Information Found in IMRaD Reports.Note: Rankings ranged from 1 (least important) to 5 (most important).3 IMRaD ¼Introduction, Methods, Results, and Discussion.

Wolfe et al. 145

at PENNSYLVANIA STATE UNIV on September 11, 2016jbt.sagepub.comDownloaded from

Tab

le4.

Ave

rage

Agr

eem

ent

(and

Stan

dar

dD

evia

tion)

Score

sof

Studen

tsin

Pilo

tSt

udy

2in

Res

ponse

toSu

rvey

Stat

emen

tsA

bout

the

Effec

tsofth

eExer

cise

son

Thei

rR

het

ori

calT

hin

king

and

Wri

ting

Skill

s

Surv

eySt

atem

ent

Studen

tsin

Com

pute

rIn

form

atio

nSy

stem

s(n¼

7)

Studen

tsin

Engi

nee

ring

(n¼

15)

Studen

tsin

Mat

han

dN

atura

lSc

ience

s(n¼

8)

Studen

tsin

Oth

erM

ajors

(n¼

25)

Tota

l(N¼

55)

Impro

ved

my

wri

ting

3.6

(0.8

)3.7

(1.0

)4.0

(0.5

)3.4

(1.2

)3.6

(1.0

)In

crea

sed

my

confid

ence

inm

yw

riting

abili

ty3.3

(1.0

)3.2

(0.8

)4.0

(0.5

)3.1

(1.2

)3.3

(1.0

)

Hel

ped

me

see

that

dat

anee

dto

be

inte

rpre

ted

3.7

(1.3

)3.8

(0.9

)4.0

(0.5

)3.8

(1.0

)3.8

(0.9

)

Seem

eda

was

teoftim

e3.1

(1.3

)2.3

(0.8

)2.3

(0.5

)2.6

(1.1

)2.5

(1.0

)Se

emed

tedio

us

4.7

(0.8

)4.2

(0.7

)4.0

(0.5

)4.1

(1.2

)4.2

(0.9

)

Not

e:1¼

stro

ngly

disa

gree

;5¼

stro

ngly

agre

e.

146

at PENNSYLVANIA STATE UNIV on September 11, 2016jbt.sagepub.comDownloaded from

In both studies, were students in engineering and scientific majors were

more likely than students in other majors to perceive the exercises as useful.

This finding is not surprising given that the IMRaD genre is pervasive in

engineering and in the sciences but is less common in majors such as com-

puter information systems or business.

Many of the students who responded to our open-ended request for com-

ments focused on the redundancy of the exercises and suggested reducing

the quantity of them, an evaluation that we came to agree with. Several stu-

dents felt that the exercises were too rudimentary for a class taken primarily

by juniors and seniors. For instance, one student commented, ‘‘I felt these

exercises would have been more helpful had I taken the class before my

upper-level technical science classes. Instead, since I already experienced

writing many tedious lab reports, I did not benefit very much from a ‘new’

way of writing.’’ Although many students made positive comments, these

were usually accompanied by complaints that there were too many exer-

cises and that the exercises were tedious. Overall, the students’ comments

pointed to a need to streamline and condense the material, particularly for

use with upper-level students. Alexander attempted to condense some of the

materials in the last two semesters of the experiment, but students continued

to complain about the exercises.

Discussion

This article contributes to research in language-focused instruction in two

ways. First, it details some of the linguistic patterns and textual structures

common to writing about data. Our finding that reports following these pat-

terns received higher evaluations from an external rater who was unfamiliar

with the study is evidence that the patterns described here do contribute to

successful data reporting. Second, we provide some provisional evidence

that a pedagogy combining sentence-combining and pattern-practice exer-

cises with rhetorical decision making can improve students’ rhetorical

awareness (making them value the same types of information that their

audience values) and help them to write more effective papers that conform

to readers’ genre expectations. This pedagogy led to improvements not only

in the coherence and style of students’ writing but also in the quality of their

analysis. Thus, the patterns taught in the exercises seem to have provided

cues for invention as well as strategies for arranging information, a finding

consistent with Witte and Faigley’s (1981) argument.

At the same time, however, our pilot studies also point to some of the

challenges and drawbacks of an exercise-oriented pedagogy that uses

Wolfe et al. 147

at PENNSYLVANIA STATE UNIV on September 11, 2016jbt.sagepub.comDownloaded from

repetition in order to help students embed particular linguistic patterns

into their memories. In retrospect, the students’ comments have persuaded

us that many of the exercises and materials could be condensed and that

more experimentation is needed to determine the best methods for deliver-

ing and reinforcing the linguistic lessons we are trying to communicate. In

addition to condensing the exercises, we may need to put more effort into

persuading students that the exercises have useful lessons to offer them.

We found, ironically, that some of the students who complained the most

about the remedial nature of the exercises were students who did not

complete them correctly. Perhaps not surprisingly, these were also some

of the students who produced the weakest writing in our classes. Although

some of this complaining may be defensive posturing, perhaps students’

willingness to engage with the exercises might be increased by some pre-

paratory activities that point out gaps in their rhetorical and linguistic

knowledge. For instance, students might be asked to complete a short

quiz that would test their knowledge of punctuation rules for various

subordination and cohesion devices, their common misconceptions (e.g.,

believing that they should never begin a sentence with because), and their

understanding of the rhetorical effects of various sentence structures, or

students might be asked to write sample papers about data. Students who

pass these assessments could be exempted from some of the exercises;

however, we suspect that students’ knowledge is not as comprehensive

as they think it is. We anticipate that once students are made aware of how

unsuccessful their writing strategies and practices are, they will be more

receptive to the exercises.

Although the results of our pilot studies are promising, they are also

highly provisional. Not only did we examine just a small number of reports,

but we also had several confounding variables. Perhaps the most major con-

founding variable was that the students in the experimental sections

received more instruction overall in the IMRaD genre, completing short

individual reports with data the instructor provided before collaborating

on their larger group reports. In addition, the instructor changed textbooks

several times over the course of the study. We did not attempt to control for

these variables because we did not want to handicap the instructor or her

students. But the instructor’s pedagogy was continuously evolving over the

course of the study, so our results are only preliminary.

Another confounding variable was the group nature of the evaluated

reports. Whether the effects we observed would be stronger or weaker if

we had analyzed reports produced by individuals rather than groups is

unclear. Ultimately, these pilot studies suggest that as a whole, the

148 Journal of Business and Technical Communication 25(2)

at PENNSYLVANIA STATE UNIV on September 11, 2016jbt.sagepub.comDownloaded from

pedagogy in the experimental classes helped students write rhetorically

effective reports, but the studies do little to tease out which elements of the

study—the detailed descriptions of the linguistic and rhetorical knowledge

needed to be successful in engineering writing, the repetition of these pat-

terns, or the increased opportunities to work with sample data—contributed

most to students’ success.

In both, students in engineering and science majors seemed to find the

exercises more useful than did students in other majors. This finding is not

surprising, given that we wrote the exercises with engineering students in

mind (though we were careful to use examples that would also appeal to stu-

dents in computing majors) and that the IMRaD genre is more central to

engineering and science than it is to other disciplines. But we wonder

whether this finding may point to a flaw in the common practice of having

one umbrella technical writing class for students from all the technology-

related fields. Students in computer information systems and other nonengi-

neering branches of computing practice have different writing needs and

may benefit from a different range of assignments from that of their more

scientifically oriented peers.

The biggest research question facing those who see the potential of rhet-

orically focused sentence-combining and pattern-practice exercises is to

determine the extent to which repetition is central to the success of these

pedagogies. Although traditional sentence-combining and (to a lesser

extent) pattern-practice pedagogies have operated on the theory that

repeated practice is necessary to embed these patterns in students’ linguistic

repertoire, perhaps simply paying close attention to a model text and how it

works is more important than repetition. Or, because much of the research

on sentence combining was performed with students at the secondary (or

younger) levels, perhaps practiced repetition is more necessary for younger

or basic writers than it is for relatively advanced ones. Thus, one of the pri-

mary questions that we should ask about rhetorical sentence-combining and

pattern-practice exercises is not whether they are effective but how much

practice students (particularly upper-level writing students) need to realize

their effects. More cost�benefits analyses are needed to determine the var-

ious trade-offs involved in assigning such exercises.

More research is also needed to investigate possible unintended conse-

quences of using formulas and guidelines in writing instruction. For

instance, the exercises we developed concentrated exclusively on the type

of writing done in various permutations of the IMRaD genre. Would this

intense instruction in the linguistic structures associated with the IMRaD

genre facilitate or interfere with students’ abilities to adapt to the rhetorical

Wolfe et al. 149

at PENNSYLVANIA STATE UNIV on September 11, 2016jbt.sagepub.comDownloaded from

demands of a different genre? Are there other potential consequences

(positive or negative) of this instructional method that have not been

investigated?

In addition, future research might examine the possibility of combining

exercises like the ones we developed with other pedagogical techniques,

such as observing another student learn how to complete the exercises. In

a recent study, Zimmerman and Kitsantas (2002) found that students who

observed peers gradually improving their technique on a sentence-

combining task outperformed students who simply examined finished

examples of successful sentence-combining activities. Further, Couzijn

(1999) found that students who observed and then evaluated peers as they

thought aloud while completing a series of structured writing exercises

(which included pattern-practice exercises) wrote stronger texts and were

better able to transfer this knowledge to a new setting than were students

who completed the exercises themselves. These studies suggest that stu-

dents may benefit from pedagogy that includes an observational component

with the exercises.

Conclusion

Our first research question asked what specific linguistic forms and rheto-

rical structures a pedagogy designed to help students write successful engi-

neering discourse would emphasize. Our initial literature review concluded

that engineering students need particular help with the rhetorical goals of

prioritizing more and less important conclusions about their data, explain-

ing the reasoning processes leading to these conclusions, acknowledging

errors and aberrant findings, and adapting the discussion of data to meet the

needs of both technical and managerial audiences. Use of various linguistic

and rhetorical structures can help them achieve these goals: That is, they

can strategically use deductive paragraph structure and various forms of

cohesion to connect data to conclusions, subordination and qualifiers to

acknowledge problems, and modifiers and voice to adapt their writing for

different audiences. Copies of the exercises that we used in the studies may

be found at http://louisville.edu/faculty/jlwolf02/writing-about-data. We

are preparing to submit a proposal based on these materials to a textbook

publisher.

Our second and third research questions asked whether rhetorical

sentence-combining and pattern-practice exercises designed to reinforce

these structures lead to meaningful improvements in students writing or

have discernable effects on students’ rhetorical knowledge. Our preliminary

150 Journal of Business and Technical Communication 25(2)

at PENNSYLVANIA STATE UNIV on September 11, 2016jbt.sagepub.comDownloaded from

results seem to suggest that they do: Compared to students in a control class,

students who completed these exercises wrote reports that were judged as

more effective and were more likely to value the same types of information

as their readers. But much more research is needed, including systematic

studies and research that investigates how much practice is needed and what

unintended consequences might result from excessive focus on an exercise-

based pedagogy.

In this article, we join our voices to others in composition studies who

have called for more attention to language and sentence-level issues in writ-

ing instruction. We have made a case that such language-focused instruc-

tion has a place in technical communication instruction and can be

particularly useful in teaching students to make verbal arguments about

numerical information—an area that has received too little attention in tech-

nical communication pedagogy. More broadly, we have begun to answer

Schleppegrell’s (2007) call for researchers and educators to better inform

themselves about the functional value of different language choices in

specific registers, and we have suggested ways that exercise-based methods

for teaching writing can be combined with context-specific rhetorical

instruction.

Notes1. To make sure that our preferences for conjunctive cohesion were not due to our

training in the humanities and social sciences, we used the search tools in Adobe

Acrobat Reader 8.0 to analyze the occurrence of the most common cohesive ties

in nine articles selected from top-tier engineering journals and compared them

with nine articles from rhetoric and composition. Although we found few major

differences between the disciplines, we did find some minor ones. Most notably,

the engineering articles used more adjuncts showing time, location, and condi-

tion (e.g., when, where, if). We also found minor disciplinary differences in

which particular cohesive ties the different disciplines preferred. For instance,

engineering articles were more likely to use therefore or hence to show conse-

quence whereas rhetoric and composition articles preferred thus. And rhetoric

and composition articles were more likely to use subordinators to show contrast

(e.g., although, whereas), while engineering articles were more likely to use the

conjunctive adverb however.

2. Part of this instruction involved giving students simple review lessons in mathe-

matics. Many students (especially those in computer information systems) did

not know how to correctly perform these percentage and ratio operations. Even

those students who did have a full command of basic math were nonetheless sur-

prised to see that the same difference in values could be reported differently

Wolfe et al. 151

at PENNSYLVANIA STATE UNIV on September 11, 2016jbt.sagepub.comDownloaded from

depending on whether the writer focused attention on the amount increased or

decreased.

3. Although Figure 2 may initially seem to suggest that the experimental-group stu-

dents are undervaluing the importance of reporting their findings, students were

asked to rank the types of information from most to least important. Thus, rank-

ing certain types of information as more important necessarily requires ranking

other types of information as less important.

Acknowledgment

We would like to thank the Engineering Information Foundation for their

generous support for this project. The views expressed here do not necessa-

rily reflect the views of the Engineering Information Foundation. We are

also grateful to Steven Smith for his help with this project.

Declaration of Conflicting Interests

The author(s) declared no conflicts of interest with respect to the authorship

and/or publication of this article.

Funding

The authors received funding from the Engineering Information Founda-

tion under the ‘‘Enhancing Communication and Use of Information in

Engineering’’ program area.

Appendix

Criteria for Evaluating the Essays

Rank each paper from 1 (lowest) to 4 (highest) on the seven scales below.

You may use fractions (e.g., 2.5) in your ranking. Unless otherwise indi-

cated, the criteria listed for each scale specify what is needed to achieve

a score of 4.

Analysis of Data

� The report foregrounds an argument about the data rather than presents

a data-dump of numbers.

� The data collected are thoroughly and persuasively analyzed.

� The data are presented in tables and figures that are well designed,

labeled, and easy to understand.

152 Journal of Business and Technical Communication 25(2)

at PENNSYLVANIA STATE UNIV on September 11, 2016jbt.sagepub.comDownloaded from

� The report clearly articulates (usually in the discussion section) why the

main trends in the data are important.

IMRaD Format Followed

� Introduction states the research question.

� Methods and results are not conflated.

� Methods include who participated, what was done, how data were

collected.

� Results include what was found.

� Methods section uses passive voice.

� Discussion section reiterates main findings and discusses their

implications.

� Discussion section mentions limitations of the study and directions for

future research.

Organization and Coherence

� Information is chunked into appropriate sections.

� The relationships between different sections are easy to understand.

� It is easy to see how different sections of the report, or different trends in

the data, support one another.

� The report avoids unnecessary repetition in various sections.

� Transitions and other cohesive strategies are used effectively to connect

ideas.

Concessions

� The report acknowledges problems/limitations with the study and

explains how these problems should influence interpretation of the

results.

� The report acknowledges and explains contradictory data or exceptions

to main trends while emphasizing a coherent and consistent story.

4 ¼ lots of concessions that are both acknowledged and explained and

feel connected to the argument

3 ¼ occasional concessions that do not distract and feel connected to

the argument

2 ¼ concessions that feel disjointed or poorly articulated

1 ¼ no concessions

Wolfe et al. 153

at PENNSYLVANIA STATE UNIV on September 11, 2016jbt.sagepub.comDownloaded from

Errors

4 ¼ report is mostly free of distracting errors and awkward phrasings

3 ¼ report has occasional distracting errors but no major errors such as

sentence fragments

2 ¼ report has one or two major errors or an abundance of awkward

phrasings

1 ¼ report has more than two major errors

Style

� Language is efficient and concise.

� Concepts are stated clearly; the reader has to do minimal work to

understand.

Overall Quality

Use your own criteria to make a holistic evaluation of the report.

References

Altman, P., Caro, M., Metge-Egan, L., & Roberts, L. (2001). Sentence combining

workbook. Boston: Heinle.

Andrews, R., Torgerson, C., Beverton, S., Freeman, A., Locke, T., Low, G., . . .

Zhu, D. (2004). The effect of grammar teaching (sentence combining) in English

on 5 to 16 year olds’ accuracy and quality in written composition. Research evi-

dence in education library. London: EPPI-Centre, Social Science Research Unit,

Institute of Education. Retrieved from http://eppi.ioe.ac.uk/cms//Default.aspx?

tabid¼56

Asher, W. (1990). Educational psychology, research methods and meta-analysis.

Educational Psychologies, 25, 143-158.

Barabas, C. (1990). Technical writing in a corporate culture: A study of the nature

of information. Norwood, NJ: Ablex.

Broadhead, G. J (1985). Rhetoric, style, and cohesion in 12 engineering research

reports. Paper presented at the Conference of the Midwest Modern Language

Association, St. Louis, MO.

Burnett, R. E. (1997). Technical communication (4th ed.). Boston: Wadsworth.

Burnett, R. E. (2005). Technical communication (6th ed.). Boston: Thomson

Wadsworth.

Charney, D., Newman, J. H., & Palmquist, M. (1995). ‘‘I’m just no good at writing’’:

Epistemological style and attitudes toward writing. Written Communication, 12,

298-329.

154 Journal of Business and Technical Communication 25(2)

at PENNSYLVANIA STATE UNIV on September 11, 2016jbt.sagepub.comDownloaded from

Connors, R. (2000). The erasure of the sentence. College Composition and Commu-

nication, 52, 96-128.

Corbett, E. P. J. (1971). The theory and practice of imitation in classical rhetoric.

College Composition and Communication, 22, 243-250.

Couzijn, M. (1999). Learning to write by observation of reading and writing

processes: Effects of learning on transfer. Learning and Instruction, 9,

109-142.

Davies, J. W., & Cousin, G. (2002). Engineering Students’ Writing Skills. Paper pre-

sented at the International Conference on Engineering Education, Manchester, UK.

DeGeorge, J., Olson, G. A., & Ray, R. (1984). Style and readability in technical

writing: A sentence combining approach. New York: Random House.

Ding, D. (2001). How engineering writing embodies objects: A study of four engi-

neering documents. Technical Communication, 48, 297-308.

Elbow, P. (1985). The challenge for sentence-combining. In D. A. Daiker (Ed.),

Sentence combining: A rhetorical perspective (pp. 232-245). Carbondale: South-

ern Illinois University Press.

Feez, S. (2002). Heritage and innovation in second language education. In

A. M. Johns (Ed.), Genre in the classroom: Multiple perspectives (pp. 43-72).

Mahwah, NJ: Erlbaum.

Ford, J. D. (2004). Knowledge transfer across disciplines: Tracking rhetorical strat-

egies from a technical communication classroom to an engineering classroom.

IEEE Transactions on Professional Communication, 47, 301-315.

Graff, G. (2008). From clueless in academe. In T. R. Johnson (Ed.), Teaching com-

position: Background readings (pp. 32-58). Boston: Bedford/St. Martin’s. (Orig-

inal work published 2003)

Graff, G., & Birkenstein, C. (2007). They say/I say: The moves that matter in aca-

demic writing. New York: Norton.

Graham, S., & Perin, D. (2007). A meta-analysis of writing instruction for adoles-

cent students. Journal of Educational Psychology, 99, 445-476.

Gurak, L., & Lannon, J. M. (2006). A concise guide to technical communication

(3 ed.). New York: Longman.

Hake, R., & Williams, J. (1979). Sentence expanding: Not can, or how, but when. In

D. Daiker, A. Morenberg, & M. Kerek (Eds.), Sentence combining and the teach-

ing of writing (pp. 134-146). Akron, OH: University Press.

Halliday, M. A. K., & Hasan, R. (1976). Cohesion in English. London: Longman.

Herrington, A. J. (1985). Writing in academic settings: A study of the contexts for

writing in two college chemical engineering courses. Research in the Teaching of

English, 19, 331-361.

Hillocks, G. (1984). What works in teaching composition: A meta-analysis of

experimental treatment studies. American Journal of Education, 93, 133-170.

Wolfe et al. 155

at PENNSYLVANIA STATE UNIV on September 11, 2016jbt.sagepub.comDownloaded from

Houp, K. W., & Pearsall, T. E. (1973). Reporting technical information (2nd ed.).

New York: Glencoe Press.

Houp, K. W., & Pearsall, T. E. (1984). Reporting technical information (5th ed.).

New York: MacMillan.

Houp, K. W., Pearsall, T. E., Tebeaux, E., & Dragga, S. (2002). Reporting technical

information (10th ed.). New York: Oxford University Press.

Knapp, P., & Watkins, M. (2005). Genre, text, grammar: Technologies for teach-

ing and assessing writing. Sydney, Australia: University of New South Wales

Press.

Kryder, L. G. (1999). Mentors, models and clients: Using the professional engineer-

ing community to identify and teach engineering genres. IEEE Transactions on

Professional Communication, 42, 3-11.

Lannon, J. M. (1979). Technical writing. Boston: Little, Brown and Company.

Lannon, J. M. (2003). Technical communication (9th ed.). New York: Longman.

Lannon, J. M. (2005). Technical communication (10th ed.). New York: Longman.

MacDonald, S. P. (2007). The erasure of language. College Composition and Com-

munication, 58, 585-625.

Macken-Horarik, M. (2002). Something to shoot for: A systemic functional

approach to teaching genre in secondary school science. In A. M. Johns (Ed.),

Genre in the classroom: Multiple perspectives (pp. 17-42). Mahwah, NJ:

Erlbaum.

McKenna, E. (1997). How engineers write: An empirical study of engineering report

writing. Applied Linguistics, 18, 189-211.

Myers, S. A. (2003). ReMembering the sentence. College Composition and Commu-

nication, 54, 610-628.

National Highway Transportation Safety Administration: The Advanced Glazing

Research Team. (1995). Ejection mitigation using advanced glazing: A status

report. Retrieved June 17, 2006, from http://ntl.bts.gov/lib/000/200/253/00253.pdf

Olsen, L. A., & Huckin, T. N. (1991). Technical writing and professional commu-

nication (2nd ed.). New York: McGraw-Hill.

Ornatowski, C. M. (1998). 2 þ 2¼ 5 if 2 is large enough: Rhetorical spaces of tech-

nology development in aerospace engine testing. Journal of Business and Tech-

nical Communication, 12, 316-342.

Paltridge, B. (2001). Genre and the language learning classroom. Ann Arbor: Uni-

versity of Michigan Press.

Penrose, A., & Katz, S. (1998). Writing in the sciences. New York: St. Martin’s.

Pfeiffer, W. S. (1991). Technical writing: A practical approach. New York:

MacMillan.

Pfeiffer, W. S. (2006). Technical communication: A practical approach (6th ed.).

Upper Saddle River, NJ: Pearson.

156 Journal of Business and Technical Communication 25(2)

at PENNSYLVANIA STATE UNIV on September 11, 2016jbt.sagepub.comDownloaded from

Reave, L. (2004). Technical communication instruction in engineering schools: A

survey of top-ranked U.S. and Canadian programs. Journal of Business and

Technical Communication, 18, 452-490.

Sams, L. (2003). How to teach grammar, analytic thinking, and writing: A method

that works. English Journal, 92, 57-65.

Schleppegrell, M. J. (2007). The meaning in grammar. Research in the Teaching of

English, 42, 121-128.

Strong, W. (1973). Sentence combining: A composing book. New York: Random

House.

Vande Kopple, W. J. (1997). Using the concepts of given information and new infor-

mation in classes on the English language. In T. Miller (Ed.), Functional

approaches to written text: Classroom applications (pp. 216-229). Washington,

DC: United States Information Agency.

Walker, K. (1999). Using genre theory to teach students engineering lab report writ-

ing: A collaborative approach. IEEE Transactions on Professional Communica-

tion, 42, 12-19.

Winsor, D. (1996). Writing like an engineer: A rhetorical education. Mahwah, NJ:

Erlbaum.

Witte, S., & Faigley, L. (1981). Coherence, cohesion and writing quality. College

Composition and Communication, 32, 189-204.

Wolfe, J. (2009). How technical communication textbooks fail engineering students.

Technical Communication Quarterly, 18, 351-375.

Wolfe, J., & Alexander, K. P. (2005). The computer expert in mixed-gendered col-

laborative writing groups. Journal of Business and Technical Communication,

19, 135-170.

Zhang, G., Anderson, T. J., Ohland, M. W., & Thorndyke, B. R. (2004). Identifying

factors influencing engineering student graduation: A longitudinal and cross-

institutional study. Journal of Engineering Education, 93, 313-320.

Zimmerman, B. J., & Kitsantas, A. (2002). Acquiring writing revision and self-

regulatory skill through observation and emulation. Journal of Educational Psy-

chology, 94, 660-668.

Bios

Joanna Wolfe is an associate professor and director of composition at the Univer-

sity of Louisville, where she teaches courses in technical communication, composi-

tion, and research methods. She is the author of the textbook Team Writing

(Bedford-St. Martin’s Press, 2010).

Cynthia Britt received her PhD in rhetoric and composition at the University of

Louisville and does research on the rhetoric of science and environmental rhetoric.

Wolfe et al. 157

at PENNSYLVANIA STATE UNIV on September 11, 2016jbt.sagepub.comDownloaded from

With Debra Journet and Beth Boehm, she is coeditor of Narrative Acts: Rhetoric,

Race and Identity, Knowledge (Hampton, IN PRESS). She currently teaches compo-

sition at Western Kentucky University.

Kara Poe Alexander is an assistant professor at Baylor University, where she tea-

ches courses in technical and professional communication. She is currently

researching the intersection of narrative and identity in stories of writing at work.

158 Journal of Business and Technical Communication 25(2)

at PENNSYLVANIA STATE UNIV on September 11, 2016jbt.sagepub.comDownloaded from