Crow, the Corpus & Repository of Writing Todays talk 1. - - PowerPoint PPT Presentation

Crow, the Corpus & Repository of Writing

Today’s talk

• 1. Introduce Crow project and research team, explain

rationale for development

• 2. Brief look at PSLW corpus
• 3. Possibilities for research, professional development
• 4. Methods for ensuring sustainability
• 5. Conversation

Crow team

• Tony Bushner, Sherri Craig, Bradley Dilger, Bill Hart-Davidson,

Michelle McMullin, and Lindsey Macdonald (Rhetoric & Composition)

• Hadi Banat, Wendy Jie Gao, Ge Lan, Shelley Staples,

Aleksandra Swatek, Ashley Velázquez, and Terrence Zhaozhe Wang (Second Language Studies)

• Samantha Pate and Louis Wyatt (Professional Writing)
• Purdue, Michigan State, Arizona State

Crow is a web-based archive for research & professional development in writing studies.

Our history

Collaborative Online Instructor’s Network (COIN) Repository Collaborative Network (Corpus + Repository) Purdue Second Language Writing Corpus Technology Initiatives Working Group

Understanding University Writing

Corpus & Repository

• f Writing

(Crow)

• Corpus of student writing from first-year courses
• Repository of first-year writing syllabi, activities, and

assignments from those and other courses

• Now international-student focused; will expand to all

sections in future

• Web-based interface to contribute to and use both

corpus and repository

Corpus and repository

Our goals (1/2)

• 1. Building a corpus and repository of student writing and

pedagogical artifacts such as syllabi, activities, & assignments

• 2. Supporting research and professional development which is

interdisciplinary and collaborative

• 3. Providing multiple interfaces which support multiple

approaches toward research and professional development

Our goals (2/2)

• 4. Developing a database which allows contributors and

researchers to articulate connections between the corpus and repository

• 5. Providing infrastructure which allows instructors and

researchers to share data (artifacts, writing samples) intra- and inter-institutionally

Why both corpus and repository?

Five reasons for making the connection

• 1. Promise of interdisciplinary approaches e.g. RGS and EAP

(Aull, 2015) or, more broadly, applied linguistics & composition studies (Silva & Leki, 2004)

• 2. Explore connections between pedagogical artifacts & related

student work

• 3. Trace development, influence of pedagogical artifacts
• 4. Research, professional development support each other
• 5. Efficient use of IT resources, larger critical mass

Example: citation research

• Initial research question focusing on integral or non-integral citation

○ Integral: Smith (2009) argued that climate change causes…. ○ Non-integral: Climate change causes … (Smith, 2009).

• First look at data: comparing two sets of papers from two different

instructors — assignments 4 and 5 in typical 106I sequence

• Counting citations and references revealed low number of references for

assignment 5 from second instructor (14 of 24 papers had 3 references; none had more than 4; first instructor was more normal distribution) — repository might help explain this variation

• New Purdue TAs are required to complete a two-semester

sequence of mentoring in their first year (“the practicum”)

• TAs use provided syllabi with little customization in first

semester, modest amount in second, then develop their own in subsequent years

• How does mentor influence development of subsequent

syllabus and assignments? Corpus could help identify, analyze change over time

Example: influences of mentoring

Crow’s corpus, PSLW

AntConc concordance tool (detail)

• Student work from 106I (international-focused FYW),

collected using scripts using Blackboard API

• De-identification is partially automated with script, followed

by review and if necessary editing of de-identified files

• Demographic data included in each file in corpus
• Many questions remain for scaling this to 106 (mainstream

FYW), where variation in assignments and use of multimedia are more common

PSLW specifics

Purdue Second Language Writing Corpus (PSLW)

Assignment # of texts # of words Average word count Narrative 888 710,356 800 Proposal 913 736,286 806 Interview Report 456 411,257 902 Synthesis Paper 881 813,123 923 Argumentative Paper 874 801,238 917 Total 4012 3,472,260 866

Purdue Second Language Writing Corpus (PSLW)

Gender % Country % Class % Major % Male 57 China 76 Freshman 86 Other 22 Female 43 Other 9 Sophomore 7 Science 18 Korea 5 Junior 2 Engineering 17 India 6 Senior 5 Liberal Arts 17 Malaysia 4 Explorers 14 Management 12

Building a sustainable, long-term project

MICUSP: Michigan Corpus of Upper-Level Student Papers

LexTutor

OCELOT (Outcome Centered Electronic Library of Teaching)

From the start, we knew Crow had to keep sustainability in mind.

Three key approaches to sustainability

• 1. Approaching rhetoric as facilitating and making

connections between participants

• 2. Using models for user-centered design which

support a high variability of users and tasks

• 3. Documenting work processes, and scaffolding

active work — deliberately

Grabill & Hart-Davidson, 2011

“We understand writing as a collective social activity, and when we treat writing in this way, we understand that writing requires infrastructure, and that the texts and technologies (and other elements of practice and standards) that comprise infrastructure are participants—they are part of the collective.”

“We would go on further to claim that the purpose of rhetoric is to serve as a type of connection between participants, and that we ought to be engaged in making—and facilitating the making—of those connections.”

We embrace a rhetoric of digital work which creates and sustains meaningful connections made between and by collaborators.

Making & facilitating connections

• 1. Team members from two disciplines (applied

linguistics, rhetoric & composition)

• 2. Balance between two purposes (research, professional

development, hybrid of both)

• 3. Diversity in expertise (students & faculty at all stages)
• 4. Reciprocity in partnerships

Quesenbery, 2004: Balancing user needs

Crow development assumes flexible frameworks will be necessary to understand our users & their tasks.

Multiple users & tasks means complex UCD

• 1. Graduate students need to learn content, get training,

help with teaching, conduct and publish research;

• 2. Faculty need tools which help students learn — but also

need to conduct and publish research;

• 3. WPAs need help with mentoring workload — but also

need to maintain shared governance of program;

• 4. Students indirectly engaged need to see benefits.

Mathieu, 2005

“While institutionalization of service learning is not evil on its face, it is risky and not necessarily beneficial, especially when universities institutionalize well-intentioned but top-down relationships. The very advantages of institutional service learning—measureable success, broad institutional presence, and sustainability—create a generic set of needs and priorities that make it difficult to respond to communities’ needs and ideas…What risks do we incur when we seek to create broad, measurable, sustainable programs that claim institutional resources and spaces?”

• 1. Build attention to partnerships and sustainability

into the design lifecycle;

• 2. Create and maintain diverse teams, and encourage

iteration and movement between them;

• 3. Share our work, and show our work — both internally

and externally — not as end goal, but as milestones along project trajectory.

Deliberate documentating & scaffolding

Building a wide, reciprocal network

• 1. Collaborate with vendors and support staff, but don’t

pigeonhole them into a service and support role;

• 2. Discover partners across campus who can cooperate — in a

manner beneficial for both in quantifiable ways;

• 3. Imagine project as inter-institutional from the start, from

research design to funding to IT support — but ensure balance and interchange between institutional partners.

Basecamp

Showing your work helps share your work

• This talk, and others like it, formal and informal
• Conference presentations (AAAL, CCCC, CWPA, C&W, SSLW,

TESOL, and others)

• Our website, writecrow.org

Again, viewing publications of all types as milestones, sharing our ideas, insights, and discoveries — not project deliverables

Writing documentation, building culture, and listening to partners deliberately isn’t busywork. It’s the most important work we do.

And there’s no easy way to do it. And that’s okay.

Bradley Dilger ~ 309-259-0238 ~ dilger@purdue.edu writecrow.org

Thank you!