Title: 2019 CSCL
Date: 2019-06-21
template: slides

class: center, middle

## Measuring the computational <br>in computational participation
### Debugging interactive stories in middle school computer science
Chris Proctor<br/>
Stanford Graduate School of Education

.refs[
chrisproctor.net/slides/2019-cscl.html
]

???
Hi, I'm Chris Proctor and I will be presenting "Measuring the computational in computational participation: 
Debugging interactive stories in middle school computer science."

---
# Introduction

How should CS fit into K-12?
- Fits well with STEM fields .cite[1]
- What about the humanities?
- (How) are we going to define and assess computational thinking?

.refs[
.anchor[1] Weintrop, et al., 2014
]

???
The theoretical CSCL questions of this paper are grounded in a question of practice: How should CS fit into K-12 education? 
As a high school English teacher, I have experience advocating for teaching reading and writing across the curriculum. 
As a middle-school computer science teacher, I had the same task. 
As I see it, this is the question the debate over computational thinking is trying to answer (following X and X, 2011)

---
# Background: Computational literacies

- Media designed for computational participation.cite[1]
- Infrastructure, out-of-school digital literacies.cite[2]
- Criticality, how code channels power.cite[3]
- Interactive storytelling.cite[4,5,6]

.refs[
.anchor[1] Burke, O'Byrne, & Kafai, 2016
.anchor[2] New London Group, 1996
.anchor[3] Freire & Macedo, 1998
.anchor[4] Ware & Warschauer, 2005
.anchor[5] Kelleher & Pausch, 2007
.anchor[6] Burke & Kafai, 2010
]

???
However, youth today engage in diverse and complex literacy practices with digitalmedia (Ito, et al., 2010), 
relying to various extents on the computational aspects of these media to engage incomputational participation 
(Burke, O'Byrne, & Kafai, 2016). If K-12 CS were contextualized within theseliteracies, its central concepts and 
practices could become an essential part of diverse youth cultures reading,writing, and analyzing digital texts. 
As a former English teacher, I recognized this when I started teaching CS. 
Simon Buckingham Shum addressed this in his keynote at last year's ICLS.

---
class: center

# Two dimensions of computational literacies

---
# Is a literacy-based approach to teaching CS a promising context for learning debugging?

.half-left[.half-frame[
## Yes
- debugging is a distributed social practice.cite[1]
- comprehension of program important for debugging success.cite[2]
- information-foraging occupies much of professional debugging time.cite[3]
- importance of context for teaching grammar.cite[4]
]]
.half-right[.half-frame[
## No
- Potentially high cognitive load.cite[5]
- Novices misunderstand programming as writing.cite[6]
- Novice misconception that computers reason about programs and interpret them intelligently.cite[7]
]]
.refs[
.anchor[1] Flood, et al., 2018
.anchor[2] McCauley, et al., 2008
.anchor[3] Piorkowski, et al., 2013
.anchor[4] Weaver, 1996
.anchor[5] Kirschner, Sweller, & Clark, 2006
.anchor[6] Bonar & Soloway, 1985
.anchor[7] Pea, 1986
]

???
I want to focus on debugging because it's one of several places where writing practices and programming practices potentially overlap. 
It also sharpens a theoretical question. There are reasons to think teaching debugging in a writer's workshop context (where everyone's 
working on and sharing personally-meaningful stories, and where new CS concepts are introduced opportunistically through mini-lessons)
could be great, but there are also reasons to think it might not work that well.

It is plausible that text-based interactive storytelling, in which authors blend prose with code toimplement discourse scenarios, could unproductively mingle the ways humans interpret text and the waysmachines interpret code. However, it is also plausible that these misconceptions could actually be amelioratedby casting programming as reading and writing. When there is real authorial meaning motivating the program,and a real audience to whom it is addressed, it could be the case that students better understand the computer'smediating role, rather than misunderstanding it as co-author or interlocutor.

---
# Research questions

## 1. Does writing interactive stories support development of debugging practices?
## 2. Does reading other stories support development of debugging practices?

---
# Context
 - Private all-girls middle school in the western United States
 - 40 (of a cohort of 67) students consented to participate
 - Study conducted over 4 months; 2 90-minute block periods per week
 - Writer's workshop-based curriculum using interactive storytelling

---
<pre style="font-family: courier; font-size: 18px;">
Welcome to California 
+ I am a 14 year old girl. 
  -> student
+ I am a 50 year old man. 
  -> teacher

=== teacher ===
As an adult you have special priviledges.
+ Fly to San Fransisco 
  -> enter

=== student === 
You can show your parent you can do something on your own!
+ Fly to San Fransisco for a short independace trip. 
  -> enter

=== enter ===
You made it to San Fransisco!!! 
+ Go to the hotel
  -> hotel

=== hotel ===
{ teacher > 0: 
    The Man at the front desk says, "Hello Sir, do you have a 
    reservation for a room?" 
    -> room 
  - else:
    The Man at the front desk says, "Why are you here? 
    Your dont have a reservation! 
    Where are your parents? Go away child!" Oh no! 
    Your stuck to sleep on the steets! 
    -> streets
}
</pre>

???
Point out diverts and choices. Note that previous research has shown students are more able to work with flow than with state here...

---
.full-bleed[
  <img src="/images/unfold_screenshot.png" width="100%">
]
.refs[
    Proctor & Blikstein, 2019
]

???
Describe what it is, how it works, prior DBR....

---
# Data: Stories
327 stories with an average of 92 lines (σ = 84) and 431 tokens (σ =400)

---
# Data: Story edits
<img src="/media/slides/2019cscl/coding_table.png" style="width:100%;">

---
# Story edits
.full-bleed[
<img src="/media/slides/2019cscl/transition_diagram.png" style="width:100%;">
]
---
# Data: Summative assessment
Chris wrote the story below, but it has some problems. This picture shows how he meant for the story to work. Please fork this story and then see if you can fix it..cite[1]

.refs[
.anchor[1] Werner, et al., 2012
]

---
# Research questions

## 1. Does writing interactive stories support development of debugging practices?
## 2. Does reading other stories support development of debugging practices?

???
Let's think through how this data might answer the questions

---
# Data: Missing pieces

- Qualitative fieldnotes
- Limited collection of literacy events 
- Analysis of students' stories

???
These are important, as they help establish how the stories were personally-meaningful, 
and the nature of the literacy practices. Better-grounded analyses have been published in 
Proctor & Blikstein, 2019 and Proctor & Garcia, 2019.

---
# Results: Does writing interactive stories support development of debugging practices?

---
# Results: Does reading other stories support development of debugging practices?
<img src="/media/slides/2019cscl/results2labeled.png" style="width:100%;">

---
# Discussion

- Literacy-based approach to teaching computer science using interactive storytelling can be an effective context for learning debugging. 
- Authors viewed stories less often during debugging
- Design challenge for future development: How might we design the literacy space so that students can learn from computational aspects of other stories as well as their content?

---
# Limitations

- Missing qualitative data
- Population

---
# Conclusion

- CSCL 2017: Should the community focus on basic research and give up trying to make tangible change in the educational system?
- How can schools attend to questions of culture, identity, power, and privilege, as well as developing students' technical skills? 
- Methodological divide 
- This paper contributes to the proposition that we do not have to choose between those goals.

---
class: center, middle

## Measuring the computational <br>in computational participation
### Debugging interactive stories in middle school computer science
Chris Proctor<br/>
Stanford Graduate School of Education

.refs[
chrisproctor.net/slides/2019-cscl.html
]

???
Unfold Studio is free and available now; looking to scale up. (Addressing CSCL '17 question on actually making stuff...)
Also on the job market now!