Orientation and The Science of Storytelling

Welcome

MC 501 · Fall 2026

• Dr. Alex Leith, Dunham Hall 1017, aleith@siue.edu
• Office hours TuTh 12:30 to 1:45 and 3:30 to 5:00, or by appointment
• We meet Wednesdays 6:00 to 8:50, Dunham Hall 1015
• The textbook is a free OER. Read it with the Graduate edition toggle on.
• That toggle reveals a required extension in every chapter, plus the assigned reading

What you will produce

• An original two-part content analysis: a qualitative phase that builds coding variables through immersion, then a quantitative phase that tests relationships
• A Research Proposal as a structured sentence outline
• An Extended Codebook and Reliability Protocol with planned sample size, target thresholds, and revision triggers
• A conference-ready White Paper, paired with an academic poster
• Work at a level approaching publishable quality

How the term is built

The Journalist runs the whole way through: a weekly journal across the 14 teaching weeks in Weeks 1 to 15, worth 150 points. It is the habit, not a stage.

The project moves through four stages, each tracking a part of the textbook:

1,025 points total.

What each stage asks of you

• Journalist: a weekly journal, 450 to 500 words, engaging both the chapter and the assigned reading. Entries that address only the chapter cap at 7 of 10.
• Architect: Syllabus Contract, GitHub Profile, Librarian Visit Report
• Builder: Annotated Manuscripts, Research Proposal (75), Topic and RQs, Definitions Practice, Extended Codebook (75), Sampling Plan and Pilot (75), CITI
• Analyst: Wrangling (50), Describing (100), Inferencing (100)
• Publisher: the White Paper (200) and its poster

The 500-level commitment

• At least one third of our meeting time is 500-level activity: the assigned readings and the discussion built on them, and the heavier deliverables
• Your ethics work includes the non-human-subjects IRB determination you would actually file, which is the protocol experience doctoral work assumes
• Statistics go deeper rather than broader: assumptions, diagnostics, interpretation
• Each week has one required reading, listed in the syllabus with access links

The dataset

One week of Twitch, November 2018

• 21,964,296 chat messages across 1,695 channels
• 590,876 stream snapshots: who streamed what, to how many people
• Public, messy, and real, collected by automated process over five and a half days
• Three roles you must not conflate: the population (1,690 channels in both logs), the working corpus (50, stratified by chat volume), the anchor set (8)
• Which role a claim rests on determines what the claim can say

The tools

• R for the analysis, VS Code as the editor, Quarto for the documents, Git and GitHub for version control and publication
• All free. Setup walkthroughs are on the course site.
• Your White Paper and your poster will both be built from Quarto sources
• Version control is not a convenience here. It is your audit trail.

What R looks like

A full analysis, start to finish:

library(v2v)
chat <- twitch_chat()

chat %>%
  group_by(channel) %>%
  summarise(messages = n()) %>%
  arrange(desc(messages))

Take the chat data, group by channel, count messages, sort descending. Code is instructions in order, and you will read far more of it than you write.

Two numbers, one week on Twitch

Chapter 1

• Bob Ross: 3,178 average viewers, streaming under Art, dead since 1995
• xQc: 17,363 average, live variety, the highest chat volume in the collection
• The move a social scientist makes: ask not what the numbers mean, but what they are
• They are data, collected on purpose so a claim can be defended later
• A corpus is a body of evidence that licenses some questions and refuses others

We are hypothesis-testing organisms

• Barrett: the brain is a prediction machine, updating its models on error
• Storr: stories are cognitive tools for cause-and-effect simulation
• Science is the same narrative machinery under rigor and a public record
• Babbie’s everyday ways of knowing, tradition, authority, common sense, and intuition, each serve daily life and each break when others must trust the result

The sacred flaw and the null

• The null hypothesis is research’s sacred flaw: nothing is happening here
• To reject it is to force the data to tell a different story than you assumed
• A p of 0.001 says the old story is untenable, not merely unlikely
• The framing matters because it puts the burden where it belongs: on the evidence, not on the elegance of your argument

The cliff that wasn’t

Reinhart and Rogoff, 2010

• Public debt above 90% of GDP crushes growth, a finding that shaped austerity policy
• In 2013 a doctoral student replicating it for a class could not match the numbers
• The original spreadsheet held an Excel error that excluded several countries
• Corrected, growth above the threshold was +2.2%, not −0.1%
• The error was caught because the work was transparent enough to retrace. That is the mechanism working, not failing.

What makes a scholarly contribution

A contribution in communication research does one of four things:

1. Tests a theory in a new context
2. Replicates a finding with new data
3. Resolves a conflict between findings in the literature
4. Introduces a testable construct

Before you design a study, name which one you are doing. If you cannot say precisely, the research question is not ready yet.

The reproducibility crisis is a design problem

• The garden of forking paths (Gelman and Loken, 2014): the many defensible analytic choices that inflate false positives, with no conscious fishing at all
• Munafò and colleagues locate a related danger in underpowered designs:

“Low statistical power increases the likelihood of obtaining both false-positive and false-negative results, meaning that it offers no advantage if the purpose is to accumulate knowledge.”

Munafò et al. (2017, p. 2)

Four pillars this course builds

• A priori power: committing to a sample size before collection
• Pre-registration: committing to hypotheses, measures, and analysis plan first
• Two-coder reliability: verifying the codebook works regardless of who applies it
• An audit trail: a version-controlled record from raw data to published figure

These are not add-ons. They are entailments of the epistemology you are adopting.

Pre-registration, precisely

“The strongest form of pre-registration involves both registering the study … and closely pre-specifying the study design, primary outcome and analysis plan in advance of conducting the study or knowing the outcomes of the research.”

Munafò et al. (2017, p. 3)

• It is what separates confirmatory from exploratory work
• Exploratory research is legitimate. Presenting it as confirmatory is not.

A study is a story, told with discipline

Anecdote and data

• Journalism makes the abstract concrete, and makes an audience care
• Science establishes generalizability: does the pattern hold across many cases?
• Anecdotes generate hypotheses; data test them. Data find patterns; anecdotes explain why they matter.
• A significant finding without human context is true and unpersuasive
• Your White Paper has to do both, which is harder than it sounds

A note on paradigms

• We work in a social scientific paradigm: hypotheses tested against data
• Interpretive work asks how people make meaning, through interviews and ethnography
• Critical work exposes and challenges structures of power
• All three have produced foundational communication scholarship
• Where would you place your own research question on the ontology to epistemology axis, and what obligations follow from that placement?

Before Week 2

• Sign the Syllabus Contract and set up your GitHub profile
• Read Chapter 2 with the graduate toggle on
• Read the assigned article: Wilson et al. (2017), “Good enough practices in scientific computing” (open access, linked in the syllabus)
• Write your journal entry, 450 to 500 words, engaging both
• Attempt the software setup. We troubleshoot together next week.