SOC 220 - November 17
Overview
Jeff Brassard introduces content analysis as a quantitative or qualitative method for examining documents, images, audio, film, mass media, and online material, emphasizing its unobtrusive nature and the importance of evaluating evidence carefully. Using Scott’s 1990 criteria of authenticity, credibility, representativeness, and meaning, he applies these tests to personal archives, government and corporate records, media productions, and changing online sources, including social media and AI-curated information.
Key takeaways
- Scott’s 1990 framework evaluates documents on authenticity, credibility, representativeness, and meaning; researchers can still use imperfect sources if they clearly document their limitations.
- Content analysis is often unobtrusive because it studies existing material rather than interacting with participants, reducing reactivity and, for public documents, often avoiding ethics review.
- Photographs are selected and staged: researchers should distinguish idealized posed images, candid natural portrayals, and demystifying images that reveal how a scene was constructed.
- Corporate and government records can be authentic while still biased; leaks and lawsuit discovery may reveal material that organizations do not publish voluntarily.
- Online research requires attention to anonymous identities, bots, disappearing archives, platform-specific language, and the source-selection biases of search engines and AI systems.
Chapters
- Jeff Brassard says the first quantitative major assignment has been marked and should be uploaded within a few days.
- The fourth mini assignment is being marked, with results expected that day or the next, before the next assignment due Sunday, November 22.
- The course has a few units remaining and may finish its scheduled content about a week early.
- Content analysis examines and codes documents, texts, and media outputs across sociology, history, political science, media studies, and other fields.
- Potential sources include corporate memos, financial statements, letters, diaries, photographs, films, television, and online material.
- Researchers ask who created a text, what it covers, where and when it was produced, why it was created, and what it leaves out.
- Quantitative content analysis often counts words, phrases, or articles, while Brassard focuses mainly on qualitative analysis in this unit.
- Documents can include legal filings, case law, newspaper articles, websites, photographs, autobiographies, and other material not created for social research.
- A meta-analysis or review of research is an exception because it deliberately examines and synthesizes scholarly studies.
- Content analysis is often unobtrusive: researchers usually study existing material without interacting with participants, reducing reactivity.
- Brassard notes that this kind of research often does not require ethics review when it uses public documents and involves no participant interaction.
- Scott’s 1990 document-quality criteria begin with authenticity: researchers should establish a document’s origin and check that it is genuine rather than forged.
- Brassard recounts how Frenchman Pierre Plantard planted forged Priory of Sion documents in archives, later used as material for claims popularized by The Da Vinci Code.
- Credibility asks whether a document is factually accurate or distorted; wartime reports and propaganda may present events to maintain public support.
- Researchers may still use documents that fail these tests, but should identify and explain their authenticity or credibility problems.
- Representativeness asks whether a document is typical of the category it is meant to represent, such as a novel within its genre or a film within the superhero genre.
- Atypical examples can still be useful, but researchers should flag them rather than treating them as representative.
- Meaning can be difficult to establish because interpretations vary; Brassard connects this to Roland Barthes’s idea that the author is “dead” once a text enters public interpretation.
- Researchers should distinguish an author’s intended meaning from how different audiences interpret the text.
- Personal documents include diaries or journals, correspondence and email, autobiographies, family photographs, videos, and public social media posts.
- Autobiographies may be ghostwritten, so researchers should assess how much the named author contributed to the final account.
- Researchers should separate factual accuracy from whether a diary or letter genuinely reflects the author’s feelings.
- Representativeness matters when personal archives or influencer posts are used to compare people across class, race, gender, or age.
- Visual sources such as photographs, paintings, drawings, and video can serve as evidence, data for analysis, or prompts in interviews and focus groups.
- Brassard describes three photographic modes: idealization through posed images, natural portrayal through candid images, and demystification that reveals staging or atypical conditions.
- Photographers choose what to frame, keep, and discard, so a photograph may not represent the typical experience of an event.
- Selection bias is difficult to detect because researchers often cannot see the images that were never taken or were later deleted.
- Government sources include census and voting information, official reports, memos, and internal emails or communications.
- Alberta’s FOIP law and the United States’ FOIA can allow citizens to request government records, although agencies may try to keep communications outside disclosure.
- Official records can have clear origins and established contexts, but governments still shape documents to support policies or present favorable assumptions.
- Brassard uses Alberta education documents and federal budget forecasts to show how researchers can analyze institutional bias and justifications.
- Company and nonprofit records, including annual and quarterly financial statements, can provide valuable evidence about organizational decisions.
- Brassard says Microsoft has shifted unprofitable products such as Surface laptops among divisions, illustrating how corporate reporting can obscure losses.
- Corporate credibility requires scrutiny because reports may emphasize positive results or frame questionable claims as benefits, as tobacco companies once did with smoking.
- Important evidence may be hidden until leaks or lawsuit discovery expose it; Brassard cites Meta’s internal research on Instagram’s effects on teen girls and tobacco-industry records.
- Newspapers, magazines, television, and films are core data sources in media studies, sociology of media, and film studies.
- Authorship can be unclear when newspaper stories lack bylines or films and television shows involve many writers, producers, and directors.
- Brassard uses films such as Frozen 2 and Justice League to illustrate how changing creative teams complicate claims about who authored a production.
- Researchers often use a director or producer as shorthand, but that may not identify who had actual creative control.
- For some media research, identifying omissions and distortions is the main purpose, as in Brassard’s work on Russian television and portrayals of Ukraine.
- Researchers can compare how outlets such as Fox News, CBC, or MSNBC frame political stories and represent particular interests.
- Audience capture occurs when a publication or podcaster avoids alienating its audience to retain viewers, advertising, and funding.
- Genre matters: satire from The Onion, The Babylon Bee, The Daily Show, or Saturday Night Live should not be interpreted like conventional news.
- Accessible websites, social media posts, discussion boards, chat rooms, listservs, emails, and YouTube comments can be analyzed as online content.
- Joining a publicly accessible account or following someone without misrepresenting oneself differs from deceiving people to enter a restricted group, which may require ethics review.
- Online authenticity is difficult to assess because users can be anonymous, impersonate others, or operate accounts with undisclosed sponsorships.
- Brassard discusses Edmonton’s YEGWave and bot activity around the war in Ukraine as examples of uncertain identity, curation, and propaganda.
- Online material changes or disappears: posts are deleted, websites go offline, and researchers may need to preserve PDFs or other archival copies.
- Search engines and AI systems expose only selected sources, so their rankings and training data can introduce representativeness and bias problems.
- Internet language varies across platforms and communities, including coded words and emojis used to evade moderation filters.
- Brassard ends by postponing detailed coding techniques and other content-analysis procedures until the next class.
Transcript excerpt
0:22 Oh, wait. I forgot something. Okay, let's go. So, three and a half weeks left of the semester. Um, back from reading week. Uh, okay. Couple of things,
1:11 um, just to very quickly mention here. Ah, so your major assignment, right? The first one, the uh quantitative one, it's all marked. I have looked through all of the marking from the TAs and it looks correct. So, I'm working on getting them to upload them in the next couple of days here. So, we should have those results back um pretty quickly. I was going through the fourth mini assignment um yesterday and I will continue doing that this afternoon and hopefully get them all back to you today or tomorrow um so that you can then you know use it for your next one which is due like Sunday I think is Sunday the 22nd. Oh my god. Time why has time no meaning? Time go fast but also slow. Um, so those
2:03 two things, um, you'll get them back this week for sure. Um, and there's a couple of late people who did the midterm late because of some special circumstances. I have to go back this week and mark all of that. You think I would have had time to do that during reading week, but you would be wrong because I spent the entire week marking sociology of youth's midterm, 160 students, five short answer questions. Um, and not to complain about life, but my TA went missing in action for the entire week and to nothing. So, I ended up doing the whole thing, which was super fun. Um, and not at all soul crushing. Mostly professor jobs are great. Marketing is not my favorite. So, that is where I think everything stands. Um, we've got a couple of units left for
This is the start of the transcript. Save this video free for the whole transcript, searchable and time-coded.
In this course
Related notes
Summary, takeaways, and chapters were generated by AI from the video's transcript and may contain errors. The video belongs to its creator, Jeff Brassard.