ADV 281 Exam 1 Study Guide

Data

Recorded info about the world that can be analyzed to answer questions.

-Numbers

-Written Text

-Videos

-Audio recordings

-Social Media Posts

-Survey responses

-Info has to be recorded so researchers can analyze it.

Ex: Saved Insta comments could be data for an advertising researcher studying reactions to a campaign.

Archives/existing data

Is an an existing collection of information or data that researchers can use to answer a question.

-old social media posts

-website analytics

-sports statistics

-company records

-Previous survey results

=Data that already exists,you don’t have to collect new information from participants-you analyze info that’s already available.

Empirical questions

A question that can be answered using observations/evidence from the real world.

-How many minutes per day do UIUC students use TikTok?

-What % of students recognize an advertisement?

-These can be investigated by gathering data,answerable using evidnece/observation.

Quantitative Data

Numbers,counting,rating,scoring,numerical measurement

-How many minutes did you spend on TikTok yesterday?

Qualitative=meanings/descriptions

Used when numbers alone can’t fully answer the question.

Example: Why do you like using TikTok?

-entertainment, social conenction, trends.

Observation

Gathered data by watching what happens, usually without intervening.

Ex:You want to know whether students buy Coke/Pepsi,you could observe purchases at vending machine.

-Actually watching the behavior.

Non-representative sample

Does not adequately represent the larger population you’re trying to understand.

Ex:

Population:All university students

Sample:Only students studying in the library.

Population

The entire group the researcher wants to understand.

Ex:How much time do UIUC students spend on Instagram?

Population:All UIUC students

-can’t study every person in a large population; sample = smaller group actually studied.

Representative Sample

A smaller group that can reasonably represent the larger population and is not systematically biased.

-The smaller group should reflect the larger group rather than systematically favoring one type of person.

Random Sample

-Process designed to avoid systematic selection bias.

EVERYONE IN THE POPULATION HAS AN EQUAL CHANCE OF SELECTION.

Bias -Systematic

systematic/predictable problem

The sample differs from the population in a consistent direction.

Ex:You want to know how much the average student studies

you survey only people in the library; your results may systematically contain more studious students.=Sampling bias

Error -Chance

Random/chance problem

-Even a good random sample won’t perfectly match the population.

By chance,your sample might have slightly more of one type of person.

=Sampling error.

Self-report measurement

Respondents provide info about themselves.

Ex:How many min did you exercise yesterday?

Which brand do you prefer?

-Self-report is useful bc some info can’t easily be observed.

But people can:

-Forget, lie, guess, misremember, misunderstand questions.

Conceptual Definition

What does the variable MEAN

-clearly defines what you mean by a concept/variable

Breaking down the meaning of smth your trying to measure.

Centurion reading=The number of minutes per day a student spends reading/looking at pictures in the Centurion newspaper.

Operational Definition

How will you measure the variable

-explains the precise procedures used to measure the variable.

-What am I actually gonna do to get a number?

Ex:How many min did you spend reading/looking at the Centurion yesterday?

Content Analysis

Systematically analyzing messages to find patterns in their content

-Instead of directly asking people questions, researchers examine existing messages/content.

Ex:

-Insta/TikTok

-Adv

-Movies

-News Articles

-Websites/magazines, conversations

Ex: You want to know how often college brands use influencers on Instagram. You could collect Instagram posts and systematically code whether each post contains an influencer.

Population of Messages

All the messages/content that you are interested in studying.

Ex:You want to study how women are portrayed in Time magazine.

Population:Every relevant Time magazine issue.

but analyzing would take forever—→ take a sample.

Sampling

Selecting a smaller group of messages from the population to analyze.

The goal is for the sample to represent the larger population.

Ex: Population=5,000 magazine issues; instead of analyzing all of them:randomly select abt 100 issues.

-Researchers can then use that sample to draw conclusions abt the larger population.

Random Sample=helps craete a representative sample.

Unitizing

identifying the individual units of content that will be coded

Insta = one Insta post = one unit.

-Deciding what counts as one unit.

Coding

Create categories and determine which category each message belongs to.

-How often do children’s social media posts describe physical activity?

-Each post could be coded Yes/No

Coding Scheme

1-Coding Instructions: Explain exactly how messages should be classified.

2-Coding Sheet:Where coders record their decisions.

Ex: Post #17

Physical activity:Yes/No

-turns messages into data reserachers can analyze.

Coders

The people who examine the messages + assign them to categories.

Ex: coder reads going for a run and records physical activty=YES

-want multiple coders,bc one person=bias.

Independent Coders

Coders should be independent.

-Largely unaware of the specific hypotheses/goals and don’t influence one another’s decisions.

Why? To reduce bias—→ If coders know the hypothesis, they might unintentionally classify messages in ways that support it.

Training Coders

Step 1:Explain coding system-tech coders+ def/rules

Step 2: Practice: give coders training posts that are NOT part of the real sample; they code them independently.

Step 3: Check reliability: see how much the coders agree. If agreement is poor—→ improve the coding instructions/provide more training.

Intercoder Reliability

How much independent coders agree with one another when coding the same content.

Ex:100 Insta posts

Coder A says 8- are physical activity

Coder B makes similar classifications.

if classifications differ=low reliability=unclear coding rules.need more training,

Krippendorff’s Alpha

intercoder reliability =measured by uisng KA

Max=1

Above 0.70=acceptable reliability

higher values=greater agreement among coders

Exhaustive Categories

Coding categories=exhaustive

=every item has somewhere to go.

-EVERYTHING FITS SOMEWHERE

Mutually Exclusive Categories

ME= an item can only belong to ONE category.

Manifest Content

Obvious,surface-level,content.

Ex:is this movie animated/live action?

-What you SEE directly

Latent Content

deeper, underlying meaning

-is this movie optimistic/pessimistic

-coder must interpret the movie’s overall themes.

Manifest

Obvious/surface level

Is there a woman in this advertisement?

Latent

underlying meaning

Does this ad portray women positively?

-requires more interpretation.

Reliability

Are we measuring consistently

-Do different coders reach similar conclusions?

Validity

Are we acc measuring what we THINK we’re measuring?

You can have a measurement that is consistent but still measure the wrong thing.

Face Validity

Does the measurement appear on the surafce to measure what it’s supoosed to measure?

Does this make sense?

You’re measuring exercise

coding scheme:running,swimming, basketball, lifting weights

=reasonable=Goof face validity.

Content Validity

Does the measurement capture the FULL breadth of the concept

-You want to measure communication,probelm communication involves other types so the measurement isn’t capturing the whole concept.

Did you cover ALL the important content.

Criterion-Related Validity

Does your measurement relate to another measure that it theoretically SHOULD relate to.

Ex: you create a messure of communication competence.

relate to things like

-public-speaking performace

-if no relatiosnhip=measure isn’t actually measuirnhh cc.

-ability to persuade others-compares/associate your measure w another relevant measure.

Observer Effects

People change their behavior when they know they’re being observed.

Ex: Given fitbits=I should walk more, may be different from what it normally would be.

Unobtrusive Measurement

Measuring people/content without people being aware they’re being measured.

analyze: existing posts, press releases, advertisements

-The content already exists.

-reduce observer effect

Qualitative Research

Experiences,meanings,perspectives,and behaviors in depth.

-asks How?,What types,or in what ways.

-complicated not Yes/No

-uses reserach questions and not hypothesis

Quantitative Research

Numbers,measurements,varibales,larger samples

-Tests relationships/effects

-Does X affect Y

-Hypotheses

-Measurement/generalization

Sampling in Qualitative Research

-Researchers want

-A diverse sample

-A broad range of experinces/perspectives

-Participants who can provide useful info

-recruit ppl from industry you want to study.

Saturation

The point when collecting additional data stops producing new info

Ex:Conduct 20 interviews

Interview #21=same themes

Interview 22=same themes

Memos

A researcher’s written reflections about the research process and/or the data.

—write them during/immediately after interviews.

Memos help:

-Remember important observations

-Notice repeated responses

-Improve future interviews

Process Memo

Abt how the reserach/interview is being conducted

Ex: I should change the order of these questions

Analytic Memos

About what the responses mean

Ex: Several participants mentioned listening to music when they feel lonely.

Interview Protocol

The planned set of questions a reseracher intends to adress during an interview or focus group.

General—→Specific

Less sensitive—→More sensitive

This helps participants become more comfortable before discussing deeper/more personal subjects.

Structured Interview Survey

Very specific

Often restricted answer choices

Little opportunity for follow-ups

Often produces data analyzed quantitatively

Ex:

How many hours do you listen to music daily?

A:0-1

B:2-3

C.4+

Structured Qualitative Interview

-Researcher has specific questions

-Questions generally follow a planned order

-Answers are open-ended

-Follow-up questions are allowed

-The researcher can ask for clarification/examples

Ex: Describe a time when music made you feel connected to other people.

Can you tell me more abt this?

Semi-Structured Interview

Core questions asked to eveyone

Researchers can go off script

Follo-ups and clarification are allowed.

planned+flexible.

Unstructured Qualitative Interview

Very littlw predetermined structure

Reseracher has a general goal

Conversation develops based on participant responses

Building Rapport

Good rapport encourages people to give honest and detailed answers.

3 major factors:

• -Professionalism -Be prepared + on time

• -Appropriate environment=Quiet+private

• -Interpersonal connection=Smile+listen carefully

• Rapport helps lead to trust.

Observation

Examining people engaging in natural behavior.

Researchers: Watch behavior, take detailed notes, try to understand what is naturally happening.

Focus Groups

A small group gathered to discuss a specific topic.

A moderator/faciliator guides the convo but encourages participation to interact with one another.

Typically:Fewer than 10 participants

60-90 minutes

-Participants build off eachothers ideas,potentially revelaing things that wouldn’t come up in individual interviews.

Dis:Less detailed info from each ind person than one-on-one interviews.

4 focus group Facilitaor Skills

1. Listening: What is said

2. Intervening: Someone dominates

3. Building Rapport:Create a comfortable atmosphere using Professioanlism,brief icebreakers

4. Creativity: Researchers can use polls, videos,adv, debates

Thematic Analysis

Analyzing qualitatibe/text data to idnetify important themes,patterns,or topics.

=what patterns and meanings keep shwoing up in what people are saying.

Ex:Intervewing college studnte abt stress—→ themes might include academic pressure,lack of sleep,finances,and social pressure.

Grounded Theory

Allowing the data to guide the development of a theory.

Instead theory: collect data to test it

GT:Collect data —→identify patterns—→develop theory

Coding

Labeling/grouping pieces of qualitative data that are similar.

Ex: I feel overwhelmed bc I have 3 exams this week

-Academix stress,workload,exams

Open Coding

The initial stage of qualitative coding, where the researcher identifies anything that seems important.

OC=Brainstorming

Ex:Hardest week of the semester=Hell Week

Descriptive Codes

Describe exactly what is happening.

Ex:Participants talk abt attending concerts—> concert attendance

Interpretive Codes

The researcher interprets the deeper meaning.

Ex: Participants describe music bringing different groups of ppl together—→ music transcends cultural differences.

Axial Coding

Combining/organizing lower-level open codes into larger, theoretically meaningful concepts.

-Theory Building

Field Notes + Memos

Field Notes=detailed notes abt observations/context

-can capture missed things recordings may miss such as

-body language, environment, nonverbal behavior, researcher observations.

Memos=Notes researchers wroe abt their developing ideas + interpretations.

-Can continue through analysis

Audit Trail

A record of important decisons made throught the study.

Ex: Researcher combines 2 codes

They documeny:When they combined them,why they combined themm,how the decision affected anaysis.

-Makes reserach process transparent.

Triangulation

Using multiple sources, methods,or groups to see whether they lead to similar conclusions.

Ex: Studying college stress using:

Interviews+observations+focus groups

-if similar themes appear across all three, greater support for findings.

-ultiple ways of checking the same data.