MMC3420 Exam 1 Study Guide

Variable

- "empirical counterpart of a concept" - Any concept that can have 2 or more values; your concept with values - Require operational definitions (measure or observe.)

Dependent Variables

- Criterion. Ex. Increases in frustration of finding a parking spot o Effect

Type I

- HO is rejected- You say there's a meaningful difference, while in face, there is not. False Positive. You're pregnant! (But not really) o To combat: Lower the significance level from .05 (1/20 chance of error) to .01 (1/100 chance) or .00001

Validity

- Measures that measure what they're supposed to. Are we being accurate? o Exists on a continuum

Independent Variables

- Predictor, antecedent. Ex. Traffic o Cause

Voluntary participation

1. Do not involve people in research without their consent

Examples of Concepts

1. Message Length 2. Cable subscriber 3. Identification with characters 4. Love

Three types of Paradigms

1. Positivist- From the hard sciences. Emphasizes quantification and objective measures a. Reality exists and we can measure it b. You can aggregate data about humans 2. Interpretive- Goal is to interpret meaning (some qualitative research) How you see the world is differently is how you see the world. Impacts how we view something. a. Reality only exists thru an individuals' lens. How I see things through my perspective. b. Reality is holistic, can't be subdivided. All snowflakes. 3. Critical- Humanities-based. Non-empirical

Intensive Interviews Advantages & Disadvantages

Advantages - Rich detail from participants - Accurate responses due to rapport Disadvantages - Minimal generalizability- nonrandom sample - Interviewer bias- rapport can reveal interviewer attitudes/biases/HRQs - Time consuming - interviewer/interviewee fatigue

Research

An attempt to discover something

Institutional Review Board (IRB)

Safeguards the rights of human participants in academic research

Field Observations Procedures

Stages of Field Observations 1. Choosing the site- where do I go? 2. Gaining access- do I need permission? 3. Sampling- what do I focus on? 4. Collecting data- how do I record the data? 5. Analyzing data- how can I use this later? 6. Exiting - how do I get out of here?

Internal Validity

(Within the study itself) - Defined: Are the study results based on expected conditions and not extraneous variables? i.e. Are we measuring what we think we're measuring? - When a study is internally valid, we have ruled out the influence of an extraneous or confounding variable, i.e., an artifact unrelated to the study influence results. - Ex. Ice Cream and Murder rise at the same time.

Concept

- "an abstract idea formed by generalization" - An abstract term for what you're studying - Should have conceptual definitions (finding your definitions; attitudes; learning)

Hypothesis

- (Stating a relationship between the two variables) Predicts an outcome. A formal, tentative statement about the relationship between independent and dependent variables o Are appropriate when there is previous theory or research to suggest an outcome o Can be associative (related) or causal

One-Tailed Test

- About a relationship between variables that predicts the direction of the suspected relationship as either positive or negative o Greater advertising exposure increases brand attitudes o Females report greater frequency of binge watching o Greater access to social media discounts increases purchasing behavior

Field Observations Advantages & Disadvantages

- Advantages: o Can help identify important variables o Don't rely on self-report o Inexpensive o Can be rich in detail in a natural environment - Disadvantages o Potential bias and specificity lead to a severe lack of external validity o Reactivity- when people are being watched...

Intensive (in-depth) interviews

- Allow for detailed background and elaborated data from participants - Take note of nonverbal - Long and often over multiple sessions - Customized to the individual

Why do we standardize scores, or create z-scores?

- Allows for comparisons on different types of scales for the same individual or different people - You can meaningfully interpret one score within a distribution by showing how many SD's it is away from the distribution

What is a paradigm?

- An accepted set of assumptions that give you theories, research areas and questions, and methods. It is a worldview of research. - One's paradigms determines RQs/hypotheses and methods approach- but not research topics.

Ethical Principles

- Autonomy- Self-determination, an individual's ability to make un-coerced, informed decisions o Ethical research is voluntary - Nonmaleficence and beneficence- Try to minimize harms and maximize benefits - Justice- People should be treated equally

Criteria of a Good Hypothesis

- Based on previous research and theory - Narrow the study's focus - Attempt to rule out confounding variables - Allow for quantification - Should be compatible with current knowledge - Are logically consistent - Are stated concisely - Should be testable

Case studies

- Combines data to analyze individuals, groups, organizations, or events - It's a free for all- methods and time - Lacks generalizability

External Validity

- Defined: How well the results of one study can be generalized to larger settings or populations - If a study lacks external validity, it can't be extrapolated to all TV viewers, all twitter users, all college students, etc. It describes nothing but the exact sample of people it studied.

Variability/Dispersion

- Dispersion measures describe how the data is spread out around a central point, usually the mean

What ethical concerns arise after you've collected data?

- Don't fabricate data - Don't plagiarize o Authorship guidelines o Yourself by sending multiple articles out o Your collaborators- acknowledge all - Be very clear on methodology - Provide appropriate interpretations of data- state suggestions, instead of correlations

Research Questions

- Exploratory. A formal question posed to guide research o Common in exploratory or preliminary research o Can be associative or causal, but are somewhat more likely to be associative o Are posed when there's a lack of theory or previous research- you have nothing to base a hypothesis

Focus Group Moderator

- Friendly way let's just take turns talking. - Introduce yourself and themselves. - Tell them amount of time. - Transition to first question - Ask How, when, what, why questions.

What are the common threats to internal validity?

- History- Events outside of your study affect the outcome - Maturation- Participants mature over the time of the study, cognitively, developmentally, emotionally, etc. - Instrumentation- Measures/equipment, coders decay over time - Statistical regression- to the mean. High and low scorers tend to move back toward a group mean in subsequent tests - Mortality- participants drop out - Demand characteristics- When participants are aware of the study goals and act in accordance (or defiance) of them - Experimenter bias- May be intentional or un - intentional - Casual time-order- The independent variable must come before the dependent variable

Ethnographies

- Immersing yourself in the research- from the participant's point of view - Variety of research techniques

What do standard deviations units tell us about scores in a normal distribution?

- In a "normal" distribution, the SD tells us what % of people fell in what percentile around the mean.

What is a Mixed-Methods research?

- Incorporating quantitative and qualitative research methods. Concurrently or sequentially o More data, broader answers o More time, effort, money, analysis issues o Triangulation

Three major measures of Central Tendency

- Mean: Arithmetic Average Ex/N o Sum of scores divided by number of scores o Most frequency use- takes all scores into account. Best deals with outliers. - Median: Divides the distribution exactly in half. It is the 50th percentile. Media Splits. (House distribution.) o X= 2,2,3,4,4,5,6 Mdn=4 o Also, X=1,1,3,4,5,6,6,8 Mdn=4.5 - Mode: Most frequently occurring (popular) score. o Appropriate for nominal data

What are active and passive online research settings that require different ethical responses?

- Passive Online Research- content widely available online. Require registration? Expectation of privacy? Identify participants? o Ex. Comments on a brand's social media page, YouTube comments, comments on news articles etc. - Individuals' social media accounts? o What are the privacy settings? o Social media company policies differ o Facebook says you must obtain consent - Active Online research o Online surveys- Consent, Debriefing, Privacy Issues

Two-Tailed Test

- Predicts a relationship (in any direction) o Advertising exposure is related to brand attitudes. o One's gender is associated with frequency of binge watching o Availability of social media discounts is related to purchasing behavior

Quantitative Research

- Puts numbers on it. Measures concepts. Standardized questioning, typically more generalizable o Assigns numerals to objects, i.e., measurement. o Ex. Surveys (Rate your Uber driver), Experiments, Content Analysis o What we'll primarily focus on in class

Major differences between quantitative and qualitative research

- Qualitative research is more concerned with description and explanation rather than measurement and quantification. - Qualitative vs. Quantitative: Flexible vs. static, a priori measures; issues of sample size and generalizability.

What are unscientific ways of knowing?

- Tenacity (it has always been true) Everything is always changing - Intuition (Self-evident) That is not a universal experience - Self-discovery - Authority

What information goes into z-score calculation, or standardizing scores?

- The formula for standardizing a score is to divide the deviation score (how much a score deviates from the mean) by the SD of the distribution Z= x-line of over x/ s Where z is the standard score S=the standard deviation of a scale X=each value in the data set Line over X=mean of all values in the data set

Central Tendency

- The overall goal of central tendency is to find the single score that is most representative or typical for the distribution; describes a center point

What information goes into standard deviation calculations?

- The typical, or standard, or average deviation b/w individual scores in a distribution and the mean of a distribution. - This measure of variability takes into account each score, which makes it more accurate than the range.

Type II

- Treatment effect really exists, but hypothesis test fails to detect it. False negative. You're not pregnant (yeah, you actually are) o To combat: Study design, larger sample size

Qualitative Research

- Uses more flexible questioning (can change the question); is interested in unique, individual stories, and cannot always be generalizable; 'text'-based (Words not numbers) o Ex. Focus Groups, Ethnographies, Interviews, Case Studies (one isolated incident related to another)

Null Hypothesis' Purpose

- We test for ruling out everything. We want to prove wrong that they aren't any relationship between our variables.

How do we try to avoid Type I and Type II error?

....

Compare and contrast academic research and applied research.

1. Academic (Basic) - Generally theoretical or scholarly. Concerned with humans interacting with media - Publicly available; Not as deadline-driven; can be cheaper 2. Private (Applied) - Concerned with specific media industry and audiences. Goal is satisfy a client who wants to make money - Often proprietary; completed quicker, more expensive 3. Both academics and private sector researchers can do both types; Both can utilize quantitative and qualitative methods. Both want to predict the future. - To predict what comes next.

Ratio

1. An interval scales with an absolute zero point (0 mintues on facebook. There can be a complete absence of something) Ex. How many times/days, do you log in on Facebook? a. Ratios of numbers reflect ratios of magnitude: you can be twice the weight or twice the height of someone b. Can be used in advanced statistical techniques, can't be (truly) negative c. Examples: 1. How many times per day do you log into Facebook? 2. On average, how many minutes do you spend on Facebook each time you log in? ... .... ...

What does qualitative data analysis entail?

1. Data Reduction a. Creating categories b. Perceive, note researcher biases 2. Draw Conclusions- a. Elaborate, refine categories b. Search for relationships among concepts and themes c. Simplify and integrate data into a theory or hypothesis

Standard Deviation Calculation

1. For each score, determine its deviation form the mean: a. Deviation score: x-c or M; where M=3 2. Square each deviation score and then calculate the mean of these. This ean is called the mean squared deviation, or variance. So, a. Variance= Sum of Squared Deviations (SS)/ # of Scores (N) 3. The standard deviation is the square root of the variance. So the square root of 2 is the SD. =1.41421 =1.41 Done! M= 3, SD= 1.41 (fall plus or minus on that score, varies from our mean 1.41 hours) - Standard Deviation, SD, is reported in the original units that a mean is reported in.

How do we determine research topic relevance?

1. Is the topic too broad? 2. Can the problem really be investigated? 3. Can the data be analyzed? 4. Is the problem significant? 5. Can't the results of the study be generalizable? 6. What costs and time are involved in the analysis? 7. Is the planned approach appropriate to the topic? 8. Is there any potential harm to subjects?

Operational Definitions

1. Measured time, in seconds, for the entire message - Or, measure time, in minutes, of the entire message 2. Survey question: "Do you pay a cable company for cable TV services in your home?" "YesNo." - Or, by contacting Spectrum, U-verse, etc. and getting subscription rates per county 3. Previously validated survey scale of "identifications with characters, "9 items, averaged for a score - # of mentions on social media about show/characters in the past month 4. Observing established behaviors related to love and counting them during interactions with others. These include # of times one says 'I love,' kisses a partner, holds hands, etc. - Survey question

Example of Variables & possible values

1. Message Length: 15 seconds; 22 minutes 2. Cable subscriber: 50 "yes's" and 40 "no's"; 60% (are cable subscribers) 3. Identification with characters: 5.5 on a 7-point scale; number of mentions on social media (3) 4. Love: 4 or zero, or 10

Interval

1. Ordered categories that are at equal distance form one another along the measurement scales (rate. So you can tell the average, numbers that mean something) Ex. On a 7-point scales, 1-not at all, how often do you be on Facebook?) a. Tells you significantly more info than Nominal/Ordinal data. Allows you use advanced statistical methods b. Differences between points are meaningful, but there is no "real" zero. Zero does not mean a total absence for the variable. Cannot be negative. c. Continuous. d. Likert Scales (midpoints; summed; 1-7; 1-10) e. Semantic Differentials - measure meanings assigned to objects; use bipolar objectives/visuals i. Likert and Semantic Differentials - 2 types of scales commonly used for surveys f. Examples: All items measured on a Strongly Disagree (1) to Strongly Agree (9) It was fun for me to watch UCF football game .... .... Avg. numbers together

What historical events/ periods encouraged the growth of mass media research?

1. Propaganda Research after World War I & II a. Hypodermic needle model/Magic Bullet: any info read is shot to us and absorbed. You sent me a message now I love it. 2. Advertising in 50s & 60s: TV being popular in homes 3. Effects on children a. Violence, sex, pro-social 4. Increased competition for money a. Audience Fragmentation b. Changing values/lifestyles demographics, image, studies, audience segmentation.

Informed consent

You have to tell participants: - Purpose of research - Expected duration and procedures - The participants' right to withdraw at any time and any possible consequences - Potential risks, discomfort - Potential benefits (incentives) - Limits of confidentiality, explain who will have access - Whom to contact with questions/ follow-up

Inferential Statistics

Makes inferences about populations using data drawn from the population.

Focus Group Procedures

Stages of focus groups 1. Define the problem/issue 2. Select a sample with shared characteristics- this important 3. Determine the # of groups necessary- goal is to achieve saturation (everything to be said is said) 4. Prepare the study mechanics- aka logistics 5. Prepare the materials- e.g. moderator's guide 6. Conduct the session 7. Analyze the data

Null Hypothesis

The logical alternative to the Hypothesis (H1); it predicts a lack relationship; Asserts any observed differences are due to chance or random error. EX. Advertising exposure has no effect on brand attitudes. - Alternative, not opposite.

Kurtosis

The peakedness (height/flatness) of a distribution.

Distributions

A collection of scores/data on a variable.

Control Variables

Eliminate unwanted influence Ex. Bad weather, construction

Focus Group Advantages & Disadvantages

Advantages - Participants often less inhibited when in a group - Inexpensive, depending on target group - Flexibility - Get a complete picture from participants from interaction Disadvantages - Group leaders take over - Answers "why" and "how", not "how many" - Moderator skill is crucial- staying on course, etc.

Statistics

Mathematical method to collect, organize, summarize, and analyze data.

What is the role of the researcher in qualitative research methodologies?

Measurement instrument

Concealment

Withholding certain information from participants

What are the six basic characteristics of the scientific method?

1. Scientific Research is public - So that it's replicability can be verified 2. Science is "objective" - Develop and follow explicit rules and procedures for collecting data. - Facts instead of interpretations of facts. 3. Science is empirical - Knowable and potentially measurable. Phenomenon which is perceivable and able to be classified - Of the world. Measurable in any way (qualitative and quantitative) - Conceptual (Constitutive) definitions are somewhat abstract; define a world/concept by substituting other words. Personal attitude that varies in favors and strengths - Operational definitions specify procedures that allow for the measurement of something. How you measure. Set of procedures that lets you measure your conceptual attitude. (Rate 1-10) 4. Science is systematic and cumulative - No single research study stands alone; nor does it rise and fall by itself - Always start with previous research - Laws are statements of fact that explain actions that are generally accepted and universal 5. Science is predictive - Predict future phenomenon. What scientific theories do 6. Science is self-correcting. - We do better over time.

What is the over-time process, or steps, in developing a research project?

1. Select a problem/question 2. Review literature and theory 3. Develop research questions/hypotheses 4. Determine appropriate methodology - Quantitative: Enumerating; Generalizable. - Qualitative: Unique; text-based; not generalizable. (Focus groups, surveys) 5. Collect relevant data 6. Analyze and interpret results 7. Present results (Visuals) 8. Replicate the study

Examples of Conceptual Definitions

1. The length of time from beginning to end of a piece of mediated information 2. Whether one pays a fee to an MSO or other provider for any cable TV services, delivered via coaxial cable into the household 3. A psychological relationship between an individual and mediated character that allows for the individual to adopt the perspective of the character and vicariously experience events from that perspective 4. A deep and tender feeling of affection or devotion to a person or persons.

What are the research phases when a new media technology is introduced?

1. The medium itself 2. Uses & Users of the medium (Who is using it and how they're using it) 3. Effects of the medium 4. How the medium can be improved

What can we do to improve external validity?

1. Use a random sample, which means everyone in the population of interest has an equal chance of being selected for participation in the research study.

Ordinal

1. Values are categories in rank order (ordered sequence) along some dimension Ex. Low, medium, or high user of Facebook user? a. Values have a greater than/ lesser than relationship, but you cannot determine the magnitude of the difference/relationships b. Example: Please rank order the following TV shows from most to least favorite.... How to Get Away with Murder Sense 8 The Walking Dead ... ...

Nominal

1. values are categories that differ in name only. (eye color, religious affiliation) Ex. Are you a Facebook user or not? a. At least 2 categories that mutually exclusive and exhaustive (if someone has 2 different eye colors and they could only pick one option, it will not be mutually exclusive) b. Many demos; treatment condition in experiments; yes/no question; often good for independent variable c. Discrete, i.e., cannot be broken down into subparts d. Example: Many options and having an 'other' section

Normal curve

A perfectly symmetry distribution.

What is a power analysis and what affects statistical power?

A statistical test that estimates the number of participants needed, or the sample size needed to generalize to your population of interest. ......

Theory

A system of ideas or set of principles intended to explain some casual phenomena or relationship among concepts.

Normal Distribution

A theoretical one or a standardized one- we rarely get an exact normal distribution.

Mass Communication

Any communication channel used to simultaneously reach a large number of people.

What is effect size?

Can be small, medium, or large. This is the magnitude of the effect observe

Deception

Deliberately providing false information o Was there no other way? o Significant benefits outweigh harm?

Methodology

Description of processes taken in conducting research. (How to conduct research)

Can a measure be reliable but not valid? Can a measure be valid but not reliable?

Yes. No. A valid measure is a valid measure today and a valid measure tomorrow. It will yield a valid measure of results.

Saturation

Everything to be said is said

Descriptive Statistics

Methods and techniques designed to reduce datasets to allow for easier interpretation (summarize your data on a variable).

Research Ethics

Moral principles and recognized rules of conduct governing the actions of researchers.

What is p value, and what does a p value have to be under to be significant in the social sciences?

Probability of the results being due to real difference rather than error, or significance level, p < .05, which means there's a 5% chance of error

What is the importance of statistical significance?

The process of analyzing data to test whether a null hypothesis is correct or false. • Tells you if there is a real difference, but does not tell you the magnitude, or size of, or meaningfulness of that difference

Skew

The symmetry of a distribution.

Reliability

When a measure consistently provides the same answer. Essential

Anonymity

When data can not be connected to individuals, privacy

Confidentiality

When data is connected to individuals, but researchers promise not to reveal that information, privacy

Examples of Validity

o Face validity: Makes sense on the face of things o Predictive validity: A measurement that can predict what you want it to o Concurrent validity: A measure that is related to other measures of the same variable o Construct validity: A measure that is related to other concepts and constructs in a theoretical way

Examples of Reliability

o Get the same results today and tomorrow and next week. o Internal consistency: The consistency of items on a scale to reliably measure a concept. Cronbach's alpha> .7 or not (Introvert vs. extrovert. Group things that measures both together, analyzing both concepts.) o Inter-Coder reliability: The consistency of coders agreeing on the presence/absence of message characteristics

3 Measures of Variability/Dispersion

o Range: The difference b/w the highest value and the lowest value of a distribution o Variance: The degree to which scores differ from the mean- A small variance means that most scores are clustered around the mean, a large variance means they are more spread out o Standard Deviation...

Normal Distribution

theoretical one or a standardized one