Week Two Notes
Week Two Notes Communication Studies 150
Popular in Methodologies in Communication Research
Communication Studies 150
verified elite notetaker
Popular in Communication Studies
This 9 page Class Notes was uploaded by Alyssa Notetaker on Saturday October 10, 2015. The Class Notes belongs to Communication Studies 150 at University of California - Los Angeles taught by PJ Lamberson in Fall 2015. Since its upload, it has received 22 views. For similar materials see Methodologies in Communication Research in Communication Studies at University of California - Los Angeles.
Reviews for Week Two Notes
Report this Material
What is Karma?
Karma is the currency of StudySoup.
You can buy or earn more Karma at anytime and redeem it for class notes, study guides, flashcards, and more!
Date Created: 10/10/15
Week Two Lecture 3 Terms 0 Unit of Analysis 0 The thing you re studying o What re you looking at What object are you studying an object that stays constant 0 It s the focus of the project Variable o What changes EX work habits in summer vs work habits in fall Unit of analysis is work habits variable season 0 Measurement 0 The process of assigning a numeral or label to the unit of analysis in order to measure it Ex unit of analysis 5 kids studied over 3 weeks 0 Allows for conceptualization carifying what you re going to do makes it operational makes it something you can actually look at and measure gives it a complete plan makes it doable Spurious correlation random chance that causes 2 things that are not causally related to be correlated 0 Correlation is just a statistical measurement tells us the numbers we observe tend to move in the same way Positive correlation when x is high y is likely to be high Negative correlation when x is low y is likely to be high Does NOT mean it s causally related Research Design cont d Theories and hypotheses cont d 0 Why do we have to make a hypothesistheory rst Why not just do it and see what happens You have to have something that you can disprove or support If you measure enough stuff something will be statistically signi cant but it could be chance correlated but not causal So we need a hypothesis to guide us to say I think this will happen because of this stuff I ve noticed and makes sensequot so it s more likely it s signi cant and not just chance 0 Theorieshypotheses are usually about the relationship between 2 variables Praba b ilitg Very UiHl39HEW abserva n39s xi M So the hypothesis is usually that there is a positive or negative correlation that is causal We can see the correlation with data and numbers But we can only try to INFER from that data that it s causal having a theory beforehand helps that inference s chances of accuracy CAUSALITY Social science research is concerned with understanding not just what happens but WHY Why Because in social research we often try to solve social issues which necessitates knowing the cause of the problem EX Are your friends making you fatquot is a study on causality 0 Uses data on a social network for a whole town Framingham to see in uences of connections family and close friends on their weight health happiness etc Found positive correlation in neighbors neighboring connections so family and close friends behavior 0 Makes clusters when graphed The correlation extended to three degrees of separation you re affecting your friend s friend We can use that correlation to infer causality infer social in uence 0 To do so have to ask two queonns 1 How do we know there are non random clusters In other words is it statistically signi cant 0 Hypothesis testing Ttest P value etc After calculating the observed correlation we use a Ttest to calculate how likely it would be to get a correlation that high randomly How the Ttest works it simulates a bunch of random connections and sees how likely it is O 0 Most likely SEi39ll39 H EFF U 39J39H Z39W I ME absewati39ms DDSEWEH are Plum 3 Fla 1 Set of Possible Results quotF A pM39EIILIE shaded gl39EiEI39I area is the prubaibilitgi if an observed or more extreme result arising bu chance to get each level of correlation The smaller the chance of getting the correlation you got or higher the more likely it is that it s causal o The clusters were signi cant not random the close connections were more likely to share behaviors How do we know the clusters were caused by in uence 0 This is hard due to several factors 0 How do we know the friends in uence the behavior rather than the behavior in uencing whether or not they re fnends If the latter happens the behavior causes the clusters rather than the friends in uencing their connections behavior An endogeneity issue L endogeneityquot means being in uenced within the systemquot so your internal preferences in uence who your friends are Possibility that x doesn t really cause y y actually causes x Hard to separate have to consider this reverse causa on o Possibility of external factors aka confounding variable Ex 2 people are friends because they re coworkers and both are sad because the company goes out of business Their friendship is not due to sadness but neither is the sadness due to their friendship Solution for this is measuring and controlling for the confounding variable 0 In short 3 causal possibilities in uence endogenous or confounding In uence friends in uencef ends behavior Endogenous Behaviors in uence connections Confounding neither of the above Some ways the study tried to prove causality in the correlation 0 Looking at the relationship Mutual friends contactperceived friends or subjectperceived friends Found that mutual friends affected each other the most but all of these connection types in uenced each other 0 Education level 0 Immediate neighbors Immediate neighbors weren t as likely to be affected by your behaviors l it s less likely there s an environmental confounding variable so it s more likely the affects are due to social connections The Bad News 0 No way to know if the correlation is truly causal in practice because we can never recognize and control for all variables 0 Best way to solve this issue randomized controlled trials Ex a trial used social media to see if more people voted when they saw online which of their friends voted Big Data 0 Answer yes 0 How controlled Some I votedquot buttons online showed friends who voted some didn t but otherwise were the same they also were shown on many many feeds randomly controlling for confounding variables 0 Checking who said I votedquot on social media and who actually voted 0 We know they were friends before voting so weren t friends because of voting so not endogeneity Does the new availability of huge amounts of data make the scienti c method obsolete make it so correlation is enough Issue with just correlation before big data you need causation if you want your ndings to hold in a different population for it to be able to be generalized But the new argument is that with big data the data set the population is EVERYONE so the correlation doesn t need to be 0 generalized This is still problematic though Polls people lie poll could show bias in polling more people from a certain demographic sample bias Data sets can be biased Questions asked can be biased Gallup tried to get a representative sample to eliminate sample bias randomly choosing enough people from each demographic to mirror the population of the US 0 Still isn t perfect Another issue might look like there are more shark attacks than ever before but look at how there are also more people in the water than ever before 0 Same idea with big data more people on the internet since its start Google Search s inception l fraction of people who search for a certain term lessens over time Predicting with web search data 0 Using search data to predict makes the prediction more accurate but only slightly 0 Google tracking the u Worked really well But then it didn t anymore 0 Used only correlation Assumed it would continue working in other time periods than the one they made the algorithm for Lecture Four Measurement What are you measuring 0 You have to de ne it clearly 0 De ne subjective and qualitative terms in ways that make it measurable quanti able Ex with a theory that your happiness affects that of your social contacts you need to de ne happiness and social contactsquot clearly and fully in quanti able ways 0 Makes it clear what you are going to look at and measure Conceptuaization and operationalizing Strictly de ning variables into measurable factors 0 De ne ALL key concepts your de nition can affect what your results are Operationalizing Versus conceptualization o Conceptualization is de ning and specifying abstract and fuzzy concepts Concepts have indicators and dimensions 0 Dimensions speci able aspects of a concept Indicators something the researcher chooses to recognize as a re ection of the variable being studied 0 Makes it more feasible to study something that is not directly observable 0 Ex using smiles to indicate happiness 0 Operationalization is the development of speci c research procedures that will result in empirical observations representing those concepts in the real world An operational de nition speci es precisely how a concept will be measured Some parts of operationalization Determine research method 0 Observations 0 Survey Remember that determining the wording of a questionnaire is important 0 Experiment 0 Looking at archival materials data or information that has already been collected 0 Case study 0 Consider the population and sampling 0 Random sampling is it representative of the population etc 0 Consider how broad the concept you want to study is 0 De ne variables and attributes make them measurable Levels of measurement 0 Binary Q g Ratio Aka dichotomous Nominal Aka categorical discrete multinomial Give categories but there s no order to the categories Ex quotWhat religion are youquot 0 Possible answers are 1 Christian 2 catholic 3 protestant 4 hindu 5 muslim 6 Jewish 7 other 8 none Ex hair color EX nationality Ordinal Aka categorical discrete multinomial Give categories but they have an order EX satisfaction levels on a scale of 1 to 5 with 1 being dissatis ed and 5 being completely satis ed Interval Aka continuous can take on any value in a range EX measuring cortisol levels temperature placing politicians on a scale of liberal to conservative IQ score 0 You can compare the variables cortisol levels politicians degrees but can t say one is twice the level of the other can t say Hillary Clinton is twice as liberal as Donald Trump 0 You can see the difference between 2 measurements but can t make it a ratio Aka continuous But in addition to being like interval there is an absolute zero 0 So you can say something is twice as much as something else EX wealth measurements age length of time MeasurementQuality 0 Reliability Consistency little variation ldea that the same data would have been collected each time in repeated observations of the same type We can increase reliability with increased range of measurement using multiple indicators EX a scale is reliable objective opinions on a topic are not Q Validity How close is your result to the truth Closer more valid Valid when the measure accurately re ects the concept it is intended to measure 0 Face validity it makes sense at face value it s reasonable adheres to common sense Criterionrelated validity the degree to which a measure relates to some external criterion 0 Construct validity asks whether the various measures for a given concept all seem to correspond to the same thing 0 Content validity the degree to which a measure covers the concept it operationalizes 0 Ex using a weight scale to measure intelligent is very reliable but not valid at all 0 Ex using cortisol levels to measure happiness is reliable but would only be valid if you re measuring the subject s happiness at that time rather than over the past year or something 0 Error 0 Systematic error lt s systematic the level of the error is always the same Ex when a scale is off and always adds an ounce to what is being measured EX when a ballot is made so it looks like to vote for someone you punch the second hole but really you should punch the third Sources of systematic error 0 Social desirability 0 Way to minimize this bias Instead of asking the subject about their own views or supposed actions give a hypothetical situation about a stranger Ex in this certain situation what do you think this person a stranger madeup person would do Anchoring o The tendency to rely too heavily or anchor on one trait or piece of information when making decisions 0 Usually the subject focuses most on the rst piece of information we get from them Acquiescence agree vs disagree 0 People are more likely to agree 0 Order effects o If you ask a series of questions or have a list to choose from people are affected by which questions comes rst Can bias the results 0 Random error Usually cancel each other out 0 Error observed value true value systematic error random error
Are you sure you want to buy this material for
You're already Subscribed!
Looks like you've already subscribed to StudySoup, you won't need to purchase another subscription to get this material. To access this material simply click 'View Full Document'