Week 3: Scientific Methods in Media Effects Research
Week 3: Scientific Methods in Media Effects Research JMC:1100:0AAA
Popular in Media Uses and Effects
verified elite notetaker
Popular in Journalism and Mass Communications
verified elite notetaker
This 10 page Class Notes was uploaded by Kylie Kreischer on Saturday March 7, 2015. The Class Notes belongs to JMC:1100:0AAA at University of Iowa taught by Patrick Wright in Spring2015. Since its upload, it has received 99 views. For similar materials see Media Uses and Effects in Journalism and Mass Communications at University of Iowa.
Reviews for Week 3: Scientific Methods in Media Effects Research
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: 03/07/15
Lecture 5 Social Scientific Methods Content Analysis and SurveysPolls Superbowl Katy Perry half time show Last Time Radio takes center stage 0 War 0 the Worlds freaks everyone out Limited Effects Theories o Dominate social science 0 ex left shark in Katy Perry s performance cool or stupid Opinion leaders and 2 step 0 Flow direct our media use Today Research Methods of Social Science ContentAnalysis Surveys 0 How do to them well Children s Shows 0 Sesame Street most studied television show in media history Because of the fascination of how children learn 0 Blue s Clues One of the most scientifically put together shows in media history Came from scientific research Content Analysis Describes the nature of particular media content 0 Helps scientists identify specific types of content before they can find effects Two types 0 Manifest natural that actually appears 0 Latent reinterpretation of media messages 0 ex Power drill pictures Benefits 1 It is objective Same standards are applied to all types of media in a study Removes the threat of bias in analysis 2 It is systematic Uses consistent and clear rules and produces for coding all study content 3 It is quantitative Content coded through number of times an event or situation occurs Allows for statistical tests and analysis to understand nature of content How to do content analysis 1 Define your study content and how to code it 2 Choose content websites movies etc 3 Start counting basically South Park Ship igh in Iransit Sucker Punch Male fantasy vs female exploitation Surveys and Polls Most familiar method Helps researchers uncover personal feelings and public opinion about content andissues KEY Allows researchers to begin to connect content to effects 0 Content analysis has nothing to do with effects Why are they important 0 They reflect our current opinions 0 Results influence our future attitudes behaviors and knowledge ex Exit polls How to do a survey 0 First we must est who we talk towe need a sample 0 Sample vs population 0 Samples The goal is to makes inferences to A larger population Other time periods Other locationssocieties Random sampling Each subject in the population has an equal chance of being selected into the sample Allows us to estimate where the real responses would lie if we could survey everyone Theoretically representative of the population Nonrandom sampling There is not an equal chance of being selected ex TV call in polls lnternet surveys American Idol etc Overloads you on a certain demographic Can t tell how accurate the results are Systematic bias too much of one population Surveys that use nonprobability sample are usually not scientific Margin of Error In order to calculate margin of error we need to use random sampling techniques Measure of a poll s precision A range of the possible poll Margin of error is the different between the sample and the actual population Usually impossible to survey a whole population Goal Bigger sample smaller MOE ex 2012 PreElection Poll in Florida MOE 3 Obama 50 4753 Romney 46 4349 Could go either way based on subtracting and adding he 3 Ranges overlap A Statistical Dead Heat CrossSectional survey Occurs at a single point in time and involves a single sample Longitudinal survey Samples collected over time Trend study looking s for patterns Panel study same people different times 0 Developing a questionnaire Question wording Openended questions What types of shows do you watch on television Pro Answer are exhaustive Con Takes a long time and requires coding Closedended questions One a scale of 1 to 10 how much do you like medical shows on tvquot Pro greater uniformity and easily processed Con answer categories must be exhaustive Types of questions to avoid Leading questions Like most students do you think dogs should be allowed in the classroomquot Push polling quotWould you vote for John Smith if you knew he was a Cyclone fanquot Double barreled questions quotDo you think cats and dogs should be allowed in the classroomquot 0 New Methods of Survey Data Collection Conclusions 0 Content analysis is great for understanding BUT it won t give us effects O Surveys and polls give us public opinion and possible sense of media effects BUT only if they are carefully worded Poor questions can lead to unreliable answer Lecture 6 Experiments The Last Best Method for Media Effects Media Today Detroiter walks 21 miles to work 0 Not anymore due to GoFundMecom 290000 raised Sucker Punch count 0 Society creates media 0 Media create society Last Time Content analysis Surveys and polls 0 Learn public opinion 0 Generalize to big populations Sampling 0 Random vs nonrandom o Margin of error Today Experiments Experiment components Correlation and causality Experiments Method designed to test for cause and effect 0 Usually done in closed conditions for accuracy 0 Uses variables something we can observe that may change from person to person group to group ex demographics Subjects and stimuli in the same location 0 Allows you to see results as they happen 0 Closed setting 0 Controlled dynamic Closed conditions are not the real world 0 Few people watch TV in a room with people watching them Experimental Designs Find subjects and assign to random groups 0 Treatment group group you are studying 0 Control group neutral group for perspective Goal groups are equal regarding types of people age race gender etc Control Groups Purpose create neutral regular group of viewers o Allows you to compare results and determine effects 0 Accounts for bias when people are observed Hawthorn Effect of Uncertainty Principle Hawthorn productivity increased while lights decreased until they turned completely off other study saw increased level of productivity not because of rewards but because of being watched Money is not everything productivity can actually come from being watched 0 Group doesn t watch study content or watches different content Experimental example Study conducted on brain function and violent video games Video from study on the brain Goals of Experiments Correlation 0 Relationship between two variables BUT no cause and effect Causality Gold Medal Result 0 A relationship between two variables such that one influences the other creates it or changes it somehow Correlation In Practice Terrorism and snowboarding 0 Both could go up in 2015 0 Doesn t mean there is a connection Causality In Practice Cats and brain damage 0 Kitty litter and brain damage in infants 0 Lack of human agency in people 3 Criteria needed for Causality 1 Logical association between variables 0 Teen surveys show a relationship between watching sex on TV and having intercourse 2 Constant time order one always occurs before the other 0 Teens start watching TV years before they start having sex 3 All other variables ruled out o Other reasons teens could be having sex Parents may be absent Teens may be lonely Teens could be bored How We Know our Experiment Works lt achieves statistical significance 0 A relationship or difference between variables that did not occur by chance 0 KEY something meaningful is happening Can conclude study results are correct ex RevengePorn sending out pictures victims and depression Experiments Valid when Replication design and method idd same result over and over again Convergence different methods yield similar results Conclusion Experiments are a great way to connect cause and effect Correlation does NOT mean causation KEY experiment must be statistically significant to draw any conclusions Chapter 2 Scientific Methods in Media Effects Research 3 Methods 1 Content analysis 2 Survey 3 Experiment Content Analysis Research technique for objective systematic and quantitative description of the manifest content of communication Can be applied to almost any type of communication Result of content analysis DO NOT allow us to makes inferences about the effects of that content Facts must be updated Coder person who examines the content and classifies it into categories Manifest content material that actually appears and required a minimum interoperation by the coder NOT latent content content that might become apparent after a coder has interpreted or read between the lines of the message before coding Random sample sample that permits every member of the population to have an equal chance of being selected in the sample 0 Crucial if you want to generalize the results Units of analysis what units of content will be coded Category scheme how the content of the message is coded Coding reliability a computing coding agreement with a special formula that makes an adjustment for the chance hits Sample Survey Valuable in helping us understand people s media habits Can explore relationship between different variables ex 911 viewers who were int he top third of the sample on TV exposure to the attacks were almost 25 times more likely to suffer from PTSD Types of surveys o Crosssectional survey of s single sample of population that occurs at a single point in time and involves a single sample 0 Longitudinal survey survey project that includes more than one sample taken at different points in time Trend study group that is studied the first time is not the same as the group studied the second time Cohort study see change through the same age cohort but not the same individuals Panel study uses the same individuals Casual relationship between media and people 0 Correlation statistical tool used to determine relationships between two variables 0 Correlation coefficient used when researchers want to know whether two variables are related to each other Positive increase in both variables 1 Negative Increase in one variable decrease in the other 1 o Statistically significant a result that is unlikely to be due by chance 0 Criteria for casual relationship 1 Determine that the correlation is statistically significant 2 Time order of the twovariables Variable doing the causing must precede the rainless that is caused Advantage of longitudinal survey you can est time order 3 Eliminate any thirdvarible explanations 0 Survey is NOT capable of satisfying all three of the criteria only 2 The Experimental Method Gathers data under controlled conditions Manipulation of a key variable 0 Independent variable manipulated o Dependent variable responds Random assignment to the experimental conditions 0 Crucial feature 0 Any differences in groups can be attribute dot the manipulation Identical treatment except for manipulation Control group Types of experiments designs 0 Pre test post test measurements before and after the experiment 0 Pre test post test control grouo added control group that doesn t do the experiment 0 Posttest only measure after experiment only Benefits avoid sensitization to the pretest questiontopic Ex Effects of Mood on Music Listening Choice 0 Moodmanagement theory states that people deliberately use media messages to alter their moods Replication refers to the fact that the same result can be observed over and over again Convergence refers to the fact that the use of different methods still leads to the same general conclusion 0 Researchers are reluctant to declare mass media findings until results from one or two studies confirms it 0 Very critical Other Methods Epidemiological approach observational science research 0 research that is often conducted in the medical arena because it is not ethical to do experiments that could cause severe harm 0 Useful for studying the potential impact of media in the natural world 0 Can t control variables 0 More compelling and complete when pared with a controlled laboratory experiment 0 ex affects of smoking on cancer Metaanaylsis examining data from other studies to have a new summary 0 Requires very precise examinations of the statistical effects 0 More reliable indication of any effect 0 Must be sure to have all information
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'