Psych 2300: Research Methods - Chapter 6 Notes
Psych 2300: Research Methods - Chapter 6 Notes Psych 2300
Popular in Research Methods In Psychology
Popular in Psychlogy
This 8 page Class Notes was uploaded by MadsSwart on Monday June 27, 2016. The Class Notes belongs to Psych 2300 at Ohio State University taught by Emanuele Rizzi in Summer 2016. Since its upload, it has received 48 views. For similar materials see Research Methods In Psychology in Psychlogy at Ohio State University.
Reviews for Psych 2300: Research Methods - Chapter 6 Notes
Report this Material
What is Karma?
Karma is the currency of StudySoup.
Date Created: 06/27/16
Chapter 6: Surveys & Observation - NOT ON EXAM WEDNESDAY Today: - 3 types of research design claims (frequency, association, and causal) - Construct validity issue in self-report [survey design] - Survey design - Define each of the four major validities and know which are most relevant to what Designing self-reports Surveys & Polls - Method of posing questions to people - Typically about social/political opinions - Phone, interview, questionnaire,, or online… Open-ended format - Information rich - Hard to code; lengthy; resource heavy Forced-choice format - Much more familiar with - Limit response range - May limit way you can contextualize complex opinions - Easier to score TYPES o Multiple choice, Response Pairs o Likert Scale 1: Strongly disagree 2: disagree 3: undecided 4: agree 5: strongly agree Questions are statements o Semantic differential Two ideas that are opposite of eachother Put on each side to show which side person is leaning more towards Male----------------------------Female Traditional--------------------Modern Casual--------------------------Formal Fancy---------------------------Plain o Example: rating a movie on Netflix: 0-5 stars Self-Report Construct Validity - Problem with Likert and Semantic Differentials o Unclear how subject interprets scale o My 1 star and your 1 star may not be the same. o What makes me disagree and what makes you disagree may not be the same. o Scale may differ between people - Question Wording Effect o Leading Questions Suggests a particular answer is desired Try to sway response Wants to confirm a belief/idea Makes one answer seem more desirable Now that you’ve seen how you can save time, would you want to buy are product? of course you will, because you think it saves time You cheated on the exam, didn’t you trying to confirm idea they already believe: affirmation of their belief; not asking a question o Double barreled questions Actually asking two questions combined into one When person responds, you don’t know which question they are answering to Do you agree that the 2 ndamendment to our US constitution guarantees your individual rights to own a nd gun and that the 2 amendment is just as important as your other constitutional rights? Wasn’t that guitar riff amazing, and weren’t the song lyrics clever? With a forced choice format; you don’t have the flexibility to elaborate on your answer o Negatively worded questions More complicated to interpret Slow subject down or confuse them Kagay, 1994: 20% of Americans were holocaust deniers Survey question: does it seem possible or does it seem impossible to you that the Nazi extermination of Jews never happened? Do I agree that it’s impossible, or do I agree that it never happened? The wording here is extremely confusing for the subject Abortion should never be restricted Agree vs. disagree People who disagree have to think about it, and may respond incorrectly o Question order Priming effect You are particularly sensitive to a particular issue after being exposed to it in some way Subconscious Contrast Effects o Feel like you have said no too much, so need to say yes Assimilation Effects o Want to seem more consistent in answers Civil union and gay marriage approval o Gay marriage first question If I disagree with gay marriage, I must agree with civil union If I agree with gay marriage, then I obviously agree with civil union o Legal agreements first question Two questions answer about the same answer turnout What is the most important problem facing the nation? Do you approve or disapprove of the way the president is handling his job? VS. Do you approve or disapprove of the way the president is handling his job? What is the most important problem facing the nation? o The second is the better order o With the first one, you think about the problem, and then don’t like the president if he is not addressing the issue o The second first asks your opinion without swaying opinion, and THEN asks what your opinion of the most important issue ASK WHAT YOUR MOST PRESSING QUESTION IS, FIRST! o If you did want them to take other things into account, and to influence their answers, then yes- put the other questions before it. People are generally not aware of this, and so fall for being swayed in surveys, marketing, advertising etc. o Wording Sensitivity Controversial material People think that their answers will look bad on them, Trying to ‘look good’ in front of others MUST PHRASE QUESTIONS IN LEAST CONTROVERSIAL WAY o Ambiguous Meaning What something means to you, may mean something different to someone else Have you received medical care recently? o What classifies medical care? o Dentist, picking up meds, visiting an ER, chiropractor, therapist? o Loaded Question Question embeds an assumption Legal question often have these If lawyer, trying to catch someone who is domestically abusive Have you stopped beating your wife? INSTEAD OF Do you beat your wife? then we know if he beats his wife based on the answer It’s hard to get the true opinion of something if the question assumed something o Leading vs. Loaded Do you agree with the rest of the class that the professor has been doing a poor job teaching the course? Makes you want to fit in with the rest of the class, and do what the rest of them are doing Do you agree that the professor will continue to do a poor job teaching this course? Already has the assumption that the professor is doing a bad job BETTER: What do you think of the professor’s performance teaching the course? - QUESTION WORDING MATTERS o Completely change perception of the event based on question you ask o Loftus and Palmer, 74 – all people watched same clip of cars crashing How fast were the cars going when they ___into each other Some people have word smashed, some have word collided, bumped, hit, and contacted Smashed – 40.8 mph Collided – 39.3 mph Bumped – 38.1 mph Hit – 34.0 mph Contacted – 31.7 mph o Another time asked two groups [smashed vs. contacted] if broken glass – smashed said yes, contacted said no Accurate Responding - Respondent Shortcuts o Response Set Aka; nondifferentiation Responding to all items the same SEI’s Solution: Reverse worded questions If paying attention to the questions, the answers to the opposite questions will be opposite of each other Beware of NEGATIVE WORDING o Acquiescence Saying yes all the time Opposite is nay-saying [saying no all the time] o Fence-Sitting Playing it safe Answer in the middle Not swaying to really one answer or another Solution: remove the neutral even numbered force-choice some people are neutral - Socially Desirable Responding [faking good] o Don’t actually share what your opinion is, but go along with popular belief o Opposite is faking bad o Faking good or bad is socially desirable responding Trying to appear a specific way to peers; to fit in o The hipster who wants to go against anything popular Says against everything, even if they like it o Solution: make survey anonymous or confidential People are less likely to lie when compared to peers Peer comparison Instead of how much do you drink Ask do you drink more or less than your friends o Friends drink 5 beers; I drink more Peer reports Their friends won’t lie about what they do Sensitive wording Did you complete your civic duty to vote? o Makes you feel pressured to lie Did you votes? o Fair; but could feel bad still Where you able to vote or did something come up preventing you? o Gives you a reasonable explanation; telling them its OK - If you ask your people to self-report ANYTHING, you have to take into account how you are asking them; because they will change what they are saying o If you don’t construct validity goes down - Accuracy of Reports o People actually struggle reporting reasons for their behavior o We struggle to explain HOW or WHY we do something Nisbett & Wilson’s 1977; Nylon stocking study Women’s favorite stockings was always the last one they felt Had a million reasons why; when none of them were different It was just the last one they remember o People have inaccurate memory of events Confidence does NOT equal accuracy of memory Can be manipulated Car crash study talked about earlier You have a high confidence of where you were, what you were doing etc. during 9/11 people say they remember seeing the 1 plane hit on the news; but the news wasn’t st covering the incident until after the 1 plane hit o An example of us actually not knowing what we are doing How many steering movements does it take to change your car from the slow to fast lane One [left], two [left, right], three [left, right, left], four [left, right, left, right] o Left; right; straighten out Good questions vs. bad questions - You find out that you ‘friend’ made something up about you… o Putting quotes around ‘friend’ makes it a leading question o Swaying you into believing your friend is bad o Assumption would be stating that they are not your friend anymore - Suppose your guy suddenly lost his job and started balding. Would it matter? o Two options give limited options o Double barrel question asks two questions in one! - You find small talk useful, fun and easy o Double barrel - You find that social situations don’t often make you feel award or comfortable o Negatively worded o Ambiguous what about it makes me awkward/what does comfortable mean, what social situations o Double barrel are we asking comfortable or awkward - I have no problem cheating on my partner if I knew I could get away with it [Likert scale] o Likert scale is a bad way to measure this; as many will answer in the middle, conform to societal answer etc. Dear News Media: When reporting poll results, please keep in mind: Statistical Validity o If two poll numbers differ by less than the margin of error, its not a news story Construct Validity o Scientific facts are not determined by public opinion polls External Validity o A poll taken of your viewers/internet users is not a scientific poll Construct Validity o What if all polls included the option ‘don’t care’? Observation Construct Validity Observational Research - Systematic recording of behavior Observer Bias - Expectations influence interpretation - Sports: always think other team deserved the flag; don’t think your team did Observer Effects - Expectancy effects - Observer changes the behavior of the observed - Reactivity; people know they are being watched so they act differently - The Truman Show - Rosenthal and Fode ‘63 o Some rats struggle with mazes some have been genetically modified to do well in mazes o Students who thought their rats had been genetically modified to do well in mazes, did well in mazes; those who thought their rats were dumb/struggle with mazes actually struggled; even though all rats were genetically the same - Pygmalion effect or Self-fulfilling effect o Teaching Have some smart students, some dumb students Kids who teacher thought was smart, performed better and vice versa Solutions - Codebooks o Strict guidelines for observers to use o Specifies how behaviors should be categorized o Allow observers to be more objective about things that could be biased o Ignore behavior that doesn’t fall in line; stick to book; go back and add behavior to book if you need to - Multiple raters o Better reliability o If everyone sees the same thing; have consistency across measurement ASSUME have more valid measure - Masked Designed o Blind: When the experimenter does not know which conditions are what o Double blind: when neither the experimenter nor the subject knew which conditions are what Dealing with Reactivity - Being observed changes participant’s behavior - Solutions o Unobtrusive observation One way mirror o Wait it out If you are part of your environment for long enough, they won’t notice you anymore – then can observe; Acclimate o Trace Behaviors You don’t have to see the actual thing you want to measure in order to measure it; by measuring something effected by the action you are looking at, can help you to measure it without affecting the participant’s behavior