Psych 225 Lecture 11Outline of Last Lecture ExamOutline of Current Lecture Current LectureSampling distribution questions are always about degrees of freedom. *piece of advice from the exam*VP StudiesSlide: Doerksen and Shimamura-Expt. 1 & 2: P’s saw 1 of 2 sets of words~emotional and neutral1) How did Expt. 1 and 2 differ?Answer: source info was spatially separated from stimulus words in second study but completely integrated in the first study 2) Measures? Answer: recognition and source memory and free recall-when people don’t know they are going to be tested on something it is called: incidental-when people were told that they were going to be tested on something it is called: intentional -recalled more emotional words in free recall than neutral; recognition of words for both neutral and emotional words was high; recognition of source memoryThese notes represent a detailed interpretation of the professor’s lecture. GradeBuddy is best used as a supplement to your own notes, not as a substitute.How did they explain results of Expt. 1 and 2? What does convergence of results across Expt. 1and 2 suggest?Which explanation of emotional memory did they address in Expt. 3?Answer: semantic clustering ~relation between the words did not facilitate memory of the source Kissler, Herbert, Peyk and Junghofer-cortical responses to emotional words during reading How quickly have P’s shown ERP reactions that indicate differentiation between words that are more/less interesting (semantic processing)?Answer: Really quick! 250 msWhat were P’s shown/stimuli?What dimensions did K match stimuli?Answer: concreteness, length, etc. -found that cortical responses were faster and larger for emotional words (+/-) words as compared to neutral words and incidental recall was higher for emotional versus neutral words -boost in recall in emotional words could be traced to early semantical processing Zimmerman and Kelley-effects of emotionality on memory predictions versus memory performanceExpt. 1~trail 1: cued recall~trial 2: cued recall Did valence of pairs affect cued recall? Calibration? Trial 1P’s were overconfident in negative words and were also a slightly overconfident for neutral words Did valence of pairs affect cued recall? Calibration? Trial 2P’s were underconfident when they were negative words and even more underconfident in neural words How did Expt. 2 differ from Expt. 1?-utilized a free recall instead of cued Chapter 14: Questionnaires, surveys, and sampling-talking about pure survey reports: called respondents-experimental research: called participants Funneling-open-ended, free response-fixed format/alternative -thus start broadly and get more narrow -this is labor-intensive work for researchers because they have to start with the free responses and categorize them-can make for better reading and provide interesting answers: that’s why they are still used even though it is a lot of work -you can also get ideas for future research-in interviews, etc. it can enhance feelings of repoir understanding; P’s may appreciate getting to express themselves-can have an orienting function: direct studyFixed Format/alternative-Likert: continuous; “to what extent to feel this way”-semantic differential/bipolar adjective: may be used for comparison warm v. cold; happy v. sad-multiple choice: ex. BEM depression inventory -much less labor intensive Advice1. Don’t reinvent the wheel~use measures with demonstrated reliability and validity Reliability of a test-extent to which we are confident that scores would recur-data reflect true scores, not error-results are consistent/tool is consistent; how confident are we that the findings will recur -as reliability increases, the more powerful the test isif effect out there, we have a better chance of demonstrating it when we have a more reliable indicator Strategies for assessing reliabilityA. test-retestP’s take test twice; expect +r (high degree of correspondence) B. split halfP’s take test once. Expect +r of scores on odd & even # items; look at degree of correspondence between scores within same testDon’t recommend doing first half and second half and see how they correlate; people may increasingly fatigued or bored; look at odd/even or randomized C. parallel forms2 tests with different, but equivalent itemsP’s take from A & B. Expect +rOr, administer A to Group 1 and B to Group . expect similar distributions; if same people you should have high correlation, if different people you should have similar distributions D. Cronbach’s alpha Internal consistency: how well do items of a measure hang togetherBased on average correlation (r) of items on a test Often, as # of times increases, Cronbach’s alpha also increasesExample: Zakrisson’s 15 item RWA: much bigger numbers because sum used; is RWA related tendency to support of jurors? The answer is yes! Consistent with anti-defendant bias~the test produced a null result when we used a less reliable