Slides - Class 6 - Sec02 - Notes.pdf

Class 6 – Science of psychology Today’s plan: - Finish up inferential statistics - Research study practice 16 Sep. 2024  Inferential Statistics IV levels Alcohol No Alcohol “Fake” Alcohol Group Experimental Control Placeb...

Class 6 – Science of psychology Today’s plan: - Finish up inferential statistics - Research study practice 16 Sep. 2024  Inferential Statistics IV levels Alcohol No Alcohol “Fake” Alcohol Group Experimental Control Placebo DV Attractiveness Attractiveness Attractiveness mean = 8.5 mean = 4.8 mean = 6.2 How do we know that 8.5 is actually “different” than 6.2 or 4.8?  Inferential Statistics We can calculate IV levels Alcoholour confidence No Alcoholthat the difference “Fake” Alcohol we observe Group in the sample is accurately representing the Experimental Control Placebo difference we would see in the population. DV Attractiveness Attractiveness Attractiveness mean = 8.5 mean = 4.8 mean = 6.2 0 10  Inferential Statistics One way to express that confidence… P –value Often misused! Be careful! If the null hypothesis (H0) is correct [there is no effect]… the probability that we mistakenly reject the H0 The chance we have a false positive. Often  threshold of.05 (or 5%) Less than that and we call it ‘statistically significant’  Inferential Statistics Statistical significance ≠ Proof  There is still some chance that this is a false positive [The null hypothesis is true, but we rejected it by accident]  Statistics Summary A. Descriptive statistics Describe characteristics of our sample Central tendency and variability B. Inferential statistics Use sample characteristics to make inferences about population Are group differences in sample the same as in overall population? Key concept t  Conclusions about Methodology in Psychology Psychologists have several goals in mind when doing research They also approach problems from multiple perspectives Different research methods are used to address each goal Each research method (describing, predicting, explaining) has its strengths & weaknesses Important to be aware of what we can/cannot conclude when using each method (and how statistics support these conclusions)  Guide to reading a research article 1. Research question 2. Hypothesis 3. Research methods 4. Independent variable(s) 5. Dependent variable(s) 6. Procedure overview 7. Methodology strengths 8. Methodology weaknesses 9. Results summary 10. New information generated 11. Real-world applications Things to look out for in assessing a method’s validity Correlational studies Lack of control group Unblinded participants (and experimenters) Roediger & Karpicke. (2006). Psychological Science: 17(3): 249-255. Research motivation  How do you like to study for an exam? Roediger & Karpicke. (2006). Psychological Science: 17(3): 249-255.  Re-watch recorded lectures Read notes Re-write notes Review class slides Flashcards/Quizlet Discussing material in a group Quizzing each other in a group Writing out answers to practice questions Something else? Research motivation  Does repeated testing lead to improved learning?  Previous studies suggested YES but…  most were lab-based studies  Was it testing or just more exposure to the material? Roediger & Karpicke. (2006). Psychological Science: 17(3): 249-255. General research approach  Following a reading passage… Self-study or post-testing  Then recall exam What would be the null hypothesis? What would be an alternative hypothesis? Roediger & Karpicke. (2006). Psychological Science: 17(3): 249-255. Research hypothesis H0: There will be no difference in retention between self-study and post-testing. HA1: There will be a difference in retention between self-study and post-testing. [two-tailed  either could be greater] HA2: Post-testing will result in higher retention than self-study. [one-tailed  only one direction is specified] Roediger & Karpicke. (2006). Psychological Science: 17(3): 249-255. Research method 120 Washington University students Read one of two passages ‘The Sun’ or ‘Sea Otters’ 30 idea units in each 7 minute reading period then 7 minute study period Self-study OR recall test Then repeated with other study method & other passage Order of learning conditions & passages were counterbalanced Roediger & Karpicke. (2006). Psychological Science: 17(3): 249-255. Research method Counterbalancing Roediger & Karpicke. (2006). Psychological Science: 17(3): 249-255. Research study design Which one is Study types LEARNING CONDITION in this study? BETWEEN-SUBJECT WITHIN-SUBJECT Different participants Same participants complete complete each condition all conditions Comparisons between groups Comparisons within groups of different individuals of the same individuals Roediger & Karpicke. (2006). Psychological Science: 17(3): 249-255. Research study design 2 x 3 mixed-factorial design Roediger & Karpicke. (2006). Psychological Science: 17(3): 249-255. Research study design 2 x 3 mixed-factorial design Learning condition = within-subject design [2 levels] Final test delay = between-subject design [3 levels] After the study/test session, there was a final recall test given 2nd independent variable = delay until final test 5 minutes, 2 days or 1 week Measurement of learning  proportion of idea units recalled Roediger & Karpicke. (2006). Psychological Science: 17(3): 249-255. Study results Roediger & Karpicke. (2006). Psychological Science: 17(3): 249-255. Conclusions Different effects on learning Immediate retention was better with the re-study group Prior testing improved performance on delayed tests Roediger & Karpicke. (2006). Psychological Science: 17(3): 249-255. But what about repeat study sessions? Experiment 2 S = STUDY T = TEST 3 studying conditions S–S–S–S S–S–S–T S–T–T–T Final test  5 minutes or 1 week later Additional DV  survey to assess… -interest level and readability of passage -# of times passage was read in time -prediction of recall 1 week later Roediger & Karpicke. (2006). Psychological Science: 17(3): 249-255. Research method 180 Washington University students 3 x 2 between subjects design 5 minute study periods, 10 minute testing periods Roediger & Karpicke. (2006). Psychological Science: 17(3): 249-255. Experiment 2 Results 77% recall 70% recall Roediger & Karpicke. (2006). Psychological Science: 17(3): 249-255. Experiment 2 Results Roediger & Karpicke. (2006). Psychological Science: 17(3): 249-255. Experiment 2 Results On 5-min test, recall was correlated with repeated studying On 1-week test, recall was correlated with number of prior tests taken Confirmed with statistical analysis (ANOVA)… repeated studying produced short-term benefits repeated testing produced greater benefits on the delayed test Roediger & Karpicke. (2006). Psychological Science: 17(3): 249-255. Experiment 2 Results More forgetting in the pure studying group Roediger & Karpicke. (2006). Psychological Science: 17(3): 249-255. Conclusions Immediate testing promoted better long-term retention of material from a prose passage. Not just due to re-exposure to material Repeat studying improved retention after 5 minutes Also inflated confidence in ability to remember Roediger & Karpicke. (2006). Psychological Science: 17(3): 249-255. Discussion questions 1. What is the take-home message of the study? 2. How would this information inform your own studying? 3. What are some strengths of the research study design? 4. What are some limitations of the research study design? Roediger & Karpicke. (2006). Psychological Science: 17(3): 249-255. Methodology Strengths Well-designed experimental study Counterbalancing to negate order effects Within subject design for study types accounts for individual variability in retention Between subject design for final test delay prevents repeat testing from contributing to retention Reported no differences across passages and study order on the initial recall test Design biased against their intervention (testing) by using free recall test [not exposed to all of the material but the re-study group was] Roediger & Karpicke. (2006). Psychological Science: 17(3): 249-255. Methodology Weaknesses Counterbalancing averages out order effects but could mask interesting trends How representative is the sample population? Participant incentivization could affect results In Exp 2 – testing periods were longer than study periods Participants tested in small groups – not a typical real-world classroom Authors don’t discuss their own ideas about study limitations Roediger & Karpicke. (2006). Psychological Science: 17(3): 249-255.  **Practice on your own** Are sleep deprivation & university performance related? 1.Determine a specific (directional) predictive hypothesis 2.Graph the proposed relationship 3.Design a study to investigate the hypothesis using the predictive method (correlational study) 4.Design a study using the experimental method. Identify the IV & DV, then operationally define them. 5.What are potential third variables that you hope random assignment will eliminate that might have been influencing your results when using the predictive method? 6.What factors must you consider in deciding what method to use to study the relationship between sleep deprivation & classroom performance? For Wednesday by 10am…  3.1  Interactive Video Quiz 3 -Look under WEEK 3  For Wednesday by 10am -It disappears at 10am.