Research Methods Final Exam Study Guide PDF

Research Method Final Exam Study Guide Introduction in a research method paper Background and research Purpose of study Hypotheses P-Value 0.05 or less is statistically significantly 0.05 or more is not statistically significantly Reliability Replication & consistency...

Research Method Final Exam Study Guide Introduction in a research method paper Background and research Purpose of study Hypotheses P-Value 0.05 or less is statistically significantly 0.05 or more is not statistically significantly Reliability Replication & consistency + reliable, the - error = better measure - reliable, the + error = poor measure Affect Reliability Confound Variables Operant definitions Physical setting Wording of questions Respondents mood Types of Reliability Test-retest if the scores are consistent or similar then it is more reliable ○ Example: Students' math test scores are 85-95 on the first test and 87-94 on the second test, showing high reliability due to consistent results. Split-Half if subset of test ○ Example: Test is split into two parts: If a student scores 85 on the first part questions and 86 on the second part questions, the test shows high split-half reliability due to consistent results. Inter-rater (inter-observer): observed the same ○ Example: Two teachers grade the same set of student essays. If both give similar scores (e.g., 90 vs. 91), the grading system has high inter-rater reliability Validity Are we measuring what we are supposed to measure? Types of Validity Construct if it is the correct concept or idea used ○ Example: measuring happiness and including joy questions / high construct Face-validity if it looks right ○ Example: a spelling test having words to spell / high face-validity Criterion-related validity if it is real world outcomes ○ Example: those who pass the driving test are good drivers / high criterion-related Increase Validity Using repeated measures Same number of participants Within- Subjects Participants would answer the same ○ Example: If participants take a memory test twice, they are likely to interpret the questions the same way each time, increasing validity. Internal Validity Cause and Effect Independent: what you manipulate Dependent: what you measure High validity: independent caused change in dependent variable ○ Example: The researcher randomly assigns participants to sleep 4 or 8 hours, keeps their diet and activities constant, and finds differences in memory are only due to sleep. Low validity: other factors explain results ○ Example: If participants who slept 4 hours also drank caffeine, it’s unclear if the caffeine, not sleep, caused the memory change. External Validity Generalized research findings to other settings, population, or times Take your sample and apply it to the outside ○ Example: When covid, there were a lot of studies conducted from Israel. Those studies extended the covid findings to the rest of the world. Threats to Internal Validity Selection Threat ○ Groups compared are not equal at the start of a study ○ If group a has more experience than group b in a training study, biased Maturation Threat ○ Change in participants over time that happen naturally, rather than as a result ○ Participants grow, learn, or develop on their own Attrition Threat ○ Participants drop out of a study ○ In a fitness program, if participants who find the workouts too hard quit, the final results will only include participants who were already fit Instrumentation Threat ○ The tools or methods used to measure change over time, influencing results ○ If a study starts with simple language but then uses a complex version, participants might not reflect their actual performance. Testing Threat ○ Taking a pretest affects performance on the posttest affecting results ○ Math problem, if participants were asked to solve it in a pretest they might improve on the posttest just because they have already practiced. Statistical Regression Threat ○ Participants are selected based on significant scores error ○ Athletic performance selected only the poor, but this could have been because of fatigue or lack of focus. Random Assignment vs. Random Selection Random assignment: different groups to ensure internal validity by reducing biased Random selection: sample representative a broader population Tuskegee Study Participants were not informed about the true nature of the study They were denied effective treatment even when it became available This study observed untreated syphilis Milgram Study Participants were deceived about the purpose of the study This study tested authority obedience to administer electric shocks Deception Intentionally misleading participants (not telling true purpose or procedures) Debrief Telling participants about purpose/benefits of the study after the study Independent What the research manipulates ○ Example: type of music, sleep time Dependent What the researcher measures or observes, the outcome ○ Example: performance, concentration, what they are looking at Confounding Variables Influence our outcome Stress How much you study Correlation ○ r T-test ○ t ANOVA ○ F Research Designs Observational Design ○ Researchers observe in their natural environment ○ No manipulation or interference ○ Example: Observing how children behave on a playground Correlational Design ○ Pearson’s Correlation ○ Looking at two or more variables and looks for relationship between them ○ Example: hours studied and test scores ○ Types of correlation: both increase or both decrease that's positive ○ Correlational design: what analysis are we going to run? Correlation ○ No correlational - not correlational r = 0 ○ r Experimental ○ Manipulating one or more independent variables to see the effect on dependent ○ Casual relationship ○ Random assignment to different groups ○ Example: testing if teaching method (IV) improves student performance (DV) Quasi-Experimental ○ Lacks random assignment to groups ○ Pre existing groups like male vs. female, pre-test vs. post-test ○ Example: comparing academic performance between male and females Longitudinal Design ○ Studying the same group of individuals over a long time to see change or develop ○ Example: Studying the long-term effects of smoking on health by tracking a group of smokers over 20 years. Archival Data ○ Existing data that has already been collected ○ Example: Analyzing historical records to understand patterns in crime rates over decades. Case Study ○ Detail and in-depth study of a single individual ○ Example: Studying a person who has a rare mental disorder to understand its symptoms and impacts. Type of Data Qualitative Data ○ Word or a statement ○ Example Word/Label: "Color," "Gender," "Favorite Food," "Mood" Quantitative Data ○ Expressed through numbers ○ Example Word/Label: "Age," "Height," "Weight," "Temperature" Run Data r correlation ○ Two variables to see if there is a relationship ○ Example: height and weight t-test ○ Comparing the means of two groups f anova ○ Comparing the means of three or more groups