Psyc 2950 Exam 2 Study Guide
Psyc 2950 Exam 2 Study Guide PSYC 2950
Popular in Experimental Psychology
Popular in Psychlogy
This 7 page Study Guide was uploaded by Marcela Leon on Friday February 26, 2016. The Study Guide belongs to PSYC 2950 at University of North Texas taught by Alexander Yu in Spring 2016. Since its upload, it has received 114 views. For similar materials see Experimental Psychology in Psychlogy at University of North Texas.
Reviews for Psyc 2950 Exam 2 Study Guide
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: 02/26/16
Exam 2 Study Guide Chapter 5 Review Reliability= consistent results The 4 types of reliability: For all of these, we want STRONG POSITIVE CORRELATIONS. 1. Test-retest: keeping the relationship between scores steady (consistent) over time Conduct one test, retest and expect the same results The longer the time interval between tests, the less steady 2. Equivalent-forms: steady scores obtained when conducting equivalent types of tests Ex: The SATs- same exam, different forms 3. Internal Consistency: How steadily a variable is measured throughout a test Ex: When taking a survey to test someone’s personality, multiple questions will be given to test certain traits that will allow us to draw accurate conclusions. Because conclusions cannot be drawn from a single question, various questions assessing the trait will be given. We expect the answers to the questions to be in agreement with one another reflecting a particular personality type. Internal Consistency Reliability is the agreement consistency of those questions. Internal Consistency Statistic- Coefficient alpha (Chronbach’s alpha)- the most common scale for measuring internal consistency Value of .70 or higher for good internal consistency 4. Interrater: how much agreement there is between 2 or more raters/judges Ex: 4 judges at a pool diving competition: we are looking for scores that look more like : 8 8 9 8 Not: 4 9 1 6 We want consistent agreement between the scorers Validity= accuracy Must be reliable in order for it to be valid. 4 Types of Validity: 1. Content validity (face/content): the question clearly reflects what it is measuring. Ex: When trying to determine if a person has depression, an example of a good content question would be, “Do you have suicidal thoughts?” as opposed to, “Do you have weird thoughts?” 2. Criterion-related validity: How reliable past performance can be in predicting future performance 2 types of criterion-related validity. Only difference between the two is: time. Concurrent: The scores of a new test compared to an established test. Predictive: The scores of a test predicting degree of future academic performance compared to the scores obtained later on actual academic performance. Ex: High school juniors’ SAT scores compared to their academic performance in college 2 years later. 3. Convergent Validity: related variables are tested and the scores are compared Ex: Testing self-worth. Variables testing will include: confidence, self-security, and self-esteem 4. Discriminant: shows that the scores measuring certain variables on one test are not related to the scores on a different test (what should not be related is actually not related) Weak and negative correlation is desired Ex: Scores measuring energy levels should not be correlated to scores on a sexuality scale. What is parameter? Parameter is a numerical feature of the population. Ex: The average salary of the population is $20.50 per hour. What is statistic? Statistic is the numerical feature of the sample. Ex: The average salary of the sample you are testing is $21.20 per hour. Sampling Error is the difference between the parameter and the statistic. So for the example above, the sampling error is $.70. 21.20 – 20.50 = .70 Sampling is important because gathering data from the entire population is very difficult. With a sample, you acquire data representative of the population. Sampling Method: HOW the sample is selected. The 4 goals: 1. Representative sample: 2. Equal Probability of Selection Method (EPSEM) 3. Proximal Similarity 4. Inferential Statistics Representative Sample: The sample reflects the population in all aspects. (Gender, age, race, etc.) Equal Probability of Selection Method: Everyone has the same opportunity at being selected to be part of the sample. o EPSEM is important because it guarantees a representative sample. Proximal Similarity: the ability to apply data to people, environments, and/or contexts similar to those tested. Inferential Statistics: the ability to make inferences based on our sample about the population 5 types of Sampling Methods 1. Simple Random Sampling: choosing samples with replacement Ex: You have a jar full of strips of papers with names on them. You pull one strip out, call out that name, then put the strip back in the jar, and repeat. 2. Stratified Random Sampling: dividing your population into groups (strata) and selecting random samples from each group Ex: dividing your population by race Proportional: ratio of samples matches population ratio Disproportional: ratio of samples does not match the population ratio Proportional Example: In our population we have 100 people total, of those 120, 40 are Black, 30 Hispanic, and 30 Asian. We expect our sample of 10 people to have 4 Black, 3 Hispanic, and 3 Asian. 3. Cluster Random Sampling: Instead of picking individuals, the researcher picks random clusters of people Ex: could include schools, neighborhoods, sport teams, etc. One-Stage: Researcher includes everyone in the cluster Two-Stage: Researcher randomly selects individuals from clusters 4. Systematic Sampling: Sampling Interval (k): population divided by Sample Size desired Steps to perform Systematic Sampling: 1. Calculate the Sampling Interval (k) 2. Randomly select an individual between 1 and k 3. Choose every k element Ex: Population size is 200. Desired sample size is 20. 1. 200/20= 10 k = 10 2. The random individual you choose between 1 and 10 is 4. 3. Now you will have person #4 as your first random individual and then you will select every 10 person after that. So your nd 2 individual will be #14, then #24, #34, etc. Warning: Periodicity- potential threat for systematic sampling if the population contains a cyclical pattern Ex: if the list of the population is in some type of order such as, by income, age, grades, etc. 5. Nonrandom Sampling: Sample is selective (weak method) o Not representative of the population, but is sometimes necessary Convenience: people that are immediately available Quota: A certain # of the types of people wanted is established Ex: You want a diverse population. You want exactly 25 Black, 25 White, 25 Hispanic, and 25 Asian people. Purposive: People who meet the criteria Ex: o Must be in college o Between the ages of 18-25 o Female Snowball: One individual leads you to identity other individuals that are more unique and harder to find Ex: split-brain patients Random Selection Random Assignment Random Selection: using a random sampling procedure to select participants for a representative sample Random Assignment: Participants are randomly assigned their roles for the study Ex: Randomly selecting individuals to be a part of the experimental group or the control group How big should a sample size be? The sample size should be as big as possible. In the case that your population is only 100 or less, you should use the entire population as your sample. Chapter 6 Review Validity = Accuracy Research Validity: How accurate the conclusions drawn from an experiment are. 4 types of Research Validity: 1. Statistical Conclusion Validity: How accurately we can conclude that with every change in the IV there is a change in the DV. This is known as covariance. 2. Construct Validity: How accurately our operations represent the population we are conducting the study for. Clearly define IV, DV, population, and context. Ex: If you are doing a study on musicians and memory, your operations should include characteristics that represent a musician. 3 sub-types of validity that contribute to Construct Validity: Content- question reflects what it is measuring Criterion- past performance can accurately predict future performance Convergent- related variables are tested and scores are compared 2 Major Threats to Construct Validity: Participant Reactivity: The participant may want to be seen as someone who performs well or will want to impress the researcher. Demand Characteristics: The participant may know or may think he/she knows what the researcher is testing and will act/perform in the way he/she believes one is expected to. 3. Internal Validity: The degree to which we can conclude accurately that the Independent Variable causes the Dependent Variable. 7 Threats to Internal Validity: 1. History: Event that may affect outcome of experimental treatment Differential History: Only one group is affected by history event 2. Maturation: physical and psychological changes that occur over time within an individual 3. Instrumentation: instrument used to measure or asses the DV is changed 4. Testing: the participant takes test that he/she has already taken before, therefore the scores are affected 5. Regression Artifact: the phenomenon in which participants’ extremes scores move towards the mean after the participants repeat the assessment 6. Attrition: incompletion of assessment by participant drop-out or no-show 7. Selection: using different criteria to place participants in different treatment groups Not random assignment 4. External Validity: The degree to which the results of a study can be applied to other people, settings, time, outcomes, and treatments. 5 Categories of External Validity: 1. Population: The degree to which the results of a study can be applied to the larger population. 2. Ecological: Degree to which the results can be applied to other environments 3. Treatment Variation: Degree to which the results can be generalized to different forms of treatment 4. Outcome: Degree to which the results can be generalized to other related DVs 5. Time: Degree to which results can be applied across time Ex: Average education level tested in 1910 in the US compared to Average education level in the US in 2010. The results of 1910 could not be generalized to the study conducted 100 years later. Limits to External Validity: When there is no random selection or random sampling Factors due to chance Relationship between IV and DV varies across different levels of other variables Ex: use of steroids (IV)- hair growth (DV) There will be a difference in results when looking at exclusively men and then exclusively women
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'