Log in to StudySoup
Get Full Access to Towson - PSYC 314 - Study Guide
Join StudySoup for FREE
Get Full Access to Towson - PSYC 314 - Study Guide

Already have an account? Login here
Reset your password

psyc 314

psyc 314


School: Towson University
Department: Psychology
Course: Introduction to Research Methods in Psychology
Professor: Brianna stinebaugh
Term: Spring 2016
Cost: 50
Name: Exam 2 Study Guide
Description: These notes cover chapter 5 (not covered on first exam) to chapter 11
Uploaded: 11/05/2016
13 Pages 137 Views 0 Unlocks

Why should we use a match design?

What Happens When You Don’t Control These Extraneous Variables?

What is a Hypothesis?

Research Methods  Exam 2 Study Guide Chapter 5: Continued I. Correlational Design 1. Degree of relationship between 2 variables a. Identify any relationship 2. No altering behavior/ variables a. Already existing 3. If we find a correlation, then any changes in one variable will be associated with  changes in the 2nd variable 4. Explains futuDon't forget about the age old question of ■ What kind of drugs do you take?
If you want to learn more check out smad 150
We also discuss several other topics like o want to know the answers to very interesting questions, -where do babies come from?
We also discuss several other topics like instinctive drift definition
If you want to learn more check out marginal private cost
We also discuss several other topics like unlv physics
re behavior A good at predicting  II. Pearson r Correlational Coefficient 1. r is the strength of the relationship a. -1.00 to + 1.00 b. Take absolute number to get the strength c. Closer to 0.00 the weaker the relationship is d. If exactly -1.00 or +1.00 it’s perfect 2. Direction of the relationship a. Positive or direct   Variables move in same direction b. Negative or inverse  Variables move in opposite directions c. No relationship has a value of 0.00 III. Scatter plots 1. Visual, dotted, graph 2. Each dot is one participant in a study 3. Lines show direction of relationship a. Positive, negative, no relationship 4. Linear regression lines a. Line of best fit b. Equation that shows the relationship between 2 variables 5. Curvelinear relationship a. U shape or inverted U shape b. As one variable increases, the second variable increases to a certain point, and then decreases 6. Multiple correlations a. Stat analysis when you have 3 or more variables b. Use R 7. Partial correlational design a. Hold one variable constant, while analyzing the relationship between other  variables IV. Causal ModelResearch Methods  Exam 2 Study Guide 1. Predicts causal relationships between variables 2. Pathanalysis a. Most commonly used b. Potential cause and effect for future research 3. Cross-Lagged Panel Analysis a. Variables over time V. Quazi-Experimental Designs 1. Set up like a true experiment, BUT lack a few essential elements that a true  experiment needs.  a. Manipulation  b. Random sampling 2. Compares groups of people a. Looks over time 3. Low in internal validity 4. High in external validity 5. Treatment conditions based on already existing antecedents/behaviors 6. No cause and effect VI. 5 Types of Quazi 1. Ex Post Facto Study a. Most common b. Looks at naturally occurring behavior characteristics  Looks at the effects that occur- predicts future behavior c. No manipulation d. Preexisting characteristics- own treatment groups  e. Very controlled f. True experiments use random assignment and no known differences g. Use when looking at  Personal characteristics (age/gender)  Different life events (marriage divorce)  Psychological functioning/personality on behavior (mental illness) 2. Non-Equivalent Groups Design a. Compare the effects from different treatment conditions on already existing  groups  Example: 2 classrooms and 2 different teaching styles b. No control over random assignment  c. No control over who gets what treatment 3. Longitudinal Designs a. Measures behavior on the same participants over a period of time  How time influences behavior b. No manipulation c. Interested in growth and development d. Can last days, weeks, months, or years e. Can have participants drop outResearch Methods  Exam 2 Study Guide 4. Cross-Sectional Design a. Very similar to longitudinal b. Looks at participants that are already at different stages and comparing them  at 1 time c. Requires more participants; faster d. No manipulation/altering e. No cause and effect  5. Pre Test/ Post Test Design a. Measuring behavior before and after an event and compare b. Use in labs and natural settings c. Low in internal validity because of 3rd variable problems d. Unethical to have a control group Chapter 6: Hypothesis I. What is a Hypothesis? 1. Hypothesis The thesis, or main idea, of an experiment or study consisting of a  statement that predicts the relationship between at least two variables.  a. Educated guess  2. Stated in a If….. Then …. Statement  a. Another way of expressing the potential relationship between the antecedents  and the behaviors to be measured. II. Non experimental vs true experimental 1. Nonexperimental hypothesis  a. Goes with non experimental research design b. A statement of predictions of how events, traits, or behaviors might be related  but not a statement on cause and effect  2. experimental hypothesis  a. Goes with true experimental research design b. A statement that is a tentative explanation of an event or behavior c. Predicts the effects of specified antecedent conditions on a measured behavior  III. Directional vs Non Directional 1. Non directional  a. State a prediction that two variables will differ but not in any direction  b. More likely used for a statistical significance (p value) 2. Directional  a. State one variable will do better than the other IV. Basic criteria for formulating a hypothesis 1. Stay away from personal beliefs or attitudes  2. Synthetic statement  a. Want your hypothesis to have a chance to be true or false  b. Overtrained athletes are prone to sicknessResearch Methods  Exam 2 Study Guide c. Stay away from non synthetic statements d. 2 Types of non synthetic statements   Analytic statement – always true  Ex. I am human, I am wearing boots   Contradictory statement- always false & variables that oppose each  other Ex. I am and I am not addict to coffee 3. Hypothesis must be testable  a. Untestable hypothesis are not useless. Sometimes technology is not  sophisticated enough (ex walking on the moon before technology was  available)  4. hypothesis must be falsifiable  a. Must be worded in a way that disproves findings  b. Worded so that failures to find the predicted effect must be considered  evidence that the hypothesis is indeed false  c. Forces us to be open minded   Ex. If you study enough then you pass your exams 5. Parsimonious statement  a. Simplest explanation possible  b. K.I.S.S. 6. A fruitful statement  a. Want your study to lead to new potential studies V. Two models to help genre a hypothesis 1. Inductive Model a. The process of reasoning from specific cases to more general principles to  form a hypothesis b. Ex. I have a friend name Kenny. Kenny wants to be a sports psychologist. All  Kenny's want to be sport psychologist. 2. Deductive Model a. The process of reasoning from general principles to specific instances b. Most useful for testing principles of a theory  c. Ex- general principle – Football Teams, specific instance – Baltimore Ravens VI. Why we should look at previous research 1. points out important issues  2. points out holes in research  3. points out variables that may serve as confounding  4. all research studies at the end point out limitations Chapter 7: Basics of Experimentation  I. True Experiments 1. High in manipulation 2. High in imposition of units 3. Has causal statements Research Methods  Exam 2 Study Guide a. Goal of experiment 4. Change in behavior caused by antecedents a. High in internal validity b. Low in external validity 5. 3 main features a. Need to manipulate antecedents  At least two different treatment conditions and compare b. Way to measure effects  Has antecedents influence behavior c. Record responses and observations  If recorded correctly, you can see difference between treatment  condition and able to compare statistically II. Independent VS Dependent Variables 1. Independent variable  a. Manipulating control of antecedents b. Minimum of 2 levels  Each level is manipulated and becomes a treatment condition  Ex: stuyinh color paper (IV: blue and yellow) on learning rate (DV) 2. Dependent variable a. Outcome of manipulation  b. What we are measuring caused by antecedents III. Hypothesis 1. Independent variable and dependent variable 2. Operational definitions a. Put in place for IV and DV b. Put in place before starting study  3. Two types of definitions a. Conceptual definition  General/factual b. Operational definition  Give each variable a meaning with your study  What to think about  Observing variables  Procedures to follow  Measurements used  Manipulating and measuring IV and DV  Replication  4. Types of operational definition a. Experimental operational definition  Giving the meaning of the IV  Treatment condition – levels of IV  Steps of manipulation  b. Measured operational definitionsResearch Methods  Exam 2 Study Guide  Gives meaning of DV  How are you assessing the effects from IV  How are you recording responses  How are you scoring the responses IV. Variables 1. Construct/ Hypothetical Variable a. IV and DV can’t be measured directly b. Behavior chosen to help measure variables 2. Non-Construct/Hypothetical Variables a. Can be directly measured  b. Do not manipulate the antecedents by creating different treatment conditions V. Levels of Measurements 1. Kinds of scales to measure responses a. Close ended questions b. Measure different types of behavior  Nominal, ordinal, ratio, and interval VI. Checking Accuracy of Operational Definitions 1. Reliability a. Consistent and dependable  2. Ways to check reliability a. Interrater reliability check  Have different observers measure same variables and see how much  agreement there is between measures b. Test – retest reliability check  Check reliability amongst participant responses by comparing 2  scores of each individual data c. Interitem reliability check  Measure internal consistency of the test items  2 stats analyses:   split-half reliability check: breaks measurements in 2 equal  halves and run correlation between the 2 halves  cronbach’s alpha: run correlation between every item 3. Validity a. Accurately measuring what we intended to measure b. Manipulation check- checks validity throughout a study 4. 5 ways to check validity a. Face validity (weakest)  Have participants rate the validity of the measures b. Content validity  Does the content of the measures accurately reflect materials I am  testing c. Predictive validityResearch Methods  Exam 2 Study Guide  If I measure what is intended; should be able to predict future behavior d. Concurrent validity  Taking the measurement of the study and compare to an already valid  measurement e. Construct validity  Most important test to run  Makes sure your study is consistent with past theories VII. 3 Threats to Internal Validity 1. Extraneous variables a. Variables that influence the study and aren’t the IV or DV 2. Confounding results a. The error produced from extraneous variables VIII. 8 classic threats to internal validity  1. History threat a. Outside events effecting results of the study 2. Maturation threat a. Any physical or psychological changes that occur throughout the study 3. Testing threat a. Effect on DV from previous testing 4. Instrumentation threat a. Part of your measuring technique breaks or stops working throughout study 5. Statistical regression threat a. Assigning people to treatment conditions based on extreme scores 6. Selection threat a. When we do not use random assignment 7. Subject/participant mortality a. Long studies dropout rates are high 8. Selection interaction threat a. Selection threat combined with any of the other threat Chapter 8: Solving Problems: Controlling Extraneous Variables I. 4 Main Extraneous Variables 1. Physical 2. Social 3. Personality 4. Context II. Physical 1. Any aspects of the testing conditions that need to be controlled a. Time of day, noises, location 2. Controlling PV a. Eliminate any variables possibleResearch Methods  Exam 2 Study Guide b. Be consistent- treat all the same c. Balancing- extraneous variables across all treatment conditions III. Social 1. Any quality of the relationship between the participants and the researcher 2. Two main types a. Demand Characteristics- any aspect of the experiment that demands a  certain behavior  Control for Demand  Single blind experiment – blind to the treatment they are  part of.  Use a cover story – false explanation of the study’s purpose b. Experimenter Bias – without realizing, an experimenter can give cues on  how participants are expected to act. Can be a confounding error  Control for Bias  Use a double-blind study 3. Rosenthal Effect a. Experimenter treats participants differently depending on the treatment  condition b. Not done on purpose IV. Personality 1. The personal characteristics the experimenter brings to the study 2. If the researcher is friendly and smiles more, then they will get more willing  participants  3. Hostile or aggressive behavior will cause a higher dropout rate and less  compliance 4. Controls  a. Consistent with interactions  Scripts, do everything the same b. Use multiple researchers c. Minimize face to face contact V. Context 1. Stem form context of the study 2. 2 types a. When participants choose what study to participate in b. When experimenters choose the participants 3. Controls a. Make title of study neutral b. Have a systematic way of choosing participants  VI. What Happens When You Don’t Control These Extraneous Variables? 1. It lowers the internal validityResearch Methods  Exam 2 Study Guide Chapter 9: Between Subjects Design I. Experimental Design 1. Structure of the experiment II. Picking the Design 1. Type of hypothesis 2. Look at prior research 3. Practical problems and limits 4. Number of independent variables (IV) 5. Number of levels per IV 6. Number of treatment conditions 7. Some group of participants vs. multiple groups III. Between Subjects Design 1. Every participant receives one treatment 2. Interesting in the differences reported between treatment groups 3. Recruiting participants a. Representative of the population b. Mirror larger population c. High in external validity d. Using random sampling e. At times, won’t be able to use random sampling f. Shouldn’t use people we know  Might not be unbiased to the true purpose of the study  Unbiased to body cues and body language  May feel obligated to participate g. Need a large sample to generalize and detect differences in treatment  conditions 4. Effect size a. Number of participants needed to see an effect from treatment b. Strong, moderate, and weak IV. Two-Group Designs 1. Use random assignment a. Every individual in your sample has an equal chance of being selected as a  participant in a treatment condition b. If we do not use random assignment, then internal validity decreases and you  will have selection threat  2. Two – Independent Groups a. Experimental group/control group  IV, 2 levels, 2 treatment condition  1 Treatment condition  Experimental condition  Experimental group  2 treatment conditionsResearch Methods  Exam 2 Study Guide  Control condition  Control group – non-treatment group; no manipulation b. 2 experimental group design  IV, 2 levels, 2 treatment conditions  No control groups – 2 experimental groups  Follow up to experimental group/control group design 3. Matching(ed) Groups Design a. Know you have extraneous variables b. 3 matching techniques  Precision matching – match based on identical scores  Range matching – match based on a predetermined range  Rank order matching – rank by score and pair each adjacent score  together; weakest c. Why should we use a match design?  Extraneous variables present – need to control   If you have a small sample V. More Than Two Independent Variable Levels 1. More than 2 treatment conditions 2. Use multiple groups design a. Multiple – Independent – Groups Design  Random assignment  Can use matching b. Pilot Study  Test experiment before your main study  Mini experiment Chapter 10: Between Subjects Factorial Design I. Factorial Design  1. Research design that has 2 or more Independent Variables (IV) a. Each IV must have at least 2 levels b. Look at multiple Independent Variables on Dependent Variables (DV) 2. Controlled  3. Each IV = 1 factor 4. 2 factorial experiment a. Simplest factorial design has 2 factors b. 4 treatment conditions/levels II. 2 Kinds of Information from a Factorial Design 1. Main effects a. Info about each IV and their levels 2. Interactions a. How different IV interact with each otherResearch Methods  Exam 2 Study Guide III. Main Effects 1. Effects of 1 IV 2. Specific changes in behavior from the effect of IV 3. Number of factors = number of main effects 4. Run an ANOVA analysis a. Test for statistical significance IV. Interactions 1. Relationships that exist between IV’s and combined effect they have together on  DV 2. Number of interactions depends on the number of factors a. 2 IV’s = 1 interaction b. 3 IV’s = 4 interactions  Higher order interaction 3. It is possible to have a significant interaction, but no main effect V. Describing Factorial Designs 1. Use short hand notation  a. Tell us number of IV b. Tells us number of levels per IV c. Number of numbers tells us the IV’s d. The numerical number tells us the levels of IV  Example: 2X2 = 2 factors, first factor has 2 levels, second factor has 2  levels, 4 treatment conditions  2X3X2 – 3 factors, first factor has 2 levels, second factor has 3 levels,  and third has 2 levels, 12 treatment condition Chapter 11: Within Subjects Design I. Within Subjects Design 1. Participants receive ALL treatment that is included in the study 2. Measure DV after EACH treatment is given 3. Factorial W/in Subjects Design a. 2 or more IV b. Example: facial expression vs. gender  Time to recognize facial expressions  Women look at photos of a woman with a happy, sad, angry, and  embarrassing expression  Men do the same with photos of a man II. Pros and Cons of W/in Subjects Design 1. Pros a. Can use same participants throughout studyResearch Methods  Exam 2 Study Guide  Fewer participants b. Constancy with instructions and interactions  Don’t need a script c. Saves time d. Good chance of detecting the effects of IV on DV e. Controls for extraneous variables f. Can track participants over time 2. Cons a. Practical Limitations  Requires to stay longer  Dropout rate, boredom, guessing hypothesis  Influence different treatment conditions on one another   Interference  Order/timing of treatment conditions III. Mixed Factorial Design 1. 1 factor is treated like a W/in Subjects Design and the other factor is treated like a  Between Subjects Design.  a. Between Subjects Design is usually a preexisting characteristic that  automatically divide participants  Gender, age IV. 3 Things to Control for with using WSD 1. Order Effects a. Use different counterbalancing to control 2. Fatigue Effects a. The decline in performance over time 3. Practice Effects a. Increase in performance over time V. Progressive Errors 1. Any increase or decrease in performance that aren’t caused by the IV 2. Can cause confounding results 3. Occur because participants receive more than one treatment condition VI. Types of Counter Balancing 1. Reverse a. Controls for progressive errors by presenting each treatment twice b. Present in one way, then present in reverse c. Example: ABC…… CBA 2. Block a. Consider a set of treatment conditions as one block b. Rearrange order of each block/treatment condition randomly c. Example: ABCD  ABCDResearch Methods  Exam 2 Study Guide  DCBA  ABDC  BACD  CDBA 3. Complete a. Want to use possible sequences of treatment conditions b. Examples  2 treatment conditions = AB then BA  3 or more treatment conditions – find factorial N!  3! = 3 x 2 1  4! = 4 x 3 x 2 1 4. Partial  a. Use when complete counterbalancing isn’t possible b. Use subset of all possible sequence 5. Two techniques to select subset a. Randomized partial counterbalancing  Participants = subset  Example- 10 participants = 10 sequences chosen randomly b. Latin square counterbalancing   Matric of sequences of treatment conditions  Each sequence appears once  Example- 4 treatments  ABCD  BADC  CDAB  DCBA

Page Expired
It looks like your free minutes have expired! Lucky for you we have all the content you need, just sign up here