Chapter 9Chapter 10Chapter 11Exam 3 Review ***Know the listed concepts. Be able to define, apply, and provide examples of them. ***Chapter 9 Properties of an experiment o Manipulate at least one IV We control the IV We measure the DVo Random assignment- different levels of our IV The easiest way to be sure that the experimental groups are roughly equivalent before manipulating the IV Equal probability of being assigned to both groups (control or experimental) Trying to eliminate our third variables so it doesn’t influence our DVo Control/eliminate extraneous variables (confounds) Types of IVs (quantitative vs. qualitative)o Environmental/ Situational (quantitative) Ex: room temperature, is associated with behavioro Invasive (quantitative) Ex: drug administration, surgeryo Instructional (qualitative) Ex: try to win (0 points, get to 100) vs. try not to lose (100 points, try not to lose those points)- This will show different types of playing styles Experimental vs. control groupso Experimental- participants who RECEIVE the IV Way to remember this; receive has three E’s, so does experimentalo Control- participants who DON’T RECEIVE the IV Pilot test (IV check)o Test manipulation BEFORE conducting the experimento A miniature experiment where we employ our manipulation and wait to see if this manipulation changed the IVo Saves money, time and energy Manipulation checko Check manipulation DURING the experimento One or more questions IV vs. DV vs. Subject variableo Independent variables The variable that is varied or manipulated by the researcher to assess its effects on participants behavior Must have two or more levelso Dependent variables The response measured in a study, typically a measure of participants thoughts, feelings, behavior, or physiological reactionso Subject variable A personal characteristic of research participants- Age, gender, self-esteem AKA: participant variable Condition Assignment (procedures, advantages, disadvantages)o Simple random assignment Each participant has an equal chance of being assigned to any condition- Examples:o Coin-flipping (heads= condition 1, tails= condition 2)o Random number tables (to assign participants to conditions) Participants in each condition share roughly the same characteristics (EX: age, personality) Limitations- Sample size- Number of conditions Posttest Design- recruit participants and randomly assign them to a condition- Measure DV once following IV administration DO NOT CONFUSE WITH RANDOM SAMPLING!o Matched random assignment Pretest’s participants for DV or related variable Researchers sometimes try to increase the similarity among the experimental groups by using Matched Random Assignment Cluster participants based on pretest Cluster size= number of conditions Randomly assign participants in clusters to conditions Example: - Pretest for risk-taking tendencies (sky diving, rollercoaster’s, gambling)- Each conditions risk-taking tendencies should match before the manipulation- 2 conditions= 2 participants/clusterso Ranked score (highest to lowest) 32, 29, 27, 26, 25, 23, 22o Clustered score (matches participants together) 32 & 29, 29 & 27, 26 & 25, 23 & 22o Flip a coin (high score) (low score)- Pretest-Posttest Designo Advantages Test effectiveness of condition assignment Pretest provides “true” baseline More powerful than a posttest design onlyo Disadvantages Pretest sensitization- patients may figure out the study Costs more and often unnecessaryo Repeated Measures (3rd in experiments?) AKA: within-subjects design All participants experience all conditions of the experiment Interested in differences in behavior across conditions within a single group of participants Advantages:- Do not need random assignmento Same participants= same characteristics- More Power (its ability to detect effects of the IV)- More powerful than a between-subjects design since within-subjects design participants are identical in every way (same individuals, DUH!)- Smaller n (don’t need as many participants) Disadvantages:- Order Effects (next section) Order Effects- participants behavior is effected by the order in which they participate in the various conditions of the experimento Practice Effects- performance improves due to completing DV multiple timeso Fatigue Effects- performance declines due to fatigue, boredom, impatienceo Sensitization- participants guess the hypothesiso Carryover Effects- effect of one condition persists while another condition begins Reduce Order Effectso Counterbalancing- presents the levels of the IV in different order to different participants Pg. 192 table example Guards against the possibility of order effects Randomize the order of the conditions Ideally use all possible combinations Limited by # of conditionso Latin Squares design Each condition appears once at each ordinal position (1st, 2nd, 3rd, ect)- May be used to control for order effects- The first condition will appear once- The second condition will appear once- The third condition will appear once- Uses the big color block thing to control the order effects of a repeated measure design- in class it was called: “research suduko” Treatment Variance vs. Confound Variance vs. Error Varianceo Treatment Variance- the portion of variance in participants scores that is due to the IV AKA: primary varianceo Confound Variance- when a variable other than the independent variable differs between the groups AKA: secondary varianceo Error Variance- result of unsystematic differences among participants AKA: within-groups variance Sources of error variance Internal vs. External validityo Internal validity- provides certainty that IV produced change in DV All confounds or alternative explanations are eliminated Confounds are threats to internal validityo External validity- provides replicable or generalized to other samples, research settings and procedures Threats to internal validity: CONFOUNDS Double-blind procedureo Neither the participants nor the experimenters who interact with them know which experimental condition a participant is in at the time the study is conducted Most effective way to eliminate experimenter expectancy Supervised by another researcher Keeps other experimenters “in the dark” Insures
View Full Document