PSYC 204 1st Edition Exam 2 Study Guide Chapters 3 9 11 3 Claims to research validity Frequency Association Casual Characteristics of Experiments IV is manipulated Common design o Pretest of DV o Intervention treatment of IV o Posttest assessment of DV Research Validity Appropriateness of inferences drawn from data o Types Internal External Statistical conclusion Construct Internal Validity extent to which you can infer that a relationship btwn 2 variables is casual or that the absence of a relationship implies absence of cause Determinants of internal validity o Internal consistency o IV cause DV effect CONFOUND CAUSE o Quality of research design controlling extraneous and confounding variables Threats to Internal Validity o History o Maturation o Testing o Mortality o Selection o Regression to the mean External Validity inference that results can generalized to and across alternate measures participants settings and times Threats to External Validity o Interaction btwn subjects and treatment Population validity o Interaction btwn setting and treatment Ecological validity o Interaction bwtn history and treatment Temporal validity Statistical Conclusion Validity appropriatness of inferences made from data as a function of conclusions drawn from statistical analysis Threats to statistical validity o Low statistical power o Violations of assumptions of statistical tests o Measures with low levels of reliability Construct Validity extent to which labels placed on what s being observed are theoretically relevant Do the results support the theory underlying the research o Threats to Construct Validity Loose connection btwn theory and experiment Diffusion of treatment Role Demands Participants expectations of what the experiment requires them to do Can lead participants to engage in socially unacceptable behaviors Participant Roles Good subject Bad subject Apprehensive subject o Solutions to participant roles Deception Separate the IV and DV in time Use unobtrusive methods measures Experimenter Bias unintentionally influencing the results of an experiment Possible causes o Experimenter attributes o Situational factors Possible solutions o Double blind study o Standardize experimental procedures o Use only one experimenter Research Methods True Experiments o Manipulations of the IV o Random Assignment Quasi Experiments o Manipulation of the IV o No random Assignment o Pre existing groups Non Experiments o No manipulation of the IV o No random Assignment True Experiments a study or research design in which the IV is manipulated Experimenter has separate control over the experiment o 1 Control over assignment of subjects o 2 Control over presentation of conditions to participants Manipulation Check presentation of one or more levels of the IV to the participant in an experiment crucial for construct validity in experiments Manipulation check additional DV added to the experiment to confirm the manipulation was perceived the way it was intended to o Rating scales o Structured observations during the experiment debriefing o Pilot study data preliminary study Characteristics of an Experiment Random Sampling o Process of choosing a rep group from an entire populationsuch that every member of the population has an equal chance of being selected to be in the study Enhances external validity This is the best way to assign random numbers to people Random Assignment o A control technique that equates groups of participants by ensuring every participant has an equal chance of being assigned to that group Basic Research Designs Between Subjects o Each subject receives one level of the IV Within subjects o Each subject receives all levels of the IV o DV measured many times Pros Fewer subjects needed Equivalence of participants across condition is certain Cons Testing effects Order and sequence effects Limited research topics o Issues with within subjects design Order effects changes in a participants performance resulting form the ordinal position in which the condition or stimuli appears in the experiement Sequence effect changes in the participants performance resulting from interactions among the conditions or stimuli themselves How to control o Counter balancing Arranging conditions so that each condition occurs in each ordinal position o Reverse counter balancing Conditions are presented in order the 1st time and then in reverse order o Block Randomization Order of conditions is randomized but each condition is presented once before any condition is repeated o Latin Square Control procedure in which each subject experiences each condition in a different order from other subjects Mixed Factorial o Mixture of the betwn subjects within subjects Pre test Measures DV o Advance measure of DV Pros o Demonstrates equivalence of groups on DV o Provides a baseline measure Cons o Effects of testing or practice effects can be problematic Matching Control procedure to ensure that experimental control groups are equated to one or more variables before the experiment Conducted before random assignment to groups Typically on subject variables o Requirments Matching variable and DV are significantly correlated Feasible to pretest on the matching variable o Cons Requires access to participants Time consuming May waste lots of participants Experimental factor research Design Single factor btwn subjects Single factor within subjects Solomon Four group design btwn subjects more superior o AVOID One group posttest only design Cons o No control group a control group is needed to know to what extent something is changing o No pretest to determine if any change has taken place One group pretest posttest design Con o You don t know if it s the intervention the caused a change Control A standard against which to compare the effect of a particular variable Reduce variability Control strategies Use lab setting Use previously validated measures Build nuisance variables as statistical controls Consider the research setting as prep o Lab vs field Differences Realism Control Threats to Internal Validity Design confound Selection effect Order effect Maturation History Regression to mean Attrition mortality Testing Placebo effect Casual inference Ability to make statements about the absence or presence of cause effect relationships o Conditions Contiguity Temporal precedence Constant conjunction o Dependent on how well or to the extent to which alternative causes or explanations are ruled out Factorial designs Experiments that