April 15 Tuesday Q1 revisited noon to 3 30 armory 270 E5 April 24th E6 May 13 May 6th assignment 1 hard copy and compass course content A2 Descriptive vs experimental research Descriptive at a particular point in time Experimental moves beyond description to determine causality goal Causality is the effect of changes in one area on one or more other areas Research to explore consumer reactions to banner ads What aspects of banner ads are most likely to influence a click through Additional experimental insights Relationships between discount codes and relevance Discount codes only important when level of relevance is low Relationship between ad size and relevance Ad size is only important when ads do not contain a discount code Limitations of survey research Survey results become less reliable for explaining behaviors when Questions require recall especially over long periods Individuals are asked for introspection in areas of behavior not normally thought about Individuals are asked to use memory to attribute cause and effect There may be complex multiple or interrelated influences on behavior Classical experiment Major components Independent and dependent variables Pre testing and post testing Experimental and control groups Independent and dependent variables Effects of x on y Where x and y are variables liable to variable or change Y is dependent x is independent DV refers to status of the effect IV refers to the presumed cause takes the form of a stimulus Pre test and post test Pretesting the measurement of a DV among participants Post testing the measurement of a DV among participants after they have been exposed Experiemental and control groups Experimental group group of subjects to whom an experimental stimulus is administered Control group group of subjects to whom no experimental stimulus is administered an who should resemble the experimental group in all other respects Steps in experimentation Identify what you need to learn Take relevant actions intervention or manipulation Observe effects and consequences Determine causality Components of an experiment At least one independent variable At least one dependent variable At least one manipulation What defines a true experiment Controlled arrangement and manipulation of the conditions treatments Random assignment of treatments to the sampling units Anything else is quasi experiment Why randomize Protect against unsuspected sources of bias No guarantee but an attempt to equalize groups Requirements for causality Events must take place in proper order Events must show an explicit relationship Alternative explanations must be reduced or eliminated Must eliminate other explanations Relationship must show strength of association Three imperatives for establishing causual relationships Covariation Changes in iv must correspond with changes in dv Temporal precedence For iv to have caused dv iv must have occurred before dv Internal validity Validity Internal validity refers to our ability to attribute the observed effect to the independent variables and not other variables External validity refers to extent to which the results of an experiment can be generalized or extended to other situations populations times Threats to internal validity History maturation Testing Instrumentation Selection Mortality History Any events or influences beyond those intentionally manipulated by researcher which have potential to affect experimental outcome Maturation People are always changing which may affect the dynamics of the experiment get older stronger wiser etc Testing Measurement is likely to influence people s behavior Testing effect pretest affects posttest practice improves performance Interactive testing effect pretest affects perceptions of treatment pretest changes the way people respond to the stimulus Instrumentation Specifics of the measures researchers employ will affect the results Selection The types of subjects selected for the study will affect the results Mortality Subjects drop out during the cource of experiement Researcher bias error Occurs when actions of experiementer bias outcome of the experiement Can be intentional or unintentional Double blind experiment Neither the subjects nor the experimenters know which is the experimental and which is the control group Quasi experimental designs One shot case study a single group of subjects is measured on a dependent variable following an experiemental stimulus One group pretest post test design a pretest is added for the experimental group but lacks a control group Static group comparison includes experiemental and control groups but no pretest True experimental designs Control group Stimulated pre test to post test Week 12 E5 4 24 Thursday o Many math items A2 submission may 6th start of class o Hardcopy stapled o Safeassign compass c ourse content a2 E6 cummulative exam 6 is may 13 Tuesday 2 3 50pm Q2 coming soon 70 pts o Math questions scantron Bring calculator Agenda Experimentation experiments o Factorial designs o Internal external validity Descriptive statistics o Mean standard deviation correlation mode Factorial designs Allows for manipulation of two or more independent variables at the same time o Each variable has two or more levels or aspects Ex Type of spokesperson Celebrity celebrity in serious commercial celebrity in humorous commercial Regular regular person in serious commercial regular person in humorous commercial Ex Effects of advertising copy on sensory thoughts and perceived taste Study 2 potato chips IN PREVIOUS NOTES WEEK 11 Study two different food item different operationalization Single our potato chips deliver taste you crave o Single sensory thought stimulus o Single condition focuses on taste only Multiple smell crunchy texture o Multiple sensory thoughts Results 1 very poor taste 9 very good taste o Multiple almost 7 o Single around 5 5 o Additional example Study 2 the experiment used a one factor between subjects subjects design with the ad multiple sense or single sense serving as the manipulated factor One factor one IV Between subjects participants administered only one treatment e g within subjects participants administered both treatments usually one after the other Sensory thoughts taste perceptions Study 3 A2 ad multiple sense or single sense x2 cognitive load yes or no between subjects full factorial design was used in study 3 Participants in the load condition were given a separate task memory roster first last name Full factorial