EDHD 306 Fall 2013 Final Exam Study Guide Experimental design o Experimental control and randomization Experimental control 1 Eliminating the influence of an extraneous variable on the outcome of an experiment by keeping the variable constant in the experiment and control groups Randomization Controlling for the effects of extraneous variable by ensuring that the variables operate in a manner determined entirely by chance Varies along with independent effects intertwined so cannot determine o Confounding variables which variable is responsible Attempts to eliminate Experimental control Randomization o Internal validity study other o Basic experiments 2 variables Ability to draw conclusions about casual relationships from results of HIGH when strong inference can be made that one variable CAUSED the o Independent o Dependent 2 levels experimental and control Posttest only design True experimental design which dependent variable is measured only once after manipulation of independent variable Researcher must o Obtain 2 equivalent groups o Introduce independent variable o Measure effect on dependent variable Must get people who do not differ in systematic way to avoid selection difference Pretest posttest design Dependent variable measured before and after o Advantages and disadvantages of having a pretest Advantages o If groups are uneven it enables researcher to assess equivalence from beginning o Sometimes necessary to select participants o Makes us focus on change from pre to post test 2 Emphasis on change incorporated into analysis o Necessary when possibility that participants may drop out Disadvantages o Time consuming and awkward o Can sensitize participant to what you are studying enabling them to figure out the hypothesis o Difficult to generalize to people who have not received pretest o Assigning participants to conditions Independent groups Participants randomly assigned to various conditions Different participants assigned to each condition using random assignment o Decision to assign individual to particular condition is Repeated measures completely random In an experiment with two conditions the same participant is measured after receiving each level of independent variable o Individuals participate in all groups o Repeatedly measured on dependent variable o Important to counterbalance Advantages o Fewer participants needed o Sensitive to finding statistically significant differences between groups bc have data from same people o Individual differences can be explained Disadvantages o Order effects Carry over effect Possible for effect from first to influence response to second Practice learning effect Performance on second task might improve bc of practice from first task When order of treatment affects dependent Order effect variable Fatigue effect Deterioration of performance as participant becomes bored tired distracted Counterbalancing o Method of controlling for order effects in a repeated measure design by either including all orders of treatment presentation or randomly determining the order for each subject 3 Matched pairs Instead of randomly assigning participants to groups you match people based on a characteristic ie GPA to achieve equivalency Helpful with small sample size o Manipulations Independent variables Straightforward manipulations Manipulated independent variable by written verbal or visual material Staged manipulations independent variable Used for Stage events during the experiment in order to manipulate the o Researcher may be trying to create psychological state such as frustration anger etc o May be necessary to simulate real world situation Confederates o Someone who acts like a participant but is working with Strength of the manipulation researcher o Dependent variables Sensitivity Particularly important when measuring human performance The ability of a measure to detect difference between groups Ceiling and floor effects Ceiling effect o When the task is so easy everyone does exceptionally well regardless of the conditions that are manipulated by independent variable o Participants quickly reach maximum performance level o When task associated is so difficult that no one can do it Floor effect o Participant expectations Demand characteristics When participant forms interpretation of experiment s purpose and unconsciously changes behavior to fit interpretation o Often controlled by Deception Filler questions in questionnaire Asking what they think the intent of the study is o Experimenter expectations Expectancy effects the expectations of the researcher affects how they administer the experiment the results Solutions to expectancy problem Well trained experimenters Simultaneous conditions 4 Use of experimenters unfamiliar with hypothesis o Pilot studies and manipulation checks Pilot study Small scale study conducted prior to an actual experiment Designed to test and refine procedures Manipulation checks Measure used to determine whether the manipulation of independent variable has had its intended effects on subject o Making experiments more complex o Factorial designs Types of designs 2x2 2x3 2x2x2 Main effects Effect of each independent variable taken by itself Overall relationship between each independent variable and the dependent variable Interactions What occurs when effect of one independent variable on dependent variable is different at different levels of the other o The effect that an independent variable has on the dependent variable depends on the level of the other independent variable Look using statistical tests and graphing levels Interaction between independent variables indicates that the effect of one independent variable depends on the particular level of the other Moderator variables Variables that affect direction of strength of relationship between independent and dependent variable Influences the relationship between two variables It may be particular situations or particular characteristics Other research designs o Program evaluation Needs assessment Assess if problem within a group needs a program Program theory assessment Does implement program address underlying problem it is supposed to be addressing Process evaluation Outcome evaluation Is program implemented correctly and to the correct group Is program resulting in intended outcomes Is program working Efficiency assessment Are costs outweighed by benefits o Quasi experimental designs 5 Type of design that approximates the control features of true