Use LEFT and RIGHT arrow keys to navigate between flashcards;
Use UP and DOWN arrow keys to flip the card;
H to show hint;
A reads text to speech;
28 Cards in this Set
- Front
- Back
Experimental Research Strategy |
Establish cause and effect relationship between two variables. Simplest form - changing one variable causes change in second variable. Four basic elements: manipulation, measurement, comparison, control. Two critical components: Independent vs dependent, Experimental vs control. |
|
Manipulation |
Manipulate the independent variable. This creates two or more conditions. Helps determine which is the cause and which is the effect. |
|
Measurement |
Measure the dependent variable, obtain scores for each treatment condition |
|
Comparison |
Scores on one condition compared with another condition. |
|
Control |
Control all other variables so they do not influence the two variables examined. Eliminate confounding variables. Control obscuring variables, hold variables constant. Matching and randomization. |
|
Independent Variable |
This is what you manipulate, Cause of your effect. At least 2 levels. |
|
Dependent Variable |
What you measured. The effect. Measured at each level of the independent variable. |
|
Experimental group |
the group exposed to the manipulation |
|
Control group |
The group not exposed to the manipulation. Used for comparison purposes. |
|
No treatment control group |
Subjects do not receive the treatment being evaluated. |
|
Placebo control group |
Participants get a placebo instead of actual treatment. Typically in clinical studies. |
|
Between subjects design |
Random assignment into groups. 2 or more treatment groups. Each group gets a different treatment. Independent groups. |
|
Within subjects design |
All subjects participate in both groups. only 1 treatment group. Subjects get both treatment levels. Repeated measures design. |
|
Between subjects design advantages |
Each measured score is independent of the other scores. Only exposed to treatment one time, measured one time. Results not influenced by other factors. Fatigue from multiple testing. Practice from exposure of other treatments. |
|
Between subjects design disadvantages |
Need more individuals. Large differences between individuals in each group will obscure differences between groups. |
|
Pretest - postest design |
Before manipulating independent variable, check if the two groups are equivalent. Goal is to compare the treatment vs control group. Pros: test equivalence of groups. Cons: time consuming, sensitize participants to procedure. |
|
Matched pairs design |
During random assignment to groups. Each condition uses different participants but they are matched in terms of important characteristics. Pros: minimize the effect of individual differences. Cons: Attrition affects both groups. Difficult to match people exactly. |
|
variability |
Variability refers to the distribution of scores. Effect of variability in between subjects design. Large differences between individuals produces large variance. Must take into account the absolute size of the difference in relation to the variance of the data points. |
|
Variance |
Statistical value for the difference of each value from the mean. |
|
Variance between subject design |
Small variance - All scores are similar to one another. Randomly assign a sample for control and sample for treatment. Difficult to see the treatment effect with large variance. large differences between individuals in each group will obscure differences between groups. |
|
Mitigating variance within treatments |
Standardize procedures and treatment setting: Keep all procedures the same. Treat each mouse the same. Sample size: Statistical feature. Using a very large sample can overcome negative effects of high variance. Group equivalency: Start by creating equal groups. Treat each group equally. Human - random assignment with matching. Animals: restrict range. |
|
Within subject design advantages |
Fewer participants are needed. Especially advantageous is participants/animals are rare or costs are high. This repeated measure design allows you to reuse the same individual. eliminate variability due to individual differences. Individuals are no longer confounding or obscuring variables. |
|
Within subject design disadvantages |
Participant attrition: when it requires extensive time and effort, participants drop out of study. Most problematic when studying changes over time, or multiple sessions. (solution: increase sample size). Cannot apply this design to all studies. Time related effects: history (environment changes over time and affects the results), Maturation (any physiological or psychological change of individuals when testing occurs over a long period of time). Instrumentation (Change to the measuring instrument lead to significant differences at testing independent of the treatment). Order (ordinal position of the manipulation can produce practice or fatigue carry over effects). Sequence (interaction between each condition in the experiment can reduce sensitivity of response). |
|
Counterbalancing |
Split the participants into two groups. advantages of testing the same individuals, need fewer research subjects, reduce within-subject response variability. Minimize order effects. Very routine practice for within group design. Goal: use every possible order of treatments with an equal number of individuals participating in each sequence. Purpose: Balance the order effect equally among the two treatment groups. Prevent order effect to accumulate in any one group. |
|
Complete counterbalancing |
Any series of treatment conditions may create a unique order effect. To completely balance all of these effects, research design must use every possible sequence. easy with 2 conditions, gets ridiculous with 3+ conditions. |
|
Partial Counterbalancing |
Does not use every possible sequence. Each treatment appears in one of the sequence positions. Latin square. |
|
Latin square |
Each treatment appears once in each ordinal position. Each sequence appears once. |
|
Placebo |
People experiencing improvement in their condition even though they were assigned a dummy pill without any active ingredients. |