Task 8 Flashcards
How do you manipulate an I.v?
You set its values to at least two different values or “levels”
What are the two ways you can manipulate a variable?
Quantitatively : changing amount if variable to which each group is exposed (29mg,39mg 3etc)
Qualitatively:. Giving participants in different groups different treatments (Prozac,lexapro,zoloft)
What are the three basic types of experimental designs?
Between subjects design
Within subjects design
Single-subject design
What does the between subject design consist?
Different groups of subjects are randomly assigned to the levels of your I.v(data is averaged and analayzed)
What does the within subjects consist of?
A single group of subjects is exposed to all levels of your I.v (databis averaged and analyzed)
What does singles subject design consist of?
Subjects are exposed to all levels of the i.v(like within group design) but you do not average data across subjects instead you focus in changes of Behaviour of a single subject (or small number of individuals) under the different treatment conditions.
What is error variance
Variability amongst scores caused by variables other than your I.v (extraneous variables or subject related variables such as age gender etc)
What are some sources of error variance?
—Subjects differ in innumerable ways
—Environmentak conditions aren’t absolutely constant
-Same subject varies from moment to moment
How can you handle error variance?
Reducing error variance
Increasing the effectiveness of your I.v
Randomizing error variance across groups
Statistical analysis
How to reduce error variance?
—>Holding extraneous variables as constant as possible by treating subjects within a group as similarly as possible (eliminate outside distractions , read instructions same way , follow same procedures)
How do you increase the effectiveness of your I.v?
—> selecting correct levels of your I.v for the experiment (if it’s too low it might not show an impact on the D.v)
How do you randomize error variance across groups?
Randomly assigning subjects to treatment conditions
Resulting in error variance distributed evenly across the groups (canceling itself)
How do you use statistical analysis to reduce error variance?for what?
-Estimating the probability with which error variance alone would produce differences between groups at least as large as those actually observed
—inferential statistics is used
— results are statistically significant when probability is low enough
What are between subject designs ?
Randomly assigning subjects to groups then exposing each groups to a single level of the i.v
What are the types of designs in between subjects design?
The single factor randomized groups design
Matched group design
What other two subtypes of designs can be found in the single factor randomized group design?
Randomized two group design
Randomized multi group design
What is the randomized two group design?
Randomly assigning subjects to TWO groups to different levels of the i.v and taking steos to hold extraneous variable
What are the advanatages of two group design?
—Simple to carry out
—require few subjects
—no pretesting or categorization of subjects
—statistical analysis is simple
What are some disadvantages of the two group design?
—you don’t learn much about the nature of the relationship, or function , relating the i and d variables
—almost no sensititivity to the effect of the i.v (if participants differ greatly, difficult to see effect of i.v)
What does randomized multi group design consist of?
More than two groups
What is a parametric design?
Manipulating the I.v quantitatively
What is a non parametric design?
Manipulating the I.v qualitatively (categorical)
What is a multiple control group design?
When just one control group is not adequate to rule out alternative explanations of results you use a variation on the single factor multi group design is used one that includes multiple control groups
What are matched groups design?
Matched sets of subjects are distributed at random, one per group into the groups of the experiment
How are subjects matched in a matched groups design?
On one or more characteristics that the experimenter believes exert influence on the deo Brent variable
What is the logic of the matched group design?
Because each of matched subject goes into different groups , the effect of characteristic in which the subjects were matched is distributed evenly across treatments.
Any disadvantage of matched groups design?
If matched characteristic has little or no effect on the dependent variable, matching won’t do any good as leads to Loss if statistics power ad reduced ability to detect effect of independent variable
What is a matched pairs design?
Matched groups equivalent to the randomized two group design
What are matched multi group design?
Matched pair design using multiple levels of a single factor (single factor multi group design) or multiple factors (factorial design)
When does a within subject design seem more appropriate?
When your design has more than 3 groups and is increasingly difficult to find three or four or more subjects with equivalent scores on variables or variables to be matched
What is a within subject design?(repeated measures design)
When each subject is exposed to all lev la of the independent variable rather than being randomly assigned to one level
Advantages of within subjects design?
-Each subject is “matched” with themselves, because they participate in the experiment across all levels of i.v (reduce error variance)
Fewer subjects can be used due to increased power of detecting effect of i.v
What are disadvantages of within subjects design?
—more demanding on subjects as each subject has to exposed to every level of the experimental treatment
—carryover effect: precious treatment alters the Behaviour observed in subsequent treatment
Sources of carryover effect?
Damn son
How to deal with carryover effects?
Counterbalancing
Taking steps to minimize carryover effect
Making treatment order an independent variable
What is counterbalancing?
Assigning various tratments of the experiment in a different order for different subjects in order to distribute any carryover effect equally across treatments
What are two variations of counterbalancing?
Complete counter balancing
Partial counterbalancing
What is complete counterbalancing?
Provides every possible order of treatments and assigns at least one subject to each ordering (practical for experiments with small number of treatments)
What is partial counterbalancing?
Includes only some of the possible treatment orders,orders to be used are chosen randomly from the total set and each treatment must appear equally often in each position
What is the Latin square design?
Partial Design used to ensure that each treatment appears an equal number of tiles in each ordinal position (ABCD then 4 times e.g)
When should you use a within subject design?
—When subject differences contribute heavily to variation in the d.v
-When number of available subjects is limited
—When you wish to assess changes in performance as a function of increasing exposure to treatment conditions.
Within subjects vs matched groups design.
Both deal with error variance by attempting to control subject related factors
If correlation between subject variables and dependent variable is weak, use randomized groups design
What are the types of within subjects design?
Single factor two level design
Single factor multi level design
What is the single factor two level design?
Simplest version of within subjects includes only two levels of the i.v
All subjects receive treatments in one order and half in the opposite order.
Twotreatment means are compared
What is the single factor multi level design?
A single group of subjects is exposed to three or more levels of a single I.v.
The order of the treatment is counterbalanced to prevent carryover effects
What is the mixed design?
A design that includes a between subjects factor and a within subjects factor
Used to lessen carryover effects unable otherwise to do with just a within subjects design