Module 10: Repeated Measures Flashcards
Repeated measure (4)
+ what it is+what do we acess+also called+ex
- When subject experience all different condition.
- We can access whether the DV differs depending on the different IV conditions
- Within-subject design
- ex: does heart rate change in rest and jump rope for a person?
In repeated measure designs, participants serve as ——— and therefore we call this approach a ——- design
- their own control
- within-subject
Between subject design
Advantage of repeated measure design (3):
- You might need fewer participants for a repeated measure design which can be helpful if it is hard to find the type of participants that you are interested in. (Ex: rather than recruiting 20 children and another 20 for autism study, you can apply both condition with the same 20 children.)
- It has more power: with similiar effect size, you are more likely to reject the null hypothesis and find statistical significance than you are with the between subject design. This is because you are doing a better job of isolating your independent variable.
- You can measure change across time: How varies factors affect human development and inform our practices and policies
—– designs can be more appropriate if a study involves special populations, and it is harder to find members of the special population, such as children diagnosed with autism spectrum disorder or students with learning disabilities.
Repeated Measure
Repeated measures have more——- then between subject
power
Unique solutions (2)
Ex:
There are some situations that people can experience only once and we cannot go bacj and create the same condition, thus we can only measure once. In these conditions it is not possible to use a repeated measures design.
For ex: In a stidy on wether gastric bypass surgery or a healthy diet would be more effect for weight loss, we need to use independent samples design as once subjects are operated on, they can no longer get back to the way they were.
Order effects (2)
+ex
Occurs when exposure to one level of the independent variable (One condition) influences the next level of the independent variable. For ex: for rating felonies and misdemeanors, if they rate the felnoes first and see the misdemeanors second, perhaps after seeing felonies, some participants will not feel so bad about misdmeanors.
Inorder to control for order effects and increase internal validity of our study, we—–
expose the participants to the conditions in different orders
4 types of order effects:
- practice effects
- fatigue effects
- carryover effects
- sensitization effects
Practice effects (2)
+ex
When participants engage in experiments that employ repeated measures designs, the more people do something, the better they may become at it.
Ex: Faster reading a book the second time
Fatigue
- perfromance decreases on subsequent levels of the IV
- By the time participants go through the first condition, they may get bored and tired and that will affect their effort and performance.
Carryover Effecr
The effects of a preceding condition spill over into the following condition. Ex: if you take a bite from a slice of cake then drink a sip of coffee, you will not get the exact taste of your coffee as your taste buds are still under the effect of their previous experience: Sweet cake.
Sensitization Effects
- After experiencing one condition. participants might react differently to other conditions.
- Specific type of carryover effect in which an experience in one condition makes participants extra sensitive to the manipulated variable in the next condition
Ex: When 2 average participants were viewed in a row they received 5.5/9. When the average participant followed 3 highly qualified applicants the average applicant received a rating of 2.5 out of 9.
Counterbalancing
2 major types of Counterbalancing:
- Within subjects counterbalancing
- across subject counterbalancing
Within Subject counterbalancing (3)
+can be achieved with….
- Each participant receives all of the condition-orders in the experiment.
- All order occurs within each patient.
- Can be achieved using reverse counterbalancing or block randomization
Across subject counterbalancing
Each participant receives a subset of the condition-orders used in the experiment.
Reverse counterbalancing (2)
+ when is it less useful?
- Each participnt received all conditions in one order and then in reversed order such as A then B followed by B then A.
*This condition is less useful when more than 2 conditions are present
The ABBA counterbalancing technique balances practice effects only when the practice effects are —-. —— are observed if participants change————- following each condition.
- linear
- Linear practice effects
- in the same way
Block randomization
participants receive a set of each experimental condition in random order.
Each participant would experience each level of the independent variable (A,B,C,D) six times.
Block randomization works best when
each condition appears at each position (first, second,etc) an equal number of times.
Block randomization balances practice effects only when….. For ex…..
the conditions are presented many times. For example, those many administrations of the conditions are needed to balance (average) practice effects across different conditions.
Complete counterbalancing+Ex (2)
Every possible order of your condition is administrated but each participant experience only one order. For ex: Participant 1 would get AB and participant 2 gets BA
Partial counterbalancing (2)
- Presents a subset of possible orders with each participant exposed to one or more subsets.
- The design requires each condition to appear in each ordinal position equally often.
Complete counterbalancing is preferable when there are
only 2 to 4 levels of the independent variable.
WIth complete counterbalancing tell me how it would be if there are 3 conditions and 6 possible orders
1/6 of the participants would be randomly assigned to each of the 6 orders.
Washout period
How do you know if it is a Latin square?
It is not a latin square if the same letter (condition) appears more than once in a given row or column.
Differential transfer
When the learning, fatugue or boredom that transfers from one level to other levels may not be the same. The ordering produces a reaction to the dependent variable for that ordering only. For example: the transfer from condition 1 to condition 2 is different from the transfer from condition 2 to condition 1.
When differntial transfer occurs, counterbalancing the orders does not help with the order effects, practice effects and fatugue; in one order you may have a huge practice effect wheas in another you may have just a minimal one. In tasks like that it is wiser to chose——
between sibjects design
Attrition
Participants gets bored bc of how long or sm and they never show up again to complete your study
Matched group design
The same participants do not go through all of the conditions but rather each one of many pairs of participants (matched according to some measure that is relevant to the dependent measure) goes through a diff condition.
Matched groups design advantage (3)
- Enjoys the advantages of the repeated measure design as well as the advanatges of between subject designs.
- Because the participants who go through diffferent conditions are matched according to some criteria, there is less individual variabiity amoung the conitions:advantage of repeated measure
- The matched group design also dodges order effects since each participant goes through only one level of the indepedent variable and is an advanatge of: inbetween subhject design
Matched group design is used when (2)
- They have limted number of particicpants
- There is great variability among participants in a variable that is relevant to the dependent variable.
Disadvantages of Repeated Measures Designs
- Unique situations can only be experienced once (Asking people about their mood every 3 months thats fine but its its seeing whawt heppens the first time someone experience X you cant have them experience something fo rthe first time every 3 month).
- Order of effects can dcrease internal validity
across-subjects counterbalancing/between-subjects counterbalancing:
A type of counterbalancing in which each participant receives a subset of the condition-orders used in the experiment. Also known as between-subjects counterbalancing.
