correlation design Flashcards
]
correlation design
a type of non-experiment
draw a sample,then measure 2 or more different variable
-see how well they hang together
-how they are related
example facebook
positive linear
up
negative linear
Down
no linear relationship
correlation (r)
a statistic that indexes the degree of relationship between two variables
-+/- = direction of relationship (posit or negative)
- Numbers= strenght of the relationship
(how close is one variable associated with the other one)
linear relationship can be summirized by a single number
r = correlations
r can range from -1 to +1
- -1=perfect negative linear relationship
- +1= perfect positive linear relationship
- 0=no relationship
the closer to the number -1 or +1 more correlation it has
correlation does not equal causation!!!!
ex
third variables vs confound varible
third varible - that causes that correlation to happen
confound variable-the problem is in the defenition of how to mesure the interest
when can we say different varible a cause of change virable
in experimental design
experimental method
it involves manipulation of one variable,controls of several other variables and then measure the outcome variable (DV)
experimental design
A must happen before B
when A is present, B must follow
when A is absent, B should not follow
when Ais the only thing change that might affect B
basic experiment design
1.sample(random assigment)
condition 1 (independable variable) condition 2= dependable varibles
independent variable
cause
is independent of the participants / is controled by the reserchers
dependent variable
effect/responde
is dependent of participants
internal validity
the extent to which you can be confident that a cause-and-effect relationship established in a study cannot be explained by other factors.
what criteria does a study need to have to have hight internal validity?
1.co-variation between two varibles
2, temporal precedes
3, eliminate plausable alternative explanation
1.co-variation
changes in one variable must cause in another (change in IV must cause change in DV)
2.temporal precedes:
IV proceds the DV
3. eliminate plausable alternative explanation >
can the difference in the DV the attributed only to the manipulated iV
(good design)
how do we achive internal validity(well design experiment has hight internal validity)
1.exeprimental control
2.rendom assigment of people to condition
1.experimental control
everythings continue the same across condition except the variable that is being manipulated
confound 2
a varible that is interwind with iv and may impact the dv
could explain all or part of the result
- random assigment
random assigment is a method of assigning subeject to experimental and control group so that each subject in the experiment has the same chance of being assign to eacgh group
allows the resercher to balance out influenceof all othersvariable except the iv
the bigger the sample the better
independent groups(between groups)
random assigment
into 2 different conditions
repeted measures design
(within group)
all participants recive both condition of the iv
compare avarage scores on the dv for each condition
pros
- needs fewer participants
- less cost
- more power to see the effect of IV on DV if there is one
cons
- order effect (fatigue,practice,contast)
- mortality(people may give up by the middle of the experiment)
- dealing with counterbalance order
- lenght of time intervals
potential problematic effects
carry-over effects/order effect
whenever a participants responses in a given condition is effected by having previosly participated in other conditions in the experiment
the effects of one condition carry over to others
carry-over effect=order effects
types of order effect
1-practice effect
2-fatigue effect
3-contrast effect:
contrast effects
Participants interprets one condition of the IV differently because of having previously participated in the other condition of the IV(violence and punishment effect)
how to conterbalance measuring order effect?
use two groups with two different orders (random assigment to order)
adressing order effects
-random assigment of order(already disscussed)
-filler task
ex: sad video>critique essey> madetition>happy video >critique paper
-passage of time (for fatigue and constrast)
seconds,minuts,hours
sad video >critique>time out>happy video>critique paper
cost/complexity
order effects
-fatigue,practice,constrast
dealing with conterbalancing orders
-4 conditions = 24 order
-can use latin square
-mortality(they dont came back)
benefits
- more power to see the effect of IV on DV
- it can help us identify the contribution of our IV to explaining variability in the DV
- need fewer participants
- less cost
balanced square
1-each tratment must occur once with each participant
2- each treat must occur the same number of times for each time period or trials
3-each treatment must procede and follow every other treatment an equal number of times
