Please enable JavaScript.
Coggle requires JavaScript to display documents.
:star:Correlation Research :star:, Participation & Instrument …
:star:
Correlation Research
:star:
Participation & Instrument
:checkered_flag:
Selection
minimum is 30
develop valid measure of variable
Data collection steps
identify variables
select sample
apply instrument
collect data
data collection
obtain predator variables before criterion variables
may lose participants in long run
when strength of predictor variable is defined, test on 1 or > new group
Data Analysis
:red_flag: :
Data analysis & interpretation
:single variable prediction equation
multiple regression equation ( Y= a+bx1 + cX2+cX3)
correlation coefficient
near <1 = positive direction
near 0 = variables unrelated
near > 1 = negative direction
/ -0.5 = useless
0.9 = very reliable
0.8 = moderate reliability
0.7 = low reliability
0.4 = bad reliability
prediction studies
if 2 variables highly related = use either one score to predict the score of other
criterion = variable to be predicted
conducted to test if variables are good predictors of a criterion
can use >1 variable to make predictions
combination of many well correlated variable to a criterion = better than a single one of them
Design & Procedure
:black_flag:
determine if causal connection between variables exits
suggest experimental studies
obtain 2 scores from each member of sample
1 score for each variables of interest
correlate pairs score
result = correlation coefficient
Problem & Solution
:fire:
Discover variables that are related to complex one
test hypothesis
select variables
logical relationship (from theory or experience)
have meaningful result