Research Flashcards
(47 cards)
Correlation
Association NOT Cause
Correlation #s
-1 to 0 to +1
0 is no correlation
+1 or -1 are perfect
Positive Correlation
Study more, grade goes up
Negative Correlation
More you brush your teeth, less you will have cavities
Bubbles
Flaws, like in a sticker
True Experiment
2 or more groups and random sample
Systematic Sampling
Every 9th or so person taken
Quasi
Cannot control the IV
+ external validity
- internal validity
Independent Variable IV
Experimental Variable
Manipulate
I manipulate the variable.
Dependent Variable DV
Control Group
I like dependency and control.
Expresses the outcome.
Null Hypothesis
All experiments have one
Means no correlation
Experimental or Alternative Hypothesis
Is there a significant difference?
Type 1 Alpha Error
Rejects the Null that is true
Rejection is the #1 thing you shouldn’t do to a person!
Type 2 Beta Error
Accepts Null when it should have been rejected
Significance Level for Social Sciences
.05 or less
.01 or .001
N=1
Single Subject Design - Case Study
Confound Experiment
Research inaccurate
Demand characteristics
Cue
Can confound an experiment
Obtrusive
Subjects KNOW they are being observed
Unobtrusive
Subjects DO NOT KNOW they are being observed
Internal Validity
High when inside the studies there are few flaws
External Validity
High when results can be generalized outside
Test to compare 2 means
T Test
ANOVA
2 or more means to compare
F Value Tables