-use of scientific method
-data-based scientific study of behavior  

     -Falsifiability 

-Framework for organizing and explaining a
 collection of data

-Spurs research that tests boundaries of a theory

Tentative statment about something that might be true

Data can show a hypothesis is false, if it is 
-makes a hypothesis testable 

The least complex theory is the most desirable one (it's the easiest to falsify)  

Definition on a variable in terms of the method used to measure/manipulate it 
 

-There is an unchanging, underlying essence of an entity that
  can be defined

-The most important ingredient

-a statistic that indexes the degree of relationship between two variables 

-•the ability to infer that the IV causes changes in the DV

Experimental Control
    -only IV changes 
    -eliminate confounds
Random Assignment
      

-Cognitive bias occurring when researcher knows participants’ condition, unconsciously manipulates experiment to find expected effect

-double-blind experiment
-get someone else to run the exp. (& lie to them)
-replicate findings

-run all Ps together
-train RAs 
-have a script or automate instructions 
 (to minimize interaction) 
 

-any feature of the experiment that might inform Ps of the study's purpose

-filler items/tasks
-deception
-misleading title
-use of confederate (adds realism)
-funnled debrief (broad to specific) 

-Ps at maximum performance on DV 
  (can't see effect of IV) 

-Ps are at minimum performance 
 (can't see effect of IV)
 

-same Ps experience all levels of the IV 
CONS:
-carry order effects (practice, fatigue, contrast)
-clues about hypothesis
-mortality
-time interval b/w conditions  
PROS:
-fewer Ps needed
-less error variance
-an effect is easier to detect  

-the order of presenting treatments affects the DV
1)practice effect - improvement in performance
2)fatigue effect - deterioration in performance (e.g. tired)
3)contrast effect - the response to the 2nd condition is altered b/c it contrasts with the 1st condition (1 carries over to 2)  
AVOID:
counterbalancing and interval b/w conditions  

-alternating the order of the conditions 
-tells you if there's an order effect
  --no difference in results = good  

-sources of variability other than IV
1)Random
2)Systematic
 

Random Error
vs.
Systematic Error
 [can separate only with repeated measures

- it removes random error]

1.(threatens reliability)
-misunderstanding a question
-time of day
-recording error

2.(threatens construct validity) 
-individual differences 
-confounds
-P response biases 

•Participants in an experiment should be informed in advance of all aspects of the research that might influence their decision to participate

-physical risk
-psychological or emotional harm
-social risk
-vulnerability  

-no direct interaction with Ps 
-questionnaires with competent adults on non-
  sensitive topics  
 

Commision - lying 
Omission - leaving out certain details
                -nothing serious left out  

-is it a stable measure?
-little error?

-test retest, interrater (.9+) and
  internal consistency (.8+) 

-are we measuring what we think we are?

-face, content
-CRITERION:
 predictive, concurrent, convergent, discriminant 

-face: a judgment 
-content: does it capture all the necessary aspects of the construct and nothing more?
 
  

1)predictive: predicts the future 
2)convergent: related scores on measures of

   similar constructs
3)concurrent: able to distinguish b/w behaviors
    measured now
4)discriminant: not related to what it shoudn't be

1)can we infer causality?
2)can we generalize our findings beyond this

    group and setting? 

-tend to see links between past and future events that are independent from each other 
-"im due for a win"