I/O Psych. Chap. 2 Flashcards
What is part of an Quasi-Experimental research design?
Manipulation but no random assignment (I.e. Using pre-existing groups ex: workplace)
What is part of a Non-experimental research design?
Surveys, observational studies (no manipulation and no random assignment)
What is internal validity?
is when you can say that the IV did cause the effect on the dependent variable, cause and effect conclusion
What do you need to have to have internal validity?
You must have correlation
Temporal= time order/ cause must happen after the effect
Elimination of 3rd variable (ex. Factoring for personality diff. By using random assignment to each condition)
ONLY GUARANTEED IN AN EXPERIMENT THAT INVOLVES RANDOM ASSIGNMENT TO CONDITIONS
What is External Validity?
To what extent you can generalize these findings to the general public
What do you need to have to have External Validity?
Represents the sample of interest
Selected from the population randomly
Extent to which the setting of the research study is representative of the actual setting you want to apply the finding to (ex: organization)
What is a correlation?
Can’t draw causative conclusions, low internal validity, relationship b/w variables can be positive (/) or negative () relationship; the association b/w two continuous variables
How to interpret correlation STRENGTH & MAGNITUDE
Can have temporal order
+ correlation increases in both the Y axis (IV) and X axis (DV)
- correlation increase in one axis and decrease in the other (IV horizontal and DV vertical)
-1 to +1 (-1 strong negative correlation) (+1 strong positive correlation)
Magnitude: the # after the sign, zero is the lowest value possible and 1 the highest
What is a multiple correlation? R^2
Used for predicting the unknown value of a variable from the know value of two or more variable also called predictors
What is incremental validity?
Extra variance explained by the set of predictors (increase in R^2) in our outcome variable (performance) by two predictors (intelligence &job related knowledge) is .4, we find that adding conscientiousness as a predictor bring the R^2 up to .5
What is Measurement?
Abstract concepts, is normally what I/O study call “constructs” (anything we are interested in such as a certain aspect of a person)
What is construct validity?
Concerns whether our methods of studying variable are accurate. Refers to the adequacy of the operational definition of variable
What is operationalization?
When you put a # to the abstract concept making it conceptual by making it quantitative and therefore measure able
What is measurement reliability?
High reliability (consistency), low reliability (non-consistent)
What is measurement validity?
High validity (on target for what we want to measure) Low validity ( we're not on target/ can't be valid w/o consistency)
What is content validity?
Extent to which the measure provides a good sample of the domain it is intended to represent
What is face validity?
Extent to which the measure looks valid (ask experts)
What is convergent validity?
Extent to which multiple measures of the same construct are highly correlated with each other
Tests the constructs that are expected to be related are, in fact, related.
Discriminant validity?
Extent to which measure of a construct are lowly correlated with measures of each construct
Tests that constructs that should have no relationship do, in fact, not have any relationship
Criterion related validity?
What kinds?
Empirical relationship between predictor and criterion
- Concurrent: predictor (IV) and Criterion (DV) measured at the same time (new Intelligence test against standard IQ test)
- Predictive: predictor (IV) measured some time before the criterion (DV) (ex: predict future performance/ high SAT scores for universities)
