Term
|
Definition
| Observing without imposing a treatment (surveys are a type) |
|
|
Term
|
Definition
| Deliberately imposing a treatment on individuals and observing their responses |
|
|
Term
| The Three Principles of Experimental Design |
|
Definition
| Conrtol, Randomization, Replication |
|
|
Term
|
Definition
| Anticipate confounding variables and control for them |
|
|
Term
|
Definition
| Use chance to assign individuals to groups or treatments |
|
|
Term
|
Definition
| Enough subjects in study to reduce effects of chance variation |
|
|
Term
|
Definition
| Explanatory variables in study, can have multiple levels |
|
|
Term
|
Definition
| Trying to find what extent the explanatory variable has on the response variable |
|
|
Term
|
Definition
| Two variables are confounded when you cannot tell which one (or both) had the effect |
|
|
Term
|
Definition
| A variable is lurking if it drives two other variables |
|
|
Term
|
Definition
| Blocks help control confounding variables. Blocking creates groups that are similar with respect to the blocking factors. Treatments assigned randomly within each block. |
|
|
Term
|
Definition
Type of blocked design. Two treatments. Each block 2 similar units/individuals.
OR
One individual has both treatments one after the other. Randomize to see which one goes first. |
|
|
