CRIM 220 Midterm: Chapter 7

Independent variable

the cause

Dependent variable

the effect

Pretesting & posttesting

participants are measured on a DV (pretest), exposed to the IV, & remeasured on the DV (posttest)

• issue = responses are different the 2nd time

Control group

the group that doesn’t experience the experimental stimulus

Experimental group

the group that experiences the experimental stimulus

• compared to the control group

Double blind experiment

when neither the observer/experimenter nor the participants kn

Random assignment

randomly assigning participants to the experimental or control group

• produces groups that are statistically equivalent

Threats: history

historical events may happen during an experiment

Threats: maturation

people are constantly growing & changing

• longitudinal studies = age, wisdom

• experiments = boredom, tiredness, etc.

Threats: testing

testing & retesting influences behaviour

Threats: instrumentation

measuring instruments may change over time

• creates a reliability problem

Testing vs. instrumentation

testing:

• changes in how participants respond to a measure

instrumentation:

• changes in the measurement process

Threats: statistical regression

regression to the mean ⇒ interest in cases that have extreme scores on a variable

Threats: selection biases

random assignment may be chosen in ways that threatens validity

Threats: experimental mortality/attrition

participants may drop out before the experiment is done

Threats: causal time order

the DV changed the experimental stimulus

Threats: diffusion/imitation of treatments

experimental group may pass elements of the stimulus to the control group

Threats: compensatory treatment

pressure to offer compensation when participants in the control group are deprived of something valuable to them

Diffusion & compensatory treatment

accidental or intentional extensions of the experimental stimulus to the control group

Threats: compensatory rivalry

participants deprived of the experimental stimulus might try to compensate by working harder

Threats: demoralization

deprivation can lead to participants giving up

Demoralization & compensatory rivalry

based on participants reactions to the experiment

Quasi-experiment

includes most, but not all, elements of an experimental group

• quasi = to a certain degree

• doesn’t use random assignment

Nonequivalent-groups design

experimental group participants matched with comparison group participants

Time series design

examining a series of observations on a variable over time