Causality
tigers2019_
Dantzker_ITK_164375_CH061.pptxChapter 6 Introduction to Research Design
Empirical Observation
Seeking answers
Direct observation to arrive at conclusions
Causality (1 of 2)
Idiographic
Examination of why a single event occurred
Nomothetic
One or a few factors provide a general understanding of the phenomena being studied
Based on probabilities
Causality (2 of 2)
Criteria
Time order
Association
Elimination
Necessary cause
Condition must occur for another to take place
Sufficient cause
Cause usually, but not always, creates effect
Classical Experiment
Random assignment
Experimental group (subject to experimental stimulus)
Control group
Pretesting and posttesting
Pretest–Posttest Design
No control group
Can compare before and after intervention
No causality, presence of external variables
Posttest-Only Design
Experimental and control groups
No pretest
Cannot assess impact of external stimulus
Factorial Design
Adds to classical design
Has two or more experimental stimuli
Different dosage, or entirely different
Experimental Design: Advantages
Isolate experimental variation
Assess impact over time
Can require little time, money, subjects
Ability to replicate
Experimental Design: Disadvantages
Artificiality
In criminal justice and criminology
Expensive
Logistically difficult
Difficult to maintain environment
Quasi-experimental Design
Offers partial control
Allows approximation of experimental conditions
No random assignment
No manipulation of stimulus
Easier to implement
Hard to separate effect of treatment
No truly equivalent groups
Levels of Measurement (1 of 2)
Nominal
Categories with no specified order
Mutually exclusive
Ordinal
Assigns rank or placement to categories
Likert scale: strongly disagree, disagree, neutral, agree, strongly agree
Levels of Measurement (2 of 2)
Interval
Scores
Expected equality in the distance between scores
No starting point
Ratio
Has an absolute zero point
Differences in scores are equal and explainable
Summary
Based on empiricism
Experimental
Classical
Pretest-posttest
Posttest-only
Factorial
Quasi-experimental
Level of measurement
Nominal
Ordinal
Interval
Ratio
