Northern Advanced Research Training Initiative

1

• Northern Advanced Research Training Initiative
• Experimental Research Methods

Nikos Bozionelos PhD e-mail nikos.bozionelos_at_durh
am.ac.uk
2
Experimental Research
3
Steps in Research
Design Implementation (e.g., data
collection) Analysis Results
valid
?
4

• Validity
• the extent to which the explanation that
• is indicated by the findings is exclusive and
• sufficient

(i.e.,whether alternative explanations can be
ruled out)
extraneous factors (i.e., factors outside the
scope of the study) may be responsible for the
results
5

• Validity
• the extent to which the explanation that
• is indicated by the findings is exclusive and
• sufficient

(i.e.,whether alternative explanations can be
ruled out)
extraneous factors (i.e., factors outside the
scope of the study) may be responsible for the
results
confounding
6

• confounding
• history
• instrumentation (measures)
• attrition (mortality)
• maturation
• causality

control for confounding to improve validity
7

• confounding
• history
• instrumentation (measures)
• attrition (mortality)
• maturation
• causality

confounding imposes threats to the validity of
research
control for confounding to improve validity
8
types of validity

• internal
• the extent to which extraneous factors
• cannot be responsible for the results
• external
• the extent to which the results can be
  generalised to other settings/environments

9
internal validity
has priority
the research method to improve/ensure internal
validity is the
experiment
10
experimenta research study conducted under
strictly controlled conditions

• only two factors/variables
• are allowed to change
• the Independent Variable (cause)
• the Dependent Variable (effect)

11
experimenta research study conducted under
strictly controlled conditions

• only two factors/variables
• are allowed to change
• the Independent Variable (cause)
• the Dependent Variable (effect)

manipulated by the Researcher
changed by the effects of the Independent
Variable
12

• aim to prove with certainty

independent variables
dependent variables
cause
allowed to change
manipulated by us
13

• types of experiments
• laboratory experiment

what do we compromise?

• field experiment/quasi experiment

greater external validity
14

• How do we prove causality?
• control group
• (no change in the Independent Variable)
• experimental group
• (change in the Independent Variable)

15

• controls in experiments

16
in experiments we must control for

• demand characteristics
• experimenter and participant effects

• how?
• single-blind double-blind procedures
• automation
• objective measures-multiple observers
• deception

17
in experiments we must also control for

• the characteristics of participants

• how?
• random selection of participants
• random assignment into experimental conditions

18
END Thank You