Extraneous Variables

Download Report

Transcript Extraneous Variables

Controlling Extraneous Variables
Environmental issues
Naturally occurring variables: usual with several
groups of people
• Noise
• Temperature
• Light
• Time of day
• Seating arrangements
Artificiality:
Classroom or Office are not “real world” or “natural”
Using a recording device or a camera
People act differently when they are aware they are
being observed
People change attitudes and actions when social
situation changes
People often want to please and be praised
People are influenced by physical environment and
personal physical state
People bring emotional states caused by extraneous
factors
Self-selection: let subjets decide which group to join
• Volunteers: different from the outsed, more highly
motivated
• Two class sections could have wide differences
Mortality (attrition, dropout):
Students elect to leave a group: leaves slant toward more
motivated
Maturation: studies conducted over longer period must
account for changes in group and individuals
People Issues
Hawthorn Effect: pleasure of being included in the
study enhances performance
Halo Effect: people respond well to those whom they
like.
Subject Expectancy: people tend to want to please
Researcher Expectancy: Subtle and uncontrolled
attitude toward subjects: like high performers, dislike
low performers
Measurement Issues
(Results are only as good as the data.)
Measures (tests, questionnaires, interviews) should
not introduce extra variables
Practice Effect: same test given repeatedly
Reactivity Effect: attitude surveys form or solidify
attitudes in response to survey
Instability of Measures and Results: degree to
which results likely to recur if study is replicated
Controlling for Validity
Logic = internal validity
Meaningfulness = external validity
Control extraneous variables
• Environmental issues: comfortable
• Grouping issues (avoid volunteer grouping): take steps to assure
that the group is representative and equivalent
• Short-term study avoids maturation
• People issues:
• try to find a double-blind technique
• Minimize obviousness to subjects
• Control expressions of attitude
• Measurement issues:
• Counterbalancing: no individual takers of the same test more
than once
• Reactivity: question subjects and gauge thought processes
Control versus Generalizability
If two or more uncontrolled extraneous variables also
influence each other…
Example: Researcher expectancy can trigger subject
expectancy, creating a practice effect.
INTERNAL VALIDITY (Logic): results due solely to variables
identified and compared.
EXTERNAL VALIDITY (Meaningfulness): Factors affect
generalizability of study to real world, i.e. artificiality.
1. Purpose: purpose of experimental research is to study cause-andeffect relationships between two variables, whereby the causal
variable can be manipulated.
2. Hypothesis: Stated at the beginning of the study.
3. Data Collection/Sources: quantitative data from test scores,
attitudinal surveys and other measurements. Researcher controls
for internal validity.
4. Data Analysis: data are coded and converted into an electronic
format and analyzed using a computer software program such as
SPSS. Researcher can use any number of analytical processes.
Differences in rwo or more groups using t tests and analysis of
variance.
5. Reporting Results: brief, clear and to the point. Description of
methodology and procedural controls. Results of data analysis
presented. Can infer external validity if threats to external validity
were controlled.