ENVI 498 Lecture Notes - Lecture 5: Sampling Error, Observational Error, Hawthorne Effect
Subject
the Sampling of units for an experiment, usually human respondents who provide measures based on
the experimental manipulation.
Experimental Conditions
- One of the possible levels of an experiments (independent) variable manipulation.
- Variables included in the statistical analysis as a way of controlling or accounting for variance due to
that variable:
Blocking Variables
Categorical Variables
Covariate
Continuos Variable
Experimental Group
Group of subjects to whom an experimental treatment is administered.
Control Group
A group of subjects to whom no experimental treatment is administered.
Main Effect
The experimental difference in dependent variable means between the different levels of any single
experimental variable. (just blue)
Interaction Effect
differences in dependent variable means due to a specific combination of independent variables. (bright
and blue)
Manipulation of the Independent Variable
- Several Experimental treatment levels (different values of the independent may be used)
- More than one indy variable may be examined
find more resources at oneclass.com
find more resources at oneclass.com
Cell
A specific treament combination associated with an experimental group
Number of cells in an experiment
K= (T1)(T2)(T3).....(Tm)
Selection and Measurement of the Dependent Variable
- Selecting dependent variables that are relevant and truly represent an outcome of interest is crucial.
- Choosing the right dep-var is part of the problem of definition process
Sample Selection And RANDOM Sampling Errors
- Systematic or non sampling error: subject selection, experimental design, and unrecognized
extraneous variables
- Overcoming Sampling Errors:
Randomization
Matching
Repeated Measures
Control over extraneous Variables
Experiment Confound
When there an an alternative explanation beyond the experimental variables for any observed
differences in the dependent variable.
Sources:
Sampling error, systematic error, later-identified extraneous variables
Demand Effect
Occurs when demand characteristics actually affect the dependent variable.
Hawthorne Effect
find more resources at oneclass.com
find more resources at oneclass.com
People will perform differently from normal when they know they are experimental subjects.
Reducing Demand Effects
One treatment / Subject
Use a blind admin
Use a disguise
isolate subjects
Establishing Control Group
Constancy of conditions (identical conditions across groups and experiments)
Counterbalancing (attempts to eliminate the confounding effects of order of presentation by varying the
order of presentation (exposure) of treatments to subject groups.
Laboratory Experiment
A situation in which the researcher has more complete control over the research setting and extraneous
variables
Field Experiment
Research proposal involving experimental manipulation that are implemented in a natural env.
Within-Subjects Design
Involves repeated measures because with each treatment the same subject is measured. (different
lights on same guy)
Between-Subject Design
- Each subject receives only one treatment combination.
- Usually advantageous although they are usually more costly.
find more resources at oneclass.com
find more resources at oneclass.com
Document Summary
Subject the sampling of units for an experiment, usually human respondents who provide measures based on the experimental manipulation. One of the possible levels of an experiments (independent) variable manipulation. Variables included in the statistical analysis as a way of controlling or accounting for variance due to that variable: Group of subjects to whom an experimental treatment is administered. A group of subjects to whom no experimental treatment is administered. The experimental difference in dependent variable means between the different levels of any single experimental variable. (just blue) Interaction effect differences in dependent variable means due to a specific combination of independent variables. (bright and blue) Several experimental treatment levels (different values of the independent may be used) More than one indy variable may be examined. A specific treament combination associated with an experimental group. Selecting dependent variables that are relevant and truly represent an outcome of interest is crucial.
