Chapter 7 - Test 3 Flashcards
Source of threats to Internal Validity
- People are different (group)
- People change (time)
- Process of studying people changes people (participant reactivity)
Artifacts
- threaten external validity
- held constant across the levels of the independent variable
Confounds
*threatens internal validity
* different for the different levels of the independent variable
Threats to Internal Validity
*Time Threats
* Group Threats
*Reactivity
Time threats
- History (external)
- Maturation (internal)
- Testing (repeatability)
- Instrumentation (different instruments to measure same confounds)
Group threats
- Selection
- Selection by maturation
- Selection by ???
- Mortality/Attrition
- Regression
Reactivity
- Diffusion/Contamination
- Demand Characteristics
History Threat
Changes due to environmental events other than the treatment
(outside the person)
Instrumentation Threat
Measuring instrument changes during the course of the study and these changes cause participants’ score to change
Selection
Groups were different to start with (before experiment is run)
Selection by maturation
Groups that are similar (but not identical) to start with may naturally grow apart
Diffusion/Contamination
Participants may change through observation, or communication with other participants
GAGEs (CONFOUNDS)
Geography
Age
Gender
Ethnicity
Social economic status
OOPS! (ARTIFACTS)
Operationalization
Occasions
Populations
Situations
Other Threats
- Demand characteristics (Hawthorne Effect)
*Participant expectancies
*Participant reactance
*Evaluation apprehension
*Response sets:
*acquiescence (agree with everything)
*nay-saying (disagree with everything)
*middle of the road (select neutral options)
Pygmalion Effect
Example: Teachers spend more time with students that are expected to get good grades. In the end, the good grades are due to the teacher spending more time with these students because of the higher expectations.
John Henry Effect
The Control group knows that they are and tries harder than the non-control group
Placebo
Telling participant they have received treatment when in actuality that have not
Nocebo
Blindly providing treatment, yet the patient does not know that they have received the treatment
single-blind
either the participant or the experimenter does not know
double-blind
both the experimenter and the participant do not know
Selection by _____________
Selection can also interact with other threats, such as X History interaction
Mortality/Attrition
Drop out of study like longitudinal overtime studies
* Homogeneous - just as likely to drop from a control group than an experimental group
* Heterogeneous - people are dropping from one group more than the other.
Overall scores change because of people dropping out of the study
Regression (to the mean)
Participants who get extreme scores have less extreme scores when retested because extreme scores are extreme, in part, due to random error; since the random error is inconsistent, chances are that random error will not make participants score quite so extreme the second time around.
SPORTS PERFORMANCE
Hawthorne Effect
People react to being studied, and productivity would naturally increases when an employer does anything with the employees (like bringing donuts for the day)