PSYCH701 QUANTITATIVE RESEARCH DESIGN.
Campbell: factors relevant to validity.
Seek to control variables in experimental designs.
Validity is placed in 2 categories: internal and external.
Internal validity: did the stimulus make some difference in the specific
context?
External validity: generalizability. To what populations, settings can this effect
be generalized to?
Effects of these variables occur at 2 levels: simple or main effects. Occur
independently or in addition to the effects of variable.
Main effects are relevant to IV.
HISTORY: during the span between O1 and O2 many events might have
occurred in addition to X.
MATURATION: natural changes of time. getting older, tireder.
TESTING: taking the test a second time changes your scores compared to
first time. distinction between reactive vs nonreactive measures.
Reactive measures modify the phenomenon studied. when subject is self-
conscious.
INSTRUMENT DECAY: a problem when ppl are used as measurement
apparatus. Ppl age or become more experienced.
STATISTICAL REGRESSION: mean will shift due to random imperfections.
Random instability.
Interaction effects in EV (representativeness).
The obtained effects are specific to the experimental population and don't
hold true to populations. We want to generalize too.
The greater the cooperation required from participants, the more they must
deviate from the normal course of events. This will create possibility of non-
representative reactions.
The longer the experiment is extended in time, the more respondents are
lost. And the less representative of the groups of the original.
Reactive arrangements.
Respondents become aware that they are participating in an experiment. Has
an interactive effect.
Direction of the effect may be negative. They will be unwilling to. Change in
persuasion.
Effects may be positive, and they will cooperate better.
READING DURRHEIM
Principles for Good/Valid Research Design.
Principle one: achieved design validity by identifying plausible rival Hypo.
Thesis and eliminating. Remove their presence or determine their
influence.
Principle 2: achieve design coherence by ensuring the research purposes
and techniques are arranged logically to fit the framework.
Example: One chance to get the experience of motherhood using a survey.
Instead of interpretive paradigm is incoherent.
Campbell: factors relevant to validity.
Seek to control variables in experimental designs.
Validity is placed in 2 categories: internal and external.
Internal validity: did the stimulus make some difference in the specific
context?
External validity: generalizability. To what populations, settings can this effect
be generalized to?
Effects of these variables occur at 2 levels: simple or main effects. Occur
independently or in addition to the effects of variable.
Main effects are relevant to IV.
HISTORY: during the span between O1 and O2 many events might have
occurred in addition to X.
MATURATION: natural changes of time. getting older, tireder.
TESTING: taking the test a second time changes your scores compared to
first time. distinction between reactive vs nonreactive measures.
Reactive measures modify the phenomenon studied. when subject is self-
conscious.
INSTRUMENT DECAY: a problem when ppl are used as measurement
apparatus. Ppl age or become more experienced.
STATISTICAL REGRESSION: mean will shift due to random imperfections.
Random instability.
Interaction effects in EV (representativeness).
The obtained effects are specific to the experimental population and don't
hold true to populations. We want to generalize too.
The greater the cooperation required from participants, the more they must
deviate from the normal course of events. This will create possibility of non-
representative reactions.
The longer the experiment is extended in time, the more respondents are
lost. And the less representative of the groups of the original.
Reactive arrangements.
Respondents become aware that they are participating in an experiment. Has
an interactive effect.
Direction of the effect may be negative. They will be unwilling to. Change in
persuasion.
Effects may be positive, and they will cooperate better.
READING DURRHEIM
Principles for Good/Valid Research Design.
Principle one: achieved design validity by identifying plausible rival Hypo.
Thesis and eliminating. Remove their presence or determine their
influence.
Principle 2: achieve design coherence by ensuring the research purposes
and techniques are arranged logically to fit the framework.
Example: One chance to get the experience of motherhood using a survey.
Instead of interpretive paradigm is incoherent.
