Random Error
Chance difference between the observed and true values of something i.e. participants not understanding the task instructions
Systematic Error
A consistent difference between the observed and true values of something i.e. a typo in the instructions
Selection Biases
-Differences exist between the two groups prior to the manipulation of the IV
- Control through random assignment
History Effect
- Occurs when an event outside the study effects the outcome of the study
- Control through conducting the experiment over a short period of time and randomisation
Maturation Effect
-Occurs when changes within participants of a study changes the outcome of study
- Control through randomisation and conducting the experiment over a short period of time
Learning/ Practice Effect
- Result in better post-test performance because participants in a group have become more familiar with some aspect of the experiment from the pre-test
- Control through a post-test only design
Experimental Fatigue
- Experimental situations can lead to physical/mental fatigue
- Control through matching the amount of required effort across all experimental conditions
Order Effects
-Occurs when the effect on the IV is different depending on which experimental condition/ DV measurement is experienced first
- Control through randomising the order in which participants experience the experimental conditions
Carry-over Effects
-Occurs when the effects of a prior experimental condition continues to influence a participants performance in a later condition
- Control through the inclusion of a washout period between experimental conditions
Attrition Effect
-Participants will often drop out of a study
- This is a problem for internal validity if attrition is different across all conditions and systematically related to the IV
- Control by providing incentives to stay in the study
Expectancy Effects
- Occurs when the experimenter subtly communicates their expectations to the participants, who alter their behaviour to conform to these expectations
-Control by training experimenters well
. Blinding experiments
. Using multiple experimenters
Motivated Reasoning in a study design
- Occurs when the experimenters own personal biases influence the design of the study to make a specific outcome more likely
- Control through peer review of study protocols
Reactivity
-Occurs when individuals alter their performance or behaviour due to the awareness that they are being observed.
- Control through .Deception
. Blinding
.Engaging tasks or habituation
Demand Characteristics
- Any cue i.e. rumors about the study, communication between participants etc. that is outside the actual experimental situation
that may guide or bias a participants behaviour - These cues may lead to the participants guessing the hypothesis and taking on a certain role in response.
Social Desirability
- Participants edit their responses so that they are viewed favorably
- Control by ensuring anonymity
. Assessing outcomes more indirectly
Sample Generalisability
Do the research participants represent the overall population or are they different in some way
Setting Generalisability
Is there something unique about where the research took place that might effect the results
Temporal Generalisability
Is there something unique about when the research took place that might effect the results
Validity Tradeoffs
- The more internal validity- the less generalisability
- The more external validity- the greater the chance of bias
