CHP1: Science skills and research methods (Sources of error and prevention) Flashcards
Error: Participant-related variables
The differing individual characteristics that may impact how a participant responds in an experiment. In large sample sizes participant differences reflect the differences in the population but if a sample is biased or too small individual participant differences can introduce a source of error, as they are not representative of the population.
Prevention methods: Participant-related variables
Sampling procedure, choosing a different experimental design.
Sampling procedure
The method of choosing an appropriate sampling procedure to reduce possible extraneous and confounding variables. Using a sampling procedure that will select a representative sample can reduce a biased sample. Or using a large sample size will help reduce participant-related variables affecting the investigation.
Choosing an experimental design
The process of choosing an appropriate experimental research design to reduce possible extraneous and confounding variables. In a repeated measures design, the same participants are used in both conditions eliminating participant-related variables.
Error: Order effects
A source of error in a repeated-measures design due to the order the participants participate in the experimental and control conditions.
Prevention methods: Order effects
Counterblanacing.
Counterbalancing
Involves systematically changing the order of treatments or tasks for participants in a balanced way to counter the biasing influence or unwanted effects on the performance of any one order.
Error: Placebo effect
When an inactive treatment has an effect due to the participant’s expectations. Participant expectations can affect their behaviour and mental processes, resulting in them behaving differently or reporting different mental processes than they otherwise would as a consequence of their expectations.
Prevention method: Placebo effect
Placebo, single-blind procedure.
Placebo
A placebo is an inactive medication or treatment, used to control for participant expectations. Administering a placebo in an experiment controls for participant expectations because if there is a significant difference in the results of the active and placebo medication, the researcher can infer that the medication had an effect.
Single-blind procedure
A procedure in which participants are unaware of which condition they have been allocated to, in order to reduce participant expectations. Participants not knowing whether they are receiving the drug or not decreases the likelihood of participant expectations affecting results.
Error: Experimenter effects
The expectations that experimenters may have about the participants or the results may influence how they conduct the research or interpret the results. Often they are unconscious but can still affect the results of the experiment and can be extraneous variables eventually turning into confounding variables.
Prevention method: Experimenter effects
Double-blind procedure
Double-blind procedure
To address experimenter bias, a double-blind procedure can be used, which addresses both the participant and experimenter bias.
Having neither the experimenter nor the participant knowing the condition that the participant is allocated to ensures that their expectations don’t affect their behaviours or the experimenter’s interpretations of the results.
Error: Non-standardised instructions and procedures
If there are multiple experimenters each conducting the experiment with varying procedures and instructions errors can be introduced when procedures and instructions are not standardised for each participant.