WEEK4- Experimental designs Flashcards
Characteristics of a basic lab experiment
What is manipulated
Control?
allocation?
What kind of relationship?
Manipulation of the IV to see the effects on the DV
An attempt to control for confounding variables
Random assignment
Casual relationship can be identified between the IV and DV
Why is randomisation important?
Avoids bias
Each outcome has an EQUAL chance of being selected (if there’s not an equal chance for every outcome it is not random)
What are the two types of experimental design?
definition
Other terms for the design?
unrelated design: different participants in different conditions
Unrelated design could also be referred to as between participants design or independent groups design
Related design: Same group of participants do the different conditions
Related design could also be referred to as repeated measures design or within participants design
What are the to main reasons for using a research lab
Practical: the equipment is easily accessible and the apparatus can be kept safe (expensive)
Experimental control: all factors, excluding the manipulated variables, can be kept constant (easy to do in a lab) e.g. temperature
Problems with lab experiments
Four types of effects?
Practice effects: If participants have already completed the task before it may lead them to improve in the task as they are familiar with the procedure from earlier conditions
Fatigue effects: Could occur if the participants become tired or bored therefore they do not perform as well on later conditions
Context effects: The environment in which a participant experienced event effects can influence how the event is remembered
Experimenter effects: if the experimenter or researcher if present during the experiment if may impact the participants performance
Experimental controls (2)
Counter balancing
- Eliminates order effects ( Group 1 does condition B first and Group 2 does condition B then switch)
Matching
Participants are matched on key characteristics that may impact their performance on the task ( eg age or intelligence) and each participant from each pair gets allocated to the experimental condition or in the control group
What is Pre-test
What does it check?
What is difficult to identify if this hasnt been done?
Pretest: checking if the use of random assigment has really equalised the experimental and control groups on the dependant variable before the experimental manipulation ( otherwise it is difficult to identify if the differences after the manipulation is because of the manipulation or already existing differences between the groups on the DV)
Measurement before manipulation
What is post-test
Measurement after manipulation
Does Using random assignment truely give accurate reasults
Using random assignment alone does not give accurate reasluts since partcipants in one condition could still have characteristics which may impact the DV therefore researchers have to use Pre test and Post test to ensure that the experimental and control groups are balanced on the DV
What is a pre-test- post-test design?
Experimental designs which include a pretest
What is a post-test only design?
Experimental designs without a pre-test
How do you identify that there has been a pre-test sensation interaction effect?
If the interaction between the pre-test and the experimental condition is statistically significant
What are the two main advantages of having a pre-test
Allows you to identify if randomisation has worked in terms of balancing the groups on the DV before the experimental manipulation
Allows us to determine if there has been a change in performance between the pre-test and post-test (only having the post-test scores will not allow us to identify any changes and what it is so without a pre-test research may incorrectly suggest that the IV is increasing the scores on the DV)
Disadvantage of having a pre-test
May suggest the purpose of the experiment to participants which influences their behaviour
ways to minimise pretest sensation
What could be done as well as the pretest?
What could be extended? What effect is reduced?
Carry out the pre-test as well as an unrelated task to disguise the true purpose of the experiment.
Have an extended period of time between the pre-test and the manipulation to try and reduce the practice in the pretest having an effect on the post test results (reduce practice effects)
