IB Psychology/External Assessment/Paper 3
3.1. Hypotheses 2 tailed = predict an effect - such as a difference or correlation (no specific direction - can go either way)
1 tailed = predict a particular direction of the effect e.g. that positive correlation will occur (directional)
3.2 Experimental methods
Laboratory - researcher deliberately manipulates the independent variable while maintaining strict control over the other variables in a controlled, artificial environment. E.g. Bandura - independent variable of exposure to aggression, dependent variable of imitation of aggression in children under controlled lab conditions by randomly allocating children to either a condition where they saw 1) an adult being violent towards the Bobo doll, or 2) where the adult showed no violence. The number of aggressive acts by each child were also measured in the laboratory.
This is the most scientific method because it demonstrates cause and effect. lab experiments increase control and accurate measurement of variables - therefore more objectivity. lab standardisation means more chance to replicate the results again. Weaknesses - total control over variables is not possible, artificial setting will cause a lack of ecological validity, results may be more biased due to sampling, demand characteristics... ethical problems may arise due to deception and mistreatment of subjects
Field experiment - researcher deliberately manipulates the independent variable, but does so in the subject's natural environment.
This type of experiment will have greater ecological validity and less bias from sampling, and ecological validity (subjects often do not know they are being observed)
HOWEVER, it is more difficult to replicate this type of study, it is difficult to record data (most data will be qualitative). Big issue of ethics - invasion of privacy, lack of consent, and deception. Infiltration of a group may prove dangerous and is unfair on the subjects. however, subject privacy is usually observed (i.e. subjects or locations are not named).
3.3 Sampling - No one under the age of 16 should be used.
Random - truly random and occurs only if every member of the target population has a chance of being chosen. the best chance of an unbiased sample, however very difficult to test randomly because of often large target population size and people who are selected from the population may not want to take part in the experiment.
Stratified - dividing the target population into subcategories (strata) and then selecting members of these subcategories in the proportion they occur in the real target population e.g. target pop. consists of 75% men and 25% women, therefore 15 men and 5 women used in the experiment.
Opportunity - selecting subjects that are available at the time. e.g. university psychologists may sample from their students. Quick and convenient, however more biased and risk of researcher choosing subjects who are 'more helpful' or demand characteristics from the participants if they have a personal relationship with the experimenter e.g. tutor/student relationship...
Volunteer - volunteers who reply to press adverts asking for participants in a study. A convenient method as volunteering is made on the basis of informed consent and is ethical. however volunteering may show an excessive keeness on the part of the participants and may lead to demand characteristics during the experiment. Biased on the part of the subject.
3.4 Experimental design - 3 main types. 1. Repeated measures - using the same subjects in each condition of the experiment. e.g. giving a group of subjects a test with no alcohol influence, followed later by the same test with alcohol influence. subjects may do better in the second condition due to learning, different tests may be needed - this makes the experiment lack ecological validity.
2. Independent measures - using different subjects in each condition
3. Matched pairs - different but similar subjects used in each condition
NB inferential statistics = Repeated measures design (Wilcoxon test); Independent measures and matched pairs design (Mann Whitney U test)
3.5 Nomothetic vs idiographic approaches to psychology
3.6 Triangulation
3.7 Types and functions of case studies
3.8 Content analysis
3.9 Observational research
3.10 Questionnaires, questions and transcription
3.11 Verbal protocols
3.12 Types of data - nominal/ordinal and descriptive statistics
3.13 Self report methods