SQA Higher Psychology
45 apuntes de estudio, creados por cookie, en 9 secciones que siguen los temas del curso.
Notes for SQA Higher Psychology (C863 76), in folders for the research and assignment area, the mandatory sleep and dreams and conformity and obedience topics, then one folder for each of the six optional topics, in the order of the course. Delete the option folders your course does not take.
Qué contiene
La primera nota
Research and the assignment / Experimental method and variables
Psychology tries to explain behaviour by testing ideas, and the most powerful way to test whether one thing causes another is the experiment. An experiment changes one factor on purpose, keeps everything else as constant as possible, and measures what happens. Because the researcher controls the change, a difference in the results can be traced back to that change, which is what allows a claim of cause and effect.
Aims and hypotheses
Every study starts with an aim, a plain statement of what the research is trying to find out. A study of whether noise affects memory, for example, has the aim of finding out whether background noise changes how many words people recall.
The aim is turned into a testable prediction called a hypothesis. A directional (one-tailed) hypothesis says which way the results will go, such as "participants who study in silence will recall more words than participants who study with background music". A non-directional (two-tailed) hypothesis says only that a difference will exist, such as "there will be a difference in recall between participants who study in silence and those who study with background music". A directional hypothesis is used when earlier research or a clear theory gives a reason to expect a particular direction.
Every experiment also has a null hypothesis, which states that any difference found is due to chance. The aim of the statistical test is to decide whether the results are unlikely enough under the null hypothesis that it can be rejected.
Independent and dependent variables
The independent variable (IV) is the factor the researcher changes or selects. In the noise example the IV is the condition in which people study. Each level of the IV is a condition: silence and music, for instance.
The dependent variable (DV) is the factor the researcher measures. It is the outcome that is expected to depend on the IV. Here the DV is the number of words recalled in a test afterwards. A good DV is measured in a way that can be counted or scored, such as a number of words, a reaction time in milliseconds or a rating on a numbered scale, and the researcher must define exactly how it is recorded. Defining a variable so that it can be measured and repeated by someone else is called operationalising it.
Extraneous variables and confounding variables
Extraneous variables are all the other factors that might affect the DV. If one group studied in a bright room and the other in a dim room, the lighting would be a confounding variable: it changes along with the IV, so the researcher cannot tell whether the IV or the lighting caused any difference. Confounds damage the conclusion of a study because they offer a second explanation for the result.
Researchers control extraneous variables in several ways. They keep the room, the time of day, the instructions and the test the same for everyone. They may use standardised instructions, which are read word for word. Participants who are not told which condition they are in, or who are tested by someone who does not know the hypothesis, reduce the effect of expectations.
Participant variables are a particular kind of extraneous variable. People differ in intelligence, mood, tiredness and prior knowledge, and these differences can change the DV without any link to the IV. Random allocation, in which each participant has an equal chance of being put in each condition, spreads these differences evenly across the groups so that they are unlikely to favour one condition.
Experimental designs
The way participants are allocated to conditions is the experimental design, and each design has strengths and weaknesses.
An independent groups design uses a different set of participants for each condition. Its main advantage is that no participant takes part in more than one condition, so practice effects and carry-over cannot happen. Its main weakness is participant variables: the two groups may differ before the study begins, which is why random allocation is needed.
A repeated measures design uses the same participants in every condition. Its advantage is that the participant variables are the same in each condition, so there is less random variation in the results and a smaller sample can be enough. Its weakness is order effects. Practice can improve performance on the second test, and fatigue or boredom can reduce it. Counterbalancing deals with this by splitting the participants so that half do the conditions in one order and half in the other, which spreads the order effect evenly.
A matched pairs design sits between the two. Participants are matched in pairs on a relevant variable, such as age or a test score, and one member of each pair is put in each condition. It keeps some control of participant variables without the order effects of a repeated measures design, but matching takes time and works only for the variables that are matched.
Cause and effect
An experiment can support a claim of cause and effect because the researcher decides what the IV is, sets it up in the same way for each participant, and then compares the DV across conditions. If the controls are good, the only systematic difference between the conditions is the IV, so any consistent difference in the DV is evidence that the IV caused it.
This claim is only as strong as the controls. A confound, an experimenter who unintentionally signals the expected result, or a sample that does not represent the population can all produce a difference that is not caused by the IV. A researcher should therefore say what steps were taken to prevent these problems, and a critical reader should ask which alternative explanations remain.
Non-experimental methods can show that two things are related, but they cannot usually show that one causes the other, because the researcher has not changed anything.
Evaluating an experiment
A strong evaluation of an experiment looks at its control and at its realism. Strengths include the ability to show cause and effect, the replicability of standard procedures and the chance to test a hypothesis directly. Weaknesses include the risk of confounds, the possibility of demand characteristics (participants working out what is expected and behaving accordingly), the limits of a small or unrepresentative sample and the question of whether the setting reflects everyday behaviour. Each of these points is most useful when it names the effect it might have on the results: a demand characteristic, for example, may make participants recall more words because they want to appear to be doing well.
Y 44 más: abre cualquiera desde la lista de arriba.
Reseñas
Todavía no hay reseñas escritas. Añade estas notas a las tuyas y podrás ser la primera persona en dejar una.