Research Methods

Question	Answer
Idiographic	Sees each person as unique and assume they should be treated as an individual - generalisations are innappropriate Research methods - interviews; case studies
Nomothetic	Assumes people are similar, direct comparisons can be made. Assumes findings from a sample can be generalised Research Methods: Experiments; Questionnaires; Content analysis
Qualitative Data	Non-numerical data, often thoughts, feelings and emotions
Strengths of Qualitative Data	- Data is detailed and rich, full explanations - People reporting on real-life behaviour so HIGH ECOLOGICAL VALIDITY. - Participants are treated as humans not numbers
Weaknesses of Qualitative Data	- Cant be represented in a graphical format, comparisons cant be made - Difficult to replicate findings - Data is highly subjective - as is interviewers opinion
Quantitative Data	Numerical data - based on behaviours which can be measured
Strengths of Quantitative Data	- Can be analysed statistically, generalisations can be made - Objective, easy to replicate
Weaknesses of Quantitative data	- Behaviour may be reduced to simple numbers, lack of detailed information - Participants may feel dehumanised as they are simply another number
Hypothesis	A testable statement or prediction about what the researcher predicts to happen. A hypothesis needs to be testable so its clear what the conditions are and what is being measured.
Null Hypothesis	This states there will be no difference or relationship.
Experiment Aspects	An experiment is an investigation in which the IV is altered in order to cause a change in the DV while trying to keep all other variables constant. These other variables are called extraneous variables.
Independent Measures	When DIFFERENT participants take part in each condition. Participants should be randomly allocated to each condition to avoid bias.
Repeated Measures	When the same participants take part in all the experimental conditions. This means that fewer participants will be needed. But there may be order effects (these can be reduced by counter balancing).
Matched Pairs Design	When a participant in condition A is paired with one in condition B
Demand Charecteristics	When the participants try and figure out the aim of the experiment
Experimenter Bias	The experimenter - deliberately or accidentally - treats a person/group of participants differently. This can be reduced with STANDARDISED instructions.
Double-Blind	Where neither the experimenter nor participants are aware which condition they are in.
Field Experiment	When the research takes place in a natural setting. This has high ecological validity. But its difficult to deal with ethical issues (due to lack of informed consent), and it can be v. difficult to control extraneous variables.
Laboratory Exeriment	Where the experiment takes place in an artificial, carefully controlled setting. It allows extraneous variables to be controlled, it is easy to establish whether the IV has effected the DV, precise measurement of the DV can be made. However it lacks ecological validity and the artificial setting may make people behave differently to how they would in real life.
Measures of Central Tendancy	Mean-takes all the scores into account, easily distorted by high values Median-middle value:not distorted by extreme values, doesn't take into account the value of each value Mode- simple to calculate, least sensitive
Measures of Variability or Dispersion	Standard Deviation-shows how far on average scores are from the mean. Range- simple to calculate, can be misleading
Target Population	The overall group of people from whom we take a sample
Opportunity Sampling	Making use of whoever is available - quick and convenient; might be unrepresentative; researchers may choose family so there would be demand charecteristics
Random Sampling	When everyone in the target population has an equal chance of being selected. This could be done by picking the names out of a hat. This is unbiased, can be time consuming to identify everyone, could be unrepresentative.
Stratified Sampling	When a random sample is taken from each group in a population. This is HIGHLY REPRESENTATIVE, the researches has no direct control so it is UNBIASED, however it is very TIME CONSUMING.
Systematic Sampling	Taking every nth person who comes along/listed on the target population. Relatively quick and simple, might not be representative.
Correlation Studies - finding the correlation between two or more variables	They allow us to see if there is a relationship which may form the basis for an experiment. We cannot see which variable is affecting the other so more research is needed.

Next up

Description

Resource summary

Similar

	Created by ellensmcgee over 10 years ago