research methods Flashcards
convenience sampling
Convenience sampling is when subjects are picked based on their availability at the time of the experiment. Quick, easy, cost effective.
Convenience sampling usually presents a biased sample. They may not be representative of the population.
random sampling
Random sampling is where every member of the population has an equal chance of being selected for the sample being used in the study.
Quick(er than stratified), relatively simple to employ. Less chance of bias, however can be open to bias if the population has under or over represented groups in it.
stratified sampling
Stratified sampling is to be used when there is diversity or under/over represented groups in the population.
Stratified sampling involves breaking the population into groups or ‘strata’ based on the characteristics you wish to control for in the sample.
This could be age, gender, ethnicity etc.
Once the population is divided into strata, participants are selected for the sample in the same proportions (ratios, percentages) that exist in the population.
If random stratified sampling is used, this just means that the selection from the already defined strata uses the equal chance methods (ie names in hat, number generator)
research designs - independent groups
participants are randomly allocated into a control or experiemental group
+ more time efficiant
- less control over individual differences
research design - matched participants
Involves first pre-testing participants on a certain characteristic (that could affect the DV) and then matching them up on the characteristics they’re very similar in (such as IQ). One member of each pair is randomly allocated to either the control or experimental conditions.
+ Eliminates potential confounding variable based on whichever characteristic has been matched & no order effect.
-Requires time to match /pre-test & if one person from pair drops out the other is also removed
research design - repeated measures
Is where each participant is exposed to both the
experimental & control conditions.
+ Almost completely eliminates potential confounding
variables associated with individual participant differences.
Counterbalancing involves changing the order of treatments or tasks for participants in a balanced way to ensure the order effects are offset. EG Half of participants receive treatment 1 then 2, while other half receive treatment 2 then 1.
conclusions
concluding statement (this is what the results mean), linking back to aims of study and theory of study, refer to data collected and or hypothesis
validity
internal - iv caused the changed in the dv
external - the extent to which the findings can be generalised out side of the study
cross sectional studies (not considered a true experiment as participants cannot be randomly assigned to control and experimental groups)
looking at data from a population at a given specifiic time
eg. look at people with different ages but at the same time
+ Can be used to study change over time.
Provides a means of conducting research on certain topics that are unethical and/or impractical to conduct through experimentation.
- Cause-effect relationship between different variables cannot be tested or determined.
Cohort effect; occurs when the researcher measures characteristics in groups of people (‘cohorts’) born at significantly different times and members of each group share life experiences associated with the period and/or place in which they grew up.
case study
A case study is an intensive, in-depth investigation of some behaviour or event of interest in an individual, small group, organisation or situation. Usually, the case is a person. It may involve the ‘normal’ or ‘abnormal’ behaviour or functioning of an individual.
+ useful way of obtaining detailed and valuable information on mental processes and behaviour, particularly in relation to rare or unusual disorders.
- Case studies, however, cannot be replicated to test the reliability of the results in the way that an experiment can.
Nor can they be used to actually test hypotheses unless combined with the results of other case studies of similar participants or used with another research method that is suitable for testing hypotheses.
observational studies
An observational study involves collecting data by carefully watching and recording behaviour as it occurs. Psychologists use observational studies to collect data when the behaviour under investigation is clearly visible and can be easily recorded. Data collection may be:
structured — a pre-prepared system is used to guide and record observations e.g. checklist of items to look for
unstructured — observations are made without a predetermined format
semi-structured — data is collected using a partly predetermined format.
+ The main advantage of observational studies, especially naturalistic observation, is that researchers can watch and record behaviour as it usually occurs, without the need for any manipulation or intervention.
- limitation of an observational study is that it can be difficult to determine the causes of the behaviour of interest that is observed, because many factors may influence that behaviour.
self report
A self-report is the participant’s written or spoken responses to questions, statements or instructions presented by the researcher.
Questionnaires, interviews and rating scales are the most commonly used self-report tools.
+ By guaranteeing anonymity, questionnaires in particular, provide a means of collecting self-report data on ‘sensitive’ or controversial topics that many people are not willing to disclose publicly, such as in an unstructured oral interview.
-they rely on the assumptions that people are actually willing to answer all questions and that they will give accurate answers.
social desirability. People may intentionally give false or misleading answers to create a favourable impression of themselves.
hypothesis
A hypothesis is a testable prediction about the relationship between two variables. An experiment looks to find out how the IV affects the DV.
In a research hypothesis, you must include 4 things:
The Population that the hypothesis will apply to
BOTH levels of the Independent Variable
The Dependant Variable
A specific prediction about what will occur
ethics - informed consent
researcher must first fully explain to the participant about the true nature and the risks of the experiment, and then obtain written permission on a consent form in order to take part. If under 18 or incapable, parental/guardian consent should be given.
ethics - confidentiality
researcher must collect, retain and dispose of all information related to participants in a manner that does not disclose their identity within the research. Participants must not be connected to their results or publically identified.
