RM year one

• Randomisation is the use of chance when designing investigations to control for the effects of bias e.g. random allocation to condition.
• Standardisation involves using the exact same, formularised procedures for all participants in a research study as otherwise differences become extraneous variables.
• Random Sampling?

  S - potentially unbiassed which helps to control extraneous variables and establish internal validity.

  W - time consuming and may not work, gaining access to complete list of target population is hard and then some participants may refuse to take part.
• Systematic Sampling?

  S - unbiassed, the first item is usually selected at random making it an objective method.

  W - takes time and effort as a complete list of the population is required and so thus you may as well just use random.
• Stratified sampling reflects proportions of people in certain subgroups within a population. Subgroups are identified and the relative proportional percentages and reflected in sample.
• Stratified Sampling?

  S - representative methods making sure characteristics of a target population are represented, makes it more generalisable than other methods.

  W - stratification is not perfect as it cannot reflect all the ways that people are different. Complete representation is not possible.
• Opportunity Sampling?

  S - convenient, quick and cheaper.

  W -inevitably biassed as sample will be unrepresentative of the target population as its drawn from a specific area. Struggle to generalise the findings.
• Volunteer Sample?

  S -participants are willing to get involved, they know the time and effort that is required because they selected themselves. Likely to engage more than people stopped in the street.

  W - volunteer bias, participants may share certain traits e.g. a willingness to be helpful. Thus they may be more respondent to cue and therefore generalisations are limited.
• Correlations?

  S - useful starting point as we can assess how variables are related. If the relationships is strong then it may suggest hypotheses for future research.

  S - economical, no need for controlled environment, can easily use secondary data, may also be less time consuming.

  W - no cause and effect, although often presented as causal in media which can lead to false conclusions regarding behaviours.

  W - untested variables may be better able to explain relationship
• Observations?

  S - capture what people do, often act differently to what the say in self report methods. Useful way of gaining insights into spontaneous behaviour.

  W - observer bias, researchers interpretation may be affected by expectations.
• Naturalistic Obs?

  S - higher external validity as in natural context behaviour is more likely to be spontaneous, more generalisable to everyday life.

  W - low control, may be extraneous variables that make it difficult to detect patterns of behaviour.
• Controlled OBs?

  S - use more standardised procedures and so can be replicated

  W - lower external validity as behaviour may be contrived as a result of the setting. Cannot be applied to everyday experiences.
• Covert Obs?

  S - reduced demand characteristics as pps do not know they are being watched, helps to increase internal validity of findings.

  W - ethically questionable, right to privacy may be affected.
• Overt Obs?

  S - ethically acceptable as pps have given their consent to be study and so have the right to withdraw if they wish.

  W - demand characteristics, knowledge of being studied may affect behaviour which would reduce internal validity.
• Participants Obs?

  S - greater insights into behaviour, researchers experiences situation as pps do which helps to enhance external validity.

  W - possible loss of objectivity, researcher may go native and begin to identify too strongly with those they are studying. Threatens internal validity.
• Non Participants Observations?

  S - more objective as researcher maintains appropriate distance so less chance of bias which helps to increase internal validity.

  W - loss of insight as researcher may be too far removed which may reduce external validity.
• Behavioural Categories?

  W - difficult to make clear and unambiguous as categories need be to self evident and not overlapping.

  W - sometimes get dustin categories, all forms of behaviour should be in the list and not grouped together and dumped behaviours go unrecorded.
• Event Sampling?

  S - useful for infrequent behaviours that do not occur at regular intervals

  W - situations can sometimes become quite complex making it harder to keep track of all details. Sometimes important details can then go unrecorded. Would affect validity.
• Time Sampling?

  S - reduces number of observations and instead of recording everything the data is recorded at certain intervals. Make the observation more structured and systematic.

  W - may be unrepresentative, researcher may miss out of key behaviours that occur outside timescale meaning the observation does not reflect the whole behaviour.
• Questionnaires?

  S - good way to gather lots of data quickly, reduces effort involved and make questionnaires more cost effective,

  S - straightforward to analyse especially with closed questions, statistical data can be easily converted to graphs and charts.

  W - responses may not always be truthful, paint themselves in positive light, social desirability bias.

  W - response bias, may favour certain response e.g. always agree. Means all respondents can tend to reply in similar way.
• Structured Interviews?

  S - standardised format means its easy to replicate, helps to reduce differences between interviewers.

  W - interviewers cannot elaborate which may limit richness of data collected.
• Unstructured Interview?

  S - greater flexibility, points can be followed up as arise meaning the researcher is more likely to gain insight into interviewees worldview and collect unexpected info.

  W - increased risk of interviewer bias, closer dialogue between interviewer and interviewee. More opportunity for unconscious cues.
• Peer Review?

  S - protects quality of research as minimises possibility of fraudulent research. Maintains psychology reputation as a science.

  W - anonymity may be used to criticise rival research due to competitiveness for fundings.

  W - publication bias, tend to publish headline grabbing findings, file drawer problem.

  W - groundbreaking research may be buried, especially if research contradicts inspectors view. Peer review may slow down the rate of change within scientific discipline.