Research Methods Flashcards
Data Analysis : Types of Data
Qualitative data is data expressed in words and an example is a dairy or an open question interview. It is based on the interpretation of language with an unstructured observation whereas Quantitative Data is nominal data such as scores or numbers and can be analysed statistically or converted into a graph
One is not better than the other and they are used in different situations
Primary data is data that is acquired from the participant themselves whether it may be from conducting an experiment or a questionnaire whereas Secondary data is data collected from sources that already exist and is also subject to statistical testing in order to identify the significance. An example is population records
Data Analysis Types of Data Evaluations Q+Q
Qualitative data gives a much more richness of detail and gives the participant time to develop their thoughts to thoughts or feelings and this means it has great external validity and also provides a more meaningful insight into the participants worldview
A limitation of Qualitative data is that it is difficult to analyse and cant be summarised statistically meaning patterns are harder to identify and is thus subject to subjectivity meaning an experimenter/ researcher may show bias
Quantitative data -
Strength- Easy to collect data and analyse since it can be graphed statistically meaning it isn’t subject to bias
Limitation- low external validity because participants cant develop their thoughts or feelings and data is less meaningful
Data Analysis : P+S Evaluations
Primary Data
Strength - data is authentic and can be designed to acquire the information needed for what the researcher inquires
Limitation - Producing Primary data takes time and effort and this could also be a limitation compared to secondary data
Secondary Data
Strength - Secondary data requires minimal effort and i inexpensive compared to primary data since desired information already exists
Limitation - Data could be outdated or incomplete and meaning it could have inaccuracies or not match what researcher is inquiring about
Self - report Designing questionnaires and interviews
When Designing questionnaires you can use open questions which have no fixed response and one can respond whichever they want to whereas closed questions provide a fixed choice of responses. The type of closed questions that can be given are…
- Likert Scales — phases in which one agrees or disagrees in a scale in response to a phrase or a question and these indicate agreement
- Rating scales — scales which represents their strength or feeling about a particular topic and uses Numbers
- Fixed Choice options — a list of possible options in which one must respond to
When Designing interviews, Interviews must be standardised and must go through a schedule in order to reduce the effect of interviewer bias. They can be group interviews or individual interviews.
one should conduct in a quiet room as this increases the chance of interviewee opening up. One should also use neutral questions in order to make the interviewee feel comfortable and relaxed but should treat answers said in confidence to prevent doubting.
Writing Good Questions
When producing questions if one uses too much jargon in the questions, one may not understand and so they must be simple
One should also make sure they are not emotive and suggestive as this may guide the respondent towards a particular answer which creates interviewer bias
Questions should also not be double barrelled as they could agree to one part of question but disagree to another part and finally they shouldn’t use double negatives but should be straightforward
Self Report Technique (Questionnaires)
Questionnaires are use in order to assess thoughts and/or feelings and this can be simple or complex however it is mainly used to assess the dependent variable
Whilst Open questions don’t have a fixed range of answers and normally produce qualitative data which is harder to analyse, Closed questions assess quantitative data and is easy to analyse but lacks depth associated with Open questions. - Closed questions which produce qualitative data can be converted into quantitative data by counting yeses or noes
Self report technique (Interviews)
Interviews can be conducted on the phone or involve a face to face interaction between an interviewer and an interviewee. The interviews can be structured or unstructured interviews. There are
Structured interviews - predetermined questions asked in a fixed order which can be conducted through any medium
Unstructured interviews - These work like conversations meaning there is no set questions and the general aim is that a certain topic is discussed in a free flowing way. Interviewee encouraged to expand and elaborate answers
Semi-Structured interviews - most common interview such as an interview. There are a list of questions to work through however interviewer allows any follow up questions when they feel it is needed
Questionnaires (Evaluation)
The strengths of the questionnaires are that they are cost effective and can easily gather large amounts of data. The number of people participating determine the volume of data and it can be completed without an experimenter which reduces the effort of researcher involved
Another Strength is that they are usually straightforward to analyse and mainly compromise of closed choice questions meaning statistical analysis can be used to compare
One limitation is that is that respondents may not be truthful so it relies on truth and a positive light. An incriminating question may lead to social desirability bias or demand characteristics
Another limitation of questionnaires is that they produce response bias in which respondents may respond in a similar way or in a favourable way which might be because respondents don’t read questions properly and complete the questionnaire too quickly. Acquiescence bias may also be common
Types of Interviews (Evaluations)
Structures Interviews
STR- one strength of structured interviews is that they are straightforward and easy to replicate through their standardised format and this reduces the difference between interviewers
LIM - one limitation of structured interviews is that interviewers can not ask to elaborate on points which could be frustrating
Unstructured Interviews
STR - One strength of Unstructured Interviews is that there is much more flexibility and interviewers can follow up or arise points in order to gain insight of interviewee’s worldview
LIM - A limitation of unstructured interviews is that analysis of results is difficult and much sift through irrelevant information which means conclusions harder to make.
Another limitation of Unstructured interviews is that they may lie due to social desirability bias but this relies on how skilled the interviewer should be able to establish sufficient rapport even with personal or sensitive topics
Observational Design (Ways of observing)
Unstructured Observation - When someone writes down what they see, produces rich detail and involves a small number of participant
Structured Observation - when one looks for target behaviour which allows researchers to quantify their observations using list of behaviours and sampling methods
Continuous recording used commonly in unstructured observation records all behaviour that occurs but not feasible for complex behaviours. This means other methods for structured observation are present such as
Event sampling - counting the number of times a particular behaviours is shown from a target individual
Time Sampling - behaviour that is recorded in a fixed time frame
Behavioural categories are categories of behaviour seen in a target. This comes from operationalisation in which target behaviours must be observable and measurable. This means, complex behaviours must be broken down into simple observable ( not inferred) behaviours and the researcher should make sure that the behaviour checklist beforehand is in depth
Inter - observer reliability
When observing behaviour, one should not miss important behaviours and confirm the hypothesis based on opinions as this causes bias.
In order to make the observation objective and unbiased, observers should be in a pair in order to make sure that any data recorded is similar
To establish inter rate reliablility, one should
- familiarise with the behavioural categories being used
-serve same behaviours at the same time as a small scale pilot study
-observers should compare data recorded and disuse differences of interpretation
- analyse data from the study and find correlating the observations made by an overall figure
Types of Observations (evaluations)
Structured observations use data that can be categorised and more likely to be numerical meaning it produces quantitative data and this means analysing behaviour between participants is more straightforward
Unstructured observations benefit from richness in detail however they are subject to observer bias meaning they may categorise behaviour not present or behaviours that catch their eye
When making these behavioural categories, one should make sure that they are
- clear and unambiguous as possible
- observable,measurable and self evident
- should ensure all target behaviours are recorded
- categories mustn’t overlap and be exclusive
Event sampling is useful when behaviour occurs infrequently but could be missed if time sampling used, complex target behaviours may be overlooked by observers using event sampling however but time sampling good with reducing number of observations but behaviour may not be representative of observation as a whole
Observational Techniques (6 types)
Observation is an non-experimental method which produce researchers of behaviour without asking them which allows moe flexibility to study more complex interactions between variables
Naturalistic observation - Watching/recording behaviour one would normally see occurring in the setting
Controlled Observation - Watching behaviour within a structured environment where one or more variables are managed
Covert Observations - participants behaviour observed without their informed constant/knowledge , allows it to be ethical
Overt observation - participants behaviour observed with their knowledge and consent
Participant observation is when the researcher becomes a member of the group whose behaviour they are watching
Non - participant observation - researcher remains outside the group of behaviour she is watching
Observational Techniques (Evaluation) n+c
A strength of Naturalistic Observations is that they have external validity since behaviour being observed occurs in the environment it would normally occur in.
A limitation of Naturalistic Observations is that they can’t be replicated because of the lack of control and this also introduces extraneous variables which make it hard to identify patterns in behaviour
In addition, a limitation of controlled Observations are that they produce findings that can’t be generalised to real life settings however extraneous variables also become less of a factor meaning strong internal validity but overall low ecological validity
Observational Techniques (evaluations) C+O
One strength of covert observations is that it removes the problem of participant reactivity and ensures behaviour observed is natural thus increasing the validity of the data gathered
A limitation of covert observations however is that one may not wish to have their behaviour recorded meaning there is an issue with the ethics fo using covert studies
On the other hand, a strength of Overt observations is that it is more ethically acceptable as participants have knowledge of being observed however a counter is that they this knowledge could influence their behaviour.
Observational Techniques (Evaluations) P+NP
A strength of participant observations is that it gives an increased insight into the lives of people being studied meaning researchers get to experience the situations
A limitation for participant observations however is that the researcher may come to identify more strongly with the ones they ate studying and this decreases the objectivity of data as the lines between participant and researcher is blurred
A strength of Non participant studies is that a researcher can maintain an objective psychological distance between the researcher and the participants meaning less chance of identifying with the participant
However, a limitation of non participant observations is that the love valuable insight compared to a participant observation since they could be too far removed from the people they are studying
Pilot studies + procedures
Pilot studies are a small scale investigating which takes place before the real investigation in order to check the procedure,materials and any variables that need to change - allows the investigation to run smoothly
Single blind procedures - in attempt to reduce the effect of confounding effects such as demand characteristics via placebo effect, participant unaware of an important detail
Double Blind procedures - procedures in which the researcher nor the participant know about the aim of the investigation - double blind drug trails
An experimental condition can be the actual drug, a control condition can be the placebo and this sets a baseline and also allows for comparison, if behaviour of experimental and control condition is vastly different, cause and effect established
Ethical issues
Ethical issues arise when conflict exists between the right of participants and the goal of research to produce authentic,worthwhile data.
Informed Consent - when the participant is aware of the aims of the research,procedure and the rights they have such as the right to withdraw. Participants should make an informed judgement if they want to take part in the study however it may also mean participant behaviour is not natural
Deception is deliberately misleading or withholding information and this means they don’t receive adequate information thus they can’t give informed consent. However it can be argued that by deceiving participants, it adds legitimacy to their behaviour
Protection from harm is making sure the participant isn’t placed in any risk that would affect their daily lives such as psychological or physical harm in addition to embarrassment or stress. The important feature is that participants are reminded of their right to withdraw from the investigation
Privacy and Confidentiality is important as this is right of the participant and is a law under the data protection act and this can extend to privatisation of info about location of study
Ways of dealing with Ethical issues
BPS code of conduct is a quasi legal statement that instructs psychologists what behaviour is acceptable and what isn’t. The BPS code of conduct is built around respect, competence , responsibility and integrity.
When dealing with informed consent, they should be given consent letters and all relevant information before taking part in the study and there are 3 types of consent, Presumptive, prior general and retrospective
Deception / protection from harm - all participants should be debriefed and be made aware of true aims not given in the study such as experimental conditions. Participants should also have the right to withhold data and also be reassured in debriefing that their behaviour was normal and if stress was induced, the participant should be given counselling
When dealing with confidentiality, personal details are protected and they are allowed to maintain priority. They can do this as well by referring to participants by their initials or numbers. In debriefing, participants must also be reminded that their data is protected through the process.
Populations and samples
Population is a group of people that are in focus of the researchers interest called the target population and from this subset, a sample is drawn which is representative of the target population. This allows generalisation to occur however it is difficult to do this due to the diverse nature which induces bias. Thus sampling techniques are used
Random Sample - Sampling in which all members have an equal chance of being selected, this is done by the assigning of numbers and then the use of a lottery method
Systematic Sample - sampling in which every nth member of the target population is selected. A sampling frame (list) is first established and then a sampling system is chosen randomly or by researchers until sampling is complete
Stratified Sample - sampling people who represent the proportion of sub groups in the target population. To this, one must first identify a strata and then proportionate them so they are representative and finally, participants from each strata are then randomly picked
Opportunity sample - sample that represent the target population are difficult to obtain so researchers may pick whoever is around at the time of their study
Volunteer sample - self selecting samples are volunteers selecting themselves t be part of the sample such as willing participants raising their hands to participate when asked.
Sample (evaluations)
Random Sample
STR: one strength of random sampling is that its free from researcher bias an has no influence over who is selected
LIM: one limitation of random sampling is that it is difficult and time consuming to conduct. In addition, it may still by unrepresentative as random isn’t 100% representative and also some of the selected participants may not want to volunteer meaning its moe like a volunteer sample
Systematic sample
A strength of Systematic Sample is that it avoids researcher bias since researcher has no influence over who is chosen. It is also fairly representative
Stratified sample
One strength of stratified sample is that it avoids researcher bias since target population is sub-divided into strata and also the sample is representative since it reflects the composition of the target population allowing generalisation to take place however, it can not reflect all the ways people are presented meaning complete representation isn’t possible
Opportunity sample
One strength of opportunity ample is that it saves time and effort in picking a sample and is also muh less costly however a large limitation of opportunity sample is that it is unrepresentative since sample comes from the same place and thus can’t be generalised and also the researcher has control over the partisans picked meaning subject to researcher bias
Volunteer sample strengths consist of having minimal input from the researcher and is also less time consuming but a limitation is that volunteers may only attract a specific profile of a person which limits its generalisation
Laboratory experiments
Laboratory experiments are experiments that take place in a controlled environment in which the researcher manipulates the IV and observes the effect it has on the DV
One strength of Laboratory Experiments are that it has high control over extraneous variables and this ensures the DV is likely to be affected from the manipulation of the IV only meaning it demonstrates cause and effect and thus has high internal validity
Another strength is that they are very replicable since extraneous variables don’t affect an experiment when repeating and this allows the results to be replicated in order to find it it was a one off or is valid
One limitation of laboratory experiments are that they lack generalisability since they usually use artificial stimuli and this may mean they behave in unusual ways which can not be generalised to real life situations and thus has low external validity
Another limitation of laboratory experiments is that participants are aware that they’re being tested in a lab and so may show demand characteristics which is a sign of unnatural behaviour.
Field Experiments + Quasi Experiments
Field experiments are experiments which take place in a natural setting and a researcher manipulates the IV and records the effect on the DV
One strength of Field experiments is that they have higher mundane realism that lab experiments since the environment is more natural and this means behaviour more valid and authentic which thus means there is high external validity
One limitation fo field experiments is that since it occurs in the natural environment, there is a lack of control over extraneous variables and thus there may be unaccounted factors at play and thus it is harder to establish and replicate studies.
Quasi Experiments are studies that have no determined IV and the variable simply exists meaning it isn’t an experiment.
A strength of Quasi-experiments is that they are often carried under controlled conditions and have the strengths of lab experiments
A limitation of Quasi experiments is they can’t randomly allocate participants to conditions and thus there are confounding variables
Natural Experiments
Natural experiments are experiments where the IV changes but is not directly observed since the researcher is not aware of it. Researchers in natural experiments record the effect the DV has on the environment
One strength of natural experiments is that they may be undertaken in which otherwise would be unethical or unpractical such as studies on institutionalised orphans
Another strength of Natural experiments is that they often have high external validity since they study real life issues such as the levels of stress when doing a specific behaviour
A limitation of Natural experiments is that it is a naturally occurring event and so it only happens very rarely thus reducing opportunity to research and this reduces the scope to generalise findings to other situations
Another limitation of natural experiments is that participants aren’t randomly allocated to experimental conditions and so the researcher is naturally less sure whether the IV really affects the DV and so there are confounding variables that may affect the DV
