Research Methods - Social Psychology Flashcards
What are self reports?
The most widely used method as they gather information from large numbers of participants.
How are self reports different from observations?
They get information direct from ppts about their behaviour or thoughts, in contrast to observations which are the researcher’s interpretations about the participant’s actions.
What are the two types of self report?
Interviews
Questionnaires (Surveys)
Describe self-reports.
They are non-experimental in design but can be used as part of an experimental design as a way of gathering information about the IV or DV
What are interviews?
Face-to-face conversations, although they can also take place over social media or telephone. They are different from questionnaires as the interviewer is contributing at the same time as the interviewee – the interviewee is also asking questions or the interviewer may have questions, clarifications or responses of their own.
What are unstructured interviews like?
Informal chats, which put the ppt at ease, proving natural responses but no two are the same (unreliable) but interviewer can stop and pursue interesting lines of conversations or go over unclear answers
What are structured interviews like?
These have pre-determined questions and don’t sound like normal conversations - more robotic with a specific sequence of questions, stilted and unnatural - but many ppts can be interviewed and their responses easily compared.
What are semi-structured interview like?
Containing predetermined questions, but also informal chats, they sound more realistic than structured interviews. They have the validity of unstructured interviews but the reliability of structured interviews but are difficult to do as interviewer have to be well-trained or quite experienced.
How do you use interviews as data?
They are recorded (audio-only or filmed) , can be transcribed later and analysed in detail.
What is a disadvantage of all interviews?
They all suffer from researcher effects: some may give directions through tome or body language or interviewees may feel hostile or intimidated by some researchers.
What are questionnaires?
They aren’t conducted face-to-face, and don’t communicate with respondents during their filling in. They are always structured but there are different types of questions that may be asked.
What are open questions?
Allow any type of answer, and the best have longer type answers (How/Why?). The respondent isn’t restricted by their means of answering but is harder to score and compare qualitative data.
What are closed questions?
They allow an answer from a list of responses, (Usually Yes/No) but more complicated ones may involve multiple choice. These are easier to score as quantitative data as all possible answers are known in advance but restrict the respondent in their answer.
What are the different types of closed questions?
Multiple choice
Ranked Scale
Likert Scale
Semantic Differential Scale
What type of data does multiple choice provide?
Nominal level data, in the frequencies that each option was ticked, to be expressed in percentages but an option can be forgotten.
What type of data does ranked data provide?
Interval/Ratio data in which each respondent get their own score (1-10)
What is the Likert Scale?
Invented by Renesis Likert, it provides a statement and respondent replies with their strength of agreement/disagreement in order to provide a score (1-5)which can be used to turn into nominal or interval/ratio level data.
What is the Semantic Differential Scale?
Similar to a Likert Scale, it allows respondents to place their level of attitude on a scale in order to obtain a score (scoring system can be hidden from respondents. These are used to turn into nominal or interval/ratio level data.
Why are questionnaires more common than interviews?
Large numbers can fill them in at any one time and researchers don’t require particular skills or training to supervise its completion. However socially sensitive research may require interviews instead as questionnaires can seem cold and disrespectful.
What is a particular type of questionnaire?
The psychometric test
What are psychometric tests used for?
Produces a score to measure psychological characteristics such as the IQ test. They produce quantitative data and measure several things at once ie. different personality traits.
What is a different approach to the psychometric tests?
Projective tests
What are projective tests like?
These have a more interview technique as respondent either describes images or draws images of their own before researcher interprets it.
What is an example of projective tests?
Rorschach Test - “Inkblot test” - inspired by Freud’s psychodynamic theory, it produced qualitative data as inner characteristics are revealed on your interpretation of the images shown
What are researcher effects?
Demand characteristics: Influences that spoil the outcome of the self-report process as the actual response is due to another aspect.
Why are researcher effects bad?
Unreliability - as some may respond and others not
Invalid results - as answers given may not be true due to researcher influence
What is the double-blind technique used for?
Is when the person asking the question doesn’t know the answer and so cannot give out hints as to what they’re looking for or help out the respondents
What are some subtle researcher effects?
When researcher is either attractive or intimidating - esp. with sensitive topics as this may make some more forthcoming or inhibit others - you can change this by matching researcher to respondent in terms of age, sex and race.
How do questionnaires suffer from researcher effects?
If researcher is present when questionnaire is being filled out, but this can be solved with postal surveys/ online questionnaires however location may also be a problem (hiding location leads to ethical questions with not letting people know who has their data).
What are the self report methods used within Biological Psychology?
Brendgan et al (2005) gave tachers a questionnaire to measure Social and physical aggression
Freud (1909) organised unstructured interviews with a 5 yr old (Little Hans) to explore his phobia of horses
What are the self report methods used within Cognitive Psychology?
Schmolck et al (2002) asked participants to complete memory tests and interviewed them about their thoughts at the same time -unstructured interview. Was recorded and scored by 14 raters for language problems
What are the self report methods used within Learning Theories?
Bandura et al (1061) gave nursery teachers a questionnaire to rate children for aggression
Becker et al (2002) used EAT-26 questionnaire to measure unhealthy attitudes and behaviours involving eating - also conducted semi-structured interviews on diet and TV-viewing
What are the self report methods used within Social Psychology?
Sherif et al (1954) used psychometric tests to measure boys’ feeling about their friends, their group and boys within the other group - also used unstructured interviews to get boys to talk about their attitudes (secretly recorded some)
Burger (2009) used psychometric tests to measure empathy (Interpersonal Reactivity Index) and need for control (Desirability of Control Scale)
What are some strengths of self report methods? (4)
Allow respondents to describe their own experiences rather than researcher interpretations
Questionnaires and interviews can study large samples of people easily and quickly
Can examine a large number of variables and ask people to reveal behaviour and feelings experienced IRL
Pilot study can test effectiveness on smaller samples beforehand
What are some weaknesses of self report methods? (4)
Respondents may not answer truthfully due to forgettable or wish to present themselves in socially desirable manners
Mass Questionnaires have low response rates
Unclear questions lead to lack of validity
Leading questions may force a particular response
Discuss the generalisabilty of self report methods
Large sample group doesn’t mean a truly representative group - is possible with postal/ internet surveys but with low response rates from those who are under-represented in research. Interviews also exclude those uncomfortable sharing things with researcher whereas questionnaires exclude those with poor literacy skill, lack of understanding or lack of expression.
Describe the reliability of self report methods
Achieved by standardized procedures, the more structure an interview has, the more reliable it is - questionnaires are more reliable than interviews as questions are written in advance ( has good inter-rater reliability) or (test-retest reliability)
Describe the validity of self report methods
Demand characteristics, researcher effects and social desirability bias can lead to a lack of validity but this can be solved by adding ‘lie’ questions in which the lie is obvious and hence can be scored in a lie scale to show trustworthiness- those over a certain limit will be removed from the study
Participant variables (literacy/ articulacy) can be controlled through semi-structured as interviewers could rephrase questions or ask for further clarification.
Describe the ethics of self report methods
Respect for privacy and dignity - confidential data
Integrity - transparent self reports, with lack of deception or if this is impossible, then to debrief afterwards with option to remove their data
Social Responsibility - clear what data used for
Don’t cause harm but need to be aware if research is socially sensitive - questions may cause embarrassment or offence and may be damaging to vulnerable groups especially if inflaming stereotypes or prejudice towards minorities
What is sampling?
A group of people who are representative of target populations with a conclusion to link back to the general population (generalization)
What is opportunity sampling?
Members of the target population that happen to be available at the time - it is a sample of convenience but has a huge risk of experimenter bias as only those near to you have nay chance of being within the study
What is volunteer sampling?
Members of the target population that select themselves- usually as a result of advertising. This sample may be more committed (useful if tasks are stressful or boring) and produces more varied samples with no experimenter bias but may still be unrepresentative as only certain people will volunteer (participant bias). Longer time taken.
What is random sampling?
Members of the target population chosen without any method of bias, sounds ideal but very difficult to identify people out of a target population (out of a hat) - usually used with a small sample and can still be unrepresentative.
What is stratified sampling?
Members of target population chosen without bias but guaranteed to be representative in certain ways. Strata = subgroups within target population - idea works by working out strata needed in sample, then calculation of how many of each people in each strata then filling it out using random sampling. More representative as each group get represented but is very fiddly and needs list of all within population - have to decide on important strata in advance (stressful/ causes problems). Some strata difficult to operationalise.
What are the strengths and weaknesses of opportunity sampling?
Quick and Convenient but has experimenter bias and may be unrepresentative
What are the strengths and weaknesses of volunteer sampling?
No experimenter bias but is slow and inconvenient as well as lack of representation
What are the strengths and weaknesses of random sampling?
No bias but but is slow and inconvenient as well as lack of representation
What are the strengths and weaknesses of stratified sampling?
No bias and is representative but is slow and inconvenient and may ignore important strata
When is opportunity sampling the best self report method?
When target population is gathered together in one location where you can approach a varied selection
When is volunteer sampling the best self report method?
When target population is likely to read the same magazine etc.
When is random sampling the best self report method?
When it’s a smaller target population and you can get hold of a master list of all members
When is stratified sampling the best self report method?
When you need random sampling and the topic to be researched suggests small numbers of obvious strata.
What is a good example of opportunity sampling?
Sherif et al Robbers Cave Experiment -sent observers into schools (Oklahoma) for white, athletic and confident boys with a good academic ability and no behavioural problems.
What is a good example of volunteer sampling?
Milgram’s 1963 obedience Experiment - Advertised for a memory study in a newspaper
What is a good example of random sampling?
Baddeley (1966b) recruits his sample from the university ‘subject panel’
What is quantitative data?
Data in the form of numbers including number scores, ranking, tally marks, percentages, statistical measures and various graphs
What is qualitative data?
Data that involves letters
What is the main advantages for quantitative data?
Is objective with a lack of need for interpretation, very reliable and scientific
Good for comparisons- esp. important for the experimental method
Can be turned into statistics
Simple
What are the disadvantages associated with quantitative data?
Reductionist, with details and meaning getting lost and so this could lack validity
What are the strengths of quantitative data?
Objective Reliable Easy to Compare Statistics can analyse large groups Short and Quick
What are the strengths of qualitative data?
Holistic - conveys meaning and individuality
Highly valid
What are the weaknesses of quantitative data?
Reductionist
Lacks Validity
What are the weaknesses of qualitative data?
Subjective Unreliable Hard to make Comparisons Harder to analyse large groups Lengthy and time consuming
What are descriptive statistics?
Straightforward calculations where you can recognise the mean, mode and median, standard deviation may be more obscure
Inferential statistics use data samples to draw conclusions about wider populations
Define the measures of central tendency?
Mean, median and mode
What is the mean?
Average number of a set of scores : calculated by adding the scores together and dividing by number of scores in the set
Easily skewed by abnormally large or small numbers
What is the mode?
The most frequent score within the set, there can be more than one and it helps us interpret the mean as if mean and mode are similar, there is standard distribution of the data but otherwise means the data is skewed…
What is the median?
The central score , in the middle : calculated when numbers in the set are ranked from highest to lowest but even numbers means median is calculated by the average of the two middle numbers
Not skewed by outliers
What are measures of dispersion?
Range and standard deviation
What is the range?
Difference between two ends of a set of scores: highest - lowest = range
What is standard deviation?
Measures the spread of data from the mean.
If standard deviation is large, scores are further from the mean and vice versa.
How do you calculate standard deviation?
- Calculate mean
- Deduct mean from each score in set
- Square result by itself ( multiply it by itself)
- Add square numbers together (results from Step 3)
- Divide by n-1 (number in set -1)
- Square root result (variance) to get your standard deviation
How do you turn frequencies into percentages?
Add total number of tallies, divide it by total number of tallies and divide it by 100
What are frequencies?
Nominal level data or scores which are interval /ratio level data
What is ‘raw data’?
each participant’s score or a table of tallies
What should you NOT do with raw data?
turn it into a graph
What can you do with a bar chart?
Display your measures of central tendency and dispersion
What are the rules with bar charts?
X axis must be labelled
Bars must be coded for what they stand for
Bar aren’t touching if they stand for different things
