Research Methods Flashcards
What is qualitative data
Information in words or pictures; non numerical
Qualitative data - pros
+ represents true complexity of human behaviour, thoughts and behaviour is not reduced to numbers — a holistic approach
+ provides rich details of how people think and behave — higher in validity as the researcher is more likely to measure the variable of interest
Qualitative data - cons
- more difficult to detect patterns and draw conclusions, large variety of information collected, words can’t be reduced to simple points
- interpreting what people mean makes it likely to be subjective, lowering credibility
What is quantitative data
Information in numbers, i.e. quantities
Quantitative data - pros
+ easier to analyse, data in numbers, using descriptive statistics or inferential statistics
+ more objective measure, more reliable, gives greater credibility
Quantitative data - cons
- May not express participants’ precise thoughts/ feelings because answers provided are fixed — low in validity
- oversimplifies reality and human experience — reductionist is to reduce human experience to numbers
What is primary data
First hand data collected for the purpose of the investigation
What is secondary data
Information collected by someone other than the researcher (e.g. books, journals, etc.)
What is random sample
A randomly collected sample
Random sample - pros
+ unbiased, all members of target population have an equal chance of selection
+ possible to choose a specific subgroup in target population
Random sample - cons
- takes more time and effort (obtaining a list of all target members, identifying them, asking consent)
- random samples aren’t always random as some might not take part, final may resemble more of a volunteer sample
What is stratified sample
Selected from different stratas (subgroups) in proportion to the population
Stratified sample - pro
+ most representative, all subgroups represented and in proportion to the numbers in the target population
+ specific subgroups can be chosen according to the variables considered to be important by the researcher
Stratified sample - cons
- deciding the subgroups may be biased
- a very lengthy process and those selected might not take part — more useful for opinion polls than psychological research
What is a volunteer sample
Participants who can volunteer to take part
Volunteer sample - pro
+ convenient way to find willing participants (gave informed consent)
+ good way to get a specialised group of participants (purposive sampling)
Volunteer sample - cons
- biased since volunteer participants are more likely to be more highly motivated (volunteer bias)
- volunteers may be more helpful, higher chance in guessing the aims
What is opportunity sample
Those most readily available during the study
Opportunity sample - pros
+ most convenient technique — takes little preparation
+ may be the only technique available since target population cannot be listed (like in random and stratified sampling)
Opportunity sampling - cons
- biased since sample is drawn from a small part of the target population might not be representative
- participants may refuse to take part, making the final sample likely to respond to demand characteristics
What are repeated measures
When the participant takes part in all conditions of the study
Repeated measures - pros
+ good control of participant variables since the same person is tested twice
+ fewer participants are needed than independent groups design
Repeated measures - cons
- order effects produced, e.g. participants might be better in the second condition after practicing or perform less since they are tired
- might make it easier for participants to guess the aim of the study
What are independent measures
Different participants are allocated to two or more experimental group representing different levels of the independent variable
There may be a control group
Independent measures - pros
+ avoids order effects since each participant is only tested once
+ avoids the possibility than a repeated measures design
Independent measures - cons
- no control of participant variables, e.g. participants from group b may be contrastingly different, scoring differently
- needs more participants than a repeated measures design
What are matched pairs design
Participants with similar variables are paired, each in a different group
Matched pairs - pros
+ controls for participants through matching — similar to repeated measures
+ avoids order effects since it is similar to independent groups design
Matched pairs - cons
- very time-consuming to match participants on key variables
- may not control all participant variables, only matching variables known to be relevant
What is an experimental hypothesis
A statement about the effect of the IV on the DV
Should include both levels of the IV
Should be precise and operationalised
E.g. people who sleep more do better on a memory test
IV: amount of sleep
DV: results of memory test
What is an one tailed directional hypothesis
States the direction of the hypothesis
E.g. people who sleep for 8 hours have a higher score on a memory test than those who sleep for 5 hours
What is a two-tailed directional hypothesis
States that there is a difference
E.g. .people who sleep for 8 hours will perform differently on a memory test than those who sleep for 5 hours
What is the null hypothesis
States that there is no difference
E.g. there is no difference between the memory test scores of people who sleep for 8 hours than those who sleep for 5 hours
What is an experiment
A research method which demonstrates casual relationships, all experiments have an IV and DV
What are IV and DV variables
IV: a factor directly manipulated by the experimenter to observe the effect of the DV
DV: measured by the experimenter to assess the effects of the IV
What is operationalisation
Variables must be operationalised
E.g. operationalising memory would be giving participants the same memory test, the DV would be the memory score
What is a lab experiment (IV and DV)
IV manipulated by experimenter: e.g. having them sleep in a lab to control the hours they sleep
DV measured in a laboratory: e.g. a test to measure memory
What is a field experiment
IV: manipulated by the experimenter
E.g. telling participants to wake up after 8 hours and conducting the study in home
DV: may be measured in the ‘field’
E.g. the participant’s own home
What are extraneous variables
Any variable other than the IV might potentially affect the DV
This includes both participant and situational variables
What are participant variables
Characteristic of the participant
What are situational variables
The environment that may affect performance
What are confounding variables
Special class of extraneous variables It changes systematically with the IV, so you cannot be sure that any change int he DV was due to the IV
E.g. study on memory — cannot be sure whether words were remembered better because they are familiar or since they were the first words on the list
What is predictive validity
The extent to which a test score is actually related to the behaviour you wanted to measure
Test score can forecast performance on another measure of the same behaviour
E.g. score on a memory test or an IQ test should be positively related to the performance in A level exams
Ecological validity - lab experiment - pros
+ high level of control, minimising confounding/ extraneous variables (increasing validity)
+ can be easily replicated because most aspects of the environment is controlled
Ecological validity - lab experiment - cons
- contrived situation where participants may not behave naturally. Low ecological validity
- demand characteristics and researcher bias/ effects may reduce validity
Ecological validity - field experiment - pros
+ less contrived, whole experience has mundane realism, higher ecological validity
+ avoids demand characteristics and researcher bias/ effects if participants’ aren’t aware they’re being studied
Ecological validity - field experiment - cons
- less control of extraneous variables, reduces validity
- may be more time consuming since experimenters have to set up a field experiment (more expensive)
- may have ethical issues if covert observation is used
What is a questionnaire
Respondents record their own answers, there are predetermined questions, prepared in a written form and there is no face-to—face contact
Questionnaire - pros
+ self-reports finds out what people think and feel
+ can be easily repeated so data can be collected from large numbers of people
+ respondents may feel more willing to reveal personal information when it is anonymous
Questionnaire - cons
- might not always tell the truth, social desirability bias
- the sample may be biased since only certain types of people would fill out the questionnaire
What are closed questions (and the type of data they collect)
Fixed number of possible answers
Collect quantitative data
Closed questions - pros and cons
+ easy to analyse quantitative data since numbers can be summarised through using statistics
+ answers more objective
- many not permit people to give their precise feelings, lacks validity
- oversimplifies reality and human experience — reductionistic
What are open questions (and the type of data they collect)
Respondents to provide their own answers
Collect qualitative data
What are ranked scales (and the type of data they collect)
To give an assessment using a scale (e.g. 1-5)
Collects quantitative data
Ranked scales - pros and cons
+ reasonably objective way to represent feelings and attitudes related to the topic being researched
+ produces quantitative data, easy to analyse
- participants may prefer to respond the same way to all scales regardless of context
- social desirability bias in responses
What are structured interviews
Predetermined questions in an interview
Structured interviews - pros and cons
+ can be easily replicated
+ easier to analyse than unstructured interviews
+ interviewer can provide extra information
- interviewer’s expectations may influence the answers the interviewee gives (researcher bias)
- participants may be reluctant to reveal personal information in face-to-face interviews
What are semi-structured interviews
When there are a mix of predetermined questions and developed questions (by the interviewer during the interview)
Semi-structured interview - pros
+ more detailed information obtained since questions can be tailored to the participant
+ can access information that may not be revealed by predetermined questions
What are unstructured interviews
No questions prepared in advance
Unstructured interview - cons
- more affected by interviewer bias, since interviewer creates questions on the spot, they may ask leading questions
- requires well-trained interviewers, may be difficult to obtain, making it more expensive
Questionnaires > interviews
Can be given out to lots of people, collecting a large amount of data (interviews takes more time to conduct interview and employ people)
Participants more willing to reveal confidential information since there is no face-to-face contact
Interviews > questionnaires
People may reveal more information with a skilled interviewer to encourage thoughtful responses
Semi-structured/ unstructured interviews: can access more information that may not be revealed by predetermined questions
What is observational research
Watching or listening to what participants do
Observational research - pros and cons
+ what people say they do is different from what they actually do, observing their behaviour possibility has greater validity
- observers may only ‘see’ what they expect to see
- observations can’t provide information on what they think
Collecting qualitative data in observational research
Often through recording everything (writing it down or recording it through tape)
Collecting qualitative data in observational research - pros and cons
+ first step in creating a structured, qualitative system for classifying observations. Thematic analysis can be used to create behavioural categories
- behaviours recorded might only be the visible or eye-catching ones
What are behavioural categories
Objective measures to separated continuous streams of actions
E.g. listing out several different kinds of behaviours and tallying them when it has been observed
behavioural categories - pros and cons
+ enables systematic observations to be made so important information isn’t overlooked, enhances validity
+ categories can be tallied and conclusions drawn
- categories may not cover all possibilities, some behaviour not recorded (low validity)
What is event sampling
Tallying behaviours in a specified time period
What is time sampling
Counting behaviour at regular intervals
Event and time sampling - pros and cons
+ both methods make observing behaviour more manageable by taking a sample
+ useful when behaviour to-be-recorded only happens occasionally, missing events would reduce validity
+ time sampling allows for tracking time-related changes in behaviour
- observations may not be representative if list of events is not comprehensive, reduces validity
- time sampling may decrease validity because some behaviour occurs outside of the observation interval
What is participant observation
Observer is a participant in the behaviour being observed
E.g. Rosenhan
Participant observation - pros and cons
+ provide special insight — greater detail
+ being on the insider means that observer may see more
- objectivity is reduced (observer bias) — observer is familiar with what’s going on, may be more subjective
- more difficult to record and monitor behaviour unobtrusively, participants are more likely to realise they are being observed, therefore may alter their behaviour for social desirability
What is non-participant observation
Observer is not a participant in the behaviour being observed
E.g. Reicher and Haslam
Non-participant observation — pros and cons
+ high in objectivity — since the observer is not an actual participant, not psychologically involved
+ can observe unobtrusively (observations likely to be covert, more natural behaviour)
- observer may misinterpret the communications within the group because they are an outsider (Reduced validity)
- the observer may see less as they are not a participant
What are structured observations
Participants are in a contained environment (often lab experiment)
They are aware that they are being observed but are randomly assigned to different independent groups by the researcher
E.g. Capafons
Structured observations - pros and cons
+ controlled environment allows focus on particular aspects of behaviour specific, so conclusions can be drawn more easily with a clearer cause and effect relation
- if participants know that they are being observed, they may respond to demand characteristics
- environment is unnatural, so participants may not behave as they would in everyday life, lowering ecological validity
What are naturalistic observations
Everything is left as usual, in an unstructured environment
E.g. learning theories practical of observing helping behaviour
Naturalistic observations - pros and cons
+ a realistic picture of a natural, spontaneous behaviour, participants more likely to behave normally – high in ecological validity
+ useful method to use when investigating a new are of research, giving the research more idea and investigation to plan
- observation more likely to be covert, raises ethical issues
- would be harder to draw conclusions as the focus would be really wide
What are overt observations
Participant is aware that they are being observed
E.g. Reicher and Haslam
Overt observation - pros and cons
+ avoids a lack of informed consent since participants know that they are being observed
+ easier to see everything that is going on as the observer doesn’t have to hide
- if participants know they are being observed, they are likely to alter their behaviour (demand characteristics or social desirability bias, depending on the study)
What are covert observations
When the researcher is ‘undercover’, participants are unaware that they are being observed
E.g. Sherif
Covert observation - pros and cons
+ participants behave more naturally because they are not aware of being observed, increases internal validity
- ethical issues, no informed consent
- invasion of privacy
What are case studies
A detailed study of a case (one person or one group of people)
E.g. HM or KF
Case studies - pros
+ can be used to investigate instances of human behaviours and rare cases
+ method produces rich, in-depth data, a more holistic approach
Case studies - cons
- difficult to generalise from individual cases, each one is unique, can’t make before and after comparisons
- necessary to use a recollection of past events which may be unreliable as memory isn’t always accurate
- researchers may lack objectivity as they get to know the case
- there might be important ethical issues like confidentiality and anonymity
Case study - Phineas Gage
1848, railway dynamite explosion resulted in a tamping iron going through his skull
He survived and functioned normally, but his personality was very different
Showed that parts of the brain could be removed without any fatal effects
Indicated that the frontal lobe is important in aspects of behaviour such as conscientiousness
Doubt in the validity of these reports on his behaviour
Case study - 2011 London riots
Studied the ‘mob’ behaviour in London riots
Observing the patterns of what they attack and don’t attack
HM case study
Had epilepsy — doctor bored 2 holes into his skull to remove his hippocampus
Resulted in significant memory loss, incapable to form new LTM
His brain was sliced after death and preserved to be studied till this day
Ethical issues: he had no memory, couldn’t give informed consent for research on him to be published, but the information was kept ‘confidential’, he was referred to as HM until he passed away
Correctional research - pros
+Looks at relationships between continuous variables and determining whether the relationship is significant
+ useful way to conduct a preliminary analysis on data, if it isn’t strong, we can rule out a casual relationship
Correlational research - cons
- cannot show a cause-and-effect relationship since there is no IV being deliberately altered
- if co-variables are correlated, one may be causing the changes in the other but we won’t know the direction of the possible effect
- may be intervening variables explaining why it is linked
- lacks reliability and validity since the method used to measure the co-variable may do so
What are CAT scans
X-rays and computer to create detailed structural images
Each image is a cross-section of the person’s brain
CAT scans - pros and cons
+ useful in revealing abnormal structures in the brain such as tumours or structural damage
- require more radiation, cannot be used often
- only provide structural information, not live brain
What are PET scans
Measures metabolic activity in the brain
Person injected with small amount of radioactive substance, will be detected scanner
PET scans - pros and cons
+ shows the brain in action, useful for psychological research
+ indicates the specific areas of the brain involved in experience
- sometimes results aren’t easy to interpret
- precise location of active is difficult to pinpoint
- ethical issues to inject radioactive glucose (may damage cells in the body)
What are fMRI scans
Uses radio waves to measure blood oxygen levels in the brain
Areas of the brain most active will show the most oxygen
fMRI scan - pros and cons
+ fMRI shows important information about which areas of the brain are being used at any one time
+ doesn’t use radiation
+ images are extremely clear and can show brain activity to the millimetre
- expensive to use
- only effective if person whose brain is being investigated stays perfectly still
- 5-second lag between brain activity and image appearing on screen, may cause interpretation problems
Other biological methods used in psychology
MRI scans - provide higher resolution if images than fMRI
EEG - measures brain activity through brainwaves, useful in diagnosing epilepsy and sleep and dream habits
Post-mortem - studies the brain after death
Lesioning - cutting lesion areas in animals to observe behavioural effects
Measuring hormones - hormones like testosterone can be measured through blood sample
Identifying genes - research on genetic influences through genes analysis
What are twin studies
When twins are compared on a specific trait to see how similar they are: MZ vs. DZ twins
Twin studies - pros and cons
+ enables researchers to investigate the influence of genes, assuming that both MZ and DZ twins share the same environment
+ although twins are unusual, information is often taken from twin registries, they hold data on thousands of twins and possible variable informations
- may overestimate genetic influence, twins still share the same environments, so it might be environmental factors as well
What are adoption studies
Genetic factors are implicated if children are more similar to their biological parents than to their adoptive parents
Adoption studies - pros and cons
+ adoption studies can remove the extraneous variables of environment, know that environment is not shared or genes are not shared
+ have been useful in showing that twin studies overestimate genetic factors
- children may be adopted to families similar to their biological families, so environmental influences may be similar
- people who adopt other people’s offsprings are unusual, so they are likely unrepresentative of the greater population
What are descriptive statistics
Measures of central tendency, frequency tables, graphs (bar chart, histogram, scatter diagram), normal distribution (including standard deviation), skewed distribution, sense checking data, measures of dispersion (range, standard deviation)
Produce, handle, interpret data-including drawing comparisons (e.g. between means of two sets of data)
Mean - pros and cons
+ ‘sensitive’ measure, reflects the values of al the data in the final calculation
- can be unrepresentative of data if there are extreme values
Median - pros and cons
+ not affected by extreme scores
- not as ‘sensitive’ as mean since not all values are reflected in the final calculation
Mode - pros and cons
+ useful when data is in categories (nominal data)
- not a useful way to describe data when there are several modes
Range - pros and cons
+ convenient way to express how spread out a data set is as both highest and lowest values are used
+ easy to calculate
- affected by extreme values
- fails to take into account the distribution of data set
Standard deviation - pros and cons
+ precise measure of dispersion since all exact values are taken into account
+ not difficult to work out with a calculator
- may hide some of the characteristics of the data set
- cannot be immediately sensed from the data, while range is fairly quick to identify
What are raw data tables
Arranging raw data in rows and columns (like chi squared data)
What is a frequency table
Displays a record of how often an event occurred
E.g. choosing favourite colour
Red III, Yellow II, Blue IIII
What are bar graphs/ charts
Height of each bar represents frequency of item
What is a histogram
Similar to bar graph, but for continuous data, area within bars must be proportional to frequencies represented
What is a line graph
Like histogram, has continuous data on x-axis, uses a dot and line to mark categories
What is a scatter diagram
Graph to display correlation data. Values are represented by dots
What is an inferential test
Procedures for drawing logical conclusions (inferences) about the target population from which samples are drawn
What is an observed value
The number (value) produced after applying an inferential test formula Sometimes called the calculated value since the researcher calculates it
What is the critical value
The number (value) which must be achieved in order for a result to be significant
What is the probability (p)
A measure of the likelihood that an event may occur
Probability given as a number between 0 and 1
The lower the probability, the more likely it was actually caused, instead of a casual relationship
Significance - inferential statistics
Statistical term indicating that the research findings are sufficiently strong to enable a researcher to reject the null hypothesis and accept the alternate hypothesis
Levels of significance - inferential statistics
Level of probability it has been agreed to reject the null hypothesis
Critical values table
List of numbers that inform whether an observed value is significant or not
There are different tables for each inferential test
What are the four types of data
Ratio: scale, includes zero (e.g. age, weight, height)
Interval: scale, excludes zero (e.g. Year, IQ, test scores)
Nominal: Tally (e.g. football team rankings)
Ordinal: Ranking based on categories (e.g. marital status, gender)
When do you use Mann Whitney U
Independent groups design
Ordinal level data
When do you use Wilcoxon
Repeated measures/ matched pairs
Ordinal level data
When do you use Spearman’s rho
Correlation
Ordinal level data
When do you use Chi-squared
Independent groups
Nominal level data
Self reports - internal validity
Whether a questionnaire does assess what it intended to assess
If a respondent doesn’t give representative answer, then there will be a low validity
Affected by ambiguous questions, social desirability bias and leading questions
Self-reports - ecological validity
Concerns the extent to which findings form a questionnaire or interview can be generalised
Depends on internal validity to an extent
Sample affects ecological validity, if sample not representative to wider population, then there will be a lower ecological validity
Self-reports - inter-rater reliability
Often more than one researcher to collect data
Findings can be cross-checked for reliability
Self-reports - test-retest reliability
Measure of whether something varies from one time to another
Same questionnaire or interview given to same participants on two different occasions to see if they have the same results
Interval between test and retest must be long enough so participant can’t remember their previous answers
What is thematic analysis
Identifying themes to impose a kind of order on the data. Ensures the ‘order’ represents the participants’ perspective
Summarises the data and enables general conclusions to be drawn
Steps to conducting thematic analysis
- Reading and rereading data, understanding the meanings of participants’ answers
- Breaking data into meaningful units: sentences or phrases which convey meaning
- Assigning a name or code to each unit, these represent themes you will be using
- Themes identified by grouping together in similar units
- Data chunks may be given more than one name/ code
- Rereading the text and ensuring that themes are correctly allocated, including all important aspects of the data
- Final report: should discuss and use quotes or other material to illustrate these themes
- Conclusions can be drawn which may include new theories
Thematic analysis - converting qualitative data to quantitative
Results obtained can be reduced to quantitative form
E.g. content analysis involves counting the content of anything
Once categories/ themes are created, instances can be counted and graphs can be used to represent the findings
If you represent each category/ theme with examples, it remains qualitative
If you count instances, data becomes quantitative
