Research methods Flashcards
Experimental method
The manipulation of an independent variable to measure the effect on the dependent variable. May be laboratory, field, natural or quasi.
operationalise
Making variables measurable
Hypothesis
A clear precise statement that states the relationship between the variables to be investigated. Stated at the outset of any study
eg, Drinking fizzy drinks cause people to become more talkative
directional hypothesis
direction of the difference or relationship
eg. people who drink fizzy drinks become more talkative then people who don’t
People who drink water are less talkative then people who drink fizzy drinks
used when there are previous studies that suggest a particular outcome
non directional hypothesis
does not state the direction
eg Eg people who drink fizzy drinks differ from people who don’t drink fizzy drinks
used when there are no previous studies to suggest a particular outcome
variable
Any “thing” that can vary or change within an investigation. Variables are generally used in experiments to determine if changes in 1 thing result in changes in another.
aim
A general statement of what a researcher intends to investigate
extraneous variable
are unwanted variables that need to be controlled for both experiments otherwise will interfere with the IV or DV eg Noise, temperature, amount of sleep, personality. It MIGHT mess with results not a very big deal.
- Participant variables: are any individual differences between participants that may affect the DV
- Situational variables: any features of the experimental situation that may affect the DV
repeated measures
same participants doing both conditions
independent group design
two groups of participants doing both designs
Confounding variables
change systematically with the IV so we cannot be sure if any observed change in the DV is due to the confounding variable or the IV, so results are meaningless
- eg if there were two groups and one group were introverts and the other were extroverts this would result in a 2nd unintended IV and so you don’t know if fizzy drink affected the 2nd group because they were already extroverted – personalities is the confounding variable
Demand characteristics
refers to any clue from the researcher or research situation that may reveal the aim of the study.
- In a study the participants will try to work out what is going on trying to find demand characteristics. The participants may act in a way they think is expected overperforming (please- U effect) or underperforming (screw-U effect).
Investigator effects
are any effect of the investigator’s behaviour on the outcome of the research (the DV)
- eg given that they researcher was expecting the energy drink group to speak more than the water group and unknowingly in our unconscious behaviour encourage a greater level of chattiness from the energy drink participants.
Randomisation
the use chance when designing investigations to control for the effects of bias and minimise the effect of extraneous/ confounding variables.
-eg a memory experiment may involve participants recalling words from a list. The order of the list should be randomly generated so that the position of each word is not decided by the experimenter.
Standardisation
Using exactly the same formalised procedures for all participants in a study
- all participants should be subject to the same environment, information and experience so all procedures are standardised.
standardised instructions example
this includes standardised instructions that are read to each participant. Example:
Thank you for participating please sign this form of consent. You have the right to withdraw at any point throughout the investigation it will take an hour of your time and you have 30 minutes to drink a liquid on the table in front and then I shall interview you for 20 minutes. Thank you for participating.
Control groups
are used for a purpose of setting a comparison and acct as a ‘baseline’ and help establish causation.
Single blind
is where a participant doesn’t know the aims of the study so the demand characteristics are reduced.
Double blind
both participants and researcher don’t know the aims of the study to reduce demand characteristics and investigator effects.
Repeated measures.
Same participants doing both conditions
Problem: They may have had practise or be too fatigued (order effects) – reduces validity of results
-participants may guess aims - reduces validity of results
Good: there are no individual differences between participants – controls important CV
-fewer participants – wastes less time
Independent groups design
The two groups of participants doing both conditions
Problems: there are individual differences between participants -reduce validity of study
-more participant – more money spent and time wasted recruiting
Good: no order effects I controls important CV
-will not guess aim – so behaviour is more natural
Matched pairs
pairs of participants are first matched on some variables that may affect the DV. Then one member of the pair is assigned to condition A and other to B.
Problems: matching is time consuming, not perfect and can’t control all relevant variables – may not address participant variables.
- more participants more time and expense
Good: Participants matched on a variable relevant to experiment – enhances validity
- No order effects, only tested once so no fatigue or practise – enhances validity
Pilot study
confederate
- A small scale trial run of a research design eg experiment or interview before doing the actual investigation
- it is done in order to find out if certain aspects of the design don’t work out and make adjustments, saving time and money.
Confederate – sometimes a researcher has to use another person to play a role in an experiment. Eg if you want to find out if people respond differently to orders from someone wearing a suit. This is a confederate.
Laboratory Experiments
what is a true experiment
– more control, less realistic
- Conducted in a controlled environment where extraneous and confounding variables can be controlled ( doesn’t have to be a lab could be a classroom)
When the IV is under direct control of the researcher who manipulates it and records effect on the DV eg only Lab and Field
strengths and weaknesses of Laboratory Experiments
+ EV’s and CV’s can be controlled so the effect on the DV can be minimised and cause and effect between the IV and DV can be demonstrated (high internal validity)
+ can be easily replicated due to standardised procedure. If the results are the same this confirms validity
- may lack generalisability. This controlled lab environment may be artificial and participants are aware they are being tested Thus, behaviour may not be ‘natural’
- demand characteristics may be a problem These are clues in the experimental situation that invite a particular response from participants. The results of the experiment may be explained by these clues rather than the effect of the IV and get participants to act in a certain way
Field experiment
less control, more realistic
- The Iv is manipulated in natural more everyday setting in the field ( not necessarily an actual field
Positives and negatives of a field experiment
+ more natural environment. Participants more comfortable in their own environment Results may be more generalisable
+Participants are unaware of being studied They are most likely to behave as they normally do so the findings can be generalised. The study has greater external validity
- More difficult to control CV’s. Observed changes in the DV may not be due to the IV, but to CVs instead. It is more difficult to establish cause and effect than in a lab
- - There are ethical issues Participants in a field experiment may not have given informed consent. invasion of privacy.
Natural experiment
The experimenter takes advantage of the pre-existing IV’s. The IV would have varied even if the experimenter wasn’t interested.
Positives and negatives of a natural experiment
+ May be the only ethical option It may be unethical to manipulate the IV. Eg studying the effects of institutionalisation of children A natural experiment may be the only way causal research can be done for such topics.
+ greater external validity. Natural experiments involve real-life issues such as the effect of a natural disaster on stress levels. This means the findings are more relevant to real experiences
- the natural event may only occur rarely. Many natural events are ‘one off’ incidents and this reduces the opportunity for research .This may limit the scope for generalising findings to other similar situations.
- participants are not randomly allocated The experimenter has no control over which participants are placed in which condition as the IV is pre-existing. May result in CV that aren’t controlled, e.g Romanian orphans adopted early may also be the friendlier ones.
Quasi Experiment
The IV is based on existing difference between people eg age or gender. Something that can’t be changed or manipulated it just simply exists.
Positives and negatives of a Quasi experiment
+ There is often high control. Often carried out under controlled conditions and therefore shares some of the strengths of lab experiments. This means increased confidence about drawing casual conclusions.
+ Comparisons can be made between people. In a quasi- experiment the IV is a difference between people eg. People with or without autism. This means that comparisons between people can be made.
- participants are not randomly allocated The experiment has no control over which participants are placed in which condition as the IV is pre existing. Participant variables may have caused the change in the DV acting as a CV.
- casual relationships not demonstrated The researcher does not manipulate/ control the IV. We cannot say for certain that any change in the DV was due to the IV.
Opportunity sample
Most available – people who are simply most available i.e ones who are nearest. How? As people nearby eg. Students in class to take part
+ a quick method – it is convenient because you just make use of the people near you and makes this a very popular method.
- inevitably biased – the sample is unrepresentative of the target population as it is drawn from a specific area and so finding can’t be generalised.
^ researcher has control over selection of participants
Volunteer sample
Self-selecting – in a volunteer sample, participants select themselves
How? Advertise. Eg an ad in a newspaper
+ participants are willing – participants select themselves and know how much time and effort is required so engage more than people selected from the streets
- likely to be a biased sample – participants may share certain traits like curiousness and keenness so generalisation is limited
Random sample
Equal chance – every person in the target population has an equal chance of being selected
How? Lottery method, all members of a target population are given a number and placed in a hat or tombola.
+ potential unbiased – the researcher has no influence over who is selected and so free from researcher bias.
- representation not guaranteed – Still possible that a random method may produce a biased sample. Limits ability to generalise.
^ time consuming and difficult to conduct
Systematic sample
Sampling frame- participants are selected using a set ‘pattern’
How? Every nth person is selected from a list of the target population
+ unbiased because the first item is usually selected randomly and so an objective method.
- Time and effort, a complete list of population is required and may as well use random sampling.
Stratified sample
Frequency- participants are selected according to their frequency in the target population.
How? Subgroups are identified such as gender or age groups. The relative percentages of the subgroups in the population are reflected in the sample.
+ representative method – The characteristics of the target population are represented and generalisability is more likely than other methods.
- Stratification is not perfect – Strata cannot reflect all the
Ethical issues
When a conflict exists between the rights of participants and the aims of the research.
- BPS code of conduct – is a quasi-legal document to protect participants based on four principles: respect, competence, responsibility and integrity.
- Ethics committees weigh up costs (e.g. potential harm) and benefits (e.g. value of potential harm) and benefits (e.g. value of research) before deciding whether a study should go ahead.
Informed consent
Participants should be able to make an informed judgement about whether to take part.
Too much information may affect participants’ behaviour so alternative forms of consent are:
- presumptive – ask a similar group
- prior general – agree to be deceived
- retrospective – get consent after the study
Deception
Deliberately misleading or withholding information so consent is not informed.
At the end of a study, participants should be given a debrief where they are advised of:
- The true aims of the investigation
- Details that were not given during the study, e.g existence of other groups or conditions.
- What their data will be used for
- Their right to withhold data
Protection from harm
Participants should be at no more risk than they would be in everyday life.
- Should be given the right to withdraw at each stage of the research process
- Should be reassured that their behaviour was typical/ normal during the debriefing
- Researcher should provide counselling if participants have been e.g distressed
Privacy/ confidentiality
privacy is the right of an individual to have some control over how his or her personal information
confidentiality is the duty to ensure information is kept secret only to the extent possible.
-We have the right to control information about ourselves. If this is invaded, confidentiality should be respected.
- If personal details are held these must be protected (a legal requirement). Usually though, no personal details are recorded.
- Researchers refer to participants using numbers, initials or false names
- Participants’ personal data cannot be shared with other researchers.
Ways of dealing with ethical issues
Debriefing – explain aims, deception, ensure no harm will be caused etc
BPS code of conduct: The British psychological society publishes a. range of guidelines on ethical issues and what is acceptable in research
Cost-benefit analysis: Where the benefits of the research are weighed up against the possible costs to participants – both are difficult to predict
Presumptive consent
rather than getting consent from the participants themselves, a similar group of people are asked if the study is acceptable. If this group agree, then consent of the original participants is ‘presumed’
Positives and negatives of correlations
+ Useful starting point for research
By assessing the strength and direction of a relationship, correlations provide a precise measure of how two variables are related If variables are strongly related it may suggest hypothesis for future research
+ Relatively economical Unlike a lab study, there is no need for a controlled environment and no manipulation of variables is required Correlations are less time consuming than experiments
- No cause and effect Correlations are often presented as casual e.g media when they only show how 2 variables are related There may be intervening variables that explain the relationship
- Method used to measure variables may be flawed E.G the method used to work out an aggression score might be low in reliability (observational) categories might have been used) This would reduce the validity of the correlational study.
Pilot study
Observational techniques
- A way of seeing or listening to what people do without having to ask them. Observation is often used within an experiment as a way of assessing the DV.
Naturalistic
– takes place where the target behaviour would normally occur. All apects of the environment are free to vary.
It is much better to study ‘interaction’ in an environment that it would usually take place.
Low control due to uncontrolled EV’s so more difficult to detect patterns
Naturalistic observation p +n
+ High external validity behaviour is more spontaneous and generalisable
- Low control due to uncontrolled EV’s so more difficult to detect patterns
Control observation
- Some control/ manipulation of variables including control of EVs.
+ Can be replicated more easily repeated due to standardised procedures so findings can be checked to see if they occur again - May have low external validity behaviour may be contrived as a result of the setting and findings cannot be applied to everyday experience.
Covert
- Participants are unaware they are being studied
p and n of covert
+ Demand characteristics reduced Participants do not know they are being watched so their behaviour will be more natural and increase validity
- Ethically questionable People do not want behaviour recorded even in public and participants’ right to privacy may be affected
overt
- Participants are aware of being studied
p and n of overt
+ more ethically acceptable Participants have given their consent to being studied so they have a right to withdraw
- Demand characteristics Knowledge of being studied influences behaviour so reduces validity
Participant
- when the researcher becomes the part of the group they are studying
p and n of participant
+ Can lead to greater insight Researcher experiences the situation as the participants do and this enhances validity
- Possible loss of objectivity The researcher may identify too strongly with those they are studying and this threatens the objectivity and validity.
non- participant
+ More objective Researcher maintains an objective distance so less chance of bias and increase in validity
- Loss of insight Researcher may be too far removed from those they are studying and may reduce validity.
Behavioural categories
the target behaviour to be observed should be broken up into a set of observable categories. This is similar to the idea of operationalisation.
- Difficult to make clear and unambiguous Categories should be self-evident and not overlap, not always possible to achieve. ‘smiling’ should be poor categories
- Dustbin categories All forms of behaviour should be in the list and not one ‘dustbin.’ ‘Dumped behaviours go unrecorded.
Time sampling
observations are made at regular intervals e.g once every 15 seconds
+ Reduces the number of observations Rather than recording everything that is seen. Data is recorded at certain intervals. The observation is more structured and systematic
- May be unrepresentative The researcher may miss important details outside of the time-scale. May reflect the whole behaviour.
Event sampling
- A target behaviour/ event is recorded every time it occurs
+ My record behaviour The researcher will still ‘pick up’ behaviours that do not occur at regular intervals. Such behaviours could easily be missed using the time sampling.
- Complex behaviour oversimplified If the event is too complex, important details may go unrecorded and this may affect the validity.
Questionnaires
- made up of a pre – set list of written questions to which a participant responds
+ can be distributed to lots of people Can gather large amounts of data quicky Reduces effort involved and makes questionnaires cost-effective
+ respondents may be willing to ‘open up’ Respondents may share more personal info than in an interview Less chance of social desirability bias compared to an interview
- Responses may not always be truthful Respondents tend to present themselves in a positive light Social desirability bias is possible
- Response bias Respondents may favour a particular kind of response eg they always agree Respondents reply in a similar way
Interviews
Face to face interaction between an interviewer and interviewee
Structured interview
– list of pre-determined questions asked in a fixed order
+ Easy to replicate Straightforward due to standardised format Reduces differences between interviewers.
- Interviewers can not elaborate Interviewers cannot deviate from the topics and points This may be a source of frustration for some.
Unstructured interview
there were no set of questions. There is a general topic to be discussed but the interaction is free-flowing and the interviewee is encouraged to elaborate
+ This is greater flexibility Unlike the structured interview, points can be followed up as they arise More likely to gain insight into interviewee’s worldview
- Difficult to replicate Such interviews lack structure and are not standardised Greater risk of interviewer bias.
Semi-structured interview
list of questions that have been worked on in advance but interviewers are free to ask follow up questions when appropriate.
Design of questionnaires
- Writing good questions – avoid jargon (technical terms only familiar to those specialised in the field)
^ Avoid double barrel questions
^ Avoid leading questions
-Closed questions – respondent has limited choices
^ Easier to analyse and can produce graphs and charts for comparison so easier to draw conclusions
^ Respondents are restricted – Forced into an answer that may not be representative so reduces validity
-Open questions – Respondents provide their own answers expressed in their own words
^ Respondents are not restricted so can provide more detail and have more validity than statistics
^ Difficult to analyse – Wider variety of answers than produced by closed questions and may be forced to reduce data to statistics
Design of interviews
Schedule – standardised list of questions that the interviewer needs to cover, can reduce interviewer bias
Quiet room – Increases likelihood that the interviewee will open up
Rapport – Begin with neutral questions to make participants feel relaxed
Ethics – Remind interviewees that answers will be treated in confidence
Quantitative data
Numerical data e.g. reaction time or number of mistakes
+ easier to analyse as you can draw graphs and calculate averages which can ‘eyeball’ data and see patterns at a glance.
- Oversimplifies behaviour eg using scale to represent feelings and this means individual meanings are lost
Qualitative data
Non–numerical data expressed in words eg an extract from a diary
+ Represents complexities so more detail included and an explanation which can also include information that is unexpected
- Less to analyse as large amount of detail is difficult to summarise so difficult to draw conclusions, many ‘ifs and buts’.
Quantitative data
Numerical data e.g. reaction time or number of mistakes
+ easier to analyse as you can draw graphs and calculate averages which can ‘eyeball’ data and see patterns at a glance.
- Oversimplifies behaviour eg using scale to represent feelings and this means individual meanings are lost
Qualitative data
Non–numerical data expressed in words eg an extract from a diary
+ Represents complexities so more detail included and an explanation which can also include information that is unexpected
- Less to analyse as large amount of detail is difficult to summarise so difficult to draw conclusions, many ‘ifs and buts’.
Primary data
positives and negatives
‘first-hand’ data collected for the purpose if the investigation.
+ Fits the job - The study designed to extract only the data needed which means information is directly relevant to research aims.
- Requires time and effort - the design may involve planning and preparation. Secondary data can be accessed within minutes.
Secondary data
positives and negatives
Collected by someone other than the person who is conducting the research, e.g taken from the journal articles, books, websites or government records.
+ Inexpensive – the desired information may already exist and requires minimal effort making it inexpensive
- Quality may be poor – information may be outdated or incomplete and challenges the validity of the conclusions
Meta-analysis
positives and negatives
A type of secondary data that involves combining data from a large number of studies. Calculation of effect size
+ Increases validity of conclusions – The eventual sample size is much larger than individual samples and increases the extent to which generalisations can be made.
- Publication bias – Researchers may not select all relevant studies, leaving out negatives or non-significant results and data may be biased because it only represents some of the data and incorrect conclusions are drawn.
mean
positives and negatives
Arithmetic average, add up all the scores and divide by the number of scores
+ Sensitive – includes all the scores in the data set within the calculation and more of an overall impression of the average than median or mode
- May be unrepresentative – one very large or small number makes it distorted and the median or mode tend not to be so easily distorted.
Median
positives and negatives
Middle value, place scores in ascending order and select middle value. If there are two values in the middle, the mean of these is calculated.
+ Unaffected by extreme scores – the median is only focused on the middle value and it may be more representative of the data set as a whole.
- Less sensitive- not all scores are included in the calculation of the median and extreme values may be important.
Mode
positives and negatives
Most frequent or common value, used with categorical/nominal data
+ Relevant to categorical data – When data is ‘discrete’ i.e represented in categories and sometimes, the mode is the only appropriate measure
- An overly simple measure – There may be many modes in a data set and it is not a useful way of describing data when there are many modes.
Range
positives and negatives
The difference between highest to lowest value
+ Easy to calculate – Arrange values in order and subtract largest from smallest and it’s a simple formula, easier the standard deviation
- Does not account for the distribution of the scores – the range does not indicate whether most numbers are closely grouped around the mean or spread out evenly. The standard deviation is a much better measure of dispersion in this respect.
Standard deviation
positives and negatives
Measure of the average spread around the mean. The larger the standard deviation, the more spread out the data it.
+ More precise than the range – includes all values within the calculation and a more accurate picture of the overall distribution of the data set
- It may be misleading – may hide some of the characteristics of the data set and extreme values may not be revealed unlike the range.
Normal distribution
symmetrical, bell-shaped curve. Most people are in the middle area of the curve with very few at the extreme ends
-The mean, mode and median all occupy the same mid-point of a curve
Skewed distribution
Distributions that lean to one side or the other because most people are either at the lower or upper end of the distribution
negative and positive skew
Negative skew – Most of the distribution is concentrated on the right of the graph resulting in a long tail on the left
Positive skew – Most of the distribution is concentrated towards the left of the graph, resulting in a long tail on the right
Ecological validity
Ecological validity is related to your ability to generalize your results.
internal and external validity
Internal validity relates to what goes on inside the study - are you measuring what you set out to measure.
External validity is the extent to which you can generalise your findings.