RESEARCH METHODS YEAR 1 Flashcards
Aim
The purpose of the experiment. Aims are to be stated before the research begins to make it clear what the study intends to investigate.
One-tailed (directional) hypothesis
states the directions of the difference or relationship. (more/less,higher/lower,faster/slower).
Is used when there is existing research in the subject
Two-tailed (non-directional) hypothesis
states there is a difference between the conditions or groups but the nature of the difference is not specified.
Is used when there is not existing research in the subject.
Null hypothesis
When we predict that nothing will happen
Bar chart
Used to represent discrete data where the data is in categories which are placed on the x-axis. the mean or frequency is on the y-axis.
they differ from histograms as columns do not touch and they have equal spacing
Histogram
used to represent data on a ‘continuous’ scale
columns touch because each one forms a single ore (interval) on a related scale
scores are placed on the x-axis
the height of the columns show the frequency of values
Scatter graph
used for measuring relationship between two variables . the pattern of plotted points show the relationship
difference between correlation and experiment:
it is not possible to establish cause and effect using correlation
you may find a strong link between things but that does not necessarily mean one causes another. instead you have found an association.
Independent and dependent variables
Independent: is manipulated
Dependent: is recorded
Operationalisation
How we measure the variables. We must define how we intend to measure the IV and DV.
Controls
Random allocation
Counterbalancing: half of the participants participate in condition A before condition B and vice versa. This means that the first and second condition is not the same for every participant.
Randomisation: means that everyone has an equal chance of doing either condition
Standardisation: everything should be as similar as possible for all of the participants. For example, instructions are the same across the same across the conditions
Extraneous Variables
Anything other than the IV which could influence your result. These should have been accounted for before the experiment takes place.
They can be controlled by randomisation or enduring that all participants are the same in a relevant way
Confounding variables
Anything other than the IV which has influenced your results which has not been accounted for before the experiment
(Ethics) Informed consent
knowing aims and giving your permission to take part in the study (Menges 1973)
(Ethics) Deception
Deliberately misleading or withholding information. BPS states that deception is only acceptable of there is a strong scientific justification for the research and there are no alternative procedures available.
(Ethics) Right to withdraw
Being able to leave when desired. Participants need to be aware of this.
(Ethics) Confidentiality
details should be kept private.
(Ethics) Protection from harm
no more harm than daily life (Glass and Singer 1972)
(Ethics) Debrief and how it deals with all other issues
it returns the participant to the state they were in before the research
(Design) independent groups
There are two separate groups of participants. One group takes part in condition A and the other in condition B.
Adv:
No order effects- only take part in one condition, don’t get bored or practised
Fewer demand characteristics- participants may only know their condition
Dis:
more participants needed
individual differences as the people take part in each condition are different- one group might be better at the task
(Design) Repeated Measures
There is only one group of participants which is put in both conditions.
Adv:
No individual differences as the same person does both conditions
Dis:
Order effects- either boredom or practice. Can be helped by counter balancing
Demand characteristics- participants know what the experimenters are expecting and may perform to meet that expectation. Also the measure has to be changed e.g. two sets of words to memorise
(Design) Matched Pairs
Two separate groups, but this time they are matched into pairs for certain characteristics. One of each pair takes part in each condition.
Adv:
no order effects
controls for individual characteristics. Can be more sure the IV caused difference in DV rather than big differences between the 2 groups
Dis:
can be difficult to make perfect matches and is costly on money and time
Field Experiments
outside of the laboratory but basic scientific procedures are still followed as closely as possible
the independent variable is manipulated
Adv:
less artificial than a lab
represents reality
can be completed in natural environments
avoids participant effects (if they aren’t aware of the study), producing more natural behaviour
Dis:
difficulty controlling extraneous variables so less able to show cause and effect
Ethical issues- participants are unlikely to know that they are being studied
Laboratory Experiments
Contolled artificial environment in which the independent variable is manipulated
Adv:
controlled environment
minimises problems caused by extraneous variables
easily repeatable
reliable
Dis:
artificial environment (mundane realism)
participants may behave differently to normal or be effected by the environment
lacks generalisability
Confederate
somebody who appears to be a participant but is aware of the aim/method of the experiment and is told how to behave
Demand characteristics
Participants trying to guess the nature and purpose of the experiment and thus behaving differently.
Single and Double Blind
Double: Neither the participants nor the researchers are aware of the aims of the investigation. This is often used in drug trials, so that the researcher doesn’t know which group gets the placebo. This avoids investigator effects (bias).
Single: Participants are not aware of the condition they are in. Attempts to control for the confounding effects of demand characteristics.
Natural Experiment
Takes place in a natural environment and the independent variable is not manipulated, taking advantage of a completely natural event.
Adv:
High ecological validity, providing opportunities for research which otherwise may not be available
enables ‘real’ problems to be studied
objective research method (little influence from the researcher)
few ethical issues
Dis:
the natural event may be one that only happens rarely, limits opportunity and generalisability
many extraneous variables which threaten ‘cause-effect’ conclusion
participants aren’t randomly selected, meaning that there could be bias
Quasi Experiement
IV is a naturally existing characteristic between people which has not been manipulated. ‘Quasi’ refers to something that is not quite real, but may work as the real thing does.
Adv:
often carried out under laboratory conditions so therefore high in controls
enables psychologists to study ‘real’ problems.
Dis:
like natural experiments, no random allocation so there may be confounding variables, meaning we cannot come to a cause-effect conclusion
Behavioural Categories
devision of component behaviours
Event sampling
counting the number of times a certain behaviour (or event) occurs in a target individual or individuals.
Time sampling
Recording behaviours in a given time frame. Might be what an individual is doing every 30 seconds
Controlled Observation
When the researcher has some measure of control over the environment.
Adv:
control over extraneous variables
inter-observer reliability (agreement between researchers)
easy to replicate
Dis:
Cannot be applied to real life settings
may be subjective towards what the researcher wants to see
Naturalistic observation
Studying behaviour in a natural setting where everything is as it normally is.
Adv:
High ecological validity
Natural environment is generalised to everyday life
fewer demand characteristics
Dis:
Replication is difficult due to lack of control
Uncontrolled extraneous variables
Covert observaiton
Participants are not aware that the are being watched
Adv:
No demand characteristics
Dis:
ethical issues as they don’t now that they are being observed
Overt Observation
Participants are aware that they being watched
Adv:
fewer ethical issues as there is no deception
Dis:
demand characteristics as they know they are being observed
Participant observation
The observer acts as part of the group being watched
Adv:
first hand experience
insight
increased validity
Dis:
lose objectivity
difficulty in recording observations
ethical issues
Non participant observation
The experimenter does not become a part of the group being watched
Adv:
more ethical
more objective
Dis:
less insight
not experiencing the same things
lower in validity
Structured obsevation
The researcher determines precisely what behaviours are to be observed and uses a standardised checklist to record the frequency with which they are observed within a specific time frame
Adv:
easier to gather relevant data as you know what to look for
Dis:
interesting behaviours could go unrecorded because they weren’t predefined as important
Unstructured observation
The observer recalls all relevant behaviour but has no system
Adv:
all behaviours will be recorded, despite not being what was originally expected
Dis:
it is harder to gather relevant data because you don’t know what you are looking for
Measures of central tendency
mean, mode, median
Mean
adding up items and dividing by numbers of items
Adv:
takes account of exact distance between all the values of the data
representative of all the data
Dis:
anomalous values can distort and misrepresent data
cannot be used with nominal data
Mode
The most common value
Adv:
unaffected by anomalies and is useful for discrete data and the only method which can be used for nominal data
Dis:
sometimes there are so many modes that the data cannot be described using this statistic
Median
middle value in an ordered list. items must be arranged in order then the central value is the median when there are two central values, they must be added together and divided by two
Adv:
not affected by anomalies
easy to calculate
Dis:
less sensitive as exact values are not reflected
Measures of Dispersion
how dispersed scores are
Range
arithmetic distance between the highest and lowest values. it is customary to add 1
Adv:
easy to calculate
Dis:
affected by extreme values
doesn’t take into account the distribution of numbers
Standard deviation
more precise. a measure of the average distance between each data item above and below the mean ignoring plus and minus values. the smaller the standard deviation, the closer together the values are
Adv:
precise method of dispersion because it takes into account all of the values
Dis:
may hide some characteristics of the data set (extreme values)
Quantitative Data
uses numbers
-Discrete: data in discrete categories: bar charts use discrete data.
-Continuous: data continues. is measured using a scale of measurement
Pilot Study
A small scale version of an investigation that taks place before the real investigation is conducted. The aim is to check that procedures, materials, measuring scales etc., work and allow the researcher to make changes or modifications if necessary. Ethics need to be checked beforehand.
Sample
The group from the population that is selected for the researcher to use in the research.
Population
The group of people from whom the sample is drawn. We use a target population if we want to investigate specific individual differences.
Random Sampling
each person has an equal chance of being selected. Chosen from a computer random generator/ picked out of a hat.
Adv:
unbiased: all members of the target population have an equal chance of selection
Dis:
The researcher may end up with a biased sample because the sample is too small
Subgroups of target population might not be selected- does not guarantee a representative sample
Volunteer sampling
Researcher advertises the study and people who are interested apply to be in the research.
Adv:
quick, convenient and ethical if it leads to informed consent
large response rate
allows more in-depth analysis and accurate results
Dis:
sample is biased because the participants are likely to be more highly motivated (volunteer bias)
Opportunity Sampling
Asking people who are available at the time to take part in the research
Adv:
the easiest and fastest method because you use the first participants you can find
Dis:
biased because the sample is drawn from a small part of the target population
unlikely to be representative of a target population
Stratified sampling
Selecting people from every portion of your population in the same proportions
Adv:
more representative than an opportunity sample because there should be equal representation of subgroups
Dis:
it is time consuming because all potential participants need to be assessed and categorised
some groups within a sample may not be represented if a small sample is used
Systematic sampling
Selecting every nth name from a list
Adv:
it avoids bias as, once the researcher has decided what number they have no control over who is being selected
Dis:
it is not completely objective because the researcher may decide on how people are listed before the selection
there is a small chance of a ‘freak’ sample which would not be representative
Informed consent
consent letter. Should include aim of the study, nature of the study, offer counselling, must be anonymous (name removed from data), must be signed, allows right to withdraw.
Presumptive consent
ask similar group of people for consent
Prior general consent
consent for different studies, including one that may involve deception
Retrospective consent
asked consent during debrief
(levels of measurement) Nominal
represented in the form of categories. is discrete and one item can only appear in one category
(levels of measurement) Ordinal
Data which is ordered in some way
does not have equal intervals between each unit
based on subjective opinion
(levels of measurement) Interval/ratio
based in numerical scales that include units of equal precisely defined size
(units of measurement for example)
Structured interview
when questions are decided in advance
adv
can be easily repeated- standardised questions
requires less skill than unstructured interviews
easier to analyse than unstructured interviews
dis
interviewer bias may still occur
social desirability may still occur
data collected will be restricted by a predetermined set of questions
Unstructured interview
the interview start with some general aims and questions and then lets the interviewee’s answers guide the subsequent questions
adv
detailed and in depth information
access information that may not e revealed from pre-determined questions
deep insights into feelings and thoughts
you can tailor questions to specific responses
good rapport, high in validity
dis
more affected by interviewer bias then structured interviews
requires well-trained interviewers
low reliability, interviewer may behave differently or ask different questions
hard to analyse answers
body language
Semi-structured interview
combination of both structured and unstructured interview techniques
(self report) Questionnaire
a set of questions used to assess a person’s thoughts and experiences. Usually gathered from large numbers of people, carried out on selected samples of people, may include a mixture of both open and closed questions
adv
cost effective
no demand characteristics
researcher does not need to be present
straightforward to analyse
dis
leading questions
social desirability bias
misunderstanding questions
Question types (questionnaires)
Likert: a number of responses to a question which often demonstrate a degree of agreement, quantitative
Open : the participant can give any answer they wish, qualitative
Closed: possible answers are determined by the researcher. various types of these and best asked when factual information is needed
Questionnaire design
clarity: questions need to be clear so that the reader understands what is being asked
bias: leading questions might lead to desirability bias (when respondents choose to portray themselves in a positive way.
sequencing for the questions: start with easier questions until respondent is more relaxed.
(self report) Interview
A research method that involves a face to face interaction with another individual and results with the collection of data.
Difference between correlation and experiment
it is not possible to establish cause and effect using correlation
you may find a strong link between things but that does not necessarily mean one causes another. instead you have found an association.
Correlation
using to or more variables (co-variables) are measured in order to identify if there is a relationship between them (e.g. height and shoe size).A single numerical value is produced that is used to describe the relationship.
adv:
-can be used when it would be unethical/impractical to conduct an experiment
-if correlation is significant, then further investigation is justified
-if correlation is not significan, you can rule out a causal relationship
dis:
-correlational analysis cannot demonstrate a cause and effect relationship between variables
-there may be other unknown variables that can explain why the co-variables are linked
-extraneous variables may lead to false conclusions
Correlation coefficient
number between -1 and 1. tells us how strong the correlation is, the nearer to 1 or -1, the stronger the relationship. It has a plus or minus sign in front of the number which tells us whether the correlation is positive or negative.
Distributions (normal and skewed)
normal distribution: a symmetrical spread of frequency data that forms a bell-shaped pattern. The mean, median and mode are all located at the highest peak.
Skewed distribution: a spread of frequency data that is not symmetrical where the data clusters to one end
negatively skewed
the long tail is on the negative (left) side of the peak and most of the distribution is concentrated on the right. The mode and the median are greater than the mean
positively skewed
the long tail is on the positive (right) side of the peak and most of the distribution is concentrated on the left. The mean is greater than the mode and median.
Validity (Internal and External)
Internal: the study measures or examines what it claimed to measure or examine
External: the extent the results of the study can be generalised to others (also known as ecological validity)
7 types of Validity
Ecological: generalisable to real life settings- generalising findings from one setting to other settings
Concurrent: results from a new test can be compared to a previously well-established test
Population: whether you can reasonably generalise the findings from your sample to a larger group of people
Temporal: assesses to what degree research findings remain over time
Face: does the test look as though it measures what it intends to measure
Content: involves asking experts in the field to check the content of the study
Predictive: if diagnosis leads to successful treatment then the diagnosis is seen as valid
Reliability (Internal and External)
reliability is the overall consistency of a measure
Internal: the extent to which a test is consistent within itself
External: refers to the ability of the test to produce the same results each time it is carried out
Ensuring/assessing reliability
Inter-rater reliability: two or more interviewers/observers must get the same outcome on 80% or more of the behaviours
Split-half method: compare an individual’s performance on two halves of a test
Test-retest method: a person repeats a test a month or after doing the same test the first time
Improving reliability
Questionnaires: test-retest
-low test-retest might require items to be deselected or rewritten
Interviews: structured interviews are more reliable due to fixed nature
Observations: operationalising the behavioural categories
Content analysis
A method of quantifying qualitative content via coding/categorisation
data is analysed as typologies, quotations and summaries
hypotheses are grounded in the data
-coding units are identified to analyse data
-see how often that code appears
used with secondary data
Evaluation of content analysis
a clear summary of the patterns in the data may be established
once a coding system has ben set up, replication is easy
this, in turn, improves reliability
can be subjective
reducing the data to coding units removes detail
Thematic analysis
Involves making summaries of data and identifying key themes and categories. These themes should be recurrent.
• researcher becomes familiar with the data
• researcher looks for different themes, reviews the themes, defines and names the themes and writes a report
Hypothesis is made after the data is collected and is grounded in the data
Evaluation of Thematic analysis
qualitative analysis preserves the details in the data
creating the hypotheses during analysis allows for new insights to develop
some objectivity can be established by using triangulation
how do you decide which categories to use and whether something fits a category?
how do you decide what to leave out of the summary
subjective decisions
Case studies
an in-depth study, using a range of methods on one person or a small group. Because it uses a range of different research methods, this increases reliability, by the process of triangulation. Using a range of different methods is a way of double-checking results.
adv:
rich data- researchers have the opportunity to study rare phenomena in a lot of detail
unique cases- can challenge existing ideas and theories and make suggestion for future research
dis:
causal relationships- cause and effect cannot be established
generalisation
ethics- informed consent
Consent form and what to include
-outline of what the experiment entails, including aims, time of the experiment, duration of experiment,general purpose of the study
-Ethical guidelines: No pressure to consent, right to withdraw, right to withdraw data, confidentiality and anonymity, they should be able to ask questions at any time, they should receive a full debrief
-Format and style of consent form:
will require participant’s agreement
could be written as a form that participants need to sign, including space to do so
space for the participant to write the date
a space to print their name
