reserch methods Flashcards
observation techniques - participate
involves the observer becoming actively involved in the experiment they are studying eg zimbardo
what is observation
measurement of naturally occurring behaviour
unstructured observation
all relevant behaviour is recorded but no system is used
structured observation
use a system to organise the data being collected eg behaviour categories sampling method
observation techniques - non-participate
involves study of behaviour from a distance in the natural environment
observational techniques - overt
has knowledge they are being observed
observational technique - covert
no knowledge of observation
observational technique - naturalistic observation
behaviour is observed in a situation where everything is left as it is normally eg animal in its natural environment
observational technique - inter-observer reliability
the extent to which there is agreement between 2 or more observers involved in observations of a behaviour
observational technique- controlled observation
usually involves the study of spontaneous behaviour but under conditions manipulates by the experimenter
observational technique - observer bias
observers’ expectations affect what they see or hear. this reduces the validity of the observers
what is a lab experiment
-Takes place in a controlled environment with an IV manipulated by the experimenter and a dependant variable which is measured by the experiment
evaluation of lab expirement
strengths
+establishes cause and effect relationships
-allows for replication
-good control of confounding variables
weaknesses
- lacks ecological validity due to artificial environment
-demand characteristics
-expensive to set up
what is a field experiment
-involves direct manipulation of an IV but done in a natural environment eg school or park still uses an IV and measure DV
evaluate field experiment
strengths
+higher ecological validity
+reduced demand characteristics
weaknesses
-less control of extraneous variables
-more time consuming
what is a natural experiment
-uses a naturally occurring IV
for example testing the effect of the introduction of Tv on aggression on St helena island
evaluate natural experiments
+higher ecological validity
+reduced demand characteristics weaknesses
-little control of extraneous variables such as individual differences personality education home life etc
-little opportunity to use this method
what is a quasi experiment
investigates the relationship between an IV and a DV when the IV cannot be directly manipulated and participants cannot be allocated to conditions for example where the IV is a trait eg gender
evaluate quasi experiment
+allows research where the IV cant be manipulated for ethical reasons
+ enables psychologists to study real world problems
- cant demonstrate cause and effect
- less control of extraneous variables
evaluate participant observation
+researcher can gain greater understanding of groups behaviour
-researcher loses objectivity becoming part of the group
evaluate non participant observation
+researcher can remain objective throughout the study
- researchers loses a sense of the group’s dynamic by being separate
evaluate overt observation
+much more ethically sound because participants are aware of the research
-people may change their behaviour if they know are being observed
evaluate covert observation
+ pps more likely to behave naturally
-gaining ethical approval is difficult
evaluate naturalistic observation
+ ecological validity behaviour natural
-extraneous variables can’t control variables that may affect behaviour
- ethics limitations of situations where u can do naturalistic observation
evaluate controlled observation
+ more controlled
-lower ecological validity and behave differently if they know they are being observed
what are questionaires
- a questionarie is set of written questions designed to collect information about a topic
-they are composed of a variety of different questions and scales to measure them
-the aim is to uncover what people think and feel about a topic
evaluate questionnaires
+partical, large volume of info quickly and cheaply
- leading questions
- social desirability bias
what are interviews
face-to-face discussion between participants and interviewer using a range of structured or open - ended questions
types of interviews
structured - follow a fix set of questions same for all participants pp’s respond verbally
semi-structures - no fixed questions but topic areas are guided order etc can vary across pp’s
unstructured - the interviewee takes the conversation in the direction that they want the interviewer often starts with a few pre-determined questions and the aim
evaluate interviews
+ pilot study - useful way to get info before a study
-social desirability bias
-impractical time consuming and skilled researchers
- data analysis can be hard
what is correlation how is it different to an experiment
-looks at relationship between 2 variables arnt manipulated may be caused by a third unknown variable or coincidence
evaluate correlational research
+ethics you can study variables that would be unethical to manipulate
-casual relationship can’t assume correlation may be caused by a third unknown variable
what is a case study and evaluate
case study - intensive descriptions of a single individual or case they allow researchers to analyse unusual cases in a lot of detail eg milner et al study
+unique cases can challenge existing ideas and suggest future ideas for research
-cause -effect relationship can’t be established
the different types of hypotheses and examples
-null hypothesis- a prediction that there will be no relationship between key variables if data doesn’t support you challenge it
alternative hypotheses- when data forces you to accept an alternative hypothesis to show variables are linked
-directional hypothesis- stating that one group will do BETTER or HIGHER data than another USED WHEN PAST RESERCH SUGGESTS THE FINDINGS WILL GO IN A PARTICULAR DIRECTIONAL
-non- directional- predict no difference eg just saying there will be a difference in exam results compared to those who revised and don’t USED WHEN THERE IS NO RELEVANT PAST RESEARCH OR CONTRADICTORY RESEARCH
-experimental hypothesis testable statement or prediction more words will be remembered when learning in silence than when TV is on
what is oppurtunity sampling and evaluate
recruiting people who are most convient or avliable eg people walking in the street or students at school and fir the criteria you are looking for
+ easiest method and less time consuming
-biased small smaple of population exculded certain populations eg people at work or children at school
what is random sampling
every member of the population has an equal chance of being chosen identifying everyone in target population and selecting amount of people needed which gives everyone equal chance
+ unbiased all members of the target population have a equal chance of selection
- takes time to create a list of all members and contact them
stratified sample
identifies subgroups in accordance to their frequency in the population and then are randomly selected in the same proportion as they occur in the real population
+more representative because there is proportional selection
-very time consuming to identify subgroups and contact them etc
systematic sample
use a predetermined to select participants for example choosing every nth person from a phonebook
+unbiased because it uses a objective system
-only random if number is selected randomly before
volunteer sample
advertise in a newspaper or on a noticeboard or on the interent and is self selected sampling
+gives access to a variety of participants who may read the advertisesment
-sample biased in some ways eg those who arehighly motivated and have time on their hands or need money this results in volunteer bias
pilot study
is a small scale test run of the reserch design to investigate if certain aspects of the experiment work to see to what extent the procedure needs fine tuning eg if participants dont understand the experiment
repeated measures
all participants recieve all levels of the IV and compare the performance of DV of the participants on the two tests eg all pp’s tries every brand of chocolate
-order effect some may be better on second test due to practice effect or get bored boredom effect or may guess what the experiment is about demand characteristics practice effect
ways of dealing - counterbalancing Each condition of IV occurs first and last equally which disrupts order effects
independant group designs
participants are placed in independant groups each group does one level of the IV we compare the performance DV of the two groups USES DIFFERENT PP’S IN EACH CONDITION EG ONE GROUP 1ST BRAND ONE GROUP 2ND BRAND
-individual differences may impact on the DV rather than the IV itself
TO OVERCOME = random allocation each pp’s a number put all in a hat draw one no at a time and allocate 1st to condition 1 continue until all are allocated
matched pairs design
match participants on key characteristics believes to affect performance of DV eg IQ
- very time consuming and difficult to match and not able to fully control eg history
ways of dealing resistrict number of variables
conduct poliet study
behavioural categories
observers agree on a list of specific behaviours they want to observe
-this ensures clear exactly what is being measured and observers are looking for same behaviour
-for example aggressive behaviour eg shouting hitting kicking and screaming
sampling procedures
event sampling - counting the number of times a certain behaviour occurs in a individual in a certain time period
time sampling - recording behaviours in a given time frame eg every 30 seconds and may tick one or more categories from a checklist
what is a closed question
in an interview where the questions have a predetermined range of answers from which respondents select one tend to produce quantitive data ed yes/no and mutiple chose
what is a open question
questions that invite respondants to provide their own answers and produce qualitative data eg describe how you feel about psychology
+avoids biasing answers
+provides richer more accurate information
+gives opportunity to respond freely
what is a confounding variable
variables that interfere with the effect of IV you could explain the results of the study change in DV with a factor other than IV
eg more intelligent better memory or eyesight
what are extraneous variables
for example reserchers dont know if participants ability is distrubted evenly which makes it more difficult to detect the effect because other factors have an effect often called nuisance variables and make it difficult to detect a significant effect
eg age, lighting, noise in background things that make it harder to detect a result they dont confound the findings of the study
types of extraneous variables
participant variables = to do with differences between the pp’s
situational variables = features of the experimental situation
what is operationalised
define a variable to specify how it is measured or manipulated make it very clear exactly what was measured and how eg intelligence IQ test score
procedure of random allocation
-obtain all the people in the population that wants testing
-put all the names in a lottery barrel or hat
-select the number of names required without looking
procedure of counterbalancing
way 1 AB or BA
divide group into 2 groups
-each particpant does A then B
-each participant does B then A
standardised procedure
a set od procedures that are the same for all participants in order to be able to repeat the study. this includes standarised instructions which tell participants how to perform the task
what is demand characteristics
a cue that makes participants unconscouisly aware of the aims of the study or helps participants work out what the resercher expects to find and may change how they would naturally behaviour
way to deal with… single blind design that participants are unaware of the researchs aim/hypothesis this prevents them from seeking cues and reacting to them
experimental realism if the experimental task is suffciently encaging participants are less likely to pay attention to the fact they are being observed
what is the investigator effects
unwanted influence that the investigator does that has an effect on a particpants performance in a study other than what is intended which may affect participants natural behaviour
ways to deal with… both the resercher and partipant dont know the aim or hypothesis and therefore less likely to produce cues about what he/she expects
The BPS code
- respect upkeeping the dignity and worth of participants eg confidentiality
-competence psychologists should maintain high standards in their professional work
-responsibility psychologists have responsibilty over physical and psychological harm eg protection from harm and debriefing
-integrity psychologists should be honest and accurate
ethical issues
-informed consent
given comprehensive info concerning the nature and purpose of research and their role in it
-deception
not given the real aims of the study eg being ied to and therfore cannot do informed consent
-the right to withdraw
participants can stop if they feel uncomfortable and can withdraw their results
-protection from psychological and physical harm
should not experience negative effects
-confidentiality
no communication of personal information and info protected
-privacy
a persons right to control the flow of information about themselves
ways to deal with ethical issues
-informed consent = gain presumptive consent
deception = being fully debriefed
right to withdraw = informed at the start they have the right to withdraw
protection from harm = stop the study is harm is suspected offer follow up counselling if neccasary
confidentiality = not record names use numbers or false names use pseudonyms
privacy = dont study anyone without their informed consent
what is peer review
the practise of using independant experts to assess the quality and validity of scientific research and academic reports
evaluation of peer review
- finding an expert in some fields is not easy
-publication bias journals tend to publish positive results less likely to agree to publish negative findings
-anonymity reviewed anonymously to enable objectivity and honesty
-preserving status quo publishes what applies to current theory science is resistant to shifts
the role of peer review
-publication of research in scientific journals and books ensures what is published is reliable and accurate prevents faulty or incorrect data enetering public domain
-allocation of research funding
paid for by government and charities
-assessing research rating of university departments rates quality of research conducted by university which influences funding received by university
intra observer reliability
the extent to which there is an agreement between 2 or more observers involved in observations of a behaviour kappa score 1 would be perfect agreement
features of a science
objectivity - being subjective
empirical - observation to reach logical judgement
replicability - being able to do the test multiple times improves reliablity
falsifiability - hypothesis can be proven wrong by the experiment
quantitative data
which represents how much or how long there are of something deals with numbers and which can be measured using behaviour categories or frequency tabled
+easy to analyse conclusions easily withdrawed
- data may over simplify reality
qualitative data
deals with descriptions concerns attitudes beliefs and emotions using collection of videos or written dairy
+ can provide rich detailed info of peoples experiences
- more difficult to analyse and draw conclusions
primary data
information observed or controlled directly from first hand experience collected for the study currently being undertaken
+data collected can be designed so it fits the aims and hypothesis of the study
-is a very lengthy and expensive process
secondary data
information used in a reaserch study that was collected by someone else or for a purpose other than the current one for example published data collected from the past
+simplier and cheaper to access someones else data significantly less time and equipment needed
-data may not the needs of the study
meta analysis
a resercher looks at the findings from a number of different studies and produces a statistic to represent the overall effect
+may increase the validity of conclusions reached because they are based on a larger sample
-different studies can vary considerably and not truly vomparable therefore calculating the effect may not be appropriate
measures of central tendency
- gives a typical value for data set where the middle of data set is
-mean adding all the data and diving by the number of numbers
-median middle value in an ordered list
-mode (most common)
measures of dispersion
- range
- standard deviation shows the amount of variation in a data set. it asscess the spread of data around the mean
bar chart
-uses data which comes in categories (discrete data)
-vertical axis = score of DV
-horizontal = individual categories or IV
-bars should be drawn separately with equal gaps
histogram
-use continuous data
-varitical axis = frequency of something
-horizontal axis = scale of something eg length or mass
-they don’t have gaps between bars
line graph
-contineous data on the x axis
-points are the top of where a bar would be
-highlights continuous nature of variable on x axis
scatter gram
-when doing correctional analysis
-involves plotting 2 scores on the x axis and one on y axis
normal distribution
population scores are evenly distributed around the mean. a symmetrical classic bell shaped curve
- the mean median and mode are all in the exact mid point
positive skewed distribution
looks like a slide mode eg is closer to the x axis
negative skewed distribution
looks like a rollacoaster mode etc are closer to the right end of the graph
when to use a sign test
-used when looking at related data
-the hypothesis states a difference between 2 sets of data
-2 sets of data are related
-nominal data
-matched pairs
why do we use 5% probability level in psychology
so we are certain that there is less than a 5% possibility that the results occurred by chance
how to calulate the sign test method
- state the hypothesis
if directional = one tailed test
non-directional = 2 tailed test - calculated difference (after-before) the directional factor away from the other factor eg After holiday - before holiday recored each score as a + or -
- find s value
added up the total number of + and - the less frquent sign is X and thefore the calculated value - Find critical value of S
N= number of participants - exclude any 0 differences and use 0.05 - results?
we can accept the hypothesis if the s value is equal to or lower than the critical value to show significance
what is the experimental method
-observe = identify problems and trends
-hypothesis
-test and collect data
-analyse results against dependant variable
-accept or reject hypothesis
define self report
- a person reports and describes their own attitudes, thoughts and feelings usually through answering questions on a questionnaire (expressing answers to open or closed questions) or via an interview which can be structured or unstructured
writing a good question
-clarity
key terms are operationalised and are non-ambiguous
-bias
be prepared for social desirability
reverse your questions so pp’s cant just tick the same end of scale regularly
-analysis
questions need to be written so easy to analyse
use similar types of questions to enable a score to be generated
questionnaire construction
-filler questions include some irrelevant questions to distract from main purpose of the study
-sequence of questions put easier / less challenging questions first
-sampling technique stratified
-pilot study test questionnaire on small group first
rank order
please indicate in rank order your preferred type of animal 1 next to you favourite 5 for your least faviourite
rabbit, dog, gerbil, cat and lion
likert scale questions
indicate the strength of agreement with a statement eg psychology is the best subject in the world do you agree or disagree 1 agree 10 disagree
checklist questions
pp’s indicate which items from a list apply eg choose 3 adjectives which best describe your personality eg grumpy happy smiley
test- retest
test - retest relatability is a measure of consistency of a psychological test or assessment and is measured by administering a test twice at 2 different points in time this type of reliability assumes there will be no change in the quality or construct being measured
face validity
the extent to which a self-report measure looks like it is measuring what it claims to measure it requires the researcher to ask people whether they intuitively think that the measure it accurate
concurrent validity
the extent to which a new questionnaire produces the same result as a pre existing questionnaire if it doesn’t it has high concurrent validity pp’s complete new and pre-existing questionnaire there scores are compared looking at similarity and correlation
interviewer bias
the effect of an interviewer’s expectations,communicated unconsciously on a respondents behaviour
social desirability bias
a distortion in the way people tend to answer questions they tend to answer in a way that presents them in a better light
correlation
a systematic association between 2 continuous variables
- a method used to analyse data not strictly a research method
-variables are simply measure no manipulation or change made
-represented using a scattergram
correlation values
-range from -1 to +1
-positive correlation co-efficient as one variable increases so does another eg taller people have bigger feet
-a perfect positive correlation would be represented by +1
-negative correlation co efficient as one variable increases other decreases perfect negative correlation= -1
evaluate correlation
+correlation can be used to investigate trends in data if correlation is significant then further investigation is justified
+the procedures in a correlation can be easily repeated to test its reliability
-correlations cannot show a cause and effect relationship no manipulation of IV
evaluate methods of central tendency
mean +uses all values in data set most sensitive to variations in the data - artificially raised or lowered by extreme values
median + not distorted by extreme values -less sensitive to variations in data
mode +summarsing frequency data -serval modes
evaluate methods of dispersion
range + easy to calculate takes into consideration extreme scores
-ignores rest of results
standard deviation +uses every score in calc +not heavily distorted by extreme scores
-laborious to calculate
stages of content analysis qualitative data —-> quantitive data
-ask a specific reserch question
-data is collected
-resercher reads through and examines data familiarising themselves
-resercher identifies coding units
-test the coding on sample material and assess reliability
-tally is made of number of times that a coding unit appears
content analysis evaluation
+high ecological validity
+ can be replicated and tested for reliability
-observer bias reduces objectivity and validity os results
-culture bias
thematic analysis
-a qualitative content analysis summarises data descriptively and identifies underlying themes in data allows themes to emerge from the data and maintain pp’s pov
steps two thematic analysis
- read and reread data transspricts
-break data down into meaningful units assign a label or code to each unit
-combine codes into larger themes
-ensure that these emerging themes represent all the data
type 1 error
FALSE POSITIVE rejecting null hypothesis that is true -eg positive covid test when you’re actually negative
type 2 error
FALSE NEGATIVE accepting null hypothesis that is not true eg negative covid test when you’re actually positive
which criteria choosing a statistical tests
-data (nominal,ordinal, interval)
-design (independent groups, repeated measures, matched pairs)
-difference (correlation or test of difference)
data levels of measurement statistical test
nominal= pp’s are categorised into groups
ordinal= data ordered in some way eg ratings or ranking often based on subjective attitudes
interval= data measured using formal units of equal distance eg weight and height
memory aid to remember the testistical tests
Carrots = chi -squared
scare = sign test
me = mann-whitney
when = wilcoxon
served = spearmans
under = unrelated t-test
roast = related t-test
potatoes = pearson’s
Data up to down = nominal, ordinal, interval
test of difference left to right design IG, design RM/MP