Research Methods, Y2 Flashcards
What is a case study
A detailed and in depth analysis of an individual/group/institution or event
What do case studies involve
Analysis of unusual individuals or events like a person with a rare disorder or sequence of event that to something. Or they may concentrate on typical cases like elderly person memories of childhood
What does conducting a case study involve
Production of qualitative data, researcher may construct case history of individual concerned using interviews, observations, questionnaires or all of them, person may be subjected to experimental or psychological testing to access what they’re capable of producing quantitative data
How long so case studies last
Take place over long period of time (longitudinal) and may involve gathering additional data from family/ friends and person themself
What is content analysis
Type of observational research where people indirectly studied via communications they have produced
What are the forms of communication in content analysis
Wide ranging but may include spoke interactions, written forms, broader examples from media lien magazines or tv shows
What is the aim of content analysis
To summarise and describe this communication in a systematic way so overall conclusions can be drawn
What is coding in content analysis
The initial stage, some data sets analysed may be very large and needs to be categorised into meaningful units
What does coding involve
Counting up number of times a particular phrase appears to produce quantitative data
What is an example of coding
Newspaper articles analysed for derogatory terms for people with mental health issues, like crazy or mad
What is thematic analysis
A form of content analysis but outcome is qualitative
What is main process in thematic analysis
Identification of themes, such as any idea, explicit or implicit, that reoccurs in whatever is being studied, likely to be more descriptive than coding units
What is an example of thematic analysis
People with mental health may be misrepresented in newspaper as threat to well being of children or drain on NHS, such themes can be put into broader categories like control, stereotypes, treatment, once researcher Stanford that themes develop cover most aspects they collect new set of data to test validity of theme and categories and if they explain new data research will write up report using direct quite to illustrate themes
What are strengths of case studies
Offer rich, detailed insights that shed light on atypical behaviour, may be preferred to superficial data from questionnaires. Eg. HM helped understanding of typical function by existence of STM and LTM stores. Case studies generate hypothesis for future studies
What are limitations of case studies
Generalisation of findings issue when dealing with small sample size, info based on subjective selection and participant may be prone to memory decay and inaccuracy especially if childhood memory being told, so evidence low in validity
What are strengths in content analysis
Useful to get around ethical issues as lots of material is already existing in public domain so no issues with getting permission. So high in external validity, can also produce both qualitative and quantitative data
What are limitations of content analysis
People studied indirectly so communication analysed outside of context so research may attribute options and motivations that writer didn’t originally intend. Content analysis lacks objectivity especially in thematic Analysis
What is reliability
A measure of consistency, if a certain measurement is made twice and produces the same result then that measurement is reliable
What would we expect if we did a test in psychology on one day and on another day
We would expect same results, unless the thing has changed
What ways do we assess reliability
Test-retest, inter-observer reliability, measuring reliability
What is test-retest for assessing reliability
Involves administering same test/questionnaire to same people on different occasions, if test is reliable results will be similar on all different occasions
What must happen when using test-retest for reliability
Must be sufficient time between the tests to ensure the participant can’t recall their answers but not too long that their attitudes/abilities have changed
test-retest method used for what studies
Questionnaires and psychological test and interviews
When using test-retest for test or interview what can be done with results
To scores can be correlated to ensure they’re similar, if correlation is significant and positive then reliability is good
When is inter-observer reliability used (for what studies) to measure reliability
Observational research, as it is subjective, bias and unreliable
How does inter rater reliability work
May involve small scale trial run of observation to check observers apply behavioural categories in same way or a comparison may be reported at end of study, observers must watch same event bud record data independently, data should be correlated to assess reliability
When is inter observer reliability also used
For content analysis and inter interviewer reliability can be used for interviews
How is reliability measured
Using a correlational analysis, in test-retest and inter observer reliability the 2 sets of scores are correlated. Correlation coefficient must exceed +.80 for reliability
How can questionnaires improve their reliability
If low test-retest reliability found may require some items to be rewritten, if some questions complex/ambiguous, they may be interpreted differently by same person on different occasion, so one solution is to change open questions to closed fixed questions which are less ambitious
How can interviews improve their reliability
The interviewer should be the same each time to ensure reliability, if not possible interviewers must be well trained and not ask leading questions, this is easily done in structured interviews
How can observations improve their reliability
Behavioural categories must be operationalised, measurable and self-evident. Categories shouldn’t overlap and all possible behaviours should be on checklist. If categories aren’t operationalised well, overlap or are absent, different observers make own subjective judgement. If low reliability observer may need more training
How can experiments improve their reliability
The procedure must be consistent to compare reliability. So if the experiment has a standardised procedure it is likely to be reliable
what is valdity
whether a psychological test, observation, experiment produces legitimate results, including whether researcher managed to measure what they intend to measure and the extent to which findings can be generalised beyond the research setting where they were found
what is an example of a study producing reliable data that is not valid
a broken set of scales may give a constant reading of weight but it is not the correct weight
what is internal validity
whether the effects observed in an experiment are due to manipulation of the independent variable and not some other factor
what is a major threat to internal validity in an experiment
if participants respond to demand characteristics and act in a way they believe is expected of them
what is external validity
factors outside the investigation like generalising to other settings, other populations of people and other eras
what is ecological validity
type of external validity concerns generalising findings from a study to other settings like everyday life
do lab studies always have low ecological validity
no, if a task that is used to measure dependant variables in a experiment isn’t like everyday life (low mundane realism) it has lower ecological validity, regardless of what experiment type it is, so must look at multiple factors to determine ecological validity
what is temporal validity
issue of if findings from a certain concept or study stay true over time,
what is an example of temporal validity
Asch experiments produced from a certain conformist era and may produce different results if repeated today
what is face validity
a basic form of validity where a measure is scrutinised to determine if it measures what its supposed to measure
what is a basic form of validity
face validity, whether a test, scale, the measure appears to measure what it’s supposed to, can be determined by eyeballing the experiment or passing it to an expert to check
how is face validity measured
eyeballing the measuring instrument or passing it to an expert to check
what is concurrent validity
extent to which a psychological measure relates to an existing similar measure
when is concurrent validity of a test or scale demonstrated
when the results obtained are very close to those obtained on another well-establsihed test, a close agreement indicated when the 2 data sets score exceeds a correlation of +.8
how is validity improved in experiments
using a control group so researcher can asses if effect on DV was due to IV, standarise procedure to minimise impact of participant reactivity and investigator effects on validity, use single-blind or double-blind trial to reduce investiagtor effect and effects of demand characteristics,
how is validity improved in questionaires
incorporate a lie scale within to assess the consistency of a response and control for effects of social desirability bias, and validity is higher if participants assured their data is anonymous
how is validity improved in observations
observer remains undetected, covert observations, meaning behaviour is authentic and natural, narrow behavioural categories which are unambiguous was improve validity too
how is validity improved in qualitative research
qualitative research produces interpretive validity of researchers conclusions, the extent to which a researchers interpretation of events match the participants. Coherance of research narrative and direct quotes from participant can improve validity and triangulation (use of a number of different sources as evidence, eg, data complied through interviews with family/friends, diaries, observations, etc)
what is statistical testing used for
to determine wether a difference/association/correlation is statistically significant which has implications for accepting or rejecting the null hypothesis
what 3 factors must be considered when deciding what statisitcal test to use
whether the researcher is looking for a difference or correlation, in case of difference what experimental design is being used, level of measurement
Why is it important to first work out if its a difference or correlation when choosing a statistical test
it relates to the aims of the investigation, it should be obvious from the wording of the hypothesis which one they are looking for. Correlation involves tests that look for associations
Why is it important to 2nd work out the experimental design when choosing a statistical test
repeated measures and matched pairs are refered to as related designs as they either use same participants or partiicpants who are matched on some varobale related to the study. As participants in each condition of independant groups are differenet the design is unrelated. So here the researcher must decide if it is related or unrelated experimental design
When is chi-squared used
nominal data, with unrelated design, or nominal datas with test of association/correlation
when is Mann-whitney usede
with ordinal data and unrelated design
when is unrelated t-test used
with interval data and unreleated design
when is sign test used
nominal data and related design
when is wilcoxon used
ordinal data with related design
when is related t-est used
interval data and related design
when is spearmans rho used
ordinal data and test or association/correlation
when is pearson’s r used
interval data with test of association/correlation
what can quantative data be divided into
different levels of measurement, the 3rd factor infulencing statistical tests
what are the 3 levels of measurement
interval, ordinal and nominal
what is nominal data
data represented in categories, It is descrete data (1 item can only appear in 1 category)
what is nominal data
data represented in categories, It is descrete data (1 item can only appear in 1 category)
what is ordinal data
data is ordered in some way, it doesn’t have equal interval between each unit like interval data, it lacks presision as based on subjective opinion. Ordinal data due to its low reliablity is not used as part of statistical testing, instead raw scores converted to ranks and ranks are used for calculations
what is interval data
based on numeric scales including units of equality and precisly defined size. Interval data is better than ordinal and ordinal is better than nominal so it is the most precise data in psychology, it is a necessary critera for parametric tests
what is sign test
statisitcal test for difference in scores between related related items, data should be nominal level or better
what is spearmans rho
a tets for correlation when data is atleast ordinal level
what is Pearsons r
a parametric test for a correlation when data is at interval level
what is wilcoxon
a test for a difference between 2 sets of scores when data is ordinal level and using related design
what is Mann-Whitney
test for difference between 2 sets of scores, data should be ordinal or above or using unreleated design
what is related t-test
a parametric test for difference between 2 sets of scores, data must be interval level with related design
what is unrelated t-test
parametric test for difference btween 2 sets of scores, data must be interval with unrelated design
what is chi-squared
test for an association between 2 variables/conditions, data should be nominal level using unrelated design
is a null hypothesis dirctional or non-directional
can be both
What is a null hypothesis
states there is no difference between the conditions
what do statistical test do
determine if null hypothesis or alternative hypothesis is true, if we accept or reject null hypothesis
what is alternative hypothesis
the opposite of null hypothesis, what we accept if we reject the null hypothesis
what are statistical tests based on
probability not certainty
what is a level of significance in statistical testing
point at which researcher can claim they found a large enough difference or correlation within data to claim an effect has been found, point at which researcher rejects null hypothesis
in psychology what is the level of significance
0.05, 5%
what does 5% level of significance mean
mean that probability that observed effect occured when there is no effect in the population is equal to or less than 5%, so even when research claims significant diffefrence still 5% change difference it isn’t true
can psychologists ever be 100% certain with statistical testing
no there is always a chance it isn’t accurate, based on probability
what is tthe calculated value
the resulting number from a statistical test
how is the calculated value found to be significant or insignificant
it is compared to a critical value, this number says whether to accept or reject null hypothesis, for some tests number must be < or equal to for others it must be > or equal to
what are the 3 criteria to know which critical values to use
ore or two tailed, number of participants in study, level of significant
what is 1 or 2 tailed in determining which critical value to use
use 1 tailed if directional hypothesis and 2 tailed for non-dirctional hypothesis, probability levels double when 2 tailed hypothesis is used as more conservative prediction
what is number of participants in study in determining which critical value to use
appears as N value or for some degreees of freedom calculated instead
what is level of significance in determining which critical value to use
0.05 is stanard level of signifcance in psychology so that is used for all statistical tests
when is 0.01 level of significnce used
when their is human cost, like in a drug trial or in studies that can’t be repeated in the future because the lower the P value is the more statistically significant the results are
what is a type 1 error
when null hypothesis is rejected and alternative hypothesis accepted when it should be other way round as infact null hypothesis is true, reffered to as false positive when researcher claims to have found a statistical difference when there is none
as psychologists cant be 100% certain with statistical tests what are the 2 errors that can be made
type 1 error and type 2 error
what is a type 2 error
null hypothesis is accepted when alternative hypothesis should have been accepted, reffered to as false negative
when are were likely to make type 1 error
if significant level is to high, 10% instead of 5%
when are were likely to make type 2 error
when significance level it too strict 1% instead of 5%
why do psychologists have 5% level of significance
it is balance between type 1 and type 2 error
What data is needed for Masnn-Whitney
independant groups, so unrelated design, levelm of measurement is ordinal and data is converted into ranks to test for a difference
what is step 1 of Mann-whitney
rank the ratings. Need to consider data from both data sets at the same time. Lowest number has rank 1. If 2 numbers have same rank add them up find the mean and give them both this number.
what is step 2 of Mann-whitney
work out value of U. Add up ranks of each data set separatly (R1) (R2), N=number of participants. add these into given formula to get U, calculate value
what is step 3 of Mann-whitney
calculated and critical values. Need to know if hypothesis is 1 or 2 tailed to find critical value. If calculate value is lower than critical value we say results are significant and reject null hypothesis and accept alternative hypothesis.
what is step 1 wilcoxon
cal;culate difference and rank difference, ranking done on difference between 2 data sets and signs ignored, if 0, not included in ranking and removed from N value
what is step 2 wilcoxon
working out value of T, T= sum of the ranks of the less frequent sign (either + or -)
what is step 3 wilcoxonq
calculated and critical values, if calculated value is lower than critical value it is significant
what is step 1 unrelated t-test
table of data, Calculate sum of data set A and data set B separately, square each value in data set and and B and calculate sum of all squared values, separatly for A and B
what is step 2 unrelated t-test
working out t, plot numbers into given formula
what is step 3 unrelated t-test
calculated and critical values, calculated value must be higher than critical value for results to be significant
what is step 1 related t-test
calculate difference between scores of condition A and B and sqaure each difference, add up values of difference and add up values of difference squared separatly
what is step 2 related t-test
working out t, plot all values into formula given
what is step 3 related t-test
calculated value must be higher than critical value to be significant
what is step 1 spearman’s rho
table of ranks, rank each set of scores separately in each conditon from lowest to highest, if 2 or more scores same add them and find there mean
what is step 2 spearman’s rho
calculate the difference, find difference between each pair of ranks and square the difference and add up squared difference, then plot these into formula to calculate rho.
what is step 3 spearman’s rho
calculated value must be higher than critical value to be significant
what is step 1 pearson’s r
table of data, calculate sum of scores for X and Y, square each X value and Y value, multiply X and Y for each participant and add XY all together
what is step 2 pearson’s r
work out r value, plot numbers into given formula
what is step 3 pearson’s r
calculated value must be higher than critical value to be significant
what is step 1 chi-squared
2 x 2 contigency table, draw 2x2 table showing observed frequencies
what is step 2 chi-squared
expected frequencies calculated (frequency we wld expect if no difference between groups). Calculate by multiplying total of row by total of column divided by grand total
what is step 3 chi-squared
put data info formula to calculate x*2
what is step 4 chi-squared
if calculated value is higher than critical value results are significant
what are the 6 sections to a scientific report
abstract, introduction, method, results, discussion, referencing
what is an abstract
short summary (200 words) that includes all major elements: aims/hypotheses, methods, results and conclusions. Psychologists often read abstracts to decide if those studies are worthy of further examination
what is an introduction
a literature of general area of research detailing relevant theories, concepts and studies that relate to currant study, research review should follow logical progression, beginning broadly and gradually becoming more specific until aims and hypotheses are presented
what is a mehtod
Method should include enough detail so other researchers can replicate it, the 5 elements are design, sample, apparatus, procedure, ethics
what is design in writing a method
design is clearly stated and justification given, e.g. independent group, naturalistic, repeated measures, covert
what is sample in writing a method
info related to poeple involved in study, sampling method and target population
what is apparatus in writing a method
detail of any assessment instruments used and other relevant materials
what is procedure in writing a method
recipie style list of everything that happened in the investigation form beginning to end including verbal records of what was said to participants (briefing, standardised instructions, debriefing)
what is ethics in writing a method
an explanation of how ethical issues addressed in study
what are resuklts in writing a report
summarise key findings from investigation, likely to include features of descriptive statistics like tables, graphs and charts, measure of central tendency and measures of dispertion. Inferential stats include reference to statistical tests + which hypothesis accepted. Any raw data collected. If any qualitative research produced results will involve analysis of themes/categories
what are discussion in writing a report
research summarises results in verbal not statistical form, should be discussed in context of evidence presented in intro and other info that may be considered relevant. Researchers should discuss limitations of investigation and how they can be addressed in future studies. Wider implications are considered, e.g. real world application and how findings added to existing knowledge
what is referencing in writing a report
includes full details of any source material cited in the report
what is format for referencing journals
author, date, artical title, journal name (italics), volume (issue), page numbers
what is format for referencing books
author, date, title of book(italics), place of publication, publisher
how can be distinguish scientifc and non-scientifc disciplines according to Kuhn
paradigm (set of shared assumptions and agreed methods in a scientifc disciploine), when many people share sam eviews such as in process of evolution in natural scinece it can be classed as scientific, but in psychology many dissagreement so no paradigm and cant be classed as scientific
what is paradigm shift
when group of scientists question an accepted paradigm and eventually so much evidence against this paradigm that a shift occurs to a new paradigm
what is a theory
a set of genersl laws or principles that have ability to explain certain events/behaviours
how does theory consturctions occur
through gathering evidence via direct observations (empirical method)
what are essential compenents of theory
scientifically tested, should suggest number of possible hypothesis and these hypotheses can then be tested using systematic and objective methods to see if we accept or reject it. If accepted provides extra support to theory, if not theory may need ediiting
what is deduction
process of deriving new hypotheses from existing theoru
what is falsifiability
principle that a theory cannot be considered scientific unless it admits possibility of being proven untrue, the strongest theories are those which have yet to be falsified after many attemots as despite researchers best efforts they have not been proved as untrue, this is why alterntive hypothesis always accompanied w null hypothesis, this is a feature of Poppers hypothetico-deductive method
what is replicability
extent to whioch scientific procedures can be repeated by other researchers in different contexts and circumstances, replication important to determine internal and external validity and reliability
why is objectivity important in science
investigator must maintain critical distance so their personal biases don’t discolour the data collected, the studies ith more control are usually lab studies and are most objective
what is objectivity the basis of
the emprical method
what is the empirical method
data collection based on direct experience and observation, experimental method and obervational method are 2 good examples of empirical method, a theory cannot claim to be scientific if it has not been empirically tested and verified
