Use LEFT and RIGHT arrow keys to navigate between flashcards;
Use UP and DOWN arrow keys to flip the card;
H to show hint;
A reads text to speech;
81 Cards in this Set
- Front
- Back
Independent Design |
comparison of 2 score sets from different participants |
|
Related Design |
comparison of 2 sets of scores from the same participants |
|
Nominal Data |
collected in categories; scores cannot be put in order on a scale e.g. conformed or not1 |
|
Ordinal Data |
scales that are made up by psychologists - subjective
e.g. ratings of aggression/attractiveness; self-report |
|
Interval Data |
where intervals between points on a scale are identical e.g. someone who took 5 seconds took half the time of someone who took 10 seconds; someone who took 4 seconds took double that of someone who took 2 seconds |
|
Ratio Data |
interval data but with an absolute zero i.e. no negative scores e.g. height or number of words recalled |
|
Chi-squared |
- tests of difference - independent design - nominal data |
|
Mann-Whitney U-test
|
- tests of difference - independent design - ordinal/interval/ratio data |
|
Wilcoxon Matched Pairs Signed Ranks Test |
- tests of difference - related design - ordinal/interval/ratio data |
|
T-test for Independent Samples |
- tests of difference - independent design - interval/ratio data |
|
Spearmans Rank Order Correlation Coefficient |
- tests of correlation - ordinal/interval/ratio design |
|
Pearson's Product Moment Correlation Coefficient |
- tests of correlation - interval/ratio data |
|
Significant Result |
- difference = significant if it is unlikely to be due to chance - can reject the null hypothesis & accept the alternate/experimental hypothesis - if a finding is significant (at p≤0.05) the probability of achieving a difference (or correlation) as strong as that found just by chance is less than 5% |
|
Hypotheses |
null alternative/experimental: - 1-tailed - 2-tailed |
|
Null Hypothesis |
"there will be no correlation/difference between..." |
|
1-tailed Hypothesis |
- directional used when: - a theory predicts the direction of a difference/correlation - when previous research found a difference in a particular direction "there will be a positive/negative correlation between..." |
|
2-tailed Hypothesis |
- non-directional used when: - different theories make different predictions - previous research is contradictory - there is no previous research "there will be a correlation/difference between..." |
|
Type 1 error |
- false positive - claiming a difference is significant when it isn't - more likely with a lenient p. value - w/ p≤0.05, probability of a type 1 error is 5% |
|
Type 2 error |
- false negative - claiming a difference isn't significant when it is - more likely w/ a stringent value |
|
Using the tests |
- calculated/observed value has to "beat" the critical value - critical value = dependent on the level of significance + whether the hypothesis is 1 or 2 tailed + either N or degrees of freedom |
|
Spearman's
|
- calculated value = correlation coefficient - calculated value has to be equal to or more than |
|
Wilcoxon's Mann-Whitney |
calculated value has to be equal to or less than |
|
Chi-squared |
- degrees of freedom = given you know the column & row totals in the table of data, how many of the actual frequencies do you need to know to work out the rest - (rows-1) - (columns-1) = degrees of freedom - calculated value has to be equal to or higher than the critical value - always use the value for a 2-tailed test even when a 1-tailed was used |
|
Ranking data |
- if ordinal data is used, then a subjective scale has been used - the best we can do is put the data in rank order - as we cannot say that one score is double that of another (e.g. GAF scale, do not know that a score of 45 is half that of 90) |
|
Quantitative data |
in the form of numbers |
|
Qualitative data |
non-numerical form e.g. a description of a clinical case, a transcript, a diary advantage: greater validity than quantitative data; truer to life disadvantages: qualitative data is subjective and ∴ more open to bias; previous opinions may make researchers bias in their observations - researchers may report / "cherry-pick" data which supports their views; a different researcher could do the same observation & arrive at a completely different conclusion |
|
Methods of qualitative data collection |
- interviews, esp. semi and un-structured interviews involving open questions - observations, esp. unstructured & participant observations |
|
Options in qualitative data analysis |
- convert qualitative data into quantitative data, e.g. content analysis - extraction of info w/out conversion into quantitative data |
|
2 methods of analysing qualitative data |
Aim: provide a systematic way of analysing findings so the analysis of qualitative data is not subjective/open to bias content analysis: actual conversion of the qualitative data into categories; may involve scoring, e.g. the amount of stress in different jobs from diaries; this converts qualitative data into quantitative data thematic analysis: aims to identify patterns/themes within the qualitative; used to analyse data from bodies of text, e.g. interviews, newspaper articles; this identifies categories but doesn't count instances within categories |
|
Thematic Analysis |
1) transcribe and read 2) divide into meaning units 3) search text & highlight themes 4) adjust as sorting continues 5) define & name themes when completed 6) write report, presenting & supporting themes |
|
Theoretical/Inductive Analysis |
theoretical: theory outlines the analysis inductive: theory only emerges from the data after analysis |
|
Reliability |
consistency/sameness of a measure, method or researcher |
|
Validity |
truth/accuracy of a measure or method |
|
Test re-test reliability |
similarity of 2 sets of scores taken on different occasions |
|
Split-half reliability |
similarity of 2 sets of scores from different halves of a questionnaire |
|
Inter-rater/observer reliability |
similarity of ratings made by different observers |
|
Intra-rater/observer reliability |
similarity of ratings make by the same observer |
|
Improving Reliability |
pilot study: - materials should be tested to ensure they yield reliable results from pps; aim to ensure questions are not affected by irrelevant factors & to eliminate items in questions that do not correlate w/ others - researchers should be trained to ensure they are reliable over time & w/ each other to ensure inter/intra-rater reliability; should eliminate inconsistencies in ratings & observations |
|
Internal Validity |
- operationalisation - controls - experimental validity (realism) |
|
External Validity |
- temporal validity - ecological validity - population validity |
|
Types of Validity of Measurement |
- face validity: extent to which a way of measuring something looks valid - concurrent validity: extent to which a score is similar to a score on another test that is known to be valid - predictive validity: extent to which a score predicts future behaviour |
|
Threats to Validity |
- demand characteristics: pps who know they're in a study may guess what the study is about & change their behaviour - social desirability effects: pps may behave/respond in socially acceptable ways, either to look good to the researcher or themselves - order effects: what pps experience earlier in the study may affect their behaviour/responses later in the study - hawthorne effect: pps who know they are in a study may try harder than they would in everyday life |
|
Improving Internal Validiy |
- ensuring extraneous variables are controlled; aim to remove all differences from the 2 conditions apart from the independent variable - enabling the testing of cause & effect - redesign studies to exclude other possible explanations of the results - studies should also have experimental validity - best way to improve validity is in advance; conducting a pilot study can identify & eliminate extraneous variables |
|
Improving External Validity |
replicate the original study: - using a very different set of pps from the original study, thus testing whether the original has population validity; particularly useful if a cross-cultural study is conducted - using a different experimental method, i.e. a more realistic setting using a field experiment; this would test the ecological validity of the original study - many years after the original; this would test the temporal validity |
|
Sampling |
Population: those who the study is about/the study applies to Representative: it's representative if those who are studied are similar to the population, so we only need to study a sample instead of the population Generalisation: claim that what is true of the sample is also true of the population |
|
3 Sampling Methods |
Random: use of sampling population/frame; all members of the population have an equal chance of being in the sample Opportunity: using whoever is available Volunteer: publicise research in an appropriate location where the population will be able to see it |
|
Bias |
- reduces generalisability
sources of bias: - deliberate & unwitting bias by the researcher in selecting pps: - use of unrepresentative sampling populations from which the samples are taken - random error, when taking samples, more likely w/ small samples |
|
Deliberate or unwitting bias |
- when selecting pps
- researcher choose pps who they believe will give them results which support his hypothesis |
|
Unrepresentative sampling population |
- from which samples are taken
- research on WEIRD (Western, Educated, Industrialised, Rich, Democratic) pps may not generalise to other cultures |
|
Random Error |
- possible to get unrepresentative samples just by chance
- more likely to occur if a small sample is being chosen - can be addresses through the use of a stratifies sampling method |
|
Stratified Sampling
|
- divide the sampling frame into groups to be represented in the sample - sample randomly from within these groups ✓ eliminates the possibility of v. unrepresentative samples which can be produced by random sampling ✗ expensive to run |
|
Quota Sampling |
- used when a sampling frame is unavailable - draw up quotas to be represented, recruit volunteers until quotas are full ✓ improves likelihood the sample is representative by ensuring proportions of each strata are representative ✗ issues associated w/ using volunteers - e.g. in attachment, few bad parents will volunteer → bias sample; if pps who do not agree to take part (after being asked to give informed consent) are unrepresentative so too will be the sample |
|
Ethical Issues |
- informed consent: pps should be made aware of anything that might affect their willingness to participate; children under 16 cannot give their own consent - must be gained from a parent on their behalf; school setting - children have to give their consent as well as the head teacher
- right to withdraw: should be clear at the onset of the study, i.e. in the consent form; pps also have their right to withdraw data after the study |
|
Contents of consent forms |
- information sheet about the study - how ethical issues will be dealt with - declaration of consent |
|
Debriefing |
- should provide pps with any necessary info to complete their understanding of the nature of the research & monitor any unforeseen negative effects or misconceptions - pps should leave the study in the same state as they entered in contents: - preliminaries - thanking pps - procedure - clarifying aim of study - ethics - dealing w/ remaining issues |
|
Writing a consent form |
- main aim = to gain informed consent information sheet about the study: - statement of purpose of study, as much as this needs to be revealed - what pps will be asked to do; procedures, as much as this needs to be revealed how ethical issues will be dealt w/: - how anonymity & confidentiality will be ensured - assurance about right to withdraw declaration of consent: - form for pps to tick & sign |
|
Writing a debriefing form |
preliminaries: - thank pps for taking part procedure: - clarify the aim of the study, inc. anything they need to know to complete their understanding of the study, i.e. anything omitted from the consent form ethics: - ask if they have any questions - identify any unforeseen discomfort, distress or other negative effects of the study - remind pps of their right to withdraw |
|
Structure of a psychological report |
- title - abstract - introduction - method/procedure - results - discussion |
|
Title |
- should give a clear idea of what the research is about - naming variables - naming theory being tested |
|
Abstract |
- summary of the key points of the research - list of key terms is oft provided separately - allow researchers to quickly decide if the research is of interest to them - also allows computerised databases to be searched quickly & efficiently for relevant research |
|
Introduction |
- introduction to background of the study - covers key: theoretical background; relevant studies - covers issues & debates about the topic - "funnel" technique - general material → more specific research/theories - leads logically towards aims & hypotheses of the study at the end of the intro |
|
Method/Procedure |
- clear comprehensive account of the method/procedure allows replication of study; finding can thus be checked; allows others o evaluate procedures - check reliability & validity of procedure sub-sections: - method & technique: lab exp, questionnaire... - design: repeated measures, longitudinal... - variables: IV & DV/co-variables - pps & sampling method: number of; characteristics; sampling method - materials: description of measuring "tools", e.g. interview schedule, questionnaire w/ full original in appendix - procedure: full description of precise procedures used, inc. standardised instructions, how techniques were implemented - controls: counterbalancing, matching - ethical issues: any possible issues & how they are dealt w/ |
|
Results |
- descriptive stats: inc. measures of central tendency & dispersion, graphs & tables, written account of findings - inferential stats: details of stat test: inc. choice of test, level of significance, 1 or 2-tailed, critical value, calculated value, relationship of results to hypothesis - identify the test to be used before to ensure there is an appropriate test for the data to be collected |
|
Discussion |
- explanation of results: how they relate to theory in intro - discussion of strengths & weaknesses: validity; reliability; generalisability - discussion of possible improvements & future research - possible practical implications: usefulness in real-life situs |
|
References |
- alphabetical list of sources used - allows others to check your accounts of research/theory are fair |
|
Appendicies |
includes: - consent form - lengthy instruction sheets - original materials - raw data (anonymised) - debriefing form would break up the flow of the text if used in the body but need to be there for replication, checking of calculations, assessment of ethical issues etc |
|
Major Features of Science |
- objectivity - replicability - theory construction - hypothesis testing - use of empirical methods |
|
Objectivity |
- impartial - could ideally be accepted by any subject bc it doesn't draw on any assumptions, prejudices or values - encourages investigators to proceed objectively, putting aside personal biases & prejudice |
|
Main threat to objectivity = bias |
- people are prone to a variety of cognitive biases, they interpret events differently & in psychological measurement there is oft room for interpretation, e.g. how to score or categorise behaviour - room for interpretation = room for biases to come into play - most people interpret events in biased ways - seeing things that aren't there, or basing ideas on prior assumptions 2 common biases: - seeing correlations between variables that aren't necessarily there - seeing cause & effect relats that aren't necessarily there |
|
Replicability |
replicability of a procedure: - a study can be repeated in the same way - others should be able to repeat the research; enabling them to check if results were a fluke, due to sampling bias replicability of results: - if the study is repeated, results will be the same - if they are not replicable the empirical claim of the research is questionable & any support for the theory being tested is undermined - a field experiment is likely to be lower in replicability than a lab experiment - an unstructured interview is likely to be fairly low in replicability |
|
Replication/Triangulation |
replication = replicability of the results of a specific study triangulation = replicability of the effect found in a study; evidence from different types of studies for the same effect |
|
Theory Construction |
- theory = explanation of why things happen the way they do - in Psych, trying to explain the human mind & behaviour; construct theories to explain why human beings behave the way they do |
|
Hypothesis Testing |
- theories → hypotheses (claims that can be tested through research) - hypothesis = a testable prediction - decent theory → a precise, testable hypothesis - ideally a theory → a hypothesis that, if supported, can only be explained by that theory |
|
Use of empirical methods |
- fundamental feature of science - observe & measure phenomenon using objective, replicable, systematic techniques for collecting data - stat tests, experimental methods - rejection of non-empirical methods |
|
Validating New Knowledge |
replication triangulation cross-cultural research reviews meta-analyses |
|
Reviews |
- summarise results from research on a topic - usually take the form of a narrative/extended account of the trends in research - allow researchers to identify overall trends in findings 2 issues: - often invalidated by cherry-picking, i.e. only including studies that support the author's views - statistical analysis is/was oft not v. systematic or rigorous; positive/negative/neutral used - ignored size of differences & sample sizes |
|
Meta-analyses |
- more sophisticated reviews ✓ deal w/ 2 problems of reviews: - less likely to cherry-pick: methodological criteria used to decide if a study should be included - results from studies are combined in a systematic way: results from studies are oft merged to give an overall score aka the effect size - a weighted average wherein larger studies count for more than small ones ✗ file drawer problem: if researchers do not find a significant result, they may not bother to submit it for publication or be less likely to get it published if they do submit it could → a misleading pattern of results |
|
Peer Review |
- designed to ensure good quality research, increasing the knowledge base of psych, is published - to be accepted as credible, research must be published in an academic journal for which there is a peer review process |
|
Stage 1 - Peer Review Process (by experts from the journal) |
- research is submitted to a journal in a standard format - then assessed by staff from the journal - editor, editorial board & external reviewer - experts in their field & competent to judge the merits of the paper - ideally, blind reviewing process used - reviewer = unaware of the identity of the researcher; methodologically reviews the study first ✓ creates set of standards for research; researchers know how to conduct research to get it published in an acceptable place ✓ ensures high methodological quality ✓ filters out poor quality research |
|
Stage 2 - Peer Review Process (by wider academic community) |
- those who conduct research in the same field respond to the publication - criticise the study, cite it in support of their own theories, replicate or adapt the study - act as a large set of peer reviewers - best research becomes part of the accepted knowledge base of Psych ✓ enables development of a body of knowledge on psychological topics ✓ ensures progress is made in the development of knowledge ✓ this knowledge base is then accessible to all |
|
Peer Review Limitations |
- doesn't solve file drawer problem - studies may not be submitted for peer review if they find no effect publication bias: - certain types of finding are less likely to be published: replications of previous research; findings that contradict the theoretical viewpoint of the reviewer or journal - some argue publication process is slow - Internet allows possibility of open publishing - no stage 1 but faster & more effective stage 2 |