A2 Psychology, Research Methods, Unit 4 – Flashcards
Unlock all answers in this set
Unlock answersquestion
-'evidence' rather than 'belief' -obtained by using 're scientific method' -make observation, develop explanation, tested -e.g lab, correlational analysis, obs -all scientific methods aim demonstrate replicability & objectivity
answer
Science definition
question
- Make observation - develop explanation (or theory) of obs -test explanation empirically i.e using scientific method -do findings fit explanation (theory)? -develop explanation (or theory) of your obs
answer
Process of scientific method diagram
question
-methods that provide evidence based on direct experience rather than reasoned by argument - seek report on world as it really is as opposed to unfounded beliefs -refers to research methods & procedures used to analyse and report this data
answer
Empirical methods
question
-distinguish between real truth / unfounded belief -variables measured in psy can be more diff to directly experience than in other sciences -f/e ^ diff measure happiness/intelligence than physical objects/ temperature -make unfounded beliefs ^ likely -more important research is empirical to combat this -methods must also be well controlled & objective to ensure measurement of variables is unbiased & accurate
answer
why empiricism important feature in science?
question
-removal bias/ human influence -research collected way thats independent of researcher -not affected by individuals views/ opinions, nothing left open to interpretation -opposite subjectivity (biased based personal opinion)
answer
objectivity
question
-only thru removing possible bias from research (e.g inv effects, D/C) can be sure results valid -builds confidence that findings represent real effect (rather than views/ interpretations) -pursuit true objectivity in research encourages R question accuracy findings/ findings others in field, helps ensure methods used in study were well controlled, whether findings have internal validity -pursuit objectivity allow researchers identify sci fraud, prevent fraud in psy
answer
why objectivity important feature science?
question
1. when planning & conducting study -clearly defining variables (operationalisation) removes researcher bias -use large sample -use methods which didn't involve inv effects -standardised instructions 2. when recording results -during investigation have more than one researcher taking down results -triangulation, f/e reaction times and brain activity 3. when writing up research report - someone else writeup report, you only write whats visible and note down all facts, no opinions
answer
ways investigators can increase objectivity
question
-ability study be repeated in same way over over again, if rep in full allows R see if sim findings can be achieved -rep important for number reasons including allowing r check wether findings reliable
answer
replicability
question
enables psy independently check whether results reliable (consistent) helps build confidence that particular finding trustworthy represents real effect rather than one-off - allows R independently check whether methods used in study well controlled, whether findings have high internal validity -extreme cases identify sci fraud, help prevent it -if study carried out in spec circumstances (e.g very specific sample) rep allow study be conducted more widely (e.g diff sample) to see if findings can be generalised more widely improving external validity
answer
why replicability important feature science
question
-what they did (aims, fully operationalised hypotheses & methods) -how carried it out (step by step procedure, e.g questions used in interview, how asked) -number P used, how selected (sampling) how allocated conditions (exp design) -where study carried out (location) -what raw data collected, how analysed
answer
in order to make research easy to replicate psychologists must be clear about...
question
-explanation of an observed phenomenon, aim constructing an effective theory is to be able to predict how people are most likely to respond in certain sit -then sci tested to gather empirical evidence -f/e useful theory of human memory should be able to predict and explain hat will be remembered/ won't in given circumstance
answer
theory construction
question
-process by which specific predictions generated from a theory can be tested -another important feature sci where validity of theory can be tested -theories modified thru hypotheses testing -if psychologist fails find support hyp, then theory requires modification
answer
hypothesis testing
question
- negative version o the research hyp -rather than stating will be an effect as hyp does, predicts be no effect -f/e if exp hyp predicts be sig diff between 2 groups on a test, null hyp predict no sig diff ''there will be no sig diff between...'' ''there will not be a sig correlation between...''
answer
null hypothesis
question
-one adv of applying scientific method in psychology is being seen as 'sci' helps to build trust -by trying to produce research that's replicable, objective, empirical, more likely view the research as valid and trustworthy, suggests psychologists should aim to apply sci method if want research be taken seriously -not possible be truly objective in psy, unlike psychical objects that might be studied in chemistry f/e, the injects of psy have thoughts/ feelings, make them react to researcher, P beh might be affected by realising they are in study or by the researcher (...And ... Effects respectively) problem, may mean it's not possible for psy to be objective and 'sci' in same way as other 'hard' sciences
answer
Evaluation of applying the 'scientific method' in psychology
question
Title Abstract Introduction Method (design, participants, apparatus/ materials, procedure) Results Discussion Conc References Appendices (Tips, written in 3rd person tense ('study aimed to...' Rather than 'we aimed to...') remember written for someone unfamiliar with the investigation but they should be able to replicate it from the description given)
answer
Reporting on psy investigations
question
Should provide clear indication of focus on study, not too general/ long/ complicated
answer
Reporting on psy investigations Title
question
-First thing to appear in report, last thing written -concise overview of entire investigation, summarises aim, methods, results, conc, about 150-200 words -allows reader decide whether or not red of report likely to be of interest
answer
Reporting on psy investigations Abstract
question
-review existing research literature in the area of interest, provide clear reasons for the work to be undertaken -should be able see logical flow ideas from general research topic thru to the research aim, find yourself persuaded about the reasons for carrying out the study -essential to be clear about aim of study, may or may not find formally writers hypothesis at end, null hypothesis usually only mentioned in relation to statistical tests
answer
Reporting on psy investigations Introduction
question
Precise details bout how investigation carried out, should be sufficient detail for somebody else to rep study, divided into subsections: -design, including, design, method, variables, controls and ethics - participants, include number, sampling, recruitment procedures -apparatus/ materials, anything used carry out research (brief, debrief, tests, QAs, standardised instructions) -procedure, describes exactly how procedure carried out, must be clear so someone else could rep it in full
answer
Reporting on psy investigations Method
question
Descriptive and inferential statistics presented in summary tables, graphical representations and figures with tiles and brief explanations
answer
Reporting on psy investigations Results
question
-used to review the findings, compare them to past research -assess quality of study (methodology, internal/ external validity) suggest what future research could be done -implications of what research suggests about area studied can be outlined, any practical apps of research can be suggested here too (more effective way delivering a therapy)
answer
Reporting on psy investigations Discussion
question
Brief summing up of main findings/ suggestings
answer
Reporting on psy investigations Conc
question
All sources that researcher used to inform the report (e.g where prev research mentioned in the introduction can be found)
answer
Reporting on psy investigations References
question
Raw data, statistical calculations, copy of QA, standardised instructions, brief, debrief etc
answer
Reporting on psy investigations Appendices
question
-1st thing appear, last thing written, gives overview of entire study, including findings and concs
answer
Writing an abstract section
question
-this review of existing research literature should follow a clear structure -should be able see logical flow of ideas from General research h topic (overview of area with any definitions) through to the research aim and find yourself persuaded about the reasons for carrying out the study and aim
answer
Writing an introduction section
question
If Q asks you write method section, useful to divide your method section into subheadings for different sections to ensure its written properly and all factors are covered (design, participants, apparatus/ materials, procedure)
answer
Writing a method section
question
1. Always be very clear on precisely what findings show -give fully operationalised names to conditions/ measurements used ( aggression score, don't just say aggression give full and precise name) quote precise names, give numbers (rather than just one higher than other) if include graphs/ tables make sure include clear and precise title and label the table or graph clearly and precisely (no interpretation of the results in this section at all, just state results
answer
Writing a results section 1
question
-when explaining what results show, give full information (including all numbers) -dont be brief -e.g 'as the observed value is less than the critical value, this means the null hypothesis can be accepted and the hyp rejected' -do quote details led you to choosing critical value (e.g level sig, whether 1/2 tailed) must remember refer null hyp -e.g 'the critical value for chi-squqared (df=1, p<0.05 (two-taield) is 3.93. As the observed value of 2.16 is less than the critical value, this means the null hypothesis can be accepted and the hypothesis rejected.'
answer
Writing a results section 2
question
-summary of results: the R in brief, some exp what show -rel to prev research: R discussed in del research reported in introduction & possibly other R not prev mentioned -consideration of methodology: criticisms may be made of methods used, improvements suggested -implications for psychological theory: the concs that can be drawn from research discussed, including what R suggests about topic being studied -suggestions for further R: based on experience & findings of conducted R described in report, researcher can make recommendations about ways in which R could be conducted in future & related topic areas that might need further investigation
answer
writing a discussion section
question
-way of assessing quality of research report whereby fellow researchers (or 'peers') comment on, criticise & veal the work -done before report published as potential articles are vetted by editorial board of experts in field -process also continues after publication as journals invite comments from readers & replies to comments from original author published in later journal issues -p.r often unpaid & anonymous to encourage objectivity & honesty
answer
peer review
question
-ensures new knowledge is subject to rigorous checking, make sure well designed, controlled, valid & meets an agreed 'sci standard' -helps prevent fraudulent research being viewed as gd evidence -by opening up work for constructive criticism for others, encourages sharing of ideas between experts, helps researchers improve each others work -alternative views may offer insight into aspects that researcher has overlooked, help move field forward
answer
why is peer review important in sci & how does it 'validate' new knowledge?
question
-strength, helps stop flawed/ fraudulent research being published, +ve as helps ensure validity of published research & encourages honesty from all researchers -prob, slow publication down, forcing articles go thru process p.r before publication may prevent important discoveries being put to use quickly, anonymous natures of review might allow some researchers use it as opportunity prevent competing researchers from publishing work -prob, growing use of internet means lot of research & academic comment is being published without official peer reviews, though systems are evolving online where everyone has chance offer opinions and police research quality
answer
evaluation of peer review process Ao2
question
-lab experiment -field experiment -natural experiment -correlation analysis -observation -interviews -questionnaires -case studies (IV distinguishes experiment from other non-experiemtnal research methods)
answer
designing your own sci research 8 methods
question
lab -when want establish cause & effect del -when want to be able to easily replicate -when reactions need to be timed -expensive equipment field -in natural env -want to control IV -reduce demand characteristics natural -important to be in real life sit -reduce demand characteristics -less investigator effects
answer
when would you choose to use a lab, field or natural experiment?
question
independent -to compare two groups based on DV repeated measures -to compare a group, P variables need to be avoided matched pairs -to compare two groups matched based on variables to prevent E.Vs
answer
which experimental design?
question
correlation -see the strength and direction of two variables/ a relationship
answer
when would you choose to use correlational studies?
question
-naturalistic -controlled -disclosed -undisclosed -participant -non-particpant
answer
types of observations
question
-observation, natural env, no attempt influence anything -used when observing natural behaviour -e.g observing children play in their first weeks of nursery
answer
naturalistic observation what is it? used when? e.g?
question
-not natural env -used when observing beh in controlled condition, bet influenced -e.g observing how children play in unfamiliar env
answer
controlled observation what is it? used when? e.g?
question
-aware being observed -when it may be (DC) unethical for the people to be unaware they're being observed -victims of abuse, public places
answer
disclosed observation what is it? used when? e.g?
question
-unaware being observed -ruin results if knew were being observed -gangs, violence
answer
undisclosed observation what is it? used when? e.g?
question
-become involved in observation -need to be involved to collect results -gang member, observing private group
answer
participant observation what is it? used when? e.g?
question
-not involved in observation -being involved may affect results -children playing, geordie shore
answer
non-participant observation what is it? used when? e.g?
question
-event sampling, observer uses beh schedule & counts each occurrence of a beh -easy to plan & implement -categories set by researcher may only count times that sup hyp, subjective -> observer bias -to avoid use time sampling (observations taken at fixed time intervals (e.g every 10 seconds) -reduce chance observer bias cos can't choose when to record, so much more random -in a beh schedule, each beh category should be fully operationalised (clearly defined, observable & mutually exclusive so ev category can be readily separated from continuous stream of events) -f.e in gambling observation 'getting excited' 'happy' would not be valid category as 'excited' & 'happiness' itself cannot be clearly measured -observable product of happiness would be ''time spent smiling in seconds' -making each 'beh category' observable & fully operationalised should allow diff observers be consistent (i.e establish inter-rater reliability)
answer
observation sampling procedures & behavioural categories
question
questionnaire when/ want: -collect info on beliefs to understand/ show trend -gather large amount data -sensitive subject e.g abuse, attractiveness -e.g how people view elderly interview when/ want: -find out someones beliefs/ views on a subject -gathering a lot of detail is important -behs like body lang, tone voice useful to study -building trust important -complexed issue, things QA couldn't cover -e.g views on drug laws, child abuse, emotional topics, new areas of research
answer
when would you choose to use interviews or questionnaires?
question
case study -to gain in depth insight to peoples lives -to observe things other experimental methods would miss -very socially sensitive topics -diff gather large sample -less data, more detail -e.g murder to understand all aspects contribute to such beh (numerous research methods can be used within case studies e.g QA/ observation to fan members & since in depth one individual, special care privacy, confidentiality, right withdraw & full informed consent, even more crucial when individual has mental health prob)
answer
when would you use case studies?
question
-group P in study called sample, selected from target pop -should represent wider group so should contain individuals with all important characteristics that appear in pop
answer
sample
question
-whoever willing available take part at time e.g often students :) quick, easy, cost efficient, practical :( often biased, not all the target pop representative at the time e.g university student ^ intelligent than average person
answer
opportunity sampling
question
-people put themselves forward to take part, advertise in newspapers, posters :) practical, quick, easy assemble (advertise & they come to you) :) allows researchers access groups people diff access thru sampling methods (people with rare condition) as they put themselves forward :( often biased, ^ likely be certain kind of person e.g confident, extrovert -> unrepresentative, lack pop validity
answer
volunteer (or self selected) sample
question
-ev person from target pop has equal chance selection, ev name given number, hat, computer generator :) most representative cos ev in pop equal chance selection :( only possible when researcher can access full target pop (e.g database) :( can still be unrepresentative e.g by chance choose all M
answer
random sampling
question
-reliability refers to consistency, 2 types: -internal, consistency of a measure within itself e.g all items on QA testing same thing -external reliability, consistency of a procedure from one occasion to another e.g exp performed on 2 diff days, diff labs, diff researchers should still produce sim results
answer
reliability
question
split-half reliability -P scores on a test, QA, interview, split into 2 halves -f.e questions divided into even numbered and odd numbered questionnaires -P scores on two halves of test should correlate if internally reliable
answer
assessing internal reliability
question
-test-retest reliability -P take same test twice at diff times -if results for two occasions correlate has high external reliability, can be used test may diff methods -inter-rater reliability -2 or more observers watch same beh (e.g on vid) -equipped with same categories & compare results to assess whether they are constant -usually used for observations, but can be used interviews, content analysis
answer
assessing external reliability
question
-tests for reliability (e.g split half, test-retest reliability -allow researcher identify individual unreliable items (e.g particular questions on QA that fail provide consistent results) -> removed/replaced -if results in improvement, stronger correlation be found when internal reliability checked again -if inter rater reliability low, improved by discussing & agreeing on new & improved operationalised definitions or behaviours -thru practice sessions where beh sequences watched, scored, discussed until agreement reached (achieved using pilot study)
answer
improving reliability
question
-face validity -whether measure appears, at face value, to test what claims to -f.e does interview about addiction to alcohol genuinely measure drinking habits, or does it simple elicit socially desirable q's -if includes q's trigger socially des response, likely have low internal validity
answer
assessing internal validity 1
question
-concurrent validity -whether new test procedures a similar measure of a variable as existing, more established tests of the same variable -if new QA that identifies risk factors in drug abuse valid, should find many of the same risk factors as existing, well known, QA of same variable
answer
assessing internal validity 2
question
-predictive validity -whether measure can accurately forecast future consequences -f.e test designed to identify risk factors for alcoholism could be followed up & if validity high, those identified as having higher risk factors will be ^ likely go on exhibit signs alcohol abuse
answer
assessing internal validity 3
question
-replication in real life settings -to test if results from lab can be gen real world -same methodology employed in real life setting -(e.g Milgram rep obedience study in run down office) -if results sim research has high eco validity
answer
assessing external validity 1
question
-replication with diff populations -to test is results from one sample can be generalise to rest of pop -same methodology replicated using diff sample (e.g diff culture) -if results sim research has high pop validity
answer
assessing external validity 2
question
-replication in modern day -to test if results from old study can be generalise to modern day -same methodology employed in modern day -(e.g replicating Milgrams 1960s study today) -if sim results are achieved the research has high temporal validity
answer
assessing external validity 3
question
-to help psychologists deal with eth issues arising in research and professional practise, the British Psychological Society (BPS) regularly updates its eth guidelines -when research conducted at institutions such as universities, the planned study must be approved by ethical committee -often research my impinge on rights of P's may be especially beneficial for soc, in these cases, cost-benefit analysis may be carried out to decide whether costs to P outweigh benefits of soc -another issue of importance to ethical committees is the perception of psychology in doc -P who are deceived/ distressed may not want P again, may portray psy in poor light, unlikely trust findings of psy research
answer
ethical considerations
question
1. introduction: public needs have confidence in psy, so way researchers treat P important as affects public perception of psy 2. general: spy's must always aim be ethical, foreseeable threats to wellbeing, dignity, health, values of P should be eliminated 3. consent: where poss, must obtain full informed consent 4. deception: avoided wherever poss, P told aim of research as soon as poss 5. debriefing: P should be given full explanation of research as soon as poss & check P leave in same state entered 6. withdrawal: P made aware have right withdraw any point & payment not affect this, P also have right withdraw their data 7. confidentiality: unless agreed with P beforehand, P details & results kept completely confidential 8. observation: risk invading privacy, if P unaware being watched, should only be done in public places where would expect be watched 9. advice: if researchers see signs harm/ threat that P unaware of should inform them 10. colleagues: if seen break any of these principles, important tell them & try persuade them to alter their conduct
answer
BPS ethical guidelines 1234567891011
question
-mode -medium -mean
answer
3 measures of central tendency
question
-range -standard deviation (working out how far each value deviates from mean, bigger the SD more spread out)
answer
2 measures of dispersion
question
-bar chart -scattergram -histogram
answer
3 methods of graphical representation
question
-nominal -ordinal -interval -ratio data (NOIR)
answer
4 levels of measurement
question
-separate named categories -such as counting 'frequencies' or people who 'pass' and 'fail' a test
answer
nominal data
question
-rank ordered in some way -intervals between values are not fixed -f.e 'agree, strongly agree' or mark on a test
answer
ordinal data
question
-measured using scale with fixed intervals -no true zero -f.e temperature
answer
interval data
question
-similar to interval data as are fixed intervals -but also there is a true zero -f.e time in seconds, or counting correct answers
answer
ratio data
question
1. nominal (very vague) 2. ordinal 3. ratio/ interval (you can convert 1 level of measurement into another but you can only covert a more precise measurement into a less precise measurement)
answer
preciseness of the levels of measurements
question
-tend used show change over time or trials -time/ number of trails usually plotted on the horizontal axis (x-axis) -measure you are interested in, presented on vertical axis (y-axis)
answer
appropriate selection of graphical representations graphs
question
-give visual impression of how often certain measurements occur -are rules do with how date measured that determine which ones to use and when -frequency is always recorder on the vertical axis -frequency of variable plotted on horizontal axis
answer
appropriate selection of graphical representations frequency diagrams
question
-best for showing frequencies of nominal data -horizontal axis is a discrete scale (i.e uses sep categories with space between each) -also used plot data in form of % and means
answer
bar charts
question
-mainly used present frequency distributions of at least ordinal data -horizontal axis is a continuous scale (i.e with no spaced between bares as the measurement is continuous and not in separate categories e.g time in seconds)
answer
histograms
question
-sim histograms -instead of bars, consists of a continuous line
answer
frequency polygons
question
-used prevent correlational analyses and they use a least ordinal data -each P measured on both co-variables and both of these measurements must be quantitative -scattergrams give initial indication whether theres is a relationship between co-variables
answer
scattergrams
question
-even if results appear gone particular way, does not necessarily mean can be confident in results -need prove that affect/ difference measured is greater than chance e.g significant -to test if results are strong enough to be confident they represent a true effect, do an inferential statistics test (based on idea of probability)
answer
inferential statistics
question
-probability ℗ expressed as number between 0 and 1 -0 means event definitely will not happen -1 def will happen -to calculate probability of a particular outcome: probability = number of particular outcomes divided by the number of possible outcomes -f.e what is p getting a head (1 particular outcome) when u toss coin, result could be head/toil, giving 2 poss outcomes, thus prob of getting head divided by 2 which is 0.5 (50%)
answer
probability
question
-inferential statistics all about assessing p that particular set of results are down to chance (opposed representing real effect) -just cos study finds results did go way hyp predicted, not mean hyp 'true' & represents real effect -a 'significant' result is essentially a result that is strong enough for us to be confident the result represents real effect (if effect stronger than specified level of sig, can be confident enough that the result is not down to chance)
answer
significance
question
-standard level sig used by psy's is p≤0.05 (5%) (standard level used in most studies) -means probability results due chance less than/ equal to 5% (1 in 20 risk results were a fluke) -some studies psy's want be more certain e.g considering effects new drug (want be v careful) or when conducting replication of prev research, use ^ stringent level sig p≤0.01 (1%) - ^ lenient level sig, conducting research new topic (never been tested before so not sure how to control) rather than just reject findings in new area, ^ lenient e.g p≤0.10 (10%)
answer
'significance level'
question
-whether result sig/ not always chance result incorrect (e.g fluke) -type 1 error, accepting hyp that is false -type 2 error, rejecting hyp that is true
answer
type 1 and 2 errors
question
-once know about level sig, decide which statistical test are going analyse your data to assess whether findings were sig or not 1. Chi- squared (x2) test 2. Wilcoxon T test 3. Mann- Whitney U test 4. Spearman's Rho (chimps will manage speech)
answer
inferential statistics tests 1234
question
1. does study investigate 'differences' or 'relationships'? -experiment used to investigate differences -correlation, will be investigating relationships 2. does study use 'nominal' or 'at least ordinal data'? -refers to 4 levels of measurements (nom, ord, int, rat) -nom or at least ord (ord, int, rat) 3. are the sets of data 'related' or 'unrelated'? -if was exp -related data, either repeated measures or matched pairs (cos 2 conditions linked/ related in some way, by using sim or the same P in each condition) -unrelated data, independent (cos 2 conditions sep/ unrelated) -if study was correlation, data is 'related' cos each P measured on both co-variables (making 2 measurements related)
answer
who do you choose which test to use?
question
-hyp predicts a difference between conditions (i.e exp) -level measurement nominal (i.e category) -data sets are unrelated (independent groups design)
answer
when to use Chi-squared (x2) test
question
-hyp predicts difference between conditions (i.e exp) -level measurement 'at least ordinal' (i.e continuous data) -data sets are related (repeated measures/ matched pairs)
answer
when to use Wilcoxon T test
question
-hyp predicts difference between conditions (i.e exp) -level measurement 'at least ordinal' (i.e continuous data) -sets data are unrelated (i.e independent groups design)
answer
when to use Mann-Whitney U test
question
-hyp predicts relationship (i.e correlation) -level measurement 'at least ordinal' (i.e continuous data) -data sets are related (each P measured on both co-variables)
answer
when to use Spearman's Rho test
question
-result of statistics test called 'observed value' (cos based obvs made in research) -often referred to as 'calculated value' (cps result calculated using statistics test) -to decide if observed (critical) value sig figure compared to 'critical value' (a number listed in a table of critical values that does not change) -fidd tables of critical values for each inferential statistics test, i need find the appropriate critical value for your study depending on type of study i did
answer
interpreting results of statistical test
question
1. 1-tailed or 2-tailed test - if hyp directional, use 1-tailed test, non-directional use 2-tailed test 2.significance level - can vary from study to study but most commonly used level usually p
answer
to find the appropriate critical value in a table you need to know
question
-for some inferential tests, observed (critical) value is sig if it is equal to or exceeds critical value, others reverse -will find this info stated underneath each table of critical values -one way remember which test requires observed value to be higher than critical value & which tests requires opposite is by seeing seeing if there is an 'R' (e.g speaRman's Rho or chi-squaRed) then the observed value should be gReateR than the critical value -if no 'R' (e.g mann-whitney and wilcoxon) then observed value should be less than critical value -if observed value sig, means accept hyp & reject null hyp, if it is not must accept null & reject hyp
answer
observed and critical values
question
-include any non-numerical data, words from written qa's, written observations, transcribed interviews, video footage, drawings -many research methods such as naturalistic observation, open ended questions, qa's, unstructured interviews, analysis diaries, notes, case study, generate qualitative data -diff ways analysing one most common content analysis
answer
qualitative data analysis and interpretation
question
-one way analysis qualitative data, turning qualitative (non-numerical) into quantitative (numerical) -gather materials to be analysed (e.g interview transcriptions/ video footage) -examine material to identify common themes -themes used generate categories, materials then analysed quantitatively to draw concs -pilot study often used as part of analysis to generate & test coding system to be employed -helps make sure categories are exclusive & discrete (i.e if no overlap) & all relevant themes included & clearly operationalised
answer
content analysis
question
-makes data easy analyse & present -rather having discuss in-depth, detailed info in further depth, allows researchers simplify info into clear categories -simplifies process of drawing clear conclusions, f.e quantitative data presented in clear & visual form of bar chart -speeds up process of analysis, makes research more accessible -over-simplifes data, by turning in-depth & detailed info into more simple form of categories -richness & detail lost -f.e may be seen as too simplistic to reduce a discussion of complex emotions down to basic categories -prob as content analysis may be unsuitable for studies were detail is essential
answer
evaluation of content analysis ao2
question
-taking quantitative data such as transcript of interview/ observational record to analyse theory/ hyp -involves organising the qualitative data into specific themes -themes then used as headings under which discussion can take place using relevant quotes to summaries the data to each heading
answer
thematic analysis
question
1. transcribe the data, if u need to, number each line & read text thru several times until know well, as read make notes of ideas that occur to you 2. divide text into 'meaning units', using a forward slash (/) between every apparent change in meaning or subject 3. search entire text for meanings that seem have sim theme & group these together, u could highlight these using different colours 4.keep adjusting themes as u continue sort thru data 5. once satisfied are no more themes to find u need define & name each theme to form categories 6. themes & categories then used draw conclusions about trends in data & what seems suggest about area being studied 7. write up report, need present a case for each theme & provide some supporting word for word quotations from text using original line number so it can be located if necessary
answer
stages in thematic analysis
question
-suitable read out to P, clear formal, courteous -explain procedures of study relevant to P including: -what required do/ how long take -make sure know free withdrawal any time -check understood instructions
answer
writing a set of 'standardised instructions'
question
-suitable read out to P, clear formal, courteous -begin by thanking them taking part -if deception used, make them aware of true aims -ask if have any q's about anything happened in study -remind details will remain anonymous -right withdraw data if wish, still be paid
answer
writing a 'debriefing statement'
question
-operationalised variables (e.g IV, DV, co-variables) -hypothesis (directional/ non-d exp hyp looking for difference including operationalised variables i.e both groups IV & DV and null hyp) -number (N) P -sampling method (i.e where get them from, how why) -experimental design (why) -controlling variables to improve validity (sampling & experimental design might help control P variables as well as another procedural or situational 'controls;) -ethical considerations (potential issues, how dealt with) -step by step procedure (chronological order) -data analysis (qualitative or quantitative including descriptive (e.g measure of central tendency & dispersion & graphs as well as inferential statistics)
answer
design your own study
