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;
131 Cards in this Set
- Front
- Back
|
Type 1 error |
Reject null wrongly. Shouldve retained. Due to high levels of significance |
|
|
Type 2 error |
Accept null shouldve reject. Low levels of significance |
|
|
Significance |
Observed diffs beyond chance |
|
|
Sign test |
Non parametric. Nominal. Rmd. Diff. Dont forget dont count no diff |
|
|
Wilcoxon |
Rmd. Diff. At least ordinal. |
|
|
Mann whitney who bitch |
Ig diff at least ordinal |
|
|
Pearsons r |
Rela at least interval rmd |
|
|
Ttest |
Rmd at least ordinal diff |
|
|
Chi2 |
Ig diff nominal. Prone to t1 errors. Two tailed |
|
|
Spearmans rho |
Rela rmd at least ordinal |
|
|
Define experimental design |
The way ppts are allocated to groups of an investigation |
|
|
Name 3 experimental designs |
Rmd ig mp |
|
|
Who is rmd |
Same do all conditions |
|
|
Pros of rmd |
Less ppts. No ppt variables. Cheap. |
|
|
Cons of rmd |
Diff stimuli material. Order effects. If ppt drops lose all data. Demand characteristics more likely. |
|
|
Who is ig |
2 Diff groups of people do one condition each. Iv manip differently. |
|
|
Pros of ig |
No order effects. Same stim material used. More time efficient. Lower chance of DC |
|
|
Cons of ig |
Ppt variables. More ppts more expensive. |
|
|
Who is mp |
Ppts matched in pair according to similar characteristics (relevant) then randomly assigned conditions |
|
|
Pros of mp |
Less ppt variables. Same stimulus material. No order effect. |
|
|
Cons of mp |
More ppts. Diff to match identically. Time and money. |
|
|
2 key features of observational design |
Temporal (predetermined times) event (every time interest behaviour performed) |
|
|
What do u need to do before performing observational research |
Devise operationalised behaviour categories: quantutative scores. Then use either TS/ES |
|
|
Anagram for questionnaire construction |
BRUSO. Brief. Relevent. Unambig. Specific. Objective. |
|
|
One form of questionnaire. Probs? |
Likert scale. Middle default or genuine |
|
|
What to consider when developing questionnaire |
Type of question. How record. Client. Interviewer. |
|
|
What to do before performing questionnaire |
Perform pilot study. Explain purpose get informed consent then ask questions |
|
|
Two type of Q used in questionnaire |
Closed open |
|
|
Features of open questions. |
More time and effort for both parties. Unbiased: more reliable and valid. Easy. Vague research q. Understand meaning. |
|
|
Closed questions features |
Gives good idea of responcs. Higher likelihood of bias and misunderstanding. More difficult to write. More obj |
|
|
What is an aim |
The purpose of a study driven by a theory to explain given observed behaviour. 'Goal' |
|
|
Define hypothesis |
Based on aims but more precise. Predictions of an investigation outcome that make specific reference to the IVs manip and DVs measured by researchers |
|
|
Directional hypotheiss |
Previous research. Predict one direction only. |
|
|
Non directional hyp |
Predictionof diff in 2 directions. Lack of previous research |
|
|
Null hyp |
Predicts a statistically significant effect or rela will not be found "no significant difference/rela between..." If supported =/= significant |
|
|
Alt hyp |
Predicts statistically significant effect of IV upon DV |
|
|
Define sampling |
Process of selecting a representative group from pop under study |
|
|
Representative sample |
Subset of target pop with similar distribution of characteristics |
|
|
What is random sampling |
Every member of population has an equal chance of selection |
|
|
Cons of random sampling |
Can't force participation so always element of volunteering. Impractical if large pop. Potential bias and non rep |
|
|
Pros of random sampling |
High pop vad. Likely to be rep - generalisable |
|
|
What is stratified sampling |
Subgroups in population identified then no. taken in proportion to representation. Representative |
|
|
Pros of stratified |
Avoids problems of misrepresentation caused by random. |
|
|
Cons of strat |
Time and resources. Care each key characteristic in each sample else bias. Access to info x. |
|
|
Opportunity |
Readily available. |
|
|
Pro opprt |
Easy cheap quick |
|
|
Cons opport |
Self selected. Biased low pop vad as highly motivated lots of free time. |
|
|
Systematic sam |
Every nth |
|
|
Pros of systematic |
Assuming random list should b rep. |
|
|
Con systemati . |
Periodic traits bias |
|
|
What is snowball sampling |
Ask one ppt and ask tgem to ask another. |
|
|
Snowball pro and con |
Bias due to limited pop. Study hard to talk about issues. |
|
|
What is volunteer sampling |
People who volunteered willing for experi |
|
|
Pros of volun |
Large sample (news ad) wide access to people. Low effort. Relatively cheap |
|
|
Cons of volunteer |
Bias af. Stable and outgoing. Dc and sdb |
|
|
What is relability. |
How consistently a method of measurement measures something. Similar each time = reliable. |
|
|
Two types of reliability. Explain |
External ability to replicate results - how consistent the method is over time and with other measures of the same thing. Internal how consistently a test/method is within itself - is it measuring the same thing |
|
|
3 ways of measuring reliability. Explain |
Split half method. Half items in method correlated with each other and compared. Measures IR Test retest. Test given to same ppts on diff occasions correlate. Strong + High ER INTER RATER RELIABILITY. Assessment of how consistent the scoring of a ppt behaviour is with another observer. Sep. Correlate. |
|
|
How to improve reliability |
>1 measurement per ppt reduce anoma Pilot studies Standardisation Check data |
|
|
Define validity |
How truly/realistically a method measures something. |
|
|
DEF IN E EXTERNAL VALIDITY |
How much results can be generalised to diff settings. |
|
|
Two types of external valid |
Temporal (successfully applies across time) ecological (generalise to real world behaviour) |
|
|
Define internal validity |
Whether results obtained solely due tO IV manipulation |
|
|
2 types of internal validity |
Construct (measure concept it is supposed to?) Concurrent (whether measurement is in agreement with pre-existing measures that are validated to test the same concept) |
|
|
What is face validity |
Whether it looks subjectively promising that it is measuring what it is supposed to. |
|
|
Content validity |
Ask expert in field if study = accurate |
|
|
Criterion validitu |
Predictive concurrent. |
|
|
Improve validity |
Control extraneous variables Standardise Eliminate dc and ie Replicate at diff times and places. Ensure measuring/operationalising things that can actually b measured |
|
|
How to "control" experiment |
Counterbalance. Randomisation in presentation of trials no systematic errors that order of trials present. Random allocation lower chance of ppt variables lower chance of individual diffs that systematically and consistently affect results. Standardisation all elements totally identical sensitive to change that can be attributed to IV manip |
|
|
Define operationalisation |
Defining key variables precisely so that they can be measured observed and quantified. Avoids ambig. Maintains IV |
|
|
What is a confounding variable |
An uncontrolled variable that brings about a change in the DV which is falsely attributed to the IV. Affects internal validity of the research |
|
|
What is an extraneous variable |
A variable that could affect DV but is identified and controlled before study begjns. |
|
|
Variable |
A factor free to vary in some quality |
|
|
Examples of some EVs and confounding Vs |
Researcher variables (consistency) situational (standardise) ppt (matched pairs) |
|
|
Investigator effects |
The ways in which the researcher unconsciously influences results. Interpretation bias. Voice. Physical. Reduce with double blind trial |
|
|
What is DC |
Ppts change natural behaviour in line with interpretation of aims of study. Please them or screw u. Reduce with blind trial |
|
|
Hawthorne effect |
Increased attention from researcher becomes a confounding variable. Try harder than normal as aware are being appraised. |
|
|
Sdb |
Present self in best light and socially acceptable fashion |
|
|
Features of science |
Forge rochs Falisfy. Objective. Replicable. Heneral laws. Empiricism. Rational. Operationalisation. Control. Hypothesis. Standardisation and systematic |
|
|
Bad science |
Bucsmuccs No Blind testing. Unrepresentative data. Cherry picked data. Sensationalised headlines. Misinterpreted results. Unreplicable results. No Control. Correlation =/= causation. Speculative language |
|
|
Define falsification |
The ability to show a theory is false. Shohkd actively try and disprove if fail strong support |
|
|
Induction |
Develop laws to fit data already observed |
|
|
Deduction |
Develop theory then test to confirm |
|
|
Why reject scientific approach |
Cant study people in the same way as physical phenomena. Lab = unnatural = dc. Control thus studies = one variable reductionist. Treated as passive info therefore superficial. Critics believe notion of objectivity is a mythas past experiences beliefs and ideas make it impossible. Only overt behaviour. Generalisation. |
|
|
Economy |
Go back to work. Disorders mother. Efficient use of public money ewt. Taxes reduce burden on public sphere |
|
|
Pilot study |
An initial run through of experimental procedures to test all aspects of the design on a small scale trial. |
|
|
Pros of pilot study |
Saves time and money by identifying flaws in procedure. Spot ambiguities or if task too hard (floor effect or ceiling effect) results sig?
|
|
|
Sections of a report |
Title. Concise but informative about. Abstract. Brief study summary research question methods findings conc. Intro. Background reasoning predictions context. Method. Design particip apparatus procedure. Results. Descriptive stats inferential stats. Discussion. Summarise prac ap limit. References. |
|
|
Design experiment |
Type of hyp. Method choice. How gathered. Target pop and sampling. Ethics. Pilot. Stat test. Expectation. |
|
|
Theory |
A proposed explanation for cause of behaviour. Must be a logically organised set of prepositions that defines events describes relas and predicts of occurence of events. Guide research by offering testable hypo |
|
|
Qualitative reports |
|
|
|
Ethics. |
Psycho harm. Anon. Confid. Rtw. Debrief. Deception. Informed consent. |
|
|
Ethics in ob research |
Consent if overt covert only where expect to be watched |
|
|
Define peer review |
Takes place before publication to ensure high quality research. Assessed by experts in same field. Provide comments and recommendations. Analyse significance validity and originality. |
|
|
Purpose of PR |
Allocation of research funding. Publication in journal validates. Assess research and rating of unis. Prevent faulty research entering public domain. |
|
|
Pros of PR |
Same field reliable accurate Maintains high research standard Prevent fraud Validation and precaution prevents damaging research from entering society. |
|
|
Cons PR |
Smith: slow expensive highly subjective prone to bias easily abused useless at detecting fraud Cant always find expert poor res passed File drawer phenomenon only stat sig published challengers and neg not published. |
|
|
Experimental method |
Designed to establish causality |
|
|
Lab study |
An experiment carried out within controlled environ |
|
|
Pros lab |
High IV. CE usually established. Control increases repicability |
|
|
Cons of lab |
Problems operationalising i and d amd in doing so measure of behaviour may become too specific. Low ev and mundane realism. Experimenter effects |
|
|
Field |
An experiment carried out in a natural environ but situation artificially set up - thus some factors/variables still controlled IV manip |
|
|
Pros of field |
Minimal experimenter effects. High ev and mr |
|
|
Cons of field |
Less control ev may still affect. Difficult to precisely replicate. Ethics IC. sample bias. |
|
|
Natural ex |
Makes use of existing IVs same pc as field |
|
|
Quasi. |
Naturally occurring IV between people that already exists e.g. gender |
|
|
Pros cons quasi |
Studies real effects eco vad. Confounding variables. Aware of study. |
|
|
Natural observation |
Plans observes and record of ppts in natural environ. |
|
|
Pros of natural observation |
High eco v. Demand characteristics |
|
|
Cons of natural ob |
Sample bias. Hard to replicate and to check observer reliability later. Observer bias. Causality. Ethics. |
|
|
Controlled observation |
Carrued out in psych lab. Controlled environ. Design and coding scheme agreed scale. Overt |
|
|
Pros of controlled ob |
Variables controlled. Quan data easy to analyse. Replicable so easier to test for reliability. |
|
|
Cons of controlled ob |
Unnaturally. Observer bias. |
|
|
Non participant pros Nd cons |
Record at time. Objective. Observer bias. Meaning behind behaviour unknown |
|
|
Participant pros and cons |
High ev. Easier to understand behaviour. Observer bias too involved retrospective unreliable. |
|
|
Correlations |
Measure to what extent 2 variables are related. Expressed as correlation coefficient. |
|
|
Cons of correlations |
Causality. Extraneous influences. Only linear relationships. |
|
|
Pros causality |
Predictions. Easy to replicate. Use as support. Allows researcher to investigate rela between variables that cant b manipulated |
|
|
Case study |
Idiographic in depth study of an individual group event or community. Uses data from range of sources questionnaires interviews observations records primary and secondary data. Provides insight into theory of psychological research too unethical to induce. |
|
|
Pros of case studies |
Identify and analyss complex interactions of many factors. Rich source of meaningful data. Stimulates more research. High ev. |
|
|
Cons of case studies |
Low pop vad. Memories. Researcher bias. Difficult to replicate cant test reliability. Causality and generalisability. |
|
|
2 self report techniques |
Interviews questionnaire |
|
|
Types of intervies |
Unstructured. Casual convo no set questions ppt given opportunity to raise topics feel relevant. Follow up qs infer deeper mean. Semi combo. Structured. Identical closed questions. Interviewer write down. Objective standardised. |
|
|
Pros of interviews |
Ease misunderstand. Clarify. Complex face to face. Quan and qual data analysis. |
|
|
Cons interview |
Ethics ic. Not articulate. Interviwer effects. Dcs. Sdb. Training time consuming. |
|
|
Questionnaires. Important to rem |
Avoid leading qs bias offence or ambig. |
|
|
Pros of questionnaire |
Replicae. Reduced investigstor effects. More willing to reveal personal info. Large no. |
|
|
Cons of questionnaires |
Low response rate. Low pop vad. Not econimal. Not as flex as interviews. Misunderstand. Time. Sdb. Superfiical issues. |
|
|
Content analysis |
Type of observational research that provides a systematic method of analysing recorded behaviour and communication through creation and identification of behavioural categories. Quan. Themes specific characteristics. |
|
|
Pros content anal. |
Unobtrusive way of communication examination. Large vol. Replicable. Infer deeper meanjng |
|
|
Cons content analy |
2+ researxhers. Cant generalise unless used a probablistic sampling method. Time con if large vol. Subject to interpretative errors. |
